University of Southern California Dissertations and Theses

Addressing unmet needs and harnessing social support to improve diabetes self-care among low-income, urban emergency department patients with diabetes

(USC Thesis Other)

Addressing unmet needs and harnessing social support to improve diabetes self-care among low-income, urban emergency department patients with diabetes

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Addressing unmet needs and harnessing social support to improve diabetes self-care among
low-income, urban emergency department patients with diabetes
By
Elizabeth Rhea Erwin Burner, MD, MPH, MS
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFONIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(PREVENTIVE MEDICINE)
May 2023
Copyright 2023 Elizabeth Rhea Erwin Burner
ii
Dedication
This dissertation is dedicated to my children, Caroline and Andrew, for their joy and patience;
to my husband Dale, for his humor and tolerance of my cockamamie plans; and to my parents,
Rita and John Erwin, for their belief that my limits are only set by myself.

iii
Acknowledgements
I would like to thank my committee members, Drs. Lourdes Baezconde-Garbanati, Dr. Cecilia
Patino-Sutton, Dr. Shinyi Wu, Dr. Wendy Mack and Dr. Ricky Bluthenthal, for your guidance and
support during this doctoral journey, and in the years preceding it when I embarked on this
path as an aspiring physician-scientist. Thank you to all the USC faculty, staff, and doctoral
students for your support and respected advice in getting me to this point. I would like to
extend a special thank you to Marny Barovich and Dr. Jonathan Lam for encouraging me to
make this jump, and to Renee Stanley for helping me get back on track after the chaos of 2020
(and 2021 and 2022!) calmed down.

This dissertation would not be possible without the patients of the Emergency Department of
LAC+USC, whose grace and joyfulness inspired me to understand and combat the inequities
they face. The expertise and guidance of Janisse Mercado and Adriana Aleman, and their years
of work to understand this community made these studies possible. The support of the
Southern California Clinical Translational Science Institute has been critical as I have gained the
skills needed to undertake these studies.

Finally, I would like to thank my family for making this dissertation a reality. My husband Dale,
who has always encouraged my plan, without regard to the extra load it places on him, my
children Caroline and Andrew, who have grown up with a professional student mother and do
not know the difference, and my parents, John and Rita, who always encouraged me to work
hard at things I love and modeled a commitment to community and justice.

This research was supported NIDDK (grant 5K23DK106538), NCATS/SC CTSI (grants
UL1TR001855-06A1 & KL2TR001854-06A1)

iv

Table of Contents
Dedication ...................................................................................................................................... ii
Acknowledgements ....................................................................................................................... iii
List of Tables ................................................................................................................................. v
List of Figures ................................................................................................................................ vi
Chapter 1: Introduction .................................................................................................................. 1
Background ........................................................................................................................ 1
Specific Aims .................................................................................................................... 10
Theoretical Model ............................................................................................................ 12
Chapter 2: Study 1 ....................................................................................................................... 15
TExT-MED+FANS: a Phase III Randomized Unblinded Trial of mHealth Augmented
Social Support vs Standard Social Support in Combination with a mHealth
Curriculum for Safety-Net ED Patients with Diabetes to Increase Social Support,
Psychological Functioning and Healthy Behaviors
Abstract ................................................................................................................ 15
Introduction ......................................................................................................... 18
Methods ............................................................................................................... 23
Results .................................................................................................................. 39
Discussion ...................................................................................................... 56
Chapter 3: Study 2 ....................................................................................................................... 66
Patient and Family Member Perceptions of a mHealth Intervention for Improving
Social Support and Self-care for Latino Patients with Diabetes in Los Angeles
Abstract ................................................................................................................ 66
Introduction ......................................................................................................... 69
Methods ............................................................................................................... 76
Results .................................................................................................................. 82
Discussion ........................................................................................................... 105
Chapter 4: Study 3 ..................................................................................................................... 121
Different people/Different needs: A subgroup analysis of a Randomized Controlled
Trial utilizing a Latent Profile Analysis of Psychosocial Indicators and a
Mixed-Methods Qualitative Analysis of Post Intervention Semi-Structured Interviews
Abstract .............................................................................................................. 121
Introduction ....................................................................................................... 125
Methods ............................................................................................................. 129
Results ................................................................................................................ 137
Discussion ........................................................................................................... 153
Chapter 5: Discussion and Summary of Implications ................................................................ 159
References ................................................................................................................................ 164

v
List of Tables
Table 1: Internet Communication Interventions to Improve Social Support for Diabetes Self-
care; modalities, study design and population .............................................................................. 5

Table 2: Patient baseline characteristics for Study 1 ................................................................... 42

Table 3: Supporter Baseline Characteristics for Study 1 .............................................................. 44

Table 4: Baseline Characteristics: Lost To Follow Up Vs Completed Six Month Assessment ....... 46

Table 5: 6-Month Change in Outcome Measures (6 month minus baseline) .............................. 49

Table 6: 12-Month Change in Outcome Measures (12 month minus 6 month) .......................... 50

Table 7: Sum of partial coefficients on A1C of randomization group and each secondary
health behavior outcome combined and individually modelled in a GSEM ................................ 55

Table 8: Initial Version of Study Codebook for Study 2 ................................................................ 82

Table 9: Final Study Codebook Study 2 ....................................................................................... 83

Table 10: Model Fit and profile size characteristics of proposed LPA models ........................... 133

Table 11: Profile Descriptive Name, Size and Rank of Profile Average of Each Scale ................ 137

Table 12: Characteristics and efficacy outcomes by predicted the latent profile ...................... 140

Table 13: Characteristics and efficacy outcomes of patients with and without primary care
access ......................................................................................................................................... 143

Table 14: Characteristics and efficacy outcomes by patient gender and supporter
relationship type ........................................................................................................................ 146

vi
List of Figures
Figure 1: Transactional Model of Stress and Coping Applied to Diabetes Self-Care. Modified
from Lazarus and Folkman ........................................................................................................... 14

Figure 2: Example of patient message and corresponding family member message .................. 28

Figure 3: Example FANS messages from each support domain ................................................... 30

Figure 4: Screening and Enrollment of patients into TExT-MED+FANS ........................................ 40

Figure 5: General Structural Equation Model: random intercept for each participant for A1C
and potential behavioral mediators ............................................................................................. 55

Figure 6: Code Co-occurrence Frequency Chart ......................................................................... 85

Figure 7: Code Cloud .................................................................................................................... 86

Figure 8: Normalized Distribution of Codes for High and Low Responders to
TExT-MED+FANS at 6 and 12 months ........................................................................................ 104

Figure 9: Screening and Enrollment of patients into TExT-MED+FANS and study flow ............. 130

Figure 10: Theoretical model of LPA analysis ............................................................................. 132

Figures 11 & 12 Andersen Aday Behavioral Model and Access to Medical Care ....................... 158

Figure 13: Integrated Transactional Model of Stress and Coping and Access to Care Model ... 160

1

Chapter 1: Introduction
Background
Although diabetes is a nationwide epidemic, US Latinos are a particularly vulnerable population.
Latinos are more likely than non-Latino Whites to develop diabetes and have higher rates of
diabetic complications and diabetes related mortality (Beckles & Chou, 2016; Geiss et al., 2014).
This disparity is prominent in Los Angeles County, where almost half the residents are Latino,
and diabetes prevalence is higher than the national average (Diabetes in LAC Adults, 2007;
Lawrence et al., 2009). Patient with diabetes who rely on the Los Angeles County + University of
Southern California Emergency Department for care are a high need population, with poor
glycemic control, low diabetes knowledge, and limited access to care (Menchine, Marzec,
Solomon, & Arora, 2013). The mean A1C --or three-month average measure of blood glucose,
also known as glycemic control-- for this population is 8.8. This is alarming, as increasing A1C is
associated with increasing rates of microvascular and some macrovascular complications
(Nathan, McGee, Steffes, Lachin, & Group, 2014; UKPDSGroup, 1998), with a significantly
increased rate at a level of 8 or higher (Huang, Liu, Moffet, John, & Karter, 2011). The patients
who resort to seeking care in the ED are not often adequately served by the current healthcare
system and require dedicated attention to address their unique health needs and barriers to
care.

The disparities in complications from diabetes are not inevitable. Good glycemic control can be
achieved with access to quality medical care combined with significant and sustained
behavioral changes. However, good disease management and glycemic control consists of
2
many daily health decisions and behaviors, impacting nearly every facet of the life of person
with diabetes. Patients might have to make changes to increase regular exercise, improve
nutrition choices, conduct regular self-monitoring of blood glucose, follow a medication
regiment, perform daily foot care, engage with healthcare systems, and practice stress
reduction (Margaret A. Powers et al., 2015; Sarkar, Fisher, & Schillinger, 2006; Toobert &
Glasgow, 1994). These behaviors (and subsequent glycemic control) in turn are linked to a
patient’s intentions, underlying self-efficacy, psychological distress and social support (Joni L.
Strom & Leonard E. Egede, 2012; van Dam et al., 2005). However, the interplay of these
behaviors must be carefully titrated as too tight of glycemic control can result in uncomfortable
and sometimes deadly episodes of hypoglycemia (low blood sugar) (Huang et al., 2011). Given
the herculean efforts to avoid diabetes complications, many patients face difficulty in
maintaining consistent good disease management, especially those from marginalized groups
and from low socio-economic positions who face additional challenges from community level
barriers to healthy choices (Gallegos-Macias, Macias, Kaufman, Skipper, & Kalishman, 2003;
Walker, Gebregziabher, Martin-Harris, & Egede, 2014).

However, culturally and linguistically appropriate interventions that emphasize improving self-
efficacy can combat this disparity, especially social support interventions (Kirk, Ebert, Gamble,
& Ebert, 2013; Steed, Cooke, & Newman, 2003). As classically formulated, social support
consists of four domains: emotional, instrumental, informational and appraisal (House,
Umberson, & Landis, 1988). Emotional support refers to expressions of empathy, love, trust
and caring which result in improved affect for the recipient, such as listening to the concerns of
3
a loved one. Instrumental support consists of tangible aid and services, such as driving a loved
one to a doctor’s appointment. Informational support generally refers to advice, suggestions,
and information provided to a loved one. Appraisal support is the provision of information
useful for self-assessment. While the American Diabetes Association recommends a mixture of
emotional, instrumental and informational support, there is value in appraisal support,
particularly in the context of behavior change interventions using motivational interviewing.
Interventions designed to increase general or disease specific social support should
intentionally target one or more of these arenas (Michele Heisler, 2007).

Harnessing social support may be the key to this recalcitrant problem, particularly for Latino
populations. There is evidence that appropriate social support is associated with improved
health behaviors and glycemic control for diabetes patients (Lindsay S. Mayberry, Rothman, &
Osborn, 2014). Training family members and peers to support patients with diabetes has been
shown to improve patient motivation, healthy behaviors and glycemic control (Hu, Wallace,
McCoy, & Amirehsani, 2014; Kirk et al., 2013; Spencer et al., 2018; Tricia S. Tang, Ayala,
Cherrington, & Rana, 2011; T. S. Tang, Funnell, Sinco, Spencer, & Heisler, 2015; van Dam et al.,
2005). These interventions have been enthusiastically accepted in multiple Latino populations
(Berg, Linden, Adolfsson, Sparud Lundin, & Ranerup, 2018; Teufel-Shone, Drummond, & Rawiel,
2005; Thompson, Horton, & Flores, 2007; Two Feathers et al., 2005). However, existing
research in developing social support interventions have had inconsistent results. Additionally,
these interventions have faced difficulties in recruiting families members and create financial,
space and personnel burdens to implement and maintain (Michele Heisler, 2007; L. S.
4
Mayberry, Berg, Harper, & Osborn, 2016). These barriers have prevented widescale
implementation of social support interventions for patients with diabetes.

To reduce these burdens, new interventions using internet and communication technologies
(ICT) to engage supporters have begun to emerge. ICT interventions -- including mHealth,
telephone based interventions and web-based interventions—can be used to train support
persons, to prompt them to provide appropriate social support and as an avenue to offer that
support to a person with diabetes (John D Piette, Kerr, Richardson, & Heisler, 2008). Employing
ICT for social support interventions could result in family members who are the most influential
being trained to provide support, rather than the ones who live closest to the patient with the
most available free time to travel. ICT may also broaden the reach of interventions that
develop new support relationships between peers, matching people with diabetes to those
with similar experiences and challenges rather than those patients that share a medical home.
Merging ICT and social support interventions could generate a solution that has the cost-
effectiveness and scalability of mobile technologies coupled with the personal touch of social
support. ICT based social support interventions could reduce the need for physical presence
and make social support interventions more accessible to populations in need.

ICT based social support interventions are currently deployed through a range of modalities. A
review of truly ICT based interventions to increase social support for diabetes self-care (defined
as 2 or fewer “in-person” encounters required), we identified 13 currently published
interventions. In these interventions, there is a broad mix of ICT modalities currently in use,
5
indicating the changing landscape of this field. (see Table 1). The most commonly used ICT
modality was peer support web-forums, with peer support offered via spontaneously
developed virtual relationships; this strategy was also used in combination with text-messaging
and internet connected medical devices to engage supporters. Clinical outcomes tended to be
neutral to positive with a general tendency towards improved psychosocial and clinical
outcomes, however these outcomes were heterogeneously reported. The most commonly
targeted social support domain was emotional support. Interventions were currently roughly
balanced between engaging current supporters of patients and developing new social support
relationships. There were few patterns of study designs that showed trends towards more
effective interventions, with the exception of including an explicit measurement of social
support (E. Burner et al., 2018; M. Heisler & Piette, 2005; M. Heisler, Vijan, Makki, & Piette,
2010; Lari, Tahmasebi, & Noroozi, 2018). In short, there is no dominant direction that
interventions are currently headed, nor consistent pattern of results to guide future studies.

Table 1: Internet Communication Interventions to Improve Social Support for Diabetes Self-
care; modalities, study design and population
Intervention name (citation) Study Design Population
Interactive Voice Recording-based Interventions
CarePartners (J. D. Piette, Rosland,
et al., 2013)
quasi experimental
(comparator groups of
patients with and without
supporter not randomized)
Midwestern VA patients
(n=285, not explicit how
many with supporter)
CarePartners (J. D. Piette et al.,
2011)
quasi experimental
(comparator groups of
patients with and without
supporter not randomized)
Honduras (n=89, 40 with
supporter)
CarePartners (J. D. Piette, Marinec,
et al., 2013)
quasi experimental
(comparator groups of
patients with and without
supporter not randomized)
Spanish speaking patients
in US, Honduras and
Mexico (n=171 with
diabetes, 91 with
supporter)
6
CarePartners (J. D. Piette et al.,
2016)
RCT (IVR to patient and family
member vs IVR to patient
alone)
Midwestern VA patients
(Subset with Diabetes,
n=39; 22 with supporter)
Reciprocal Peer Support (M.
Heisler et al., 2010)
RCT (peer support vs nurse
case manager support)
Midwestern VA patients
(n=244, 126 with supporter)
Reciprocal Peer Support (M.
Heisler & Piette, 2005)
Quasi experimental (no
comparator)
Midwestern VA patients
(n=40, all with peer
supporter)

Short-Message-Service/Text message-based Interventions
TExT-MED+FANS (E. Burner et al.,
2018)
RCT (SMS to supporter and
patient vs SMS to patient
alone)
CA, Low-income, Latinos
(n=44; 22 with SMS to
supporter)
FAMS (L. S. Mayberry et al., 2016) Quasi experimental (no
comparator)
Nashville, Low-income
(n=19, 7 with supporter)
PM+FF (Ramirez & Wu, 2017) RCT (SMS to supporter and
patient vs SMS to patient
alone vs usual care)
CA, Low-income, Latinos
(n=42; 14 with SMS to
supporter)

E-mail based Interventions
CarePartners (Aikens, Rosland, &
Piette, 2015; Aikens, Trivedi, Aron,
& Piette, 2015; Aikens, Zivin,
Trivedi, & Piette, 2014)
Quasi experimental
(comparator groups of
patients with and without
supporter not randomized)
Midwestern VA patients
(n=301; 118 with
supporters)
SUPPORT (Reese et al., 2016) RCT (supporter receives email
about patient's medication
adherence vs usual care)
US patients with Humana
insurance (n=200, 150 with
supporter)

Web Forum-based Interventions
MoDIAB (Berg et al., 2018) Quasi experimental (no
comparator)
Swedish women with
gestational diabetes, post-
partum (n=81, 61 used peer
support web forum)
MoDIAB (Linden, Berg, Adolfsson,
& Sparud-Lundin, 2018)
RCT (web-based support with
peer forum vs usual care)
Swedish women with
gestational diabetes, post-
partum (n=174, 83 in
intervention group)
D-NET (Glasgow, Barrera, McKay,
& Boles, 1999)
quasi experimental (no
comparator)
International population
(n=221)
D-NET (Glasgow, Boles, McKay,
Feil, & Barrera, 2003; McKay,
Glasgow, Feil, Boles, & Barrera Jr,
2002)
RCT (Peer support vs self-
management coach vs usual
care)
Pacific Northwest US
(n=160; not stated
randomization ratio used)
7
Virtual Clinic (Jennings, Powell,
Armstrong, Sturt, & Dale, 2009;
Powell, Jennings, Armstrong, Sturt,
& Dale, 2009)
Quasi experimental (no
comparator)
UK patients on insulin
pumps (n=17)

Mixed Modality-based Interventions
mDAWN (Ho, Newton, Boothe, &
Novak-Lauscher, 2015)
quasi experimental (no
comparator) of SMS, web-
forum and internet connected
medical devices
International general public
(n=56, 28 agreed to enroll a
supporter)
CD training+SMS (Lari et al., 2018) RCT (SMS to patient and
supporter with CD training in
home vs Usual care)
Southwest Iran (n=76, 40 in
intervention)
Family Defeating Diabetes
(McManus, Miller, Mottola,
Giroux, & Donovan, 2018)
RCT (web-forum & emails for
patient and male partner vs
usual care)
Canadian women with
gestational diabetes, post-
partum (n=170, 89 in
intervention)
Internet-based Mentoring
Program (Suh et al., 2014)
RCT (SMS to peer mentor vs
no peer mentor assigned,
patients upload glucose
readings to web)
Korea (n=57, 26 with peer
mentor)

While ICT based interventions are generally promising, mHealth specific interventions may be
more feasible for low-income Latino patients, such as those that seek care in the ED of
LAC+USC. Research indicates these patients are more likely to access the internet and health
information specifically via mobile devices, and maybe less likely to have a “land phone” and/or
personal computer per studies from the Pew Hispanic Center (Livingston, 2011). mHealth
interventions improve disease management of chronic illnesses, including diabetes (Bell, Fonda,
Walker, Schmidt, & Vigersky, 2012; Quinn et al., 2008; Quinn et al., 2011; Seto et al., 2012). It is
important to consider that many existing ICT and mHealth interventions require personal
computers or frequent engagement with mobile applications. These are not widely used in
resource-poor communities such as safety-net emergency departments (EDs) (Arora et al.,
2013). Additionally, Latino patients from low-resource settings have enthusiastically joined text-
8
message and interactive voice-recording interventions to improve diabetes self-care indicating
this is an acceptable modality for the population (Arora, Peters, Burner, Lam, & Menchine,
2014; E. Burner et al., 2018; Hay et al., 2018; Ramirez & Wu, 2017).

Importantly, ED care presents a unique opportunity to reach patients and families during a
health crisis, when they are susceptible to behavior change (Boudreaux, Bock, & O'Hea, 2012).
As emergency care is more expensive than healthcare in other sites (Galarraga, Mutter, &
Pines, 2015), ED interventions should maximize the benefit of these visits through improved
behaviors and health outcomes. ED-based interventions can also bridge patients until primary
care can be established for those lacking primary care, or during the wait for the next visit for
patients with limited access to a primary care provider. Those seeking care in the ED also
represent a group of patients with greater health needs, with higher rates of hospitalization and
worse glycemic control (Birtwhistle et al., 2017). By incorporating mHealth and other ICT
interventions into ED-based chronic disease management interventions, this behavior change
opportunity window may be prolonged.

TExT-MED+FANS (Trial to Examine Text-Messaging in Emergency patients with Diabetes+ Family
and friends Network Support) is an intervention designed for ED patients with diabetes at
LAC+USC. It adds a mobile social support module for family members and friends, FANS, to the
existing patient focused TExT-MED intervention. I developed the FANS intervention by
developing messages that corresponded to the existing patient messages (TExT-MED), focusing
on instrumental, emotional and informational support. This social support focused intervention
9
will engage family members in the goal-setting and behavior change process that the patients
are undergoing. TExT-MED+FANS uses mHealth to overcome the transportation and time
obstacles that social support solutions face by offering social support training via a mobile
platform. Using mobile training, a patient can select the most influential person to support
them, rather than the most proximate. TExT-MED+FANS will allow for a greater understanding
of the role of baseline social support in disease management, the effectiveness of ICT and
mHealth based interventions to improve diabetes control, and changes in perceptions and
motivations of disease management that take place through the trial. I propose three studies
to investigate these questions.

In the first study, I conduct a trial of TExT-MED+FANS compared to TExT-MED for patients with
a more traditional pamphlet based social support intervention for family members. In the
second study, I conduct a qualitative analysis of the changes in perceptions and motivations of
disease management and social support experienced by intervention group patients and
supporters. In the third study, I conduct a mixed-methods analysis utilizing a Latent Profile
Analysis of psychological and social factors and themes from the qualitative findings to identify
distinct subgroups of patients enrolled in the study and perform a subgroup analysis of the
TExT-MED+FANS trial to identify future directions for this research. Through these studies, I will
better understand the patient population and their specific needs and explore innovative
options to impact the psychosocial factors that lead to healthy behaviors and overall good
disease management

10
Specific Aims

Aim 1: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED +
pamphlet social support on clinical, behavioral and psychosocial outcomes among emergency
department patients with diabetes at the end of the six-month intervention and after a six-
month washout phase via longitudinal SEM modeling.

Aim 1a: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED
+ pamphlet social support on glycemic control (A1C) among emergency department patients
with diabetes at the end of the six-month intervention and after a six-month washout phase.
Hypothesis 1a: A text message module delivered to a patient’s FANS (Family And friend
Network Supporter) will improve A1C at six months compared to that of patients whose
FANS receive similar support information in a pamphlet by mail.

Aim 1b: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED
+ pamphlet social support on diabetes related clinical, behaviors and psychosocial well-being
among emergency department patients with diabetes at the end of the intervention and after
six-month post intervention phase.
Hypothesis 1b: A text message module delivered to a patient’s FANS (Family And friend
Network Supporter) will improve specific diabetes related clinical, behaviors and
psychosocial at six months compared to that of patients whose FANS receive similar support
information in a pamphlet by mail.
11

Aim 2: Evaluate experience with TExT-MED+FANS and impact on perceptions of disease, social
support and motivation for behavior change among intervention patients and supporters via
qualitative analysis of semi-structure individual interviews.
Research Question 2a: What were the user experience with and perceived benefits of
TExT-MED+FANS for supporters and patients?

Research Question 2b: What changes in perceptions of disease, motivations, social
support resources and coping strategies occur with participation in TExT-MED+FANS.

Aim3: To identify subgroups of the study participants with distinct psychosocial patterns and
examine behavioral and clinical properties of the patients in these subgroups via latent profile
analysis methods.
Exploratory Research Question 3a: What are the distinct latent psychosocial profiles
among patients?

Exploratory Research Question 3b: Do these latent profiles of psychosocial factors
predict response to TExT-MED+FANS?

Exploratory Research Question 3c: Do participant subgroups identified by qualitative
themes predict response to TExT-MED+FANS?
12
Theoretical Model

Diabetes self-care is a complex process that involves a patient navigating hundreds to
thousands of decisions daily. Managing self-care requires response to each of these potential
stressor and coping mechanism to make healthy choices. The Transactional Model of Stress
and Coping (Lazarus & Folkman, 1984) incorporates initial reactions or appraisals of stressors,
modifications of these stressors based on individual characteristics and situation, coping
responses to manage these stressors and eventual outcomes. In the primary appraisal, an
individual responds to a potential stressor by identifying it as of no relevance, positive or
helpful, or harmful or threatening. If the initial stressor is perceived as potentially harmful or
threatening, an individual judges whether they have the capacity to cope with this threat with
their intrinsic abilities (e.g. self-efficacy) and extrinsic resources (e.g. social support). The
individual then responds to the threat with beneficial or maladaptive coping strategies, which
fall into two categories: 1) Problem solving or 2) Emotional response. A third strategy of
Avoidance coping has been distinguished from emotional response, however this also may be
considered a sub-category (NS Endler, 1990). Importantly, coping strategies may impact an
individual’s perception of their capacity to cope with a stressor in the future.
The Transactional Model of Stress and Coping can be applied to diabetes self- because it
can incorporate both individual factors, environmental resources and how coping mechanisms
can influence future appraisals (Samuel-Hodge, Watkins, Rowell, & Hooten, 2008). The model
has been applied to several disparity populations including low-income African American
church attendees in the US south (Samuel-Hodge et al., 2008), poor whites in Appalachia
13
(Carpenter, Theeke, Mallow, Theeke, & Gilleland, 2017) and publicly insured Latinos in New
Mexico (Shah, Gupchup, Borrego, Raisch, & Knapp, 2012).
For the three proposed studies, the measures used can be attributed to each of the
components of the model (See Figure 1). The stressor is the continual need for diabetes
management. Primary appraisal, or the patient’s initial appraisal of disease management, is
measured by the Diabetes Distress Scale. Intrinsic resources involved in secondary appraisal
are measured by the Diabetes Empowerment Scale and the Diabetes Fatalism Scale, which
extrinsic resources involved in secondary appraisal are measured by several social support
measures: the Diabetes Family Behavior Checklist, the Diabetes Care Profile support questions
and the Norbeck Social Support Questionnaire emotional sub score and tangible sub score.
Coping is measured as problem solving strategies the involve controlling the stressors
(Summary of Diabetes Selfcare Activities and the Wilson 3-item medication adherence scale)
and emotional responses that involving changing the emotional and cognitive response to the
stressors (Diabetes Fatalism Scale and Depression-PHQ-9). The distal health outcomes are
represented by the clinical outcomes (glycemic control/A1C; BMI; Systolic blood pressure) and
patient centered measure (WHO Quality of Life -5).
In Study 1, an analysis of the randomized trial of TExT-MED+FANS, SEM will be used to
conduct the analysis, so that the inter-connectedness of the primary appraisal, secondary
appraisal and coping (healthy behaviors) and the relationship to the subsequent health
outcomes can be modeled. In study 2, the qualitative follow-up to TExT-MED+FANS, individual
semi-structured interviews and inductive analysis will allow for a better understanding of how
the intervention changes coping strategies and appraisals by exploring patient and supporter
14
perceptions of disease, motivations for change and strategies for disease management. In Study
3, the patient factors associated with primary and secondary appraisal and themes from the
qualitative analysis will be used to describe unique subgroups of patients with specific need
profiles and responses to the intervention, which will aid future intervention design.

Figure 1: Transactional Model of Stress and Coping Applied to Diabetes Self-Care. Modified from Lazarus and Folkman

15
Chapter 2: Study 1: TExT-MED+FANS: a Phase III Randomized Unblinded Trial of mHealth
Augmented Social Support vs Standard Social Support in Combination with a mHealth
Curriculum for Safety-Net ED Patients with Diabetes to Increase Social Support, Psychological
Functioning and Healthy Behaviors

Abstract:
Background: Patients with diabetes often seek acute and chronic care in safety-net emergency
departments (EDs). Mobile health (mHealth) – the use of mobile phones to provide medical
care and education – is a low-cost method to increase healthy behaviors for these patients,
increasing ED care’s value. ED-based mHealth programs have improved diabetes behavior, but
we do not know if a social support component will further improve diabetes management and
glycemic control, or if it increases the long-term effects of an intervention.

Trial Design: Unblinded parallel equal allocation randomized phase III trial

Methods: We recruited patients with A1C≥8.5%mg/dL and a text-capable phone during their
visit to a safety-net emergency department in 2017 and 2018. Patients selected a relative or
friend to be their supporter. All patients received 6 months of TExT-MED, an existing and
effective mHealth program for patients with diabetes consisting of motivational, educational
and healthy lifestyle challenge text messages in English and Spanish. While all patients received
TExT-MED, supporters were randomized 1:1 to 1) Family and Friends Network Support (FANS)
intervention, which consisted of daily text messages encouraging and instructing supporters on
providing diabetes-related social support for their loved one, or 2) active control: paper
augmented social support, the same FANS content in a pamphlet mailed to their home. We
16
measured change in glycemic control (A1C), self-report of diabetes self-care activities
(Summary of Diabetes Self-Care Activities) and medication adherence (Wilson 3-Item Scale). We
examined differences in hypoglycemic events and deaths between the groups at 6 and 12
months with Fisher’s exact test. We utilized mixed effects linear regression models to analyze
differences between the intervention group and active control at the end of the intervention
and after a 6 month post intervention phase for all outcomes. We conducted planned
subgroup analysis by gender, race and ethnicity, language preference, health literacy, baseline
frequency of mobile usage, physically proximity to supporter, and a post-hoc subgroup analysis
by baseline social support.

Results: 166 patients enrolled; 97 (58%) patients followed up at 6 months (45 FANS and 52
comparison patients) and 106 (64%) patients followed up after the 12 post-intervention phase
(50 FANS and 56 comparison patients). Combined intervention groups patients showed equally
large, clinically significant improvements in A1C (combined group change 1.36%mg/dl (0.87 to
1.83), between group (0.14%mg/dl (-0.83 to 1.11)). Patients in the pamphlet social support arm
had a mean decrease in A1C of 1.42%mg/dl (95%CI 0.82 to 2.02), while patients in the FANS
arm had a mean decrease in A1C of 1.28%mg/dl (95% CI 0.48 to 2.09), with a combined sample
change of. Self-care behaviors similarly improved equally across groups. There was no
difference in rates of hypoglycemia or patient deaths.

In prespecified subgroup analysis, we found that among patients with a new diagnosis of
diabetes, FANS group patients improved their A1C to a greater degree than patients in the
17
pamphlet social support arm, mean between group difference of 1.96%mg/dL(95%CI -3.81 to -
0.125, p between group= 0.036). We found no differences at 6 months based on gender, race and
ethnicity, language preference, health literacy, baseline frequency of mobile usage, physically
proximity to supporter, and baseline social support.

Conclusions: In this randomized controlled trial of an mHealth augmented social support
curriculum added to an existing patient focused mHealth study in ED patients with diabetes, we
did not find a difference between groups; the intervention groups each improved glycemic
control and healthy behaviors and maintained this improvement in the post intervention phase.
Patients who were newly diagnosed with diabetes may benefit the most from the mHealth
augmented social support and special attention to support person training in this time period
may be critical.

Funding: NIDDK Award 5K23DK106538

18
Introduction:

Although diabetes is a nationwide epidemic, US Latinos are a particularly vulnerable population.
Latinos with diabetes have higher rates of retinopathy, nephropathy, peripheral vascular
disease and uncontrolled hypertension than the general population with diabetes (Spanakis &
Golden, 2013). Latinos are twice as likely as non-Latino Whites to develop diabetes and are 50%
more likely to die from it (Dominguez et al., 2015). It is estimated that one in two Latino
children born in the year 2000 will develop diabetes during their lifetime (Diabetes in LAC
Adults, 2007; Gregg et al., 2014; Narayan, Boyle, Thompson, Sorensen, & Williamson, 2003).
Nowhere in the country are the effects of this disparity seen more prominently than in Los
Angeles County, where almost half the residents are Latino and the prevalence of diabetes is
50% higher than the national average (Torres Mosst, 2017).

Social support interventions utilizing family members and peers to provide emotional and
informational support for patients with diabetes have shown improvements in patient
motivation, healthy behaviors and glycemic control (Tricia S. Tang et al., 2011; van Dam et al.,
2005). Latinos in particular have rated these interventions highly (Teufel-Shone et al., 2005;
Thompson et al., 2007; Two Feathers et al., 2005). However, social support interventions
require: 1) in-person training of family and friends at a specific physical location to be
supporters 2) coordinating schedules and physical location between the patient and their
supporter and 3) the cost of providing physical space and personnel to train these supporters.
As this training and support usually occurs face-to-face, social supporters are often limited to
people who are proximate to the patient and have the time available to be trained, rather than
19
the most influential person in the patient’s life regardless of location. Due to this requirement
for live training, social support interventions have limited scalability in populations with limited
transportation and financial resources.

Mobile health (mHealth) is the use of mobile phones to provide public health and medical
solutions. mHealth interventions have been successful in improving disease management in
chronic illnesses, including diabetes(Ahmed et al., 2009; Bell et al., 2012; Seto et al., 2012).
While initial interventions were computer based, mobile phone based interventions are more
efficacious(Pal et al., 2018). However, many of the current interventions require smartphones,
which while widely available in resource-poor communities, are not frequently used at the
maximum technical capacity of the phone. In prior work with this population, less than 10% of
Latino patients at LAC+USC know how to use smartphone applications, but over 80% have text
message capable mobile phones (Arora et al., 2013). TExT-MED (Trial to Examine Text Message
for Emergency Department (ED) Patients with Diabetes), a scalable, low-cost, bilingual text-
message based intervention designed to increase diabetes-specific knowledge and change
health behaviors in patients with poorly controlled diabetes was completed with this
population. TExT-MED was enthusiastically accepted by patients, and resulted in positive
behavior change, and improved glycemic control among Latinos and Spanish speakers (Arora et
al., 2014).

mHealth can overcome the transportation and time commitment obstacles that existing social
support solutions face. Family member social support training via mHealth will increase the
20
scalability of social support interventions by eliminating the need for in-person training by
moving social supporter training to a mobile platform. This allows the most influential rather
than the most proximate and available supporter to be trained and reduces the time and travel
burden on the supporter. Using mobile training, a patient can select anyone from their social
support network, regardless of physical location to be a “FANS” (Family And friends Network
Supporter). Adding this mobile social support module, TExT-MED+FANS builds on the success of
the original TExT-MED intervention by adding an emotional and highly personal touch to
enhance results.

The ED provides a unique environment to reach patients and their social supporters during a
health crisis when they are highly susceptible to behavior change. As ED care is more expensive
than care in other sites, this intervention can maximize the benefit of this visit through
improved behaviors and health outcomes. TExT-MED+FANS can bridge patients until primary
care can be established for those lacking primary care, or until the next visit for patients who
already have a primary care relationship (up to 6 months in the LAC system).

Integrating the reach of mHealth with the potential benefits of augmenting social support is a
nascent field of research and has the potential to create scalable, effective interventions that
can be translated into clinical care. By studying such an intervention in a safety-net population
with poor disease control and significant time and travel limitations, TExT-MED+FANS can be
tested in a high need population with a greater chance to show potential benefit. This
intervention to improve self-care behaviors may result in covarying changes in psychological
and behavioral measures and requires an analysis that will allow these multiple interactions to
21
vary over time and to possibly change in magnitude of relationship. Structural equation
modeling allows for this type of complex mediation analysis and will model how appraisals and
coping strategies are associated with clinical outcomes. In this study, we examine the effect on
ED patients’ diabetes outcomes in a randomized controlled trial of (1) an mHealth augmented
social support curriculum synchronized to a an existing patient focused mHealth study
compared to (2) a paper based social support curriculum for family members in dated pamphlet
form added to the same patient focused mHealth curriculum.

This study had the following primary and secondary aims:
Aim 1: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED +
pamphlet social support on clinical, behavioral and psychosocial outcomes among emergency
department patients with diabetes at the end of the six-month intervention and after a six-
month washout phase via longitudinal SEM modeling.
Aim 1a: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED
+ pamphlet social support on glycemic control (A1C) among emergency department patients
with diabetes at the end of the six-month intervention (primary aim) and after a six-month
washout phase (secondary aim).
Hypothesis 1a: A text message module delivered to a patient’s FANS (Family And friend
Network Supporter) will improve A1C at six months compared to that of patients whose FANS
receive similar support information in a pamphlet by mail.

22
Aim 1b: Compare efficacy of TExT-MED+FANS mHealth augmented social support to TExT-MED
+ pamphlet social support on diabetes related clinical, behaviors and psychosocial well-being
among emergency department patients with diabetes at the end of the intervention and after
six-month post intervention phase.
Hypothesis 1b: A text message module delivered to a patient’s FANS (Family And friend
Network Supporter) will improve specific diabetes related clinical, behaviors and psychosocial
at six months compared to that of patients whose FANS receive similar support information in a
pamphlet by mail.

23
Methods:

Study Design: This study was a unblinded, randomized, parallel, active controlled intervention
with 1:1 allocation ratio. All patients received a SMS text-message based mHealth curriculum
for diabetes self-care. Support people for half of the patients were randomized to receive a
supporter curriculum delivered by SMS text-message; the other half of supporter were
randomized to receive the curriculum by pamphlet organized by week of the curriculum.

Ethical considerations and Trial Registration: IRB approval for this study was obtained prior to
study initiation from the USC Health Sciences Institutional Review Board (HS-17-00406). The
trial was registered with ClinicalTrials.gov (NCT03178773); protocol is accessible on
ClinicalTrials.gov.

Patient Screening, Eligibility Criteria and Recruitment:
Trained research assistants (RAs) conducted screening and enrollment from 7am to 10pm on
weekdays and 11am to 5pm on weekends in the LAC+USC ED from July 2017 to October 2018.
They surveyed the ED electronic patient tracking system for patients with diabetes. Inclusion
criteria were age 18 or greater, A1C 8.5%mg/dl or greater, and able to consent (were not on a
psychiatric involuntary hold, in police custody, or too clinically unstable to consent and
complete baseline assessments). Patients were excluded if they did not have stable ownership
of a mobile phone for 30 days or more, were not able to send and receive text message, did not
read English or Spanish, and could not identify a support person who could be contacted within
2 weeks to enroll. Only subjects with A1C ≥8.5%mg/dl were enrolled in this trial, as these
24
patients have the greatest need for intervention and the potential to demonstrate a beneficial
intervention effect, as lower baseline A1C are subject to floor effects with behavioral
interventions.(Harkness et al., 2010) The A1C-based eligibility requirement was verified in the
emergency department during the patient’s visit using the Afinion AS-100 capillary point-of-
care A1C meter (Axis-Shield PoC AS, Oslo, Norway). Patients who reported both type 1 and 2
diabetes were enrolled, as prior work with this population has shown that 30% of patients are
unsure which type of diabetes they have (Menchine et al., 2013). RAs explained the purpose of
the study and obtained written informed consent in the language of the patient’s preference.
To be eligible, patients had to identify a family member or friend to agree to serve as a FANS.
Only one FANS was enrolled per patient, as the intervention instigates communication between
the patient and supporter, and we did not wish to overwhelm patients with suggestions for
healthy living from multiple loved ones. Patients were aware at enrollment that the designated
supporter could receive multiple text-messages per day and would be prompted to offer
increased support.

Patient Enrollment & Randomization
The goal of this investigation was to study the effect of augmenting existing social support via
mHealth; patients were only eligible to complete enrollment if a supporter agreed to
participate as well. Contacting, consenting and enrolling a supporter took up to three weeks.
Given the potential lag, all potentially eligible patients were registered in the SHERPA platform
used by Agile Health while in the ED. Occasionally, patients lacked cellular service in the ED and
were not able to text back the Federal Communications Commission-required YES message to
25
opt into the patient-focused side of the intervention during their initial ED visit. For these
patients, we texted and called daily for up to one week until they texted back in YES. The
supporter was not contacted until after the patient was fully registered in the system. If a
supporter was not eventually enrolled as well, the patient still received the patient text-
messaging program, however they were excluded from further participation in the study.
Randomization group was assigned by sealed envelope allocation. Randomization sequence
was generated by the senior study biostatistician, and the randomization envelope was opened
by a research assistant after a supporter consented to participate in the trial.

Supporter Enrollment & Randomization
Supporters were enrolled in the FANS program either in the ED during the initial contact with
the patient or remotely by telephone if they were not present in the ED. Each patient was
instructed to select the person who would provide them the most support and ranked up to
three potential family member or friends to serve as a FANS. We collected multiple contact
numbers for each potential FANS from each patient. We called daily for up to three weeks to
enroll supporters. Enrolling supporters consisted of verbally consenting the support person,
confirming age>18 and the ability to send and receive text messages. Supporters then had to
complete baseline survey instruments and registration in the SHERPA platform for the
intervention arm. Registration in the SHERPA system required a YES text back from the
supporter. If the supporter did not respond with the required text-in YES message, we called
them daily until the start date of text-messages for patients to remind and assist them with
completing registration. After supporter enrollment was completed, the dyads were
26
randomized to the FANS mHealth-augmented social support or pamphlet-based social support
education for the supporter. All patients received the TExT-MED patient program.

Intervention Description

Patient Intervention – TExT-MED – All Patients Received
The original TExT-MED curriculum description and development were previously described
(Arora et al., 2014). TExT-MED was developed from National Diabetes Education Program
(NDEP)(Diseases) messages adapted to the character constraints of text messages (160
characters); messages emphasized education and behavior change. TExT-MED was a six-month,
fully automated, text message-based program designed to increase knowledge, self-efficacy
and subsequent disease management and glycemic control. The twice-daily text messages for
patients consisted of 1) educational/motivational messages 2) medication reminders 3)
diabetes trivia questions and 4) healthy living challenges.

After the original TExT-MED study was completed, the program was purchased, modified, and
commercialized by Agile Health into a new, enhanced program called MyAgileLife. MyAgileLife,
consists of three messages a day and has a greater focus on skills than the original TExT-MED
intervention – including setting goals, tracking progress, enabling social support, creating
environmental cues, and celebrating success. MyAgileLife also contains messages designed to
increase engagement by including trivia questions and patient self-assessments of motivation
and disease management which request a response. Due to technical issues related to the
27
timing of the messages in this study, we used a slightly locally modified version of the
MyAgileLife program.

FANS Family member/Supporter Intervention
FANS Curriculum Development
The FANS support messages were initially developed from National Diabetes Education
Program and American Diabetes Association (ADA) recommendations for family members to
offer emotional, informational, and instrumental support. The existing patient curriculum of
text messages was reviewed, and of the 3 daily patient text messages, 2 messages were
selected to develop a coordinating supporter FANS text message. FANS text messages were
then translated into Spanish by a native Spanish speaker, and back translated by a native
Spanish speaker of a different origin country to ensure retention of meaning. Four family
members of patients with diabetes, all bilingual native Spanish speakers, then reviewed all
English and Spanish FANS and TExT-MED patient messages to confirm retention of meaning.
The messages were also tested and refined with a group of Spanish-speaking promatoras—
community health workers with special training in health education who are valued opinion
leaders in their neighborhood. The promatoras identified messages that needed further
development. In particular, they noted translations that had negative connotations or nuances.
For example, a message in English stated, “to get things in order”; the Spanish translation of
“sigue en linea” reminded some promotoras of current political discourse regarding waiting for
one’s turn to cross the border. Due to the potential negative connotations, we were careful to
reword such messages.
28

Theoretical basis of FANS curriculum
The FANS messages for supporters mirrored the patient messages, and the coordinating
messages were sent to the patient and the supporter synchronously (see Figure 3). This
synchronous message delivery was designed to instigate conversation between the patient and
the supporter, increasing the “stickiness” of the message. The FANS messages were constructed
on a model of four arenas of social support: 1) Instrumental support (tangible goods and
actions), 2) Informational support (knowledge), 3) Emotional support and 4) Appraisal support
(feedback regarding accuracy of beliefs and appropriateness of actions).(Langford, Bowsher,
Maloney, & Lillis, 1997) The NDEP and American Diabetes Association (ADA) recommend a
combination of emotional, informational, and instrumental support from loved ones. Given the
financial constraints of many of the patients and family members of this safety-net ED
population, the FANS messages emphasized emotional, informational, and non-financial forms
of instrumental support.

FANS Supporter Curriculum and Delivery
Figure 2: Example of patient message and corresponding family member message
29
Half of supporters were randomized to receive the FANS curriculum by SMS text-messaging and
half were randomized to receive the FANS curriculum by a pamphlet organized by curriculum
week. The FANS supporter curriculum consists of six months of twice daily messages for the
enrolled supporters and synchronizes in time and content with the patient messages. The
messages are 160 characters in length or less, to conform to short-message-service text
message requirements. A few messages require that two separate texts be sent to encompass
the entire content. The messages are designed to allow personalization with the supporter’s
name, or the patient’s name (see Figure 2 for example), as this was a requested feature in prior
qualitative evaluation of mHealth user experience.(Burner, Menchine, Kubicek, Robles, & Arora,
2014) Most messages were passive, however approximately one FANS message per week was
an active support challenge message that encouraged contact with the patient, prompted the
supporter to engage in a specific care behavior or challenged the FANS perform the same
health behavior the patient was challenged to do, and to communicate that effort to the
patient. In total, the FANS curriculum consists of 381 messages of educational and motivational
content with an emphasis on inspiring appropriate social support. 39% of messages focus on
informational support, 42% of messages focus on emotional support and 18% focus on
instrumental support. Appraisal support was omitted as supporter’s baseline knowledge of
appropriate diabetes measures was not known. (Figure 3 for examples of messages in each
domain)

30
Figure 3: Example FANS messages from each support domain
Support domain Emotional Informational Instrumental
Example FANS
message from
support person
curriculum
Celebrate every time
John does something
for their health. Feel
great about your
progress together.
The A1C test (A-one-C)
shows what Jackie's
blood sugar has been
over the last 3 months.
The A1C goal for most
people is 7.
Challenge: Learn the
names and doses of
Gio’s medications.
Write them down and
keep the list handy for
health appointments

Efforts to support Non-English speaking and bilingual dyads
In order to support Spanish and English-speaking patient and supporters, we created versions
of the intervention in each language. Additionally, we allowed patients and supporters to
select different languages, as patients and supporters may have different language preferences
for texting. This required an additional layer of programming in the platform to synchronize
messages between dyads. As an additional step to support Spanish-speaking participants
required all research staff using Spanish with patients were required to pass a language
certification test.

Participant Safety Monitoring
There are two areas of risk to patients in this intervention. The first is the risk of hypoglycemia
as patients improve their medication adherence, as we anticipate that up to half of patients will
be on insulin or oral insulin secretagogues. Patient knowledge of symptoms and treatment for
hypoglycemia is low and is the first focus of educational messages sent to both supporters and
patients. Additionally, on enrollment, patients were instructed to report episodes of
hypoglycemia to the research team (either by text or voice message) and to call their primary
31
doctor and to inquire about medication adjustment. If the patient did not have a regular
primary care doctor, the research team instructed the patient to visit the urgent access center
at LAC+USC where a safety-net system exists for patients lacking a medical home. A clinical
pharmacologist working with a family medicine specialist conducts same day/next day
appointments to make medication adjustments for diabetes, hypertension, and anticoagulation
medications. We evaluated for a difference in patient reported hypoglycemic events between
the two groups at three, six and twelve months.

Retention Efforts
A safety-net ED population is more transient and can be more difficult to follow up with than
clinic-based populations; additionally, this trial recruited from primarily low-income patients
whose jobs often do not allow phone calls on duty and who work irregular hours. We made
systematic efforts to increase the completion of over-the-phone surveys and remind patients of
their in-person appointments. Patients were texted through the text-messaging platform and
called on their primary and backup phone numbers. If the patient did not respond, we
contacted the supporters directly by texting and phone calls. We also collected an alternative
phone number to contact the patient (i.e., another family member or close friend) so that
patients receive reminders about study appointments from multiple sources.

“Opting out” of the intervention only required a single text of “STOP”, so we confirmed each
opt-out messages with patients and supporters to be sure that the opt-out was intentional. In
32
this intention-to-treat analysis, we attempted to follow up with patients at six and twelve
months unless they explicitly dropped out of the study in addition to stopping messages.

Data collection procedures, schedule, and outcome measures for patients
Patient assessments occurred at enrollment, three, six, nine, and twelve months on behavioral
and psychosocial outcomes and at baseline, six and twelve months for clinical outcomes.
Trained RAs conducted in-person assessments using standardized protocols and equipment, in
the language of the patient’s preference. For assessments at three and nine months,
participants had the option of an in-person, mail, or phone appointment.

At baseline, patient age, self-reported race and ethnicity, self-reported language preference,
Health literacy (3 item Brief Health Literacy Screen(Chew et al., 2008)), and mobile technology
use measured by questions modeled after the Pew Hispanic Center survey (Livingston, 2011)
were collected.

Clinical measures collected at baseline, 6 months, and 12 months consisted of glycemic control
measured by hemoglobin A1C, systolic and diastolic blood pressure averaged over three
readings, weight measured on the same scale at all time points and abdominal circumference.
We measured patient height only at baseline, and used this to calculate BMI. Healthcare
utilization was collected by review of patient’s electronic medical record for clinic
appointments, ED visits and hospitalizations within the 6 months prior to each study timepoint
was collected. Initially, patient self-report was planned for health care utilization to capture
33
visits outside of the Department of Health Services electronic medical record (EMR), however
patient underreporting led us to select EMR review by an RA instead.

Patient measures collected at baseline, 3 months, 6 months, 9 months and 12 months consisted
of (1) healthy behavior measures: Summary of Diabetes Self-care Activities (Toobert, Hampson,
Glasgow, & RE, 2000), the Wilson 3 item Medication Adherence scale (Wilson et al., 2014); (2)
Psycho-social measures: (a) Diabetes Empowerment Scale Short Form(self-efficacy) (R. M.
Anderson, Fitzgerald, Gruppen, Funnell, & Oh, 2003), (b) Diabetes Distress Scale (Diabetes
related distress) (Polonsky et al., 2005), (c) Depression (PHQ-9, (Kroenke, Spitzer, Williams, &
Lowe, 2010), (d) the Diabetes Fatalism Scale (Egede & Ellis, 2010), and (e) Quality of life (World
Health Organization WHO-5 Well Being Index) (Topp, Ostergaard, Sondergaard, & Bech, 2015)
(3) Social support measures were collected: (a) Diabetes Family Behavior Checklist supportive
and non-supportive sub-scores, (b) the Diabetes Care Profile Support Questions, (Fitzgerald et
al., 1996), (c) Diabetes Care Profile support questions Support wanted subscore, and support
received subscore, and (d) the Norbeck Social Support Questionnaire Emotional and Tangible
subscales.(Norbeck, Lindsey, & Carrieri, 1981). Report of frequency of patient-supporter
contact and proportion of communication that is about diabetes was also collected at these
time points by self-report.

Data collection procedures, schedule, and outcome measures for supporters
34
Supporter assessments occur at baseline, six and twelve months. Supporters had the option of
an in-person, mail, or phone appointment. All assessments were with trained RAs in the
language of the participant’s preference.

At baseline, support person age, self-reported race and ethnicity, self-reported language
preference, Health literacy (3 item Brief Health Literacy Screen(Chew et al., 2008)), and mobile
technology use measured by yes/no questions modeled after the Pew Hispanic Center survey
(Livingston, 2011) were collected. At baseline and six months, support people report of (1)
frequency of patient-supporter contact and proportion of communication that is about
diabetes, and (2) supporter diabetes-related distress (Partner Distress Scale (Polonsky, Fisher,
Hessler, & Johnson, 2016)).

Potential effect modifier measures for patients
Potential modifiers at baseline include mobile technology use and health literacy. Mobile
technology use was measured by questions modeled after the Pew Hispanic Center survey
(Livingston, 2011) with the addition of questions about frequency of contact between the
patient and supporter, and the proportion of communication that is about diabetes. Health
literacy was measured by the Brief Health Literacy Screen (Chew et al., 2008).

Engagement with the intervention was measured by percentage of quizzes and assessments
that patients and supporters responded to at the end of the six-month intervention. These
responses were tracked by the SHERPA platform.
35

Data analysis
Primary Outcome:
0-6 month efficacy of TExT-MED+FANS vs TExT-MED+ pamphlet social support on A1C
The primary outcome in was the change in A1C from baseline to six-months, with the exposure
of interest of TExT-MED+FANS. Normality of the outcome variable (six-month change) was
graphically evaluated. We employed longitudinal methods with a mixed effects regression
model to account for correlated outcome data (zero-six month and six-twelve month changes)
and loss to follow up. Analyses was conducted by intent-to-treat, with participants analyzed
according to their randomized intervention regardless of adherence. Participants who
completed the twelve-month study provided two outcome measures of six-month change: a
zero-six month measure of treatment efficacy, and a six-twelve month measure of sustainability
of treatment effect. The linear mixed effects model included a random intercept term for
participants. Fixed effects included treatment allocation, initial level of A1C (zero month
measure for treatment efficacy, six-month measure for sustainability), and a covariate of study
period (zero-six month, six-twelve month). The main effect of treatment tested for group
differences over both zero-six and six-twelve month periods. An interaction term of treatment
by study period tested for differences in treatment effects by study period; treatment effects
were estimated and tested for differences by study period in this interaction model. Model
assumptions including normality of model residuals and homogeneity of variance were
evaluated. We examined for confounding by the change is estimate method with a cut off of
20% change. Predicted mean differences between the groups in A1C at 6 months were
36
examined with the margins and contrast postestimation tools in Stata at the 6 month
timepoint. A sensitivity analysis confined to adherent participants (those who have not opted
out of messages and have received 75% or greater of messages confirmed by message delivery
platform) was planned, but was not possible due to limitations in the cellular service providers
most patients had.

Secondary Outcomes:
6-12 month efficacy of TExT-MED+FANS vs TExT-MED+pamphlet social support on A1C
With the same mixed effects linear regression model used, an estimate of the 6-12 month post
intervention maintenance phase change in A1C was estimated with the margins and contrast
postestimation tools in Stata at the 12 month timepoint.

0-6 month efficacy of TExT-MED+FANS vs TExT-MED+ pamphlet social support on Clinical
Outcomes and Diabetes Self-Care Behaviors
With the same mixed effects linear regression model used for the primary outcome, an
estimate of the 0-6 month efficacy change in A1C was estimated with the margins and contrast
postestimation tools in Stata at the 6 month timepoint.

A Priori Subgroup Analysis
To determine if subgroups of participants are differentially affected by the intervention,
secondary analyses evaluating intervention moderators were planned a priori. For A1C and
37
each of the secondary outcomes, interaction terms (randomized intervention-by-moderator
product terms) were added to the mixed effects linear models described above. Variables
evaluated as moderators included patient and supporter gender, race and ethnicity, language
preference, health literacy, years with diabetes, baseline frequency of mobile usage as high or
low based on latent profiles,(Treacy-Abarca et al., 2022) and physically proximity to supporter.
All measures were dichotomous. Intervention effects were estimated by levels of the
moderator for significant moderators only at a p value of 0.05 for the interaction term between
randomization group and moderator.

Mediation Analysis with Generalized Structural Equation Modeling
Secondary analyses of the primary A1C outcome employed structural equation modeling to
evaluate the secondary self-care behavior outcomes as mediators of the TEXT-MED+FANS
intervention. Initial analyses evaluated the associations of changes in behavior and efficacy
variables with change in A1C using mixed effects models as detailed above. The A1C mediating
model then included a term for randomized intervention, baseline A1C and change in all of the
self-care behaviors as possible mediators. Mediation was tested with bootstrapped samples,
evaluating the direct and indirect (mediated) effects of changes in behavior and efficacy
outcomes (Bentler & Chou, 1987; Chou & Bentler, 1995). Using the Stata lincom procedure to
sum the partial coefficients of randomization group and each self-care behavior, we examined
for mediation relationships. We repeated this analysis for each self-care behavior individually as
well as in a combined model.

38

Post Hoc Subgroup Analysis
After completion of enrollment, substantial differences in baseline support and contact
between patients and their selected supporters became evident, as some supporters took up to
three weeks to complete enrollment procedures. We then decided to conduct a subgroup
analysis based on supporter immediate availability for enrollment versus delayed enrollment as
a baseline moderator.

Sample size
We planned to enroll 166 patient-supporter dyads, which assuming a 30% loss to follow up,
gave a sample size of 116 total dyads. With power of 0.8 and two-sided alpha of 0.05, using the
standard deviation of final A1C of 1.6, (the value of our prior trials), this gave this trial the ability
to detect a mean difference in change of A1C of 0.84 between the two groups at six month
follow up.

39
Results:
Screening and Recruitment
RAs identified nearly 4000 ED patients with diabetes via electronic medical record real-time
searches. Over half of these patients (2004) were screened for eligibility (see Figure 4 for
CONSORT style diagram of screening and enrollment for reasons for ineligibility). Of the 2004
patients screened for eligibility, 209 patients were initially recruited, 173 (9%) met criteria and
agreed to enroll. However only 166 were also able to have a support person consented and
enrolled. The most common reason to not meet eligibility was not using text-messages at
baseline (31%, 613/2004 patients), followed by not having a stable mobile phone number (21%,
427/2004 patients). A substantial portion of patients had A1C levels below the threshold (20%,
394/2004 patients.) Less than 10% of patients could not identify a potential supporter or
identified a potential supporter who could not be reached. Of note, 65 patients (3% of
screened patients) did not believe they had diabetes, despite having a diagnosis of diabetes in
the EMR. After randomization, 7 supporters failed to complete the initial process of enrollment
in the study (i.e. initially answered call from RA, agreed to participate, but then ended call
without completing assessment and did not answer any further calls or texts.) We ended
enrollment once the final cohort of 166 patient-supporters dyads was reached.

40
Figure 4: Screening and Enrollment of patients into TExT-MED+FANS
41
Participant characteristics
The characteristics and baseline measurements of the patient cohort are displayed in Table 1.
The enrolled patient cohort was 51% female, 70% Spanish-speaking, 79% born outside US, with
a median age of 48.2 years. Patients self-reported the type of diabetes, with 67% (111)
reporting type II diabetes (111), 8% (14) reporting type I diabetes and 25% (41) who did not
know which type of diabetes they had. Of these 166 patients, 50% (83) used insulin and 10%
(16) were managed with diet and exercise without medications. Their mean A1C was 10.8, they
averaged mild hypertension, and an overweight BMI. Medication adherence and diabetes
selfcare behaviors were poor. Depression scores averaged in the mild-moderate depression
range, while WHO well-being scores were low, but not in the “depressed range”. The
intervention group consisted of more males and fewer Spanish speakers. Full details of the
baseline characteristics are displayed in Table 1.

42
u Higher value indicates clinically worse value
£ Lower value indicates clinically worse value

Table 2: Patient baseline characteristics

Total (n=166) Intervention (n=80) Pamphlet (n=86)

Mean
or %
95% CI
Mean or
%
95% CI
Mean or
%

95% CI
A1C (%mmHg/dL) u 10.84 10.58-11.10 10.90 10.52-11.29 10.78 10.43-11.14
BMI (kg/m2) u 30.07 28.90-30.93 29.34 27.74-30.93 30.78 29.05-35.51
Age (years) 47.60 46.00-49.20 46.91 44.68-49.13 48.25 45.92-50.57
Systolic BP (mmHg) u 134.6 130.8-138.4 133.8 128.1-139.6 135.4 130.3-140.4
Male Gender 49%

56%

43%

Race & Ethnicity

Latino 92%

90%

94%

Non-Hispanic white 1%

3%

0%

Asian/Pacific islander 1%

0%

2%

Black 5%

8%

3%

Spanish Speaking 70%

65%

74%

Foreign Born (n=165)

78%

70%

86%

Acculturation
(Marin short form)
2.01 1.83-2.37 2.10 1.83-2.37 1.92 1.67-2.17
Health literacy
(Brief Health Literacy
Screen) £
4.60 4.07-5.12 4.75 3.96-5.54 4.45 3.74-5.17
Depression (PHQ-9) u 9.16 8.14-10.17 9.44 8.03-10.84 8.90 7.41-10.38
Self efficacy u
(DM Empowerment Scale
– Short Form)
3.85 3.75-3.95 3.89 3.75-4.02 3.81 3.66-3.96
Distress due to DM u
(DM distress Scale)
2.48 2.32-2.64 2.63 2.38-2.87 2.34 2.14-2.58
Quality of life (WHO-5) £ 60.9 55.72-64.53 58.0 51.42-64.53 61.9 56.21-67.61
Fatalism u
(DM Fatalism Scale)
34.94 33.43-26.46 35.86 33.67-38.05 34.09 31.97-36.21
Medication adherence £
(n=165, Wilson 3 item)
66.5 62.05-71.13 65.8 59.57-72.13 67.29 60.63-73.95
Summary of Diabetes Self Care Activities

General diet £ 3.23 2.85-3.61 3.07 2.51-3.63 3.38 2.85-3.90

Specific diet £ 3.87 3.58-4.16 3.74 3.32-4.16 3.98 3.76-4.39

Glucose monitoring £
2.65 2.20-3.10 2.76 2.07-3.44 2.55 1.95-3.15

Foot care £ 4.06 3.61-4.51 4.11 3.48-4.75 4.01 3.37-4.66

Carb spacing £ 2.89 2.49-3.28 3.00 2.45-3.59 2.78 2.21-3.35

Exercise £ 2.47 2.08-2.86 2.43 1.87-2.99 2.51 1.94-3.07
43
u Higher value indicates clinically worse value
£ Lower value indicates clinically worse value

Supporter Characteristics and Baseline Support
The supporters were 70% female and 57% Spanish-speaking, 43% English speaking with 66% of
supporters born outside the US. Their mean age was 43.7 years. 28% (n=46) patient-supporter
dyads were language discordant in language they preferred to receive text-messages. (see
Table 2) Of these supporters, 21% (34) also had diabetes; 68% (23) with type II diabetes, 6% (2)
with type I DM, and 26% (8) who did not know the type of diabetes they had. The supporters
were predominantly family members: 31% spouses (51), 14% siblings (24), 23% adult child of
the patient (39), 16% other relatives (28), 12% friends (20) and 4% of patients did not wish to
disclose the nature of their relationship with their supporter. Baseline supporter measures are
presented in Table 3; the intervention and pamphlet control group had similar baseline
supporter measures.
Table 2 Continued: Patient Baseline characteristics
Total (n=166) Intervention (n=80) Pamphlet (n=86)
Mean 95% CI Mean 95% CI Mean

95% CI
Supportive family
behaviors £ (DM Family
Behavior Checklist)
23.90 22.54-25.25 24.46 22.46-26.46 23.37 21.49-25.26
Non-supportive family
behaviors (DM Family
Behavior Checklist) u
18.23 17.21-19.24 18.70 17.21-20.19 17.78 16.38-19.19
Support needs (Diabetes
Care Profile) u
23.61 22.46-24.77 24.04 22.36-25.72 23.22 21.61-24.83
Support received £
(Diabetes Care Profile)
18.54 17.18-19.91 19.16 17.15-21.18 17.98 16.10-19.86
Support attitudes £
(Diabetes Care Profile)
6.52 5.66-7.24 6.39 5.35-7.43 6.51 5.31-7.72
General Emotional
support (social support
questionnaire, n=152) £
13.8 13.32-14.36 13.7 12.96-14.40 14.0 13.25-14.75
General Tangible support
(social support
questionnaire n=152) £
6.99 6.79-7.26 7.15 6.79-7.51 6.83 6.43-7.23
44

Study follow up
We successfully obtained measures of our primary outcome, patient A1C at 6 months, the end
of the intervention phase, in 52% of the FANS intervention group, and 60% of the pamphlet
support control group. In the FANS intervention group, 8 patients had dropped out of the study,
and an additional 27 were lost to follow up after multiple attempts at contact. In the pamphlet
support group, 5 patients dropped out of the study and an additional 29 were lost to follow up
after multiple attempts at contact. At 12 months, after the maintenance phase with no text-
messages to patients or supporters, we were able to obtain A1C measurements for 50 patients
(58%) in the FANS group, and 56 of the pamphlet support group (65%). In the FANS group, an
additional 7 patients withdrew from the study during the maintenance phase, but we were able
to reach 5 more patients than had been available at 6 months. In the pamphlet support group,
an additional 5 patients withdrew from the study during the maintenance phase, but we were
able to reach 4 more patients than had been available at 6 months. Comparison of the patients
who completed follow up at 6 months vs those who were lost to follow up or discontinued the
intervention (Table 2) showed patients who did not complete 6 month assessments reported
negative attitudes about their baseline social support (Diabetes Care Profile Support Attitudes –
Negative sub-score 2.25 (95% CI 1.95 to 2.55) vs 1.56 (1.36 to 1.76), group difference -0.69 (-
Table 3: Supporter Baseline Characteristics

Total (n=166) Intervention (n=80) Pamphlet (n=86)

Mean (standard
deviation) or %
Mean (standard
deviation) or %
Mean (standard
deviation) or %
Supporter has diabetes 21% 20% 21%
Age (years) 43.69(14.54) 42.89 (13.79) 44.45 (15.26)
Male Gender 49% 56% 43%
Spanish Speaking 57% 60% 53%
Foreign Born 66% 65% 66%
45
1.03 to -0.35)). Patients who were unavailable for follow up were not statistically younger
(mean age 45.78 (43.13 to 48.43) vs 48.80 (46.80 to 50.80), group difference 3.01 (-0.23 to
6.26)), or more acculturated (Marin Acculturation Scale – Short form 2.17 (1.86 to 2.49) vs 1.90
(1.68 to 2.12), group difference -0.27 (-0.64 to 0.10).)

46
Table 4: Baseline Characteristics: Lost To Follow Up Vs Completed Six Month Assessment
Measure Completed Follow-Up
Mean (95% CI) or %
Loss to Follow-Up
Mean (95% CI) or %
Group difference
Mean (95% CI)
A1C u 10.86 (10.50 to 11.21) 10.82 (10.44 to 11.20) 0.34 (-0.50 to 0.54)
Race/Ethnicity Latino 92.99% 91.50%
Non-Hispanic white 1.08% 1.31%
Asian 1.08% 1.31%
Black 4.85% 5.88%
Foreign Born 78.92% 77.5%
English Languge preferred 29.38% 30.72%
Low health literacy 53.91% 55.56%
BMI u 30.21 (28.67 to 31.75) 29.87 (28.01 to 21.73) 0.34 (-2.06 to 2.74)
Age 48.80 (46.80 to 50.80) 45.78 (43.13 to 48.43) 3.01 (-0.23 to 6.26)
Quality of Life £ 59.48 (53.84 to 65.12) 60.82 (54.06 to 67.57) -1.34 (-10.12 to 7.45)
Marin Acculturation 1.90 (1.68 to 2.12) 2.17 (1.86 to 2.49) -0.27 (-0.64 to 0.10)
DM distress score u 2.55 (1.01 to 2.75) 2.37 (2.11 to 2.63) 0.18 (-0.14 to 0.50)
Fatalism score u 35.48 (33.47 to 37.49) 34.14 (31.80 to 36.48) 1.34 (-1.76 to 4.44)
Medication Adherence £ 66.52 (60.73 to 72.31) 66.71 (59.19 to 74.21) -0.18 (-9.50 to 9.14)
SDSCA: general diet £ 3.33 (2.82 to 3.08 (2.50 to 3.67) 0.24 (-0.53 to 1.02)
SDSCA: specific diet £ 3.9 (3.53 to 4.27) 3.82 (3.34 to 4.30) 0.08 (-0.51 to 0.68)
SDSCA: glucose
monitoring £
2.41 (1.85 to 2.96) 3.02 (2.26 to 3.78) -0.62 (-1.53 to 0.29)
SDSCA: foot care £ 4.33 (3.75 to 4.90) 3.66 (2.94 to 4.38) 0.67 (-0.25 to 1.58)
SDSCA: carb spacing £ 3.86 (2.36 to 3.36) 2.92 (2.28 to 3.57) -0.64 (-0.87 to 0.74)
SDSCA: exercise £ 2.49 (1.97 to 3.01) 2.44 (1.82 to 3.06) 0.05 (-0.76 to 0.86)
Supportive Family
Behaviors £
24.04 (22.25 to 25.83) 23.68 (21.54 to 25.82) 0.35 (-2.43 to 3.14)
Non-Supportive Family
Behaviors u
18.31 (16.97 to 19.65) 18.10 (16.52 to 19.68) 0.21 (-1.87 to 2.28)
Support needs
Diabetes Care Profile u
24.1 (22.70 to 25.50) 22.88 (20.88 to 24.88) 1.22 (-1.13 to 3.57)
Support received
Diabetes Care Profile £
18.57 (16.74 to 20.40) 18.52 (16.43 to 20.60) 0.05 (-2.74 to 2.85)
Support attitudes £
Diabetes Care Profile
7.36 (6.46 to 8.26) 5.08 (3.66 to 6.50) 2.28 (-0.70 to 3.87)
Support attitudes: £
positive sub-score
Diabetes Care Profile
4.01 (3.74 to 4.29) 3.94 (3.61 to 4.28) 0.07 (-0.36 to 0.50)
Support attitudes:
Negative sub-score
Diabetes Care Profile u
1.56 (1.36 to 1.76) 2.25 (1.95 to 2.55) -0.69 (-1.03 to -0.35)
Emotional support £ 13.77 (13.04 to 14.49) 13.95 (13.22 to 14.68) -0.18 (-1.24 to 0.87)
Tangible support £ 6.94 (6.59 to 7.29) 7.05 (6.62 to 7.48) -0.10 (-0.65 to 0.44)
u Higher value indicates clinically worse value
£ Lower value indicates clinically worse value

47
Safety Data
The TExT-MED FANS and TExT-MED+pamphlet support had a low adverse event profile. One
patient died in the TExT-MED FANS arm during the post-intervention maintenance phase of
kidney failure. The cause of death was urosepsis and kidney failure and was determined to be
unrelated to the intervention after review by the local institutional review board. Severe
hypoglycemic events (blood glucose <70) were not statistically different between the groups,
with 20% of FANS arm patients reporting at least one episode during the intervention phase,
and 37% in the pamphlet support arm reporting at least one hypoglycemic event in the
intervention phase )p=0.051). With a six month follow up of 98 patients who completed safety
reporting, and a baseline rate of hypoglycemia of 37%, we had 42% power to detect a clinically
meaningful difference.

Primary Efficacy Data: Efficacy: 6 month Change in A1C
The TExT-MED+FANS and TExT-MED+pamphlet social support interventions were similarly
efficacious in improving glycemic control (A1C) Patients in the pamphlet social support arm had
a mean decrease in A1C of 1.42%mg/dl (95%CI 0.82 to 2.02), while patients in the FANS arm
had a mean decrease in A1C of 1.28%mg/dl (95% CI 0.48 to 2.09), with a combined sample
change of 1.36%mg/dl (0.87 to 1.83) (See Table 4 for change at 6-month results). When we
examined for potential confounders in a mixed effects model with individual participants with
random intercepts and controlling for baseline A1C, we found no confounders of the
intervention effect of FANS vs pamphlet social support on A1C.

48
Secondary Outcome: 6-12 month efficacy of maintenance of TExT-MED+FANS vs TExT-
MED+pamphlet support on A1C
Overall, the FANS intervention and pamphlet social support arms maintained improved A1C,
with a combined mean change of 0.06%mg/dL (95%CI 0.47 to -0.34) after the 6-12 month post
intervention maintenance phase, with no clinically meaningful difference between groups
(FANS patients with a mean decrease of 0.29%mg/dL (95%CI 0.27 to -0.85) vs. pamphlet social
support with a mean increase of 0.36%mg/dL (95%CI 0.93 to -0.22), between group difference
of -0.65%mg/dL (95%CI -1.45 to 0.16). (See Table 5 for change at 12 month results) Using the
mixed effects model described above allowing for random intercept by individual patient, we
found similar results: patients maintained their improved A1C, with a combined sample non-
clinically significant increase of 0.212%mg/dL (95%CI -0.30 to 0.55). There was no difference in
maintenance of A1C change, with a non-clinically significant between group mean difference of
0.23%mg/dL (95%CI -0.88 to 0.41).

49
Table 5: 6 Month Change in Outcome Measures (6 month minus baseline)
Pamphlet Support
Mean D (95% CI)
FANS Support
Mean D (95% CI)
Group difference
Mean D (95% CI)
Combined group
Mean D (95% CI)
A1C £ -1.42 (-0.82 to -
2.02)
-1.28 (-0.48 to -2.09) -0.14 (0.83 to -
1.11)
-1.36 (-0.87 to -
1.83)
BMI £ 7.91 (17.52 to -1.70) 3.84 ( 10.65 to -2.97) 4.07 (15.88 to -
7.74)
5.96 (11.84 to 0.08)
Systolic BP £ 6.14 (13.44 to 1.14) 10.06 (20.83 to 0.72) -3.91 8.74 to -
16.56)
8.02 (14.32 to 1.72)
Medication
Adherence £
15.73 (24.32 to
7.13)
12.39 -21.63 to 3.15) 3.34 (15.78 to -
9.11)
14.16 (20.35 to
7.97)
SDSCA: general
diet £
1.55 (2.36 to 0.74) 0.17 (1.97 to -0.38) 0.37 (1.50 to -0.75) 1.37 (1.93 to 0.81)
SDSCA: specific
diet £
0.42 (1.01 to- 0.17) 0.78 (1.35 to 0.21) -0.36 (0.46 to -
1.18)
0.60 (0.10 to 0.18)
SDSCA: glucose
monitoring £
0.91 (1.75 to 0.08) 0.10 (0.94 to -0.75) 0.82 (1.99 to -0.36) 0.53 (1.12 to -0.06)
SDSCA: foot care
£
1.25 (2.08 to 0.42) 1.36 (2.16 to 0.56) -0.11 (1.03 to -
1.25)
1.30 (1.87 to 0.73)
SDSCA: carb
spacing £
0.71 (1.92 to -0.50) 0.20 (1.20 to -0.80) 0.51 (2.09 to -1.06) 0.47 (1.26 to -0.31)
SDSCA: exercise £ 0.39 (1.23 to -0.44) 0.90 (1.84 to -0.31) -0.51 (0.72 to -
1.74)
0.63 (1.25 to-0.02)
Self-efficacy £ 0.17 (0.35 to 0.00) 0.05 (0.25 to -0.16) 0.12 (0.39 to -0.14) 0.11 (0.25 to -0.02)
DM distress score
£
-0.65 (0.38 to -0.92) -0.60 (-0.26 to -0.94) -0.05 (0.37 to -
0.48)
-0.63 (-0.42 to -
0.84)
Depression £
(PHQ9)
-3.86 (-2.12 to -
5.60)
-2.33 (-0.49 to -4.17) -1.54 (0.96 to -
4.04)
-3.13 (-1.88 to -
4.39)
Quality of life £ 9.70 (18.45 to 0.93) 4.96 (13.83 to -3.91) 4.74 (17.08 to -
7.61)
7.47 (13.62 to 1.32)
Fatalism score£ 1.44 (4.13 to -1.25) -0.58 (2.57 to -3.73) 2.02 (6.09 to -2.05) 0.47 (2.50 to -1.56)
Supportive £
Family Behaviors
-0.58 (1.65 to -2.80) -0.12 (2.58 to -2.81) -0.46 (2.96 to -
3.88)
-0.36 (1.34 to -2.06)
Non-Supportive
Family Behavioru
-0.58 (1.16 to -2.32) -0.56 (1.48 to -2.60) -0.02 (2.61 to -
2.65)
-0.57 (0.74 to -1.87)
DCP support needs
u
-4.44 (-1.44 to -
7.45)
-5.80 (-3.11 to -8.49) 1.36 (5.39 to -2.66) -5.08 (-3.08 to -
7.08)
DCP support
received £
0.08 (2.77 to -2.62) 3.65 (6.36 to 0.95) -1.91 (0.21 to -
7.36)
1.76 (3.67 to -0.16)
DCP support
attitudes £
-0.96 (0.85 to -2.77) 0.87 (2.57 to -0.83) -1.83 (0.64 to -
4.31)
-0.10 (1.14 to -1.34)
Emotional
Support£
-0.64 (0.54 to -1.82) 0.22 (1.61 to -1.17) -0.86 (0.93 to -
2.64)
-0.24 (0.65 to -1.13)
Tangible Support
£
-0.06 (0.52 to -0.65) 0.29 (0.93 to -0.34) -0.36 (0.50 to -
1.21)
0.10 (0.53 to -0.32)
Supporter u
Diabetes Distress
-0.45 (-0.24 to -
0.66)
-0.28 (-0.02 to -0.54) -0.17 (0.15 to -
0.50)
-0.37 (-0.21 to -
0.54)
u Higher value indicates clinically worse value
£ Lower value indicates clinically worse value
50

Table 6: 12 Month Change in Outcome Measures (12 month minus 6 month)
Pamphlet Support
Mean D (95% CI)
FANS Support
Mean D (95% CI)
Group difference
Mean D (95% CI)
Combined group
Mean D (95% CI)
A1C £ 0.36 (0.93 to -0.22) -0.29 (0.27 to -0.85) 0.65 (1.45 to -0.16) 0.06 (0.47 to -0.34)
BMI £ 3.38 (10.14 to -3.38) -4.02 (4.28 to -
12.31)
7.40 (17.82 to -
3.03)
0.44 (5.26 to -5.17)
Systolic BP £ 4.95 (14.99 to -5.09) -2.09 (10.78 to -
14.96)
7.04 (22.87 to -
8.79)
1.71 (9.59 to -6.17)
Medication
Adherence £
-1.74 (3.88 to -7.36) -3.22 (4.10 to -
10.54)
1.48 (10.45 to -
7.48)
-2.43 (2.02 to -6.88)
SDSCA: general
diet £
-0.23 (0.50 to -0.97) -0.96 (-0.09 to -
1.84)
0.73 (1.84 to -0.39) -0.57 (-0.01 to -
1.13)
SDSCA: specific
diet £
-0.29 (0.26 to -0.83) -0.45 (0.14 to -1.04) 0.16 (0.95 to -0.63) -0.36 (0.03 to -0.75)
SDSCA: glucose
monitoring £
0.66 (1.60 to -0.28) 0.18 (0.94 to -0.59) 0.48 (1.70 to -0.73) 0.43 (1.04 to -1.70)
SDSCA: foot care
£
0.81 (1.59 to 0.03) -0.01 (0.61 to -0.63) 0.82 (1.82 to -0.18) 0.43 (0.93 to -0.07)
SDSCA: carb
spacing £
-0.11 (1.21 to -1.43) -0.31 (0.95 to -1.57) 0.20 (2.03 to -1.63) -0.20 (0.71 to -1.10)
SDSCA: exercise £ -0.11 (0.77 to -0.98) -0.81 (0.11 to -1.73) 0.71 (1.96 to -0.55) -0.43 (0.19 to -1.06)
Self-efficacy £ -0.13 (0.02 to -0.27) 0.03 (0.21 to -0.14) -0.16 (0.06 to -0.38) -0.05 (0.06 to -0.16)
DM distress score
£
-0.06 -0.21 to -0.33) 0.01 (0.23 to -0.22) -0.07 (0.29 to -0.42) -0.03 (0.15 to -0.21)
Depression £
(PHQ9)
0.18 (2.24 to -1.88) 0.13 (1.78 to -1.53) 0.05 (2.70 to -2.59) 0.15 (1.47 to -1.16)
Quality of life £ -0.09 (9.90 to -10.07) -6.30 (3.26 to -
15.86)
6.21 (19.99 to -
7.56)
-2.94 (3.91 to -9.80)
Fatalism score£ 1.30 (4.02 to -1.43) 1.20 (-4.00 to 1.59) -0.10 (-3.96 to 3.76) -1.26 (-3.17 to 0.66)
Supportive £
Family Behaviors
-0.70 (-3.02 to 1.61) -0.98 (3.96 to -2.01) -0.27 (3.39 to -3.94) 0.83 (2.64 to -0.99)
Non-Supportive
Family Behavioru
0.85 (2.64 to -0.94) 1.13 (3.35 to -1.09) -0.28 (2.50 to -3.06) 0.98 (2.36 to -0.40)
DCP support needs
u
0.39 (2.78 to -2.01) -0.40 (2.90 to -3.70) 0.78 (4.72 to -3.16) 0.02 (1.98 to -1.93)
DCP support
received £
1.38 (3.60 to -0.83) -2.70 (-0.09 to -
5.31)
4.08 (7.4 to 0.73) -0.49 (1.22 to -2.21)
DCP support
attitudes £
0.77 (2.11 to -0.57) -0.08 (1.21 to -1.36) 0.84 (2.69 to -1.01) 0.38 (1.30 to -0.54)
Emotional
Support£
0.48 (1.68 to -0.72) -0.69 (0.61 to -1.99) 1.17 (2.91 to -0.57) -0.06 (0.81 to -0.93)
Tangible
Support £
0.36 (0.10 to -0.28) -0.48 (0.19 to -1.14) 0.84 (1.75 to -0.07) -0.02 (0.44 to -0.48)
u Higher value indicates clinically worse value
£ Lower value indicates clinically worse value

51
Secondary Outcome: Secondary Clinical Outcomes 0-6 months
The TExT-MED+FANS and TExT-MED+ pamphlet social support interventions were similarly
efficacious in improving clinical outcomes. The two groups had similar changes in systolic blood
pressure (see Table 5). There was no change in abdominal circumference combined or between
groups at the 6 month efficacy time point. We were unable to examine weight or BMI as
outcomes consistently, as a substantial portion of patients had full or partial limb amputations,
and weight loss was not a useful metric across patients. We also examined clinical secondary
outcomes using the mixed effects modelling described above allowing for random intercept by
individual patient. We found patient language preference and gender to be substantial
confounders the effect of intervention group on change in abdominal circumference. We found
no difference in physiologic outcomes by FANS vs pamphlet support after adjusting for
potential confounders.

Secondary Outcome: Diabetes Self-Care Behaviors
The TExT-MED+FANS and TExT-MED+pamphlet social support interventions were similarly
efficacious in improving self-care behaviors. The two groups had similar changes in all self-care
measures (see Table 5).Lastly, we examined diabetes self-care behavior outcomes using the
mixed effects modelling described above allowing for random intercept by individual patient.
We found patient language preference to be a substantial confounder of intervention group on
general diet plan adherence and also to be a substantial confounder of intervention group on
disease specific diet plan adherence. We found no difference in self-care behaviors by FANS vs
pamphlet support after adjusting for potential confounders of intervention group.
52

Subgroup Analyses of Primary Outcome: 0-6 month efficacy of TExT-MED+FANS vs TExT-
MED+pamphlet social support on A1C:
In prespecified subgroup analysis, we found that among patients with a new diagnosis of
diabetes, FANS group patients improved their A1C to a greater degree than patients in the
pamphlet social support arm, mean between group difference of 1.96%mg/dL(95%CI -3.81 to -
0.125, p between group= 0.036). We found no differences at 6 months based on gender, race and
ethnicity, language preference, health literacy, baseline frequency of mobile usage, physically
proximity to supporter, and baseline social support.

Subgroup Analyses of Secondary Outcome: 6-12 month efficacy of maintenance of TExT-
MED+FANS vs TExT-MED+pamphlet social support on A1C:
In the same prespecified subgroup analysis as the 6 month timeframe, we found that among
patients with a new diagnosis of diabetes, FANS group patients improved their A1C to a greater
degree than patients in the pamphlet social support arm, with a predicted between group
difference of 2.40%mg/dL (95%CI -4.33 to -0.47, p between group= 0.002) at 12 months. We also
found that patients who preferred texts in English maintained glycemic control better, with a
predicted between group difference of 2.53%mg/dL(95%CI -4.15 to -0.91, p between group= 0.015)
at 12 months. We found no differences at 12 months based on gender, race and ethnicity,
health literacy, baseline frequency of mobile usage, physically proximity to supporter, and
baseline social support. (See Table 6)

53

Subgroup Analyses of Secondary Clinical Outcomes:
In the same pre-specified subgroup analysis as with the primary outcome, we found the FANS
intervention to be less efficacious in reducing systolic blood pressure in male with high health
literacy (contrast of predicted systolic blood pressure of male high health literacy receiving
FANS vs pamphlet social support 47.27mmHg higher, CI95% 21.96 to 72.58, p between group<0.001).
(See Table 5) We found the FANS intervention to be more efficacious at decreasing for patients’
abdominal circumference for patients’ who had supporters with high mobile technology use at
baseline, with lower baseline support and who preferred English (contrast of predicted
abdominal circumference at 6months for patients with high technology use supporters
receiving FANS vs pamphlet social support 12.90cm smaller, 95%CI -23.40 to -2.40, p between
group=0.016; contrast of predicted abdominal circumference at 6months for patients with low
baseline support receiving FANS vs pamphlet social support -12.77cm, 95%CI -20.24 to -5.30
p between group<0.001; contrast of predicted abdominal circumference at 6months for patients who
prefer English receiving FANS vs pamphlet social support 10.89cm smaller, 95%CI -19.70 to -
2.08 p between group=0.015).

Subgroup Analyses of Diabetes Self-care Behaviors:
In the same pre-specified subgroup analysis as with the primary outcome, we found the FANS
intervention to be less efficacious in improving self-monitoring of glucose in patients with lower
baseline support (contrast of predicted days self-monitoring glucose for patients with lower
baseline support receiving FANS vs pamphlet social support +1.74 fewer days per week, CI95%
54
3.19 to .029 fewer days per week, p between group=0.018). We found the FANS intervention to be
more efficacious in improving exercise spacing in patients who preferred English (contrast of
predicted days exercising of English preference patients receiving FANS vs pamphlet social
support at 6 months 2.38 days, 95%CI 0.73 to 4.03 p between group=0.005).

Mediation Analysis of Randomization Group and A1C by Diabetes Self-care Behaviors
We examined the secondary health behavior outcomes as potential mediators of the effect of
FANS vs pamphlet social support with the GSEM function in Stata, modeling random intercept
for each participant for A1C and each potential mediator. We then used the lincom procedure
to sum the partial coefficients of randomization group and each self-car behavior. We found no
evidence of mediation of the relationship between intervention group and A1C by any of the
secondary health behavior outcomes when examined together (diagrammed in Figure 5.) We
repeated this analysis for each self-care behavior individually; there was also no evidence of
mediation when examining each secondary health behavior outcome alone in an individual
GSEM with the each secondary health behavior outcome individually. (See Table 7)

55
A1C ε
1
Medication Adherence
ε
2
General Diet
ε
3
Specific Diet
ε
4
Glucose Monitoring
ε
5
Foot Care
ε
6
Carb Spacing
ε
7
Exercise
ε
8
Participant ID
Participant ID
Table 7: Sum of partial coefficients on A1C of randomization group and each secondary health
behavior outcome combined and individually modelled in a GSEM
Health behavior GSEM with all secondary health
behavior outcomes combined in
one model: point estimate (95%CI)
GSEM with each secondary health
behavior outcome in an individual
model: point estimate (95%CI)
Medication Adherence .0019 (-.48 to .49) .0388 (-.44 to .52)
SDSCA: general diet .0286 (-.53 to .47) -.1402 (-.65 to .37)
SDSCA: specific diet .0104 ( -.50 to .52) -.0934 (-.61 to .42)
SDSCA: glucose monitoring -.0658 ( -.56 to .43) -.1111 (-.60 to .38)
SDSCA: foot care -.0236 ( -.51 to .46) .0398 (-.46 to .54)
SDSCA: carb spacing -.0015 (-.49 to .49) .0726 (-.43 to .58)
SDSCA: exercise .0062 ( -.48 to .50) .0841 (-.42 to .59)

Figure 5: General Structural Equation Model: random intercept for each participant for A1C and
potential behavioral mediators

2

56
Discussion:

In order to test if augmenting existing social support via mHealth for safety-net ED patients
receiving an mHealth curriculum improved diabetes self-care behaviors and glycemic control
compared to a pamphlet based support intervention, we conducted the randomized TExT-FANS
study. We designed a randomized trial with some pragmatic features to increase potential
translation to clinical care: a focus on enrollment of patients while in the emergency
department and not requiring their identified supporter to also be physically present. The study
population was a highly vulnerable group, with low access to medical homes, and had a high
potential to benefit from both the patient focused and social support interventions. While all
patients improved their glycemic control to a clinically significant degree and selfcare
behaviors, in our main analysis, we found that the FANS curriculum sent via text-message to
patient identified supporters did not provide additional benefit over a dated pamphlet based
support curriculum sent to supporters. However, in prespecified exploratory subgroup analysis,
we found the FANS intervention was superior to pamphlet social support for patients for
improving glycemic control in the first year of a diagnosis with diabetes, and that patients who
preferred to receive text-messages in English maintained improvement in glycemic control in
the post-intervention phase to a greater degree. TExT-MED for patients with or without a social
support component has the potential to improve the long-term health outcomes of all of these
vulnerable patients and has high potential to be translated to a system wide intervention, give
the possibilities of remote and even self-enrollment by patients.

57
The small differences in improvement of glycemic control, self-care behaviors, psychological
outcomes and social support measures between the FANS supported group and pamphlet
social support group were not clinically or statistically different. Prior pooled analysis of
mHealth interventions to improve diabetes self-management have shown mean A1C
improvements of 0.5% across types of diabetes and mHealth modalities(Hou, Carter, Hewitt,
Francisa, & Mayor, 2016), while a prior meta-analysis for all types of social support
interventions for diabetes self-management showed an improvement in A1C of 0.25% at 3
months (95% CI −0.40 to -0.11), I2=12%, 9 RCTs)(Spencer-Bonilla et al., 2017). There is not
sufficient literature to generate pooled estimates of mHealth/eHealth based social support for
diabetes self-management (Vorderstrasse, Lewinski, Melkus, & Johnson, 2016). As there was no
inactive control group in this trial, the potential benefit of TExT-MED plus social support
augmentation via mHealth may be subject to a floor effect In this study, we found the overall
combined sample A1C improvement of 1.42% at 6 months (95% CI 0.82 to 2.02) and the
maintenance of that improvement at 12 months with a washout period change of 0.06% (95%CI
-0.47 to 0.34.) If this augmented social support intervention with social support messaging that
coordinates with patient curriculum is utilized in a large-scale setting, the additional time and
expense needed to recruit a supporter for each patient may not be cost effective. However, as
an optional feature, it was enjoyed by patients and was requested in prior iterations of the
patient focused intervention.

Patients who benefitted the most from the mHealth augmented social support of the FANS
curriculum to supporters were those who were diagnosed with diabetes in the year prior to
58
enrollment. Diabetes and other chronic medical conditions require complex lifestyle changes
for the patient, and for their family members. Patients and family members may be largely
unfamiliar with best self-care practices and stressed by the new diagnosis.(Elizabeth Burner et
al., 2018; Kovacs Burns et al., 2013; L. S. Mayberry & Osborn, 2014) Cochrane analysis on
diabetes self-management education interventions versus standard care for patients with
newly diagnosed diabetes showed mean differences in A1C (-0.21%, 95% CI -0.38 to -0.04) 12
months after initial intervention of educational materials.(Tanaka, Shibayama, Sugimoto, &
Hidaka, 2020) National and international guidelines recommend that diabetes self-
management education and support be provided to patients with diabetes upon their initial
diagnosis and this benefit is covered by many medical insurers.(Beck et al., 2017; M. A. Powers
et al., 2020) However, less than 10% of patients receive this critical training internationally, with
even lower rates in the US and in under resourced environments.(Li et al., 2014; Rinker et al.,
2018; Strawbridge, Lloyd, Meadow, Riley, & Howell, 2015) Additionally, there are disparities in
barriers to attendance at these trainings by type of medical insurance, socioeconomic status,
language and mental health conditions.(Horigan, Davies, Findlay-White, Chaney, & Coates,
2017; Li et al., 2014; Peyrot, Rubin, Funnell, & Siminerio, 2009) Educational interventions for
those with newly diagnosed diabetes need to reach patients and their families at a critical time
when their health behaviors require drastic change. mHealth based training that incorporates
family members and close friends, such as TExT-MED+FANS, can overcome some of these
barriers.

59
Despite TExT-MED+FANS’ development to specifically address the needs of a predominantly
Spanish population at the LAC+USC Emergency Department, including careful translation of
messages and extensive pre-testing for cultural and linguistic congruency, patients who
preferred to receive text-messages in English maintained their improvements in glycemic
control to a greater degree. This disparity by language speaks to the continued obstacles that
Spanish preference patients experience in accessing providers who are culturally competent
and able to speak their preferred language, which results in poor health outcomes. Spanish
speakers in the United States are disproportionately overburdened and are considerably less
successful in managing type 2 diabetes than their English-speaking counterparts, especially
when they have a non-language congruent primary healthcare provider.(Aceves et al., 2022;
Fernandez et al., 2011; Lasater, Davidson, Steiner, & Mehler, 2001) However, the Spanish-
speaking and medically underserved patients who enrolled in the Vida Health Diabetes
Management Program (a novel, culturally adapted, Spanish-language mHealth diabetes
program for glycemic control based out of continuity care clinics) showed an impressive
decrease of -1.23% at 1 year post enrollment.(Edwards, Orellana, Rawlings, Rodriguez-Pla, &
Venkatesan, 2022) The initial TExT-MED patient focused curriculum showed a 0.8% decrease
among Spanish-speaking Emergency department patients at 6 months after enrollment.(Arora
et al., 2014) The critical link in addressing language and cultural barriers to adequate continuity
care for diabetes management may not be fully addressed in individual level interventions such
as SMS text-message, and integration into medical homes may be required to improve patient
outcomes in linguistically underserved communities.

60
A relatively unique feature of this study in the mHealth-based social support for diabetes self-
management literature is the requirement that a family member or friend be enrolled as a
supporter. A diabetes mHealth social support pilot intervention for physical activity in a similar
population restricted enrollment to patients who had an available supporter and exhibited
increased perceived social support for the social support arm, but no difference in physical
activity recorded by a pedometer.(Ramirez & Wu, 2017) Several trials of social support have
shown improvements in diabetes behaviors and glycemic control when a patient elects to enroll
a supporter or informal caregiver.(Aikens, Rosland, et al., 2015; Aikens, Trivedi, et al., 2015; L. S.
Mayberry et al., 2016) However, having higher baseline social support is generally associated
with improved diabetes control and selfcare behaviors in cross sectional studies.(Koetsenruijter
et al., 2016; Mohebi et al., 2018; J. L. Strom & L. E. Egede, 2012) By requiring a supporter, we
were better able to isolate the effect of adding a social support module rather than measuring
the moderation of intervention efficacy by a patient being willing to identify a social supporter,
as well as having a supporter that would be willing to be enrolled. We did not identify a
significant benefit between the augmented social support intervention and the pamphlet
support comparison groups. The ideal method of including a social support person may not
require complex designs of apps to coordinate messaging. Being identified as the support
person for a patient and receiving very basic education on how to best support a loved one with
diabetes may be sufficient.

Despite the importance of this study give the innovation and high-needs population, there are
several limitations. Assessing “dose” of messaging was difficult, as there were no messages
61
that bounced back, and participants used their own devices, limiting ability to determine how
many messages were actually read and thus determining a threshold for effectiveness. This
decision to utilize participants own phones and to utilize universally compatible SMS text-
messaging was an important decision to maintain the pragmatic nature of this trial.
Additionally, we did not include a social desirability measure in baseline data, which may
impact the self-reported behavior measures. However, given that there was no inactive
comparison group, this likely had limited impact. Health behaviors were measured by self-
report given financial constraints of the study; remote pill monitoring, mobile-connected
pedometers and extensive diet records were not possible. The population of this study was
constrained to ED patients with poor control of their diabetes and limits generalizability to
primary care populations and patients with a more modest need for improvement in their
diabetes management. A potentially serious limitation is the significant loss to follow up in this
highly transient patient population. However, the patients who followed up were generally
similar to those who did not, with the exception that patients who stayed in the study had
slightly higher perceived attitudes that the social support they received was negative,
potentially attenuating the estimated effect of the FANS curriculum.

62
Conclusions:
In this randomized controlled trial of an mHealth augmented social support curriculum added
to an existing patient focused mHealth study in ED patients with diabetes, we did not find that
the extra effort required to engage family members via mHealth resulted in overall benefit
compared to the patient focused mHealth curriculum with a standard pamphlet sent to family
members. Patients who were newly diagnosed with diabetes may have benefited the most
from the mHealth augmented social support, with greater and more persistent improvements
in glycemic control. We did not find any mediation effects of the augmented social support by
specific self-care behavior. However, our exploratory findings that the activation of social
support by mHealth is most helpful in patients with newly diagnosed diabetes suggest that the
first years of a diabetes diagnosis are the time when family members and friends are the most
“activatable” and special attention to support person training in this time period may be
critical.

63

Section/Topic
Item
No Checklist item
Reported
on page
No
Title and abstract
1a Identification as a randomized trial in the title 1
1b Structured summary of trial design, methods,
results, and conclusions (for specific guidance see
CONSORT for abstracts)
1
Introduction
Background and
objectives
2a Scientific background and explanation of
rationale
2-4
2b Specific objectives or hypotheses 2-4
Methods
Trial design 3a Description of trial design (such as parallel,
factorial) including allocation ratio
5
3b Important changes to methods after trial
commencement (such as eligibility criteria),
with reasons
N/A
Participants 4a Eligibility criteria for participants 7-8
4b Settings and locations where the data were
collected

Interventions 5 The interventions for each group with
sufficient details to allow replication, including
how and when they were actually
administered
8-12
Outcomes 6a Completely defined pre-specified primary and
secondary outcome measures, including how
and when they were assessed
14
6b Any changes to trial outcomes after the trial
commenced, with reasons
N/A
Sample size 7a How sample size was determined 19
7b When applicable, explanation of any interim
analyses and stopping guidelines
N/A
Randomisation:
Sequence
generation
8a Method used to generate the random
allocation sequence
8
8b Type of randomisation; details of any
restriction (such as blocking and block size)
8
Allocation
concealmen
t
mechanism
9 Mechanism used to implement the random
allocation sequence (such as sequentially
numbered containers), describing any steps
taken to conceal the sequence until
interventions were assigned
8
CONSORT 2010 checklist of information to include when reporting a
randomised trial
64
Implementation 10 Who generated the random allocation
sequence, who enrolled participants, and
who assigned participants to interventions
8
Blinding 11a If done, who was blinded after assignment to
interventions (for example, participants, care
providers, those assessing outcomes) and
how
N/A
11b If relevant, description of the similarity of
interventions
9-10
Statistical
methods
12a Statistical methods used to compare groups
for primary and secondary outcomes
17-18
12b Methods for additional analyses, such as
subgroup analyses and adjusted analyses
18-19
Results
Participant flow (a
diagram is
strongly
recommended)
13a For each group, the numbers of participants
who were randomly assigned, received
intended treatment, and were analysed for
the primary outcome
19
13b For each group, losses and exclusions after
randomisation, together with reasons
19
Recruitment 14a Dates defining the periods of recruitment and
follow-up
6
14b Why the trial ended or was stopped N/A
Baseline data 15 A table showing baseline demographic and
clinical characteristics for each group
22
Numbers
analysed
16 For each group, number of participants
(denominator) included in each analysis and
whether the analysis was by original
assigned groups
22
Outcomes and
estimation
17a For each primary and secondary outcome,
results for each group, and the estimated
effect size and its precision (such as 95%
confidence interval)
30
17b For binary outcomes, presentation of both
absolute and relative effect sizes is
recommended
N/A
Ancillary analyses 18 Results of any other analyses performed,
including subgroup analyses and adjusted
analyses, distinguishing pre-specified from
exploratory
30
Harms 19 All important harms or unintended effects in
each group (for specific guidance see CONSORT for
harms)
25-26
Discussion
Limitations 20 Trial limitations, addressing sources of
potential bias, imprecision, and, if relevant,
multiplicity of analyses
40-41
Generalisability 21 Generalisability (external validity,
applicability) of the trial findings
40-41
65

Funding Acknowledgements:
Funding for this trial was provided by NIDDK Award 5K23DK106538.

Interpretation 22 Interpretation consistent with results,
balancing benefits and harms, and
considering other relevant evidence
36-40
Other information
Registration 23 Registration number and name of trial
registry
6
Protocol 24 Where the full trial protocol can be accessed,
if available
6
Funding 25 Sources of funding and other support (such
as supply of drugs), role of funders
46
66
Chapter 3:
Study 2: Patient and Family Member Perceptions of a mHealth Intervention for Improving
Social Support and Self-care for Latino Patients with Diabetes in Los Angeles

Abstract:
Background & Objectives:
Diabetes and its complications result in over 2 million emergency department (ED) visits
annually. This burden on patients, communities, and the healthcare system is particularly
pronounced in safety-net hospitals. Social support and mobile health (mHealth) interventions
can improve glycemic control and health behaviors while reducing ED visits in vulnerable
populations. In this study, we used qualitative methods to examine acceptability and
satisfaction with a social support mHealth intervention designed for low-income, inner-city
Latinos with diabetes receiving care in the ED.

Methods:
We conducted semi-structured interviews with 50 patients and 56 supporters at the end of a
twelve-month randomized controlled trial of mHealth augmented social support for ED patients
with poorly controlled diabetes (TExT-MED+FANS). The patient interviewees were 72% (n=36)
Spanish speaking, and 40% (n=20) female. The supporter interviewees were 59% (n=33)
Spanish speaking and 69% (n=38) female. The interviews were designed to identify program
aspects that were most persuasive and reasons for selecting a particular supporter. Every
patient and supporter who completed the 12 month trial of the TExT-MED+FANS intervention
was asked if they would complete a separate semi-structured interview. We imported verbatim
transcripts into a qualitative analysis program, Dedoose. A rigorous text-based coding system
67
was used. Transcripts were analyzed iteratively. Broad categorical themes arose from the initial
codes and were developed into a paradigm of barriers and facilitators to good diabetes self-
management and the impact of the mHealth intervention on healthy self-care behaviors.

Results:
Through four rounds of iterative co-coding of 17 interviews by three coders, we developed a set
of 27 codes. Intercoder reliability was excellent (pooled Kappa >0.94). Interview participants
were overwhelmingly positive about the program; the text messages inspired conversations
about health in real-time and increased supporter and patients’ consciousness of the
importance of diabetes management. Intrinsic motivators were desire to fulfill ideal family roles
and to feel healthy; Extrinsic motivators were fear of diabetic complications and family
members providing support and encouragement. The most well-regarded text-messages were
those that had specific calls for action for healthy living or specific support behaviors.
Supporters were selected for their perceived emotional strength, prior experience living with
diabetes or potential to offer instrumental support. Through our inductive analysis, we also
identified social barriers to diabetes management that were not addressed in the intervention,
and should be considered in future interventions: access to primary care, food costs and
medication affordability.

Conclusions:
Augmenting a mobile health diabetes intervention with social support messages to family
members was well received, and the most impactful messages were calls to action. Social
68
support delivered via mobile health may be a powerful and easy to implement and scale
addition to ED-based behavioral interventions. However, social determinants of health
including access to care and medication affordability may need to be simultaneously addressed
to obtain maximal benefits.

69
Introduction:

Although diabetes is a nationwide epidemic, US Latinos is a particularly vulnerable population.
Latinos are more likely than non-Latino Whites to develop diabetes and have higher rates of
diabetic complications and diabetes related mortality. (Beckles & Chou, 2016; Geiss et al., 2014)
This disparity is prominent in Los Angeles County, where almost half the residents are Latino
and diabetes prevalence is 14.4%; much higher than the national prevalence of 8.6%. (Hales,
Carroll, Kuo, & Simon, 2019; "QuickFacts: Los Angeles County, California," 2022) Patient with
diabetes who rely on the Los Angeles County + University of Southern California Emergency
Department for care are a high need population, with poor glycemic control, low diabetes
knowledge, and limited access to care. (Menchine et al., 2013) The mean A1C --or three-month
average measure of blood glucose, also known as glycemic control-- for this population is 8.8;
normal A1C are less than 5.7 and good glycemic control for people with diabetes is less than 7.
This is alarming, as increasing A1C is associated with elevated rates of microvascular and some
macrovascular complications, (Nathan et al., 2014; UKPDSGroup, 1998) with a significantly
increased rate at a level of 8 or higher. (Huang et al., 2011) The patients who resort to seeking
care in the ED require dedicated attention to address their unique health needs and barriers to
care.

Importantly, ED care presents a unique opportunity to reach patients and families during a
health crisis, when they are susceptible to behavior change. (Boudreaux et al., 2012). Those
seeking care in the ED represent a group of patients with greater health needs, with higher
rates of hospitalization and worse glycemic control. (Birtwhistle et al., 2017). As emergency
70
care is more expensive than healthcare in other sites, (Galarraga et al., 2015) ED interventions
should maximize the benefit of these visits through improved behaviors and health outcomes.
ED-based interventions can also bridge care to patients until primary care can be established
for those lacking primary care, or during the wait for the next visit for patients with limited
access to a primary care provider.

This opportunity for behavior change is critical because the disparities in complications from
diabetes are not inevitable. Good glycemic control can be achieved with access to quality
medical care combined with significant and sustained behavioral changes, and continuous
monitoring.(American Diabetes, 2022; UKPDSGroup, 1998) This knowledge is not new,
however, translation into population level improved health outcomes has lagged, with rates of
diabetes and complications from diabetes increasing annually.(Hales et al., 2019) Good disease
management and glycemic control consists of many daily health decisions and behaviors,
impacting nearly every facet of the life of person with diabetes. Patients might have to make
changes to increase regular exercise, improve nutrition choices, conduct regular self-monitoring
of blood glucose, follow a medication regiment, perform daily foot care, engage with
healthcare systems, and practice stress reduction. (Margaret A. Powers et al., 2015; Sarkar et
al., 2006; Toobert & Glasgow, 1994) These behaviors (and subsequent glycemic control) in turn
are linked to a patient’s intentions, underlying self-efficacy, psychological distress and social
support (Joni L. Strom & Leonard E. Egede, 2012; van Dam et al., 2005) and impact health
outcomes. However, the interplay of these behaviors must be carefully titrated as too tight of
glycemic control can result in uncomfortable and sometimes deadly episodes of hypoglycemia
71
(low blood sugar). (Huang et al., 2011) Given how difficult it is to avoid diabetes complications,
many patients face challenges in maintaining consistent good disease management, especially
those from marginalized groups and from low socio-economic positions who face additional
community level barriers to healthy choices, and language barriers. (Gallegos-Macias et al.,
2003; Walker et al., 2014)

However, culturally and linguistically appropriate interventions that emphasize improving self-
efficacy can combat this disparity, especially social support interventions. (Kirk et al., 2013;
Steed et al., 2003) As classically formulated, social support consists of four domains: emotional,
instrumental, informational and appraisal. (Cohen, 1988; House et al., 1988) Emotional support
refers to expressions of empathy, love, trust and caring which result in improved affect for the
recipient, such as listening to the concerns of a loved one. Instrumental support consists of
tangible aid and services, such as driving a loved one to a doctor’s appointment. Informational
support generally refers to advice, suggestions, and information provided to a loved one.
Appraisal support is the provision of information useful for self-assessment. While the
American Diabetes Association recommends a mixture of emotional, instrumental and
informational support, there is value in appraisal support, particularly in the context of behavior
change interventions using motivational interviewing. Interventions designed to increase
general or disease specific social support should intentionally target one or more of these
arenas. (Michele Heisler, 2007)

72
Harnessing social support may be the key to this recalcitrant problem, particularly for Latino
populations. There is evidence that appropriate social support is associated with improved
health behaviors and glycemic control for diabetes patients in ethnic minority patient
populations, posited to be due to less individualistic tendencies that the majority US
population. (Lindsay S. Mayberry et al., 2014) Training family members and peers to support
patients with diabetes has been shown to improve patient motivation, healthy behaviors and
glycemic control. (Hu et al., 2014; Kirk et al., 2013; Spencer et al., 2018; Tricia S. Tang et al.,
2011; T. S. Tang et al., 2015; van Dam et al., 2005) These interventions have been
enthusiastically accepted in multiple Latino populations, potentially tapping into cultural
strengths such as familismo. (Berg et al., 2018; Teufel-Shone et al., 2005; Thompson et al.,
2007; Two Feathers et al., 2005) However, existing research in developing social support
interventions have had inconsistent results. Additionally, these interventions have faced
difficulties in recruiting family members and create financial, space and personnel burdens to
implement and maintain. (Michele Heisler, 2007; L. S. Mayberry et al., 2016) These barriers
have prevented widescale implementation of social support interventions for patients with
diabetes.

To reduce these burdens, new interventions using internet and communication technologies
(ICT) to engage supporters have emerged. ICT interventions -- including mHealth, telephone
based interventions and web-based interventions—can be used to train support persons, to
prompt them to provide appropriate social support and as an avenue to offer that support to a
person with diabetes. (John D Piette et al., 2008) Employing ICT for social support interventions
73
could result in family members who are the most influential (regardless of proximity or location
from the patient) trained to provide support, rather than the ones who live closest to the
patient with the most available free time to travel. ICT may also broaden the reach of
interventions that develop new support relationships between peers, matching people with
diabetes to those with similar experiences and challenges rather than those patients that share
a medical home. Merging ICT and social support interventions could generate a solution that
has the cost-effectiveness and scalability of mobile technologies coupled with the personal
touch of social support. ICT based social support interventions could reduce the need for
physical presence and make social support interventions more accessible to populations in
need. US Latinos have the technological capacity to engage with ICT interventions, and may
even prefer these modalities for health information.(Livingston, 2011; Lopez, Gonzalez-Barrera,
& Patten, 2013)

While ICT based interventions are generally promising for US Latinos, mHealth specific
interventions may also be more feasible for low-income Latino patients, such as those that seek
care in the ED of LAC+USC. Research indicates these patients are more likely to access the
internet and health information specifically via mobile devices, and maybe less likely to have a
“land phone” and/or personal computer. (Livingston, 2011) mHealth interventions improve
disease management of chronic illnesses, including diabetes. (Bell et al., 2012; Quinn et al.,
2008; Quinn et al., 2011; Seto et al., 2012) It is important to consider that many existing ICT and
mHealth interventions require personal computers or frequent engagement with mobile
applications. These are not widely used in resource-poor communities such as safety-net
74
emergency departments (EDs). (Arora et al., 2013) Additionally, Latino patients from low-
resource settings have enthusiastically joined text-message and interactive voice-recording
interventions to improve diabetes self-care indicating this is an acceptable modality for the
population. (Arora et al., 2014; E. Burner et al., 2018; Hay et al., 2018; Ramirez & Wu, 2017)

Importantly, explanations for inconsistent results from predominantly quantitative studies
include heterogeneous interventions, dissimilar patient populations, and the mixed effect of
family for different individuals. If social support strategies will be incorporated into mHealth
interventions, these discrepancies must be addressed. A qualitative, inductive approach allows
for a more nuanced understanding of the role of family and close friends in a patient’s self-care
for diabetes. With this in mind, we designed a qualitative study to augment the analysis of an
RCT of a social support and behavior change intervention. TExT-MED+FANS (Trial to Examine
Text-Messaging in Emergency patients with Diabetes+ Family and friends Network Support) is
an intervention designed for ED patients with diabetes at LAC+USC. It adds a mobile social
support module for family members and friends, FANS, to the existing patient-focused TExT-
MED intervention. This social support-focused intervention will engage family members in the
goal setting and behavior change process that the patients are undergoing. TExT-MED+FANS
uses mHealth to overcome the transportation and time obstacles that social support solutions
face by offering social support training via a mobile platform. Using mobile training, a patient
can select the most influential person to support them, rather than the most proximate.

75
We used a sequential mixed-methods study design, analyzing transcripts of semi-structured
individual interviews with patients and family members who had completed patient focused
mHealth diabetes intervention in the mHealth augmented social support arm, detailed
interviewer notes taken while patients and supporters completed the self-report measures at
the conclusion of the trial and the text-messages sent in to the mHealth platform by patients
and supporters. Through this qualitative analysis, we explored potential mechanisms that
contribute to changes in behaviors and outcomes, including changes to self-efficacy, perceived
barriers, perceived threat of diabetes and perceived benefits of glycemic control. Additionally,
we are examined the user experience as captured in the Text-MED+FANS application of both
patients and supporters, focusing on content of messages.

Aim 2: Evaluate experience with TExT-MED+FANS and impact on perceptions of disease, social
support and motivation for behavior change among intervention patients and supporters via
qualitative analysis of semi-structure individual interviews.
Research Question 2a: What were the user experience with and perceived benefits of
TExT-MED+FANS for supporters and patients?

Research Question 2b: What changes in perceptions of disease, motivations, social
support resources and coping strategies occur with participation in TExT-MED+FANS.

76
Methods
Interview guide development
The interview guide was developed from prior work by this research group,(E. Burner et al.,
2018) with a focus on motivation, barriers to care, and perceptions of disease and disease
management using a “think-back” approach to attempt to have participants recall their disease
management before the intervention and then compare to the time of interview. The guide
went through 2 rounds of testing with RAs in the department as well as experienced qualitative
researchers. The guide was then translated into Spanish by a native Spanish speaking RA and
back translated by two other RAs (one of who is ancestrally from a separate country than the
initial translator) who had not yet read the initial English version. Issues with Spanish
translations were resolved by consensus.(Brislin & Freimanis, 2001) (See Appendices 1-4 for
final question guide)

Data Collection
This study was approved by the local Institutional Review Board prior to initiation. Data
collection and analysis were designed in accordance with the COREQ Guidelines (see Appendix
5).(Tong, Sainsbury, & Craig, 2007). Patients were consented for interviews at their initial
enrollment in the randomized controlled trial, and reminded at time of interview that they
could refuse to participate at any time.

Patients and supporters who completed the experimental arm of the TExT-MED+FANS RCT
were invited to complete semi-structured interviews after their final study visit to complete
77
specimen collection and psychosocial and health behavior measures. Patients and supporters
could either complete their interview on the same day or return on another day. All
participants elected to complete semi-structured interviews on the same day as their final
study measures were collected. Interviews were conducted in the office of the RAs located in
the LAC+USC Medical Center, with the exception of two interviews which were conducted in a
participant's home due to severe mobility issues. Participants completed only one semi-
structured interview each. Participants were compensated $100 at the completion of their
semi-structured interview in addition to the compensation for their time for the RCT study
completion.

All interviews were conducted by JM, NRG, and AA, three natively Spanish-English bilingual
females with bachelor’s degrees. They were professional research coordinators and research
associates at the time of the study. JM and AA had previously conducted semi structured
interviews and focus groups after training with EB. Additionally, the first 15 transcripts were
reviewed by EB with feedback on strategies to increase responses from participants without
creating the perception of approving or disproving statements made by participants. JM, AA
and NRG had previous experience with the participants as they conducted or oversaw every
patient follow up visit in the TExT-MED+FANS study. All three had interest in understanding if
TExT-MED+FANS improved diabetes management. EB had interest in understanding the
mechanism of action of TExT-MED+FANS in building social support, and initially developed the
FANS curriculum.

78
Patients and supporters were interviewed in their primary language, either English or Spanish.
In the case of bilingual patients, they participated in the language in which they received their
text messages. A natively fluent Spanish-speaking RA conducted the Spanish language
interviews. At least one other RA was present during each interview to assist with logistics and
note taking. Interviews were audio recorded to accurately capture both the words and the
context of statements made by participants. Recordings were transcribed in the original
language, then reviewed and corrected by a research coordinator. Spanish language
transcriptions were translated by RAs who are native Spanish speakers and who have passed an
internal language competency assessment. Transcripts were not given to participants for their
review for accuracy.

Additional data was collected from the spontaneous text messages that patients and
supporters sent to the mHealth platform. While not encouraged, a fair number of patients sent
in messages to the system on a regular basis. Only texts that consisted of more than one word
was aggregated with the interview transcripts (i.e., ok, or “thumbs up” icon alone would be
excluded).

Analysis
The analysis of the transcripts was conducted from the Constructivist Grounded Theory
Framework which emphasizes generating themes and theories from the data gathered from
participants rather than pre-existing frameworks. (Charmaz, 2014) The analysis of the interview
was transcript-based to enhance rigor. The study team developed an initial set of codes after
the first reading of 15 transcripts and line by line coding. This coding scheme was further
79
developed through the analysis process. As further participant transcripts were coded and
analyzed, concepts and phenomena were compared to one another in a constant, iterative
process to ensure the coding structure continued to accurately represent the experiences of
the patients. The transcripts were analyzed using Dedoose™. Dedoose™ is a computer-based
qualitative analysis tool that contains data organization tools necessary to complete an analysis
of different subgroups.("Dedoose," 2012) After the development of the preliminary set of
codes, we conducted multiple rounds of co-coding with iterative redefinition of the codes until
consensus on codes definitions was achieved. The final tests of coders’ agreement on code
application reached Cohen’s pooled Kappa coefficients of greater than 0.9, showing excellent
inter-coder reliability.(De Vries, Elliott, Kanouse, & Teleki, 2008) NRG and JM then coded all
transcripts through focused coding with the final codebook. We reviewed these transcripts as
they were coded, developing overarching phenomenon categories (“axial coding”) as per
standard grounded theory techniques by generating memos and meeting regularly to discuss
developing themes. These themes were reanalyzed and expanded with new categories that
comprise the experience with TExT-MED+FANS (Charmaz, 2006). Major themes were selected
for salience in addition to frequency, and both were considered.

Strategies to maintain integrity and enhance validity of the qualitative findings were used,
including consistent use of the discussion guide, audiotaping, employing of RAs from multiple
ancestral backgrounds for data analysis, standardized coding and analysis processes, and used
of a pooled Kappa measure to ensure coders were adequately trained. We did not conduct
member-checking of final themes.
80

At the conclusion of coding, qualitative data was combined with the quantitative results of the
TExT-MED+FANS RCT to gain a better understanding of which patients benefited the most from
the intervention. Patients were coded as “high responders” (A1C drop of 1 or greater) or “low
responders.” By entering this quantitative information into Dedoose™, we were able to
organize the quotes and corresponding themes into these two categories, as was planned
during the trial design. We re-examined the distributions of codes and themes with this
dichotomous grouping to see if patterns of motivational and behavioral change became
evident.

Results:
Participant characteristics
50 interviews with patients and 56 interviews with supporters took place. The patient
interviewees were 72% (n=36) Spanish speaking, and 40% (n=20) female. The supporter
interviewees were 59% (n=33) Spanish speaking and 69% (n=38) female. All RCT intervention
group patients and supporters who completed their 12 month follow up measures elected to
stay for the semi-structured interview after the final RCT visit. No participants elected to return
on a following day. One English language supporter audio file was corrupted and not able to be
analyzed. Patient interviews lasted a mean of 7:56 minutes and supporter interviews lasted a
mean of 7:15 minutes?

Codebook
81
The three interviewers and the PI conducted initial coding on 15 interviews, iteratively
developing a codebook (see Table 8 for first draft and Table 9 for final version of the codebook).
Data saturation was reached at the 17
th
interview, however interviews were continued for all
intervention arm participants who had been promised the opportunity to share their views and
to ensure we captured a broad and rich representation of the experiences of patients and
supporters. Themes were derived from repeated analysis of the data, generating memos with
the Charmaz techniques and examination of the code co-occurrence chart (Figure 6). We
qualitatively examined the frequency of code application by viewing the “code cloud”, with
more frequently applied codes represented as larger text side. (Figure 7)

82
Table 8: Initial Version of Study Codebook
1. New Diagnosis/recent diagnosis
2. Comorbidities: heart/htn/depression
3. Causes of disease
4. Alternative (to west med – tx & causes)
5. Info sources
I. Family
II. Internet
III. Friends
IV. Texts
6. Balance,
I. Controlled vs. Uncontrolled
II. Managing many aspects at once
7. Benefits of the program patient specific:
(co code if a code exists for benefit)
I. Includes text msgs,
II. Coming in and talking
III. Changes in motivation,
IV. Increased knowledge,
V. Increased attention to disease,
VI. Increased awareness, increase
communication with subject.
VII. Decreasing fatalism,
VIII. Controlling sugar and disease
IX. Support relationship better
8. Benefits to patient via supporter:
I. Supporter behaviors,
II. Supporter knowledge (what is good
self-care),
III. Supporter motivation to support.
9. Types of Support:
I. Emotional
II. Positive regard, encouragement
III. Physical
IV. Negative Support/Nagging
V. Appraisal support: includes
feedback, advice.
VI. Instrumental
VII. informational: sharing information.
VIII. Ensuring good self-care
10. Supporter Health (supporter’s self-care,
support’s assessment of risk)
11. Witnessing diabetes: in family, friends
12. Advice: tips and imparting knowledge
13. Motivation:
I. Family
II. Self
III. From daily messages
IV. Fear of consequences (co-code)
14. LACK OF action/agency
I. Unaware of disease/ No me han
dicho nada
II. Denial
III. Lack of knowledge
IV. Resistant to advice
V. Careless/ “on/off your mind”
VI. Inattentive/forgetting.
15. Food/nutrition
16. Forget to check sugar.
17. Consequences/ what happens if do or
don’t control the disease/prognosis.
I. Understands consequences.
18. Self-Care: actual behaviors (co-code
with specific behavior)
19. Checking sugar: doing it or not,
20. Nutrition
21. Exercise
22. Going to the doctor/accessing care
23. Medication (subcode insulin)
24. Stress/emotional context
25. Difficult Social situation (homeless ,
home environment, financial)
26. Barriers/Challenges (need to co code
with a self-care behavior)
27. Impactful msgs: which messages stuck
out, or caused change (Co code with
self-care behavior)
I. Reminders
II. Advice/tips
28. Feedback on program (to improve):
I. Referral to social work,
II. Expand to more people,
III. More sites
IV. Able to refer in friends and family.
29. Before Code (must be co-coded)
30. Physical symptoms(+/-),NV,energy level,
feeling better, physical control of DM)
83

Table 9: Final Study Codebook
Code Definition
New/recent diagnosis Diagnosis referenced as recent or new, how these impacts?
Comorbidities Includes Heart, kidney dz, depression (clinical depression) when
referenced as disease, not mood
Causes of disease Perceptions of what causes diabetes: include things like susto, emotional
events, also obesity, genetics
Alternative Treatments used outside western med, or what doctors have prescribed
or recommended
Info sources How patients and supporters learn about diabetes causes, treatments,
strategies
Texts
Social Non-family, friends, acquaintances
Internet
Family
Health Care Personal Medical Doctors, Nurse, diabetes educational classes, other types of
health care providers
Benefits to pt specific Perceived benefits from patients own changes in intentions, actions,
motivation, thoughts
Support from messages
Support from staff
Changes in motivation
Increased knowledge
Increased attention to dz Includes increased awareness on a regular basis
Self-care behaviors
Self-efficacy Co-code
Benefits to pt via supporter: Perceived benefits from changes in supporters’ intentions, actions,
motivation, thoughts
Supporter behaviors,
Supporter knowledge (what is good self-care)
Supporter motivation to provide support
Increase communication
Types of Support Social support provided by the supporter or other people
Emotional Positive regard, encouragement, compliments; teasing (co-code with non-
supportive)
Appraisal support includes feedback, advice, nagging (co-code as non-supportive)
Instrumental Actual support activities: i.e., Picking up/paying for meds; driving to
doctor; cooking or buying food
Informational sharing information; fear tactics (co-code as non-supportive)
Non- supportive Nagging, teasing, ignoring
Physical presence
Witnessing diabetes in family, friends; can be applied to patients or supporters
Healthcare experience describing experience or relationship with provider
Self-efficacy/Fatalism Belief that one can manage their own diabetes, or belief that nothing they
can do matters
Motivation Source of motivation to make healthy choices
Family Children, spouse, close friends, etc
Self
From daily messages
Fear of consequences Co-code
84
Unaware Of diagnosis or of seriousness of consequences (co-code with
consequences) “no me han dicho nada”; unaware how to manage DM;
New dx if applies
Negative Coping
Denial Disbelieve that pt has DM
Resistant To advice, to behavior change
Avoidance Careless, “on/off your mind”, inattentive, forgetting, not thinking about it
Self-Care: actual behaviors (select the subcode with specific behavior)
Checking sugar doing it or not
Checking their feet
Nutrition
Exercise
Accessing care Going to the doctor/accessing care
Medication Pills vs insulin
Consequences What happens if do or don’t control the disease/prognosis ; Understands
consequences
Physical symptoms (positive or negative symptoms), nausea/vomiting; energy level, feeling
better, physical control of DM
Balance Controlled vs. Uncontrolled, Managing many aspects at once
Emotional Stress/emotional context /perseveration on disease
Difficult Social situation homeless, home environment, financial
Barriers/Challenges What keeps one from making the healthy choice; need to co code with a
self-care behavior, motivation, or other code
Strategy Explicitly described behaviors, plans and actions to overcome challenges
Impactful msgs which messages stuck out, or caused change (Co code with self-care
behavior)
Reminders
Advice/tips
Challenges
Feedback on program Negative aspects or areas to improve
Social work Referral to social services or social worker built in
access to care Add ability to refer to find a primary care doctor or specialists
Expand More sites, or ability to refer in family and friends
More contact More text msgs, info sources, communication with team
Supporter Health (supporter’s self-care, support’s assessment of risk)
Supporter’s own DM Supporters experience having diabetes
Before Code (must be co-coded); references to disease management, behaviors,
thoughts or beliefs before program
85
Figure 6: Code Co-occurrence Frequency Chart
Frequency of code cooccurrences, with heat map indicating frequency of overlap: dark blue
indicating low co-occurrence, red indicating highest co-occurrence
Alternative
Balance
Barriers/Challenges
Before Code
Benefits to pt specific
Changes in motivation
Increased attention to Dz
Increased knowledge
Self-care behaviors
Self-efficacy
Support from messages
Support from staff
Benefits to pt via supporter
Increase communication
Supporter behaviors
Supporter knowledge
Supporter motivation
Causes of disease
Comorbidities
Consequences
Difficult Social situation
Emotional
Feedback on program
Expand
More Contact
Social work
access to care
Healthcare experience
Impactful msgs
Advice/tips
Challenges
Reminders
Info sources
Family
Health Care Personnel
Internet
Social
Texts
Motivation
Family
Fear of consequences
From Daily Messages
Self
Negative Coping
Avoidance
Denial
Resistant
New/recent diagnosis
Physical symptoms
Self-Care
Accessing care
Checking sugar
Checking their feet
Exercise
Medication
Nutrition
Self-efficacy/Fatalism
Strategy
Supporter Health
Supporter’s own DM
Types of Support
Appraisal support
Emotional
Informational
Instrumental
Non supportive
Physical presence
Unaware
Witnessing DM complications
Witnessing Pt DM management
Totals
Alternative 0 0 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2 1 2 3 1 1 0 0 0 1 0 0 0 0 3 1 0 1 1 1 0 0 0 0 0 3 2 0 1 0 1 9 4 1 0 3 6 8 0 3 0 0 7 3 1 2 4 3 2 0 1 1 91
Balance 0 0 23 3 7 0 4 4 2 0 3 1 1 1 0 0 0 0 6 8 6 10 1 1 0 1 0 3 2 2 2 2 7 0 4 1 0 4 7 2 2 2 3 2 2 0 2 1 10 19 6 2 0 6 10 11 4 5 0 0 7 4 4 1 1 3 1 2 4 4 238
Barriers/Challenges 7 23 0 19 20 3 6 3 1 2 6 2 11 4 7 2 3 6 35 25 54 44 3 1 0 3 0 20 7 5 4 7 12 2 8 1 1 2 46 20 5 5 18 32 21 0 18 2 37 166 79 27 2 34 89 90 10 25 6 1 70 22 21 8 23 26 15 6 13 21 ###
Before Code 0 3 19 0 15 2 9 5 4 0 3 0 11 1 3 7 5 0 3 43 3 6 0 0 0 0 0 4 1 0 0 1 10 1 4 1 2 5 26 7 12 3 7 23 15 2 9 3 6 52 6 6 0 10 19 38 4 6 2 2 22 12 9 8 7 4 5 34 9 10 550
Benefits to pt specific 0 7 20 15 0 35 93 38 56 7 105 31 13 5 3 3 3 0 4 22 6 12 2 1 0 0 0 8 69 46 29 59 31 0 5 2 0 25 64 13 14 37 21 7 4 1 2 1 6 109 15 11 2 37 56 64 10 17 4 0 27 11 12 4 5 6 7 5 3 6 ###
Changes in motivation 0 0 3 2 35 0 16 9 10 2 19 3 0 0 0 0 0 0 0 3 1 1 0 0 0 0 0 0 12 9 9 12 3 0 1 0 0 2 29 8 3 19 10 2 1 0 0 0 1 16 3 3 1 7 9 8 2 3 0 0 3 0 2 1 0 0 0 0 1 0 286
Increased attention to Dz 0 4 6 9 93 16 0 22 31 4 29 5 5 2 2 1 1 0 3 18 1 5 0 0 0 0 0 3 20 14 8 16 12 0 3 1 0 9 26 2 10 17 9 5 3 0 2 0 4 49 7 8 1 17 25 31 7 9 2 0 9 4 6 0 2 1 3 3 2 4 611
Increased knowledge 0 4 3 5 38 9 22 0 11 3 15 3 1 0 0 0 1 0 1 6 0 1 0 0 0 0 0 1 8 6 5 6 14 0 2 1 0 12 12 0 3 10 3 0 0 0 0 1 3 18 3 2 1 6 7 14 3 3 0 0 3 1 2 0 0 1 0 3 0 0 278
Self-care behaviors 0 2 1 4 56 10 31 11 0 1 27 1 2 1 1 0 0 0 0 6 0 2 1 1 0 0 0 1 24 19 10 21 6 0 0 0 0 6 15 0 0 13 6 3 1 1 0 0 1 55 5 4 2 23 35 33 1 9 1 0 2 2 0 0 0 0 0 0 0 4 464
Self-efficacy 0 0 2 0 7 2 4 3 1 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 4 1 0 2 4 0 0 0 0 0 0 4 1 0 0 0 0 4 6 4 0 0 0 0 0 0 0 0 0 0 0 0 55
Support from messages 0 3 6 3 105 19 29 15 27 2 0 11 6 2 1 2 1 0 0 6 2 4 1 1 0 0 0 1 60 41 27 55 16 0 0 1 0 15 40 6 6 30 8 3 2 0 1 0 1 52 7 3 2 21 33 28 2 9 1 0 14 7 6 3 3 2 4 0 2 0 763
Support from staff 0 1 2 0 31 3 5 3 1 1 11 0 1 0 0 0 1 0 0 0 2 4 1 0 0 0 0 3 2 1 1 2 2 0 1 0 0 1 5 1 0 3 2 0 0 0 0 0 0 8 2 1 0 1 4 3 1 2 1 0 3 0 1 0 0 2 1 1 0 0 124
Benefits to pt via supporter 0 1 11 11 13 0 5 1 2 0 6 1 0 39 51 50 32 1 1 11 1 7 3 1 2 0 0 0 31 23 14 21 24 0 4 2 0 22 41 10 5 26 5 7 3 0 4 1 4 75 3 7 3 20 17 65 2 25 16 4 84 36 39 15 37 7 15 7 7 10 995
Increase communication 0 1 4 1 5 0 2 0 1 0 2 0 39 0 7 9 2 0 0 1 0 5 2 1 1 0 0 0 16 12 8 13 7 0 0 1 0 7 13 2 0 10 1 1 0 0 1 0 1 17 1 2 2 3 4 12 0 1 4 0 28 13 16 7 5 0 4 0 1 3 301
Supporter behaviors 0 0 7 3 3 0 2 0 1 0 1 0 51 7 0 13 8 1 1 7 0 2 0 0 0 0 0 0 10 5 5 7 6 0 0 1 0 5 14 3 4 7 3 2 0 0 2 0 2 39 0 2 1 13 9 35 1 23 8 4 46 16 19 7 31 6 8 2 4 6 465
Supporter knowledge 0 0 2 7 3 0 1 0 0 0 2 0 50 9 13 0 5 0 1 5 1 1 0 0 0 0 0 0 14 10 5 7 20 0 4 1 0 18 9 1 2 7 1 4 2 0 2 1 2 29 1 3 2 7 4 28 2 8 7 1 23 9 11 4 9 2 3 6 3 3 377
Supporter motivation 0 0 3 5 3 0 1 1 0 0 1 1 32 2 8 5 0 1 0 2 0 1 1 0 1 0 0 0 6 5 3 5 1 0 0 0 0 1 25 8 4 15 4 1 1 0 0 1 1 17 2 2 0 5 7 14 0 4 5 1 19 13 12 2 6 2 5 2 1 2 270
Causes of disease 1 0 6 0 0 0 0 0 0 0 0 0 1 0 1 0 1 0 2 2 2 2 0 0 0 0 0 4 0 0 0 0 1 1 1 0 0 0 6 2 1 1 6 1 1 0 0 0 4 9 6 1 1 3 4 5 0 2 2 1 3 1 1 0 0 0 1 0 2 2 91
Comorbidities 2 6 35 3 4 0 3 1 0 0 0 0 1 0 1 1 0 2 0 14 18 19 2 1 0 2 0 6 0 0 0 0 4 2 2 0 1 1 19 12 3 2 8 6 3 0 5 1 8 36 28 11 1 9 24 12 3 4 1 1 19 4 4 2 8 9 1 0 4 2 382
Consequences 1 8 25 43 22 3 18 6 6 0 6 0 11 1 7 5 2 2 14 0 8 13 1 1 0 0 0 5 2 2 1 1 15 3 5 1 4 5 40 12 34 5 12 12 7 0 7 0 13 49 15 7 2 17 22 34 8 7 7 5 21 11 6 8 8 10 2 9 23 13 673
Difficult Social situation 2 6 54 3 6 1 1 0 0 0 2 2 1 0 0 1 0 2 18 8 0 30 2 1 0 2 0 10 1 0 0 1 0 0 0 0 0 0 17 10 4 0 6 10 8 0 8 2 9 52 38 8 1 9 35 18 5 6 1 0 26 4 3 1 10 12 6 0 5 5 474
Emotional 3 10 44 6 12 1 5 1 2 0 4 4 7 5 2 1 1 2 19 13 30 0 3 1 0 3 0 10 4 4 4 4 1 0 1 0 0 0 23 15 6 4 7 11 6 0 10 1 14 52 32 7 0 15 34 20 8 11 3 0 29 8 13 1 10 9 8 2 10 5 584
Feedback on program 1 1 3 0 2 0 0 0 1 0 1 1 3 2 0 0 1 0 2 1 2 3 0 12 31 5 3 0 4 3 4 3 4 0 0 1 1 2 2 0 0 1 1 5 2 1 3 1 1 7 1 1 0 2 2 5 0 0 1 0 0 0 0 0 0 0 0 0 1 1 135
Expand 1 1 1 0 1 0 0 0 1 0 1 0 1 1 0 0 0 0 1 1 1 1 12 0 2 1 0 0 2 2 2 2 2 0 0 0 1 1 1 0 0 0 1 3 2 1 1 1 0 2 1 0 0 0 1 1 0 0 1 0 0 0 0 0 0 0 0 0 1 1 57
More Contact 0 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 1 0 0 0 0 0 31 2 0 1 1 0 2 1 2 1 3 0 0 1 1 1 1 0 0 1 0 1 0 0 1 0 0 2 0 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 59
Social work 0 1 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 0 2 3 5 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 2 1 0 2 1 1 3 1 0 0 1 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 35
access to care 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5
Healthcare experience 1 3 20 4 8 0 3 1 1 0 1 3 0 0 0 0 0 4 6 5 10 10 0 0 0 0 0 0 0 0 0 0 8 1 7 0 1 1 9 5 3 0 4 1 1 0 0 0 7 33 25 4 1 4 20 14 4 4 0 0 16 4 2 1 4 8 4 2 3 3 284
Impactful msgs 0 2 7 1 69 12 20 8 24 1 60 2 31 16 10 14 6 0 0 2 1 4 4 2 2 0 0 0 0 93 57 97 30 0 1 2 0 30 41 5 3 36 7 2 1 0 0 0 0 78 4 5 3 32 38 50 1 7 7 1 30 15 17 5 6 3 5 0 3 4 ###
Advice/tips 0 2 5 0 46 9 14 6 19 1 41 1 23 12 5 10 5 0 0 2 0 4 3 2 1 0 0 0 93 0 49 68 24 0 1 0 0 24 30 3 1 27 5 1 1 0 0 0 0 58 3 3 2 31 23 43 1 7 7 1 19 12 11 4 3 2 3 0 1 4 780
Challenges 0 2 4 0 29 9 8 5 10 0 27 1 14 8 5 5 3 0 0 1 0 4 4 2 2 0 0 0 57 49 0 49 14 0 0 2 0 14 19 5 2 16 5 0 0 0 0 0 0 39 1 1 0 23 16 30 0 4 4 0 12 6 6 3 3 1 2 0 3 3 536
Reminders 0 2 7 1 59 12 16 6 21 0 55 2 21 13 7 7 5 0 0 1 1 4 3 2 1 0 0 0 97 68 49 0 17 0 0 0 0 17 34 4 1 31 6 2 1 0 0 0 0 66 4 5 3 28 38 39 0 4 6 0 23 12 14 4 4 3 4 0 2 3 840
Info sources 3 7 12 10 31 3 12 14 6 0 16 2 24 7 6 20 1 1 4 15 0 1 4 2 3 0 0 8 30 24 14 17 0 8 26 10 8 58 24 2 11 12 6 5 4 0 1 0 5 55 6 4 3 16 13 47 1 9 9 3 26 15 9 8 6 4 4 8 4 8 738
Family 1 0 2 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 2 3 0 0 0 0 0 0 0 1 0 0 0 0 8 0 1 1 2 0 3 1 2 0 1 1 1 0 0 0 2 7 2 0 0 4 3 7 0 1 1 0 2 1 0 1 1 0 0 0 1 2 67
Health Care Personnel 0 4 8 4 5 1 3 2 0 0 0 1 4 0 0 4 0 1 2 5 0 1 0 0 0 0 0 7 1 1 0 0 26 1 0 0 0 6 4 1 3 0 2 3 2 0 1 0 3 18 4 2 1 5 3 13 1 5 3 2 9 7 3 4 1 2 3 4 1 3 200
Internet 1 1 1 1 2 0 1 1 0 0 1 0 2 1 1 1 0 0 0 1 0 0 1 0 1 0 0 0 2 0 2 0 10 1 0 0 0 6 3 0 1 2 1 1 1 0 0 0 0 3 0 0 0 0 1 3 0 1 0 0 3 1 0 2 1 0 0 0 0 0 62
Social 1 0 1 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 4 0 0 1 1 1 0 0 1 0 0 0 0 8 2 0 0 0 0 4 1 4 0 0 0 0 0 0 0 0 3 1 0 0 2 1 3 0 0 1 0 1 0 0 0 0 1 0 1 2 1 49
Texts 1 4 2 5 25 2 9 12 6 0 15 1 22 7 5 18 1 0 1 5 0 0 2 1 1 0 0 1 30 24 14 17 58 0 6 6 0 0 15 0 4 11 3 2 1 0 1 0 2 31 1 3 2 9 7 28 0 6 6 1 15 9 7 3 3 1 2 4 2 4 487
Motivation 0 7 46 26 64 29 26 12 15 4 40 5 41 13 14 9 25 6 19 40 17 23 2 1 1 0 0 9 41 30 19 34 24 3 4 3 4 15 0 85 53 76 60 9 6 0 4 1 13 111 30 12 4 33 53 74 10 22 13 6 61 29 30 11 20 11 11 4 18 16 ###
Family 0 2 20 7 13 8 2 0 0 1 6 1 10 2 3 1 8 2 12 12 10 15 0 0 0 0 0 5 5 3 5 4 2 1 1 0 1 0 85 0 18 11 22 4 3 0 1 0 9 42 15 7 1 9 26 25 4 8 6 2 28 13 14 3 9 5 3 1 10 4 551
Fear of consequences 0 2 5 12 14 3 10 3 0 0 6 0 5 0 4 2 4 1 3 34 4 6 0 0 0 0 0 3 3 1 2 1 11 2 3 1 4 4 53 18 0 4 15 2 1 0 2 1 4 24 6 1 2 5 14 18 3 4 4 4 11 6 5 3 3 5 2 3 13 6 395
From Daily Messages 0 2 5 3 37 19 17 10 13 2 30 3 26 10 7 7 15 1 2 5 0 4 1 0 1 0 0 0 36 27 16 31 12 0 0 2 0 11 76 11 4 0 10 2 1 0 1 0 1 37 4 2 1 14 14 25 2 4 2 1 22 10 14 4 6 1 4 1 2 3 640
Self 0 3 18 7 21 10 9 3 6 4 8 2 5 1 3 1 4 6 8 12 6 7 1 1 0 0 0 4 7 5 5 6 6 1 2 1 0 3 60 22 15 10 0 3 2 0 2 0 3 35 9 3 1 13 15 22 7 10 6 2 15 7 6 1 5 3 2 1 7 3 477
Negative Coping 3 2 32 23 7 2 5 0 3 0 3 0 7 1 2 4 1 1 6 12 10 11 5 3 1 2 0 1 2 1 0 2 5 1 3 1 0 2 9 4 2 2 3 0 50 5 38 1 8 45 13 8 0 11 29 28 5 7 5 1 31 19 11 10 8 13 8 6 5 12 561
Avoidance 2 2 21 15 4 1 3 0 1 0 2 0 3 0 0 2 1 1 3 7 8 6 2 2 0 1 0 1 1 1 0 1 4 1 2 1 0 1 6 3 1 1 2 50 0 2 12 1 5 27 6 5 0 7 18 19 2 3 2 0 14 9 3 3 1 5 5 6 2 7 327
Denial 0 0 0 2 1 0 0 0 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 2 0 1 0 1 1 0 0 0 0 1 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 19
Resistant 1 2 18 9 2 0 2 0 0 0 1 0 4 1 2 2 0 0 5 7 8 10 3 1 1 2 0 0 0 0 0 0 1 0 1 0 0 1 4 1 2 1 2 38 12 1 0 1 4 24 9 4 0 5 14 13 3 4 3 1 24 13 10 8 8 11 6 2 4 7 323
New/recent diagnosis 0 1 2 3 1 0 0 1 0 0 0 0 1 0 0 1 1 0 1 0 2 1 1 1 0 1 0 0 0 0 0 0 0 0 0 0 0 0 1 0 1 0 0 1 1 0 1 0 1 2 1 0 0 0 2 1 0 0 0 0 1 1 1 0 0 0 1 4 0 0 38
Physical symptoms 1 10 37 6 6 1 4 3 1 0 1 0 4 1 2 2 1 4 8 13 9 14 1 0 0 1 0 7 0 0 0 0 5 2 3 0 0 2 13 9 4 1 3 8 5 1 4 1 0 39 19 7 1 8 24 18 5 7 2 0 18 6 9 1 6 5 4 5 13 3 401
Self-Care 9 19 166 52 109 16 49 18 55 4 52 8 75 17 39 29 17 9 36 49 52 52 7 2 2 3 0 33 78 58 39 66 55 7 18 3 3 31 111 42 24 37 35 45 27 1 24 2 39 0 121 54 11 137 213 316 15 82 25 8 164 66 57 29 68 36 32 18 27 50 ###
Accessing care 4 6 79 6 15 3 7 3 5 1 7 2 3 1 0 1 2 6 28 15 38 32 1 1 0 1 0 25 4 3 1 4 6 2 4 0 1 1 30 15 6 4 9 13 6 0 9 1 19 121 0 21 2 23 76 38 6 18 2 0 45 10 13 3 15 17 6 1 10 8 867
Checking sugar 1 2 27 6 11 3 8 2 4 0 3 1 7 2 2 3 2 1 11 7 8 7 1 0 1 0 0 4 5 3 1 5 4 0 2 0 0 3 12 7 1 2 3 8 5 0 4 0 7 54 21 0 0 9 25 18 2 8 0 0 20 7 8 2 9 5 3 1 3 7 398
Checking their feet 0 0 2 0 2 1 1 1 2 0 2 0 3 2 1 2 0 1 1 2 1 0 0 0 0 0 0 1 3 2 0 3 3 0 1 0 0 2 4 1 2 1 1 0 0 0 0 0 1 11 2 0 0 1 2 7 1 1 1 1 3 2 0 1 1 1 0 0 1 0 85
Exercise 3 6 34 10 37 7 17 6 23 0 21 1 20 3 13 7 5 3 9 17 9 15 2 0 0 1 0 4 32 31 23 28 16 4 5 0 2 9 33 9 5 14 13 11 7 0 5 0 8 137 23 9 1 0 54 92 3 29 10 3 41 14 15 7 22 4 11 4 6 14 ###
Medication 6 10 89 19 56 9 25 7 35 0 33 4 17 4 9 4 7 4 24 22 35 34 2 1 0 2 0 20 38 23 16 38 13 3 3 1 1 7 53 26 14 14 15 29 18 1 14 2 24 213 76 25 2 54 0 103 6 30 7 4 72 29 30 11 22 19 19 4 11 18 ###
Nutrition 8 11 90 38 64 8 31 14 33 4 28 3 65 12 35 28 14 5 12 34 18 20 5 1 1 2 0 14 50 43 30 39 47 7 13 3 3 28 74 25 18 25 22 28 19 1 13 1 18 316 38 18 7 92 103 0 11 65 24 7 120 53 42 26 52 25 25 14 15 38 ###
Self-efficacy/Fatalism 0 4 10 4 10 2 7 3 1 6 2 1 2 0 1 2 0 0 3 8 5 8 0 0 0 0 0 4 1 1 0 0 1 0 1 0 0 0 10 4 3 2 7 5 2 1 3 0 5 15 6 2 1 3 6 11 0 7 1 0 8 2 2 1 4 4 1 0 3 0 207
Strategy 3 5 25 6 17 3 9 3 9 4 9 2 25 1 23 8 4 2 4 7 6 11 0 0 0 0 0 4 7 7 4 4 9 1 5 1 0 6 22 8 4 4 10 7 3 0 4 0 7 82 18 8 1 29 30 65 7 0 8 2 41 12 15 7 31 7 6 1 5 10 689
Supporter Health 0 0 6 2 4 0 2 0 1 0 1 1 16 4 8 7 5 2 1 7 1 3 1 1 0 0 0 0 7 7 4 6 9 1 3 0 1 6 13 6 4 2 6 5 2 0 3 0 2 25 2 0 1 10 7 24 1 8 0 7 21 10 14 6 6 4 4 0 5 7 322
Supporter’s own DM 0 0 1 2 0 0 0 0 0 0 0 0 4 0 4 1 1 1 1 5 0 0 0 0 0 0 0 0 1 1 0 0 3 0 2 0 0 1 6 2 4 1 2 1 0 0 1 0 0 8 0 0 1 3 4 7 0 2 7 0 8 4 3 4 0 3 2 1 2 4 108
Types of Support 7 7 70 22 27 3 9 3 2 0 14 3 84 28 46 23 19 3 19 21 26 29 0 0 0 0 0 16 30 19 12 23 26 2 9 3 1 15 61 28 11 22 15 31 14 0 24 1 18 164 45 20 3 41 72 120 8 41 21 8 0 94 128 39 88 54 73 10 16 27 ###
Appraisal support 3 4 22 12 11 0 4 1 2 0 7 0 36 13 16 9 13 1 4 11 4 8 0 0 0 0 0 4 15 12 6 12 15 1 7 1 0 9 29 13 6 10 7 19 9 0 13 1 6 66 10 7 2 14 29 53 2 12 10 4 94 0 35 26 21 20 13 4 4 12 816
Emotional 1 4 21 9 12 2 6 2 0 0 6 1 39 16 19 11 12 1 4 6 3 13 0 0 0 0 0 2 17 11 6 14 9 0 3 0 0 7 30 14 5 14 6 11 3 0 10 1 9 57 13 8 0 15 30 42 2 15 14 3 128 35 0 9 22 9 31 3 7 15 818
Informational 2 1 8 8 4 1 0 0 0 0 3 0 15 7 7 4 2 0 2 8 1 1 0 0 0 0 0 1 5 4 3 4 8 1 4 2 0 3 11 3 3 4 1 10 3 0 8 0 1 29 3 2 1 7 11 26 1 7 6 4 39 26 9 0 11 10 3 1 4 6 361
Instrumental 4 1 23 7 5 0 2 0 0 0 3 0 37 5 31 9 6 0 8 8 10 10 0 0 0 0 0 4 6 3 3 4 6 1 1 1 0 3 20 9 3 6 5 8 1 0 8 0 6 68 15 9 1 22 22 52 4 31 6 0 88 21 22 11 0 10 19 0 6 7 683
Non supportive 3 3 26 4 6 0 1 1 0 0 2 2 7 0 6 2 2 0 9 10 12 9 0 0 0 0 0 8 3 2 1 3 4 0 2 0 1 1 11 5 5 1 3 13 5 0 11 0 5 36 17 5 1 4 19 25 4 7 4 3 54 20 9 10 10 0 5 4 5 6 438
Physical presence 2 1 15 5 7 0 3 0 0 0 4 1 15 4 8 3 5 1 1 2 6 8 0 0 0 0 0 4 5 3 2 4 4 0 3 0 0 2 11 3 2 4 2 8 5 0 6 1 4 32 6 3 0 11 19 25 1 6 4 2 73 13 31 3 19 5 0 2 5 9 433
Unaware 0 2 6 34 5 0 3 3 0 0 0 1 7 0 2 6 2 0 0 9 0 2 0 0 0 0 0 2 0 0 0 0 8 0 4 0 1 4 4 1 3 1 1 6 6 0 2 4 5 18 1 1 0 4 4 14 0 1 0 1 10 4 3 1 0 4 2 0 1 2 206
Witnessing DM complications 1 4 13 9 3 1 2 0 0 0 2 0 7 1 4 3 1 2 4 23 5 10 1 1 0 0 0 3 3 1 3 2 4 1 1 0 2 2 18 10 13 2 7 5 2 0 4 0 13 27 10 3 1 6 11 15 3 5 5 2 16 4 7 4 6 5 5 1 0 11 340
Witnessing Pt DM management 1 4 21 10 6 0 4 0 4 0 0 0 10 3 6 3 2 2 2 13 5 5 1 1 0 0 0 3 4 4 3 3 8 2 3 0 1 4 16 4 6 3 3 12 7 0 7 0 3 50 8 7 0 14 18 38 0 10 7 4 27 12 15 6 7 6 9 2 11 0 450
Totals 91 238 ### 550 ### 286 611 278 464 55 763 124 995 301 465 377 270 91 382 673 474 584 135 57 59 35 5 284 ### 780 536 840 738 67 200 62 49 487 ### 551 395 640 477 561 327 19 323 38 401 ### 867 398 85 ### ### ### 207 689 322 108 ### 816 818 361 683 438 433 206 340 450 0
86
Figure 7: Code Cloud: Larger text indicates more frequently applied code. Different colors are
for legibility.

Through the analysis of the transcripts, we found that patients and supporters found the
program positive and motivating, inspiring real-time conversations and increasing their joint
attention to diabetes management. Several major themes emerged as important to the
87
experience of patients and supporters with managing diabetes and the TExT-MED+FANS
intervention: 1) decreases in maladaptive coping strategies, 2) intrinsic and extrinsic sources of
motivation and perceived capacity to make healthy choices 3) persistent barriers to disease
management due to social context and 4) patient’s selection of supporter driven by supporter
needs as well as patient’s needs.

Positive perceptions of the text-messages in TExT-MED+FANS
Supporters and patients noted that the text-messages called attention to their loved one’s
diabetes, and also provided the initiative and content areas to begin conversations about
health. Through the unprompted reminders, patients and supporters were more conscious of
the need to constantly manage diabetes. Supporters felt that the messages were of sufficient
variety in content and timing to shake the patient and the supporter out of inattentiveness to
diabetes management. For example, supporters noted:

“There’s times when one… forgets about the disease. And days pass by and nothing, but
when the texts came that said ‘Yeah I’ll check’ ‘Yes I will remind her. Yes’ in that form
well, it is good help.”
144S, Supporter, age 56, male, Spanish language preference

“It changed our relationship in that it caused us both to be more aware. Every message
we received was different, but it was something good for him.”
88
236S, Supporter, age 18, female, English language preference

Patients and supporters noted that the synchronized content in the text-messages was key to
instigating conversations about health, initially superficial check-ins, but also transitioning to
deeper discussions about health, social connections and behavior change.

One noted the real time check-ins that the supporter messages prompted:
“When I would receive the messages, I would tell her, ‘look at the what I got.’ ”
143S, Supporter, age 30, female, Spanish language preference

A patient reported:
“She would constantly send me a text, even though I would get the same text message
from you guys. She would text me afterwards and be like ‘Hey, did you this today. Did
you do that? Are you going to do this today?’ So, you know, that prompted us to text
more and more, through the texting came more conversations and then from
conversations, it became more like ‘Hey mom, this is what I did today.’ So it became me
opening up to her, as oppose to her having to come to me and check up on me.”
282P, Patient, age 31, male, English language preference

While his supporter described a similar transition from “checking in” after a message to truly
discussing health:
89
“We’re conversing more than we have in the past with regards to his health and I think
with the study it has to do with it, because the messages. And then I would always get
text messages and I would let him know I got a text message ‘Was it the same as
yours?” and I would explain it if he did or what mines said to help him out. So, he is
starting to see the importance.”
282S, Supporter, age 54, female, English language preference

Decrease in maladaptive coping.
When patients and supporters describe diabetes self-management both before and after the
TExT-MED+FANS intervention, there is common theme of negative or maladaptive coping
mechanisms preventing positive behavior choices and action, which include: 1) inattention to
disease, 2) avoidance and 3) resistance to advice. Patients perceived that these maladaptive
strategies may persist after the program, but were also identified as decreased with the
intervention, with adaptive coping mechanisms taking the place of maladaptive strategies.

When patients reflected back to before the program, they identified avoidance, lack of action
and fatalism as their predominant mechanism for dealing with the distress of diabetes. The
changes in fatalism and avoidance of diabetes distress are part of the process of patients
choosing healthy coping strategies, including good self-care.

90
“[with diagnosis of diabetes] I thought that my life had changed; that I had nothing else
to do in this world.”
130P, Patient, Spanish Language Preference, Male, age 45

“I didn’t do anything. I would just think that I was sick, and that I was sick, -and that I
was sick.”
159P, Patient, Spanish Language Preference, Female, age 61

Patients and supporters identified resistance to informational and appraisal support as negative
coping mechanisms to the threat of diabetes diagnoses and complications, especially prior to
the program. Through participation in TExT-MED+FANS, patients and supporters believed
patients became more open to advice from loved ones and health professionals, and some
supporters were selected because patients knew they would be able to persist against that
resistance.
“Um, yes, I did know he was diabetic because I am an ICU at county, an ICU nurse at
county and he had all the symptoms of diabetes, but he was stubborn as any fifty-year-
old would be and he never listened to me until he really felt it and his sugar was over 600
when he checked it with my mom’s machine.”
189S, Supporter, English Language Preference, Female, age 41

[Interviewer] “ Thinking back before participating in the study TExT-MED+FANS, and now
after the program, how have you changed the way you take care of yourself?“
91
[Participant] “Quite much, before well I wouldn’t even listen. (nervous laugh)”
183P, Patient, Spanish Language Preference, Male, age 47

[Interviewer] “That is why you chose him?”
[Participant] “Yes, because I know I’m stubborn and I don’t want to listen and I know
that he’s going to get, like, ‘Leave that to others because I don’t want you to eat it, but
you know that you’re going to relapse.’ That’s why.”
246P, Patient, Spanish Language Preference, Female, age 53

“After [TExT-MED+FANS] talking to him, he would tell me what he would feel the
frustration of tingling on his hands, his feet, his temper level getting high because his
sugar levels were not control. After that I kinda understood him a little more. At the end
of the day all you can do talk to them. You can’t even give them their medicine without
them wanting it or anything but the support that I did give him is just letting him know
‘you know what, it is not gonna change if you keep your same eating habits. Nothing is
going to get better if you are not putting your part.’ ”
185S, Supporter, English Language Preference, Female, age 45

Participants (patients more than supporters) also identified ways that that participating in TExT-
MED+FANS changed the attention or conscious thought placed into disease self-management,
reducing the negative coping strategies of avoidance and denial. Patients have noted an
increased “consciousness” after the program, and that they made more intentional health
92
behavior choices.
Patients noted:

“It makes me aware of where I am at and what I am doing, and it is kind of a conscious
thing.”
103P, Patient, English Preference, Male, age 49

“Well, I wasn’t conscious, [now] I’m a little more conscious of my diabetes, of the
medications that I need to take on time.”
113P, Patient, Spanish Preference, Male, age 38

Extrinsic and Intrinsic Motivations & Perception of Capacity
Patients and supported identified multiple sources of intrinsic and extrinsic motivation and
described changes in the balance of intrinsic and extrinsic motivation throughout the study.
Extrinsic motivation described was generally a fear of consequences that they had witnessed in
family members or close friends, and the direct encouragement and praise they received from
family members. Intrinsic motivations revolved around living up to the identities patients held
for themselves as well as valuing their own health and feeling stronger and healthier when their
diabetes was managed. Text-messages were perceived as either building perceived capacity to
make healthy choices, or as prompting patients and supporters to make that choice in the
moment, which also built their perception of capacity. Extrinsic motivators important to the
participants were the encouragement from family members and subsequent approval when
93
healthy behavior choices were accomplished, and the fear of experiencing the consequences
witnessed in family members who had long-standing poor control of diabetes.

Extrinsic Motivations
Fear of consequences was motivating to patients as they had witnessed loved ones lose limb,
sight, be started on dialysis and many other complications of long-standing uncontrolled
diabetes. The physical toll of the disease was frightening, and the threat of having the same fate
was a substantial motivator for many patients, and for supporters who did not want to lose
another loved one to diabetes.

“And that is what I say, "I don't want to get to that point." Eh, it motivates my health,
my life. I don't want to… inject insulin. It's my own health. My wife, I tell her, "We are
going ...no matter what, we fight." And sometimes, like, the...treatment that you guys
offered me, all the reminders, the dates, to put, like, challenges, which is what I would
like to continue to do again.”
Participant 302, Male, Patient, Spanish Language Preference, age 45

“I tell her -that how I don’t want you to get sick, I don’t want that one day they cut your
leg and that the wound won’t heal. I don’t want you to lose your vision. I don’t want to
have you here in the bed sick because of diabetes. That you can’t walk I tell her. You
have to get it together.”
Participant 129, Male, Supporter, Spanish Language Preference, age 55
94

“My aunt passed a few years ago, um, but my grandma was able to reverse it, so she
doesn’t have [diabetes] anymore.”
[Interviewer] Did she die because of complications?
“My aunt, uhh, yeah pretty much. She had, I think more, it became a heart related um
but yeah one thing led to another. She had amputations, she neuropathy, um, her feet
swell because of the heart, so I believe diabetes is pretty much what sealed the deal on
that one. I… kinda saw the same pattern happening and I personally use them as like
the, for the lack of a better word, as my motivation to not become that. So, it literally
scared me because I heard about it growing up and to actually see it in person was one
thing and to hear about it was different so hearing about that and seeing it actually from
fruition, from getting in was pretty frightening for me as a diabetic.”
Participant 282, Male, Patient, English Language Preference, age 31

Family members and loved ones were frequently mentioned sources of extrinsic motivation,
providing both encouragements to make healthy choices and also providing positive feedback
and approval when patients succeeded. The TExT-MED+FANS program was noted to increase
this extrinsic source of motivation, as some supporters were more highly engaged in offering
emotional and instrumental support during the program.

“My family. My children. My husband. Everyone motivated me to continue so that I
would not have problems later on.”
95
Participant 129, Patient, Female, Spanish Language Preference, age 60

“She would always send me texts and after getting out of work checking on me to see
how I was doing. Because she did take it to heart being my mentor.”
Participant 178, Patient, Female, Spanish Language Preference, age 47

Transition of Extrinsic to Intrinsic Motivation
Some patients identified the TExT-MED+FANS program and the behavior challenge messages as
prompts to help inspire them to make a healthy choice for themselves, and to begin to turn the
motivation from an extrinsic fear of consequences to an intrinsic motivation to show
themselves they can make healthy decisions and be “strong” or “normal” despite having
diabetes.

[Interviewer]: “What text messages made you reflect or think about your health?”
[Participant]: “More than anything the consequences that one has. My mom, suffered from
diabetes for many years and had the same symptoms that I have from the cataracts. [She] was
losing her vision, the movement of her legs and from the same she became depressed, she
became depressed in herself, and I always tried to cheer her up. That was what made me think a
lot, regulating the sugar and what I’m doing now.”
[Interviewer]: “So, it was more about what you were living (seeing) than the messages?”
[Participant]: “Yes, no. But also, in the form of motivation because, like, like, how do you call it,
day-to-day motivation that you should do, that should be, including what one should eat. The
96
challenges! [They] challenge you to do this, and it helped me look at the text and I say, I’m going
to do it. And they’re challenging that one can do. If someone proposes to do it, you can do it.”
Participant 292, Patient, Male, Spanish Language Preference, age 39

[Interviewer]: “In general, what motivates you to take control of your diabetes?”
[Participant]: “I don’t’ want to be sick. I just want to be a normal person and you can be
normal and still have diabetes. I know that now [after TExT-MED+FANS].”
Participant 237, Patient, Female, English Language Preference, age 49

[Interviewer]: “Before the program, a year ago, what did you think would happen if you
didn’t manage your diabetes?”
[Participant]: “I was almost certain, not that I’d be dead now, but that I would definitely
be close to what my aunt and uncle were at, and they were pretty bad. So, to hear that
my A1C is down, to see that my cholesterol is back down to normal or is, has been
normal, is definitely given me more of that definite, like, sign of victory, so my confidence
has gone up significantly since, you know, since a year ago.” [before starting TExT-
MED+FANS]
Participant 282, Patient, Male, English Language Preference

Intrinsic Motivations
Intrinsic motivation sources were most commonly the desire to fulfill their perceptions of being
a good family member, and most saliently the realization that the desire to improve their own
97
health with better diabetes management generates a more profound change. The desire and
motivation to make changes to diabetes management were often driven by patients desires to
live up to responsibilities in their families or the expectations of their family members.
Generally perceived as a very positive force in their lives, patients believed their changes in
diabetes self-care would results in happier and longer time together.

“Well one day my doctor told me, “Do you want to see your children grow? You need to
take care of yourself because diabetes can –can finish with your life if you do not control
it.” And then, it was a wakeup call for me, and then I started to -to try to improve my
diabetes for my children.”
Participant 178, Patient, Female, Spanish Language Preference, age 47

“And in general, what motivates you to take control of your diabetes?
In general, well my family and also the family that I yet don’t have, that’s to say family I
don’t have yet like grandchildren. And I want to be there when they come along.”
Participant 298, Patient, Male, Spanish Language Preference, age 49

While participants often noted that succeeding in their family roles was important to them,
they ultimately desired improvements in their own health, or identified the need to make one’s
own health decisions as the mechanism of lasting behavior change.

“Eh, my kids have always motivated me and if one takes care of oneself … you feel
98
better too. But honestly the program is really good because it reminds you of your own
care.”
Participant 109, Patient, Female, Spanish Language Preference, age 57

“My [motivation to change comes from] family. Uh huh. And myself because I don’t want
to get a leg amputated or end up with no eyesight or be receiving dialysis.”
Participant 125, Patient, Male, Spanish Language Preference, age 46

Social Context
Another important theme in patient’s perceptions of the program is the social context in which
they try to manage their diabetes. These ED patients and their family members are managing
financial difficulties, and housing stress including homelessness, and barriers to accessing
insurance and medical care. While the TExT-MED+FANS program provided a significant benefit
to patients with diabetes during the course of the study, several limitations highlighted the
barriers to basic medical care experienced by this specific study population. Many patients
reported significant efforts made just to make the study appointments (ex. Taking several buses
for hours and walking long distances for a single trip). Additionally, the difficulty with which to
find and see a PMD consistently was highly cited among our study participants. Lastly, the
financial burden of diabetic medication and equipment posed an enormous stressor to many of
our study participants who often had to decide between paying for their medications or basic
99
living needs like groceries and rent. Supporters and patients noted these stresses impacted
their care, preventing patients from making the lasting behavior changes needed to improve
their glycemic control. While the TExT-MED+FANS intervention instigated motivation and
encouragement, it did not change the patients’ perceived or actual capacity to overcome the
persistent barriers related to their financial and social constraints. For example:

“My brother [name removed], has been homeless, and he was homeless when enrolled
in the program. And right now, he’s at a shelter, so that’s improving, but because he’s
homeless, and because even at the shelter, it’s very hard for him to take care of his food,
right?”
Participant 114, Supporter, English Language Preference, Female, age 64

“It’s good, the program is good, it helps us, but like I mentioned earlier some of us don’t have a
primary care doctor…hopefully I can get a primary care doctor right now so that my diabetes
can be better. “
Participant 104, Patient, Spanish Language Preference, Female, age 34

From interviewer notes on small talk prior to recorded interview:
Patient does not have primary care provider and has not been taking her insulin for more than 5
months. She noted that she was tired of having a new doctor every 3 months.
Participant 293, Patient, Spanish Language Preference, Male, age 43

100
“I got sick and then she [wife] got sick too and we were both sick, well there’s just one more
thing on top of it all.” (referring to losing home when both could not work per RA notes; now
lives in family member’s living room with their 2 dogs)
Participant 112, Patient, Spanish Language Preference, Male, age 48

While they have the desire to care for themselves or their loved ones in the best way possible,
the interviewees perceived social stresses to greatly impact how well patients can make these
choices. Interviewees identified nutrition challenges from homelessness and lack of regular
access to medical care. Additionally, disability from diabetes creates additional stress when a
patient is too ill to work and does not have the security of state disability in the informal jobs
that they work.

Selection of Supporter
Patients selected supporters because they felt the identified supporter could offer
informational support (especially if the supporter also had diabetes), emotional
support/encouragement or instrumental support (i.e., assistance with food preparation,
transportation). Each patient was specifically asked why they selected their support person.
Patients expressed different reasons for selecting a supporter, although all received the same
instructions; to pick the person that they thought would be the most supportive in managing
diabetes. When selecting a supporter for the duration of the program, many patients preferred
choosing a close relative or friend who already had some experience with diabetes. Many cited
a personal lack of knowledge or confidence as a significant barrier to taking on their own
101
diabetes care. Choosing a supporter who has “already been through it” offered an established
avenue of guidance and experience and provided them with an encouraging success story.
Many patients also explained that they highly valued emotional strength when choosing their
supporter. Some supporters were chosen if they have diabetes, could either benefit from the
program themselves, or that the patient and supporter would provide mutual support. Patients
explained their reasoning for their supporter selection:
“He is a man that also has diabetes and it helped him a little too and I think that being in
the same problem both of us.”
Participant 109, Patient, Female, Spanish Language Preference, age 57

“I choose him because he has the same illness, and we support each other. We say to
each other “How are you? How’s your sugar levels? What is it that you’re doing?"
Participant 132, Patient, Female, Spanish Language Preference, age 46

[Interviewer]: Why did you pick your daughter XXX, to participate with you in the
program? Why her?
[Participant]: Why? Because it's possible because she's very grouchy and she's young and
-and this helps her see you (diabetes team) the way you are, share with her what you
guys know with her so she can learn. And like she is very grouchy, I think later she will
also deal with diabetes herself, even though I wouldn't want that for her, right. But her
dad is diabetic, so am I, so what can she expect. Therefore, this will reminder her to
102
always take her medicine, she has to take care of herself. Right? The food she is going to
eat for it to be nutritious. The good thing is that she's not a big eater, that helps.

Participant 159, Patient, Female, Spanish Language Preference, age 61

Female supporters were often selected for the support they provided the patient even prior to
the FANS intervention, either emotional or instrumental support. Specifically, male patients
often selected wives or life-partners as supporters for the study. These patients identified their
wife or life-partner as the person in their life that actually drives many of the self-care
behaviors:

“[I selected her because] she is a stronger person; she believes in herself.”
Participant 103, Male, Patient, Spanish Language Preference, age 49

“Because she's the woman who’s with me in the house and she is the person who helps
me with everything. I didn’t choose her; it's how we live.”
Participant 152, Male, Patient, Spanish Language Preference, age 37

“Since the first time that she (patient’s wife) came, she was taught to make food a little
more healthier.”
Participant 125, Male, Patient, Spanish Language Preference, age 46

103
Differences in Codes and Themes for Low and High Responders to TExT-MED+FANS
In examining the distribution of codes and themes between the participants who had a change
of A1C of 1 or greater (High Responders) compared to those who had minimal or no change
(Low Responders), there was not a noticeable difference in distribution of codes between the
two response groups.
104
Figure 8: Normalized Distribution of Codes for High and Low Responders to TExT-MED+FANS at
6 and 12 months
Normalized Distribution of Codes for High and Low Responders to TExT-
MED+FANS at 6 months
Normalized Distribution of Codes for High and Low Responders to TExT-
MED+FANS at 12 months
0 100 200 300 400 500 600
Alternative
Balance
Barriers/Challenges
Before Code
Benefits to pt specific
Changes in motivation
Increased attention to Dz
Increased knowledge
Self-care behaviors
Self-efficacy
Support from messages
Support from staff
Benefits to pt via supporter
Increase communication
Supporter behaviors
Supporter knowledge
Supporter motivation
Causes of disease
Comorbidities
Consequences
Difficult Social situation
Emotional
Feedback on program
Expand
More Contact
Social work
access to care
Healthcare experience
Impactful msgs
Advice/tips
Challenges
Reminders
Info sources
Family
Health Care Personnel
Internet
Social
Texts
Motivation
Family
Fear of consequences
From Daily Messages
Self
Negative Coping
Avoidance
Denial
Resistant
New/recent diagnos is
Physical symptoms
Self-Care
Accessing care
Checking sugar
Checking their fe et
Exercise
Medication
Nutrition
Self-efficacy/Fatalism
Strategy
Supporter Health
Supporter’s own DM
Types of Support
Appraisal support
Emotional
Informational
Instrumental
Non supportive
Physical presence
Unaware
Witnessing DM complications
Witnessing Pt DM management
Change
No change
0 100 200 300 400 500
Alternative
Balance
Barriers/Challenges
Before Code
Benefits to pt specific
Changes in motivation
Increased attention to Dz
Increased knowledge
Self-care behaviors
Self-efficacy
Support from messages
Support from staff
Benefits to pt via supporter
Increase communication
Supporter behaviors
Supporter knowledge
Supporter motivation
Causes of disease
Comorbidities
Consequences
Difficult Social situation
Emotional
Feedback on program
Expand
More Contact
Social work
access to care
Healthcare experience
Impactful msgs
Advice/tips
Challenges
Reminders
Info sources
Family
Health Care Personnel
Internet
Social
Texts
Motivation
Family
Fear of consequences
From Daily Messages
Self
Negative Coping
Avoidance
Denial
Resistant
New/recent diagnos is
Physical symptoms
Self-Care
Accessing care
Checking sugar
Checking their fe et
Exercise
Medication
Nutrition
Self-efficacy/Fatalism
Strategy
Supporter Health
Supporter’s own DM
Types of Support
Appraisal support
Emotional
Informational
Instrumental
Non supportive
Physical presence
Unaware
Witnessing DM complications
Witnessing Pt DM management
Change
No change
105

Discussion

In this mixed method study of predominantly Latino adult patients and their family members who
completed an mHealth diabetes self-management intervention with mHealth activated social
support, we identified themes of positive components of the intervention, barriers to changes in
diabetes self-care, facilitators of positive change, and factors in supporter selection in the TExT-
MED+FANS intervention. These include 1) decreases in maladaptive coping strategies, 2) changes
in intrinsic and extrinsic sources of motivation 3) social barriers to disease management and 4)
supporter selection by patient driven by supporter needs as well as patient’s needs. We did not
identify differences in among high and low responders to the intervention.

In the current study, patients identified increased “consciousness” of diabetes and motivations to
behavior change as mechanisms of successful change in diabetes self-care. Increasing awareness
of the disease and the daily reminder about self-care decisions changed patients' intrinsic
motivations and emphasized extrinsic motivations such as family expectations. Cues to action, as
conceived in the Health Belief Model, have traditionally been thought to increase a person’s
perceived risk of a disease or complication. (Janz & Becker, 1984, 1985) As the patients in this
study are all people with poorly controlled diabetes, their perceived risk of diabetic complications
may have been optimistically low prior to the intervention, due to coping strategies of avoidance
and denial. The daily messages may serve as novel cues to action that overcome these
maladaptive coping strategies. SMS-text message interventions have also been conceptualized as
a moderator of the relationship between motivation and behavior change by lower “activation
energy” to make a new and uncomfortable health decision. (Burner et al., 2014; Fogg, 2009) By
increasing perceived threat of diabetic complications, the TExT-MED+FANS intervention may
106
increase intrinsic motivation for change. In addition to motivation for themselves, patients in this
study also reported being more aware of how poor diabetes control would impact their families.
The needs and expectations of their family were a preeminent extrinsic motivation for behavior
change. While patients all report increases in motivation, the reasons that some patients
internalize that desire for change and others continue to rely on external motivators such as fear
of consequences and family approval are not fully defined. Future interventions may benefit from
targeted programming to emphasize intrinsic motivation over extrinsic motivation.

Many diabetes experts and clinical diabetes educators encourage support from family members to
improve self-care behaviors, acknowledging the powerful influence of loved ones over
patients.(Denham, Ware, Raffle, & Leach, 2011; Fisher et al., 1998; Weiler & Crist, 2009)
Traditionally, familial and social support has been viewed as positive, with associations between
increased social support and improved self-care behaviors and glycemic control.(Aikens, Trivedi, et
al., 2015; Hansen, Jones, Zander, & Willaing, 2015; Khosravizade Tabasi, Madarshahian, Khoshniat
Nikoo, Hassanabadi, & Mahmoudirad, 2014) In addition to encouraging good self-care, strong
relationships with family have been found to improve psychological and emotional well-being of
patients, which independently are associated with improved self-care and glycemic control.(R. J.
Anderson, Freedland, Clouse, & Lustman, 2001; Scollan-Koliopoulos, Schechter, Caban, & Walker,
2012) In this study, patients shared that they selected a supporter who would either provide
emotional or instrumental support to them or who could provide peer support and potentially
benefit from the program in addition to the patient.

107
In this supporter selection process, we see a potential gender pattern in supporter selection, with
women selecting supporters for emotional support or to benefit the supporter, while men
identified supporters based on instrumental support. Prior research into diabetes behavioral
intervention for Latino immigrants have found similar patterns, with women identifying more
social and emotional barriers to motivation and selfcare, while men identified structural barriers
that require a concerted effort to overcome. (A. Cherrington, Ayala, Scarinci, & Corbie-Smith,
2011) Supporters selected were overwhelming females; this may be due to gendered social
expectations regarding health and attending to the health of family members. Latina women are
more likely to informally share diabetes knowledge and have higher baseline disease knowledge
than men,(Daniulaityte, 2004) and may be seen as more effective supporters for this reason.
Gender differences in the type of social support desired by a patient and in the type of support
they expect to receive from a supporter of a particular gender make sense in this context. Men
and women perceive different support needs in themselves and are also perceived to have
different strengths in offering social support. Additionally, women and men may respond
differentially to mobile interventions to increase social support.(Igarashi, Takai, & Yoshida, 2005)
Intervention designers may need to consider that activating the family of a female patient may
require different strategies than the family of a male patient.

Social support is likely critical in the self-management of these patients, given the high level of
social stress they face daily. The patients and supporters interviewed in this study identified social
stressors as significant barriers to managing their diabetes, specifically housing and access to
regular medical care. Achieving diabetes control requires intense daily efforts at self-care to
108
maintain a healthy diet, achieve a high level of physical activity, and follow a complicated
medication regimen. Alone, diabetes self-care can be an overwhelming task, even without the
added burden of socioeconomic stressors. Patients who face additional challenges of securing
sufficient food and crowded housing are at higher risk of poor disease outcomes; people who live
in economically distressed neighborhoods are more likely to have severe complications from
diabetes, with higher rates of renal failure, dialysis, and amputations after osteomyelitis. (Moghani
Lankarani & Assari, 2017; Stevens et al., 2014; Venermo et al., 2013; Vijayan et al., 2016) These
social stressors can play a role in nutrition choices, particularly for those with food insecurity or
lack of adequate housing to prepare nutritious foods.

This study has several limitations. We do not have an entirely comprehensive view of the
experience of patients and supporters in the intervention. Patients and supporters were recruited
for interviews after completing a 6-month educational intervention and 6-month post intervention
phase study that was delivered via mobile text messages to encourage and challenge patients to
make positive behavior changes. We did not interview the active controls from this trial, and those
who returned for interviews after the completion of the trial may represent a biased sample of the
initial study population. Patients and supporters who come to the interviews are likely those who
are most engaged in the intervention. The overall positive assessment of the intervention must be
understood in that context. Also, these interviews were conducted with an inductive aim and
purposive sampling; the findings are not generalizable in isolation and require an understanding of
context. This approach, however, allowed us to explore the complicated relationship between
social support and diabetes self-care. Additionally, there is a concern for a biased representation
109
of the user experience; in contrast to our prior studies with focus groups, patients have been
reluctant to open up about difficulties and have required significantly more prompting. We found
that the detailed interviewer notes often contained a more negative self-assessment by patients,
as well as barriers to disease management that the patients did not wish to bring up while being
recorded. This may be related to the power dynamics involved (excessively educated PI and RAs
conducting interviews in the hospital) which has been suggested in the past to benefit focus
groups over interviews (Magill, 1993). The perceptions of the programs shared by patients and
supporters may be overly complimentary given this potential power differential between the
research staff and interviewees. We did not conduct member checking at the end of the analysis,
which would have provided participants an opportunity to correct misperceptions or add their
own interpretation to our findings.

This qualitative analysis of a novel mHealth-based social support intervention for diabetes selfcare
is critical to understanding the underlying mechanisms of behavior change and how patients’ lives,
and social contexts influence the effects of the intervention. The interplay of self-care, motivation,
social support, and barriers to care faced by a high-need population suffering from excessive
health disparities is complex. This qualitative analysis is key to understanding this and future ICT
and mHealth based interventions, as the psychosocial factors and behaviors these interventions
are targeting interact on multiple levels and integration of quantitative and qualitative data may
clarify these interactions. By using an inductive approach to coding, we discovered areas of
intervention impact and barriers to success that were not anticipated in the initial intervention or
trial design. And lastly, this analysis directs areas of intervention improvement so that TExT-
110
MED+FANS can be further refined before it is implemented in a larger population and so similar
interventions can be designed with these considerations.

111
Appendix 1: Interview Guide: Patient, English

Introductory/Ice breaker questions: Name and….
-How many years have you had diabetes? How many years have you been coming to LAC+USC for care?

Main questions:
1. Thinking back to before Text-MED FANS, and now after the program, how have you changed the way
you take care of yourself?

a. Sub-prompts if not addressed:
i. Who did you see to help you take care of it?
ii. (prompts if necessary: Medications, doctors, alternative medicine)
1. What has been the hardest part?
2. How do you learn about what you need to do to take care of yourself?
(if necessary, prompt on diet, exercise, taking medicines)

2. Thinking back to before the program, what kinds of things inspired you to take control of your
diabetes?
a. Sub-prompts if not addressed:
i. What made it easier to manage it well?
ii. What made it more difficult to manage it well?

3. Before you participated in the program, what did you think would happen if you didn’t manage your
diabetes?
a. How about now? Reword if necessary: What did you think would happen if you did not control
your diabetes?

4. What kinds of things have helped you to change how you manage your diabetes?

a. Sub-prompts:
i. Has it worked for you?
ii. Does it actually get you to change your behavior?
iii. What else have you tried?
1. How successful have these strategies been for you?

5. How has TextMED FANS affected how you manage your diabetes?

6. -What types of messages made you think the most?
a. Prompt if not addressed:
i. About your diabetes,
ii. about how you manage your diabetes,
iii. and about your health in general

7. How did you decide on the supporter you selected to work with you in TexT-MED FANS?
112
a. Has TExT-MED FANS changed your relationship with your supporter?

8. -What was your favorite part?

9. If not addressed earlier in session -How could TExT-MED be better?

113
Appendix 2: Interview Guide, Supporter English

Ice breaker: Tell us your name and what is your relationship to the person who chose you to be their support
partner in the Test-MED FANS?

Main questions:
1. Before the program, did you know your loved one had diabetes? What did you know about diabetes?

2. [If yes to the first part of the question or only of those who responded yes to the first part of question
#1] Before TExTMED FANS, did you think about your loved one’s diabetes? How often?

3. How have your thoughts about their diabetes changed?

4 . Before the program, how did you offer support to your loved one?
a. How has this changed with TExT-MED FANS?

b. Prompt on:
i. instrumental support: offered to take them places, buy things.
ii. informational support: gave advice, helped them learn things.
iii. emotional support: someone to talk to when feeling blue.
iv. appraisal support: helped them think about situations, decide if something that
happened really was stressful.

5. Which text messages made you call/text or talk to your loved one the most?

6. Which text messages made you think the most?
a. About your loved one, about their diabetes, about their health or your own health

7. Which text messages moved you to actually do something to help your loved one? Like call and check
in on them, help them go to the store, or anything else helpful.

8. -What was your favorite part of participating in the program?

114

9. To help us improve the program – what could we change in program to make it better?

a. Was there a least favorite part for you?
b. Would you encourage others to participate in such a program if we made the changes you
suggested?

115
Appendix 3: Interview Guide, Patient, Spanish

Introducción/Romper el Hielo: Díganos cuál es su nombre y….
-¿Cuántos años ha tenido diabetes? ¿Cuántos años ha estado viniendo al hospital General para su cuidado
medico?

Main questions:
1. ¿Antes de comenzar el programa de Text-MED FANS, y ahora después del programa, como ha
cambiado la manera en que se cuidan?
a. Sub-prompts if not addressed:
i. ¿A quién veían para su cuidado médico?
ii. (prompts if necessary: Medications, doctors, alternative medicine)
1. ¿Cuál ha sido la parte más difícil?
2. ¿Cómo aprenden acerca de lo que necesitan hacer para cuidarse?
(if necessary, prompt on diet, exercise, taking medicines)

2¿Pensando en el periodo antes del programa, que tipo de cosas los inspiraban para tomar control de su
diabetes?
b. Sub-prompts if not addressed:
i. ¿Qué situaciones los hacía dirigir la diabetes bien?
ii. ¿Qué situaciones los hacia que no dirigieran tan bien la diabetes?

2. ¿Antes de participar en el programa, que pensaban pasaría si no dirigían bien su diabetes?
a. ¿Y ahora que piensan?
Reword if necessary: ¿Qué pensaban pasaría si no controlaban bien su diabetes?

3. ¿Qué tipos de cosas los han ayudado a cambiar la manera en que dirigen su diabetes?
a. Sub-prompts:
i. ¿Les han funcionado estas cosas?
ii. ¿Realmente hacen estas cosas que cambien su comportamiento?
iii. ¿Cuáles otras cosas han intentado?
1. ¿Qué tan exitosas han sido estas estrategias para ustedes?

4. ¿Cómo ha afectado el programa de TextMED FANS en la manera que dirigen su diabetes?

5. ¿Qué tipos de mensajes de texto que mando el programa hicieron que ustedes reflexionaran más?
a. Prompt if not addressed:
i. Acerca de su diabetes
ii. Acerca de la manera en que dirigen su diabetes
iii. Y acerca du su salud en general

6. ¿Cómo decidieron escoger a su compañero/a de apoyo para trabajar con ustedes en el programa de
TexT-MED FANS?
116
a. ¿Ha cambiado la relación con su compañero/a de apoyo después de participar en el programa
de TExT-MED FANS?

7. ¿Cuál fue su parte favorita del programa?

8. If not addressed earlier in session -¿Que pudiéramos cambiar en el programa para mejorarlo?
c. ¿Hay alguna parte del programa que no les gusto?
d. ¿Animarían a otras personas que participaran en un programa como este si hiciéramos los
cambios que nos sugirieron?
Appendix 4: Interview Guide, Supporter, Spanish

Introducción/Romper el Hielo: ¿Díganos su nombre y cuál es su relación a la persona que lo escogió para ser
la compañero/a de apoyo en el proyecto de Text-MED FANS?

Main questions:
1. ¿Antes que comenzara el programa, sabían ustedes que su ser querido tenia diabetes?
¿Qué sabían ustedes acerca de la diabetes?

2. [If yes to the first part of the question or only of those who responded yes to the first part of
question #1] Antes de TExTMED FANS, pensaban ustedes en la diabetes de su ser querido? ¿Qué tan
frecuente?

3. ¿Cómo han cambiado sus pensamientos acerca de la diabetes que sus seres queridos tienen?

4 . ¿Antes del programa, como les ofrecían apoyo a sus seres queridos?
a. ¿Cómo ha cambiado esto con el programa TExT-MED FANS?
b. Prompt on:
i. INSTRUMENTAL SUPPORT: ofrecían llevarlos a lugares, comprarles cosas
ii. INFORMATIONAL SUPPORT: les daban consejos, los ayudaban que aprendieran cosas
iii. EMOTIONAL SUPPORT: les brindaban consuelo cuando se sentían tristes
iv. APPRAISAL SUPPORT: les ayudaban a reflexionar sobre situaciones, deducir si algo que
paso les estaba causando estrés

3. ¿Cuáles mensajes de texto que mando el programa hicieron que ustedes se comunicaran
(llamada/texto/hablar en persona) con su ser querido más?

4. ¿Cuáles mensajes de texto que mando el programa hicieron que ustedes reflexionaran más?
a. Acera de sus seres querido, de la diabetes que ellos tienen, de la salud de ellos, o acerca de SU
propia salud

5. ¿Cuáles mensajes de texto que mando el programa los conmovieron a hacer algo para ayudar a sus
seres queridos? Cómo llamar y chequearse con ellos, ayudar a ir a la tienda, o hacer otra cosa util

6. ¿Cuál fue su parte favorita de participar en el programa?
117

7. Para ayudarnos a mejorar el programa - ¿Que pudiéramos cambiar en el programa para mejorarlo?
a. ¿Hay alguna parte del programa que no les gusto?
b. ¿Animarían a otras personas que participaran en un programa como este si hiciéramos los
cambios que nos sugirieron?

118
Appendix 5: COREQ Guidelines Checklist
COREQ (COnsolidated criteria for REporting Qualitative research) Checklist.

A checklist of items that should be included in reports of qualitative research. You must report the page number
in your manuscript where you consider each of the items listed in this checklist. If you have not included this
information, either revise your manuscript accordingly before submitting or note N/A.

Topic Item
No.
Guide Questions/Description Reported
on
Page No.
Domain 1: Research team
and reflexivity
Personal characteristics
Interviewer/facilitator 1 Which author/s conducted the interview or focus group? 10
Credentials 2 What were the researcher’s credentials? E.g., PhD, MD 10
Occupation 3 What was their occupation at the time of the study? 10
Gender 4 Was the researcher male or female? 10
Experience and training 5 What experience or training did the researcher have? 10
Relationship with
Participants
Relationship established 6 Was a relationship established prior to study
commencement?
10
Participant knowledge
of
the interviewer
7 What did the participants know about the researcher? e.g.,
personal?
goals, reasons for doing the research

10

Interviewer
characteristics
8 What characteristics were reported about the inter
viewer/facilitator?
e.g., Bias, assumptions, reasons and interests in the research
topic

10&11

Domain 2: Study design
Theoretical framework
Methodological
orientation and Theory
9 What methodological orientation was stated to underpin the
study? e.g., grounded theory, discourse analysis, ethnography,
phenomenology,
content analysis

11

Participant selection
Sampling 10 How were participants selected? e.g., purposive, convenience,
consecutive, snowball

10

119
Method of approach 11 How were participants approached? e.g., face-to-face,
telephone, mail,
email
10

Sample size 12 How many participants were in the study? 13
Non-participation 13 How many people refused to participate or dropped out?
Reasons?
13
Setting
Setting of data
collection
14 Where was the data collected? e.g., home, clinic, workplace 10
Presence of non-
participants
15 Was anyone else present besides the participants and
researchers?

10

Description of sample 16 What are the important characteristics of the sample? e.g.,
demographic?
data, date

13

Data collection
Interview guide 17 Were questions, prompts, guides provided by the authors? Was
it pilot?
tested?
Appendix 1

Repeat interviews 18 Were repeat inter views carried out? If yes, how many? N/a
Audio/visual recording 19 Did the research use audio or visual recording to collect the
data?
11
Field notes 20 Were field notes made during and/or after the interv iew or focus
group?
11
Duration 21 What was the duration of the interviews or focus group? 13
Data saturation 22 Was data saturation discussed? 14
Transcripts returned 23 Were transcripts returned to participants for comment and/or No
120
Topic Item
No.
Guide Questions/Description Reported
on
Page No.
correction?
Domain 3: analysis and
findings
Data analysis
Number of data coders 24 How many data coders coded the data? 12
Description of the
coding
tree
25 Did authors provide a description of the coding tree?
16-17

Derivation of themes 26 Were themes identified in advance or derived from the data? 12
Software 27 What software, if applicable, was used to manage the data? 12
Participant checking 28 Did participants provide feedback on the findings? No
Reporting
Quotations presented 29 Were participant quotations presented to illustrate the
themes/findings?
Was each quotation identified? e.g., participant number

20-33

Data and findings
consistent
30 Was there consistency between the data presented and the
findings?
20-33
Clarity of major themes 31 Were major themes clearly presented in the findings? 20-33
Clarity of minor themes 32 Is there a description of diverse cases or discussion of minor
themes?
20-33

Developed from: Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative
research (COREQ): a 32-item checklist for interviews and focus groups. International Journal for
Quality in Health Care. 2007. Volume 19, Number 6: pp. 349 – 357

Once you have completed this checklist, please save a copy and upload it as part of your submission.
DO NOT include this checklist as part of the main manuscript document. It must be uploaded as a
separate file.

121
Chapter 4: Study 3: Different people/Different needs: A subgroup analysis of a Randomized
Controlled Trial utilizing a Latent Profile Analysis of Psychosocial Indicators and a Mixed-
Methods Qualitative Analysis of Post Intervention Semi-Structured Interviews

Abstract
Background:
mHealth based interventions for diabetes self-management have shown moderate efficacy in
improving self-care activities and some efficacy in improving glycemic control. Better
understanding the subgroups of patients who benefit from mHealth interventions may inform
future interventions and direct the implementation of the interventions in real world settings.
Additionally, identifying the patients who would benefit from social support augmented via an
MHealth mode in addition to a patient directed curriculum will allow for directed efforts to
engage family members and supporters for patients that need the additional support. In this
study, we conduct post trial analyses of subgroups identified by latent profile analysis of
psychosocial indicators and a mixed-methods qualitative analysis of post intervention semi-
structured interviews.

Methods:
The study population includes all patients from the TExT-MED+FANS RCT (n=166), those who
completed 6 month follow up (n=100) and those who completed 12 month follow up (n=106).
We used Latent Profile Analysis of patient who initially agreed to the study to identify
subgroups of patient with different psychosocial indicators (n=173). We examined the
qualitative analysis with 50 patients and 56 supporters at the conclusion of the trial, identifying
access to primary care and patient gender & supporter relationship as important to successful
122
behavior change; we identified accessing a continuity clinic and selecting a spouse as a
supporter as quantitative variables that captured these themes.

Using subgroups identified by the latent profiles, baseline access to primary care, and the
patient gender & supporter relationship, we examined subgroup differences in demographic
characteristics, clinical measures and baseline health behaviors with ANOVA or Fischer’s exact
test as appropriate. We conducted a mixed-effects analysis of the RCT, using these subgroup
variables (Latent Profile, Access to Primary Care and Gender & Supporter Relationship) as
interaction terms to support intervention mode group (mHealth or pamphlet). We estimated
predicted changes in A1C across support intervention mode groups at 6 and 12 months after
enrollment, and also estimated predicted changes in A1C between intervention groups at 6 and
12 months after enrollment for each subgroup variable (predicted latent profiles, baseline
access to primary care, and the patient gender & supporter relationship).

Results:
4 latent profiles of psychosocial indicators were identified from the patients initially eligible for
the RCT: (1) Alone, depressed & fatalistic; (2) Supported & fatalistic, not depressed; (3)
Supported, depressed & fatalistic and (4) Highly supported, depressed, not fatalistic.

Qualitative analysis of semi-structured interviews identified 2 important social determinants of
perceptions of change in social support and diabetes self-care: access to primary care, and a
gender difference in expectations of a spouse as an identified supporter in the intervention.
123

In sub analysis by latent profile, the predicted mean A1C change from 0 to 6 months was twice
as large for patients in the “Alone, depressed and fatalistic” Profile #1 compared to the “Most
support, most depressed, not fatalistic” Profile #4, -1.6%mg/dL (95%CI -2.4, -0.8) vs -0.8 (95%CI
-2.1, 0.4) p=0.304. There was similar efficacy with similar between intervention mode group
A1C changes at 0 to 6 months. In maintenance of the intervention effects, the highly supported
but most depressed Profile #4 showed the largest difference between support modes in post
intervention change in A1C, with the FANS mHealth support mode arm with a rebound in
predicted A1C 1.49% mg/dL higher compared to the pamphlet support mode group in months 6
to 12, p=0.236.

In sub analysis by access to primary care, the mean A1C change from 0 to 6 months was twice
as large for patients without prior access to primary care (-2.1 %mg/dL (95%CI -3.0 to -1.2) vs -
1.2%mg/dL (95%CI -1.6 to -0.7), p=0.061), with a larger rebound in the 6 to 12 month period as
well, (0.8%mg/dL (95%CI -0.1 to 1.9) vs 0.04%mg/dL (95%CI -0.5 to 0.4)p=0.961 ). The FANS
mHealth support mode arm vs pamphlet support mode showed similar efficacy between the
groups at 6 months regardless of access to primary care, and similar small rebound of A1C at in
the 6-12 month post intervention maintenance phase.

In sub analysis by patient gender & supporter relationship, the mean A1C change from 0 to 6
months was largest in males with non-spouse supporters (-1.9 (95%CI -2.8 to -1.1) %mg/dL),
and smallest in males with spouse supporters (-0.8 (95%CI -1.7 to 0.1) %mg/dL). The FANS
124
mHealth support mode arm vs pamphlet support mode arm showed the largest difference in
A1C between intervention arms at 6 months for female patients with spouse supporters (-0.8
(95%CI -2.7, 1.1) %mg/dL compared to the unaugmented social support arm).

Conclusions:
Patients who may benefit most from TExT-MED plus with the FANS intervention augmenting
social support via mHealth or the pamphlet augmented social support are those with
depression and low baseline support or without existing access to primary care. Patients that
may benefit the most from the additional mHealth support mode of the FANS support
intervention were men without spousal social support. Identifying these groups may allow
design of future interventions to reach the groups that may have the great potential for benefit.

125
Background & Rationale
Patients with diabetes who seek care in the Emergency Department represent a unique group
of patients. Compared to those who seek care in outpatient settings, ED patients with diabetes
are more likely to be hospitalized and exhibit worse glycemic control (Birtwhistle et al., 2017).
EDs are fruitful sites for screening for diabetes, and also for treating those patients with
diabetes who are not achieving good disease control.(Arora et al., 2014; Charfen et al., 2009;
Danielson et al., 2023)

At LAC+USC in particular, the patients face socioeconomic challenges, have poor access to
primary care and frequently have low English proficiency and low health literacy, all of which
are associated with poor diabetes outcomes (Menchine et al., 2013). Understanding the
specific psychosocial barriers faced by these patients and how these psychosocial barrier
impact intervention efficacy will improve future interventions.

Patients with different psychosocial and behavioral patterns may require differently tailored
interventions. Self-care consists of many decisions and behaviors; the interventions needed to
address one subgroup of patients with specific needs may be very different than the
intervention required for other groups. Underlying self-efficacy, fatalism, psychological distress
and depression, as well as differences in social support, may all impact appraisals of the stress
associated with disease management. These all play a role in a patient’s daily experience with
diabetes and are all associated with changes in disease management and glycemic control,
which are in turn associated with decreased short and long term complications.
126

Latent Profile Analysis (LPA) is one strategy to explore subgroups of patients and their
psychosocial profile. LPA is an analysis method which sorts individuals into groups based on a
set of observed variables (Oberski, 2016). These groups are not directly measured, nor are they
defined a priori, and are latent variables only apparent though analysis techniques. Latent Class
Analysis (LCA) (a form of Latent Profile Analysis utilizing only categorical observed variables) has
been used successfully to identify subgroups of patients with diabetes with distinct stages of
change, risk factors for complications of diabetes and healthy behavior patterns (Gao et al.,
2017; Gariepy et al., 2012; Geiss et al., 2014; Jiang et al., 2012; Wang et al., 2017). However,
little work has been done on processes underlying behavior such as self-efficacy, fatalism,
psychological distress, perceived quality of life and perceived social support. However, LCA
limits analysis to variables which can easily be dichotomized or may lead to artificial
dichotomization of variables that are ordinal or linear. LPA allows for finer representations of
the psychosocial and behavioral profiles that underlie disease management. These Latent
Profiles can then be used to examine the results of intervention studies, exploring the
subgroups of patients who might benefit the most from the interventions.

An alternate strategy to identify which subgroups of patients who may benefit most from
interventions is to conduct qualitative research with intervention participants. Qualitative
research allows for participants voices and perspectives to be more clearly heard, as opposed
to the restrictions on their perspectives that are placed by survey instruments and quantitative
analysis. Mixed-methods approaches that combine qualitative and quantitative findings allow
127
for a richer understanding of the mechanism of an intervention.(Creswell, Fetters, Plano Clark,
& Morales, 2009) When themes from the qualitative analysis are triangulated with quantitative
data, sub group analysis can be conducted in exploratory analysis to identify patients who
might benefit most from future disseminated versions of the intervention, especially for
interventions that require complex social and behavioral changes, such as diabetes support and
management education programs.(Aschbrenner, Kruse, Gallo, & Plano Clark, 2022; E. Burner et
al., 2018; Andrea Cherrington et al., 2008; Sahin & Naylor, 2017)

In this study, we are utilized both LPA subgroup and qualitative triangulation strategies to
conduct post-intervention sub-analysis. We conducted an LPA of the enrolled patients from
TExT-MED+FANS to investigate the potential psychosocial patterns that may underlie latent
subgroups of the study population, using the identified subgroups. We also conducted a mixed
methods analysis utilizing the findings of the semi-structured interview conducted after the
final RCT study visit to identify subgroups of patient who benefited most from the mobile
health intervention. With this two-pronged strategy, we aim to improve on the TExT-
MED+FANS intervention and inform the dissemination of the program in future real world
settings.

Aims and Research Questions
Aim 3: To identify subgroups of the study participants with distinct psychosocial patterns and
examine behavioral and clinical properties of the patients in these subgroups via latent profile
analysis methods.
128

Exploratory Research Question 3a: What are the distinct psychosocial profiles among trial
participants?
Exploratory Research Question 3b: Do these latent profiles of psychosocial factors predict
response to TExT-MED+FANS?
Exploratory Research Question 3c: Do the persistent psychosocial barriers identified in post-
intervention participants through semi-structured interviews identify patients who
respond differently to the TExT-MED+FANS intervention?

129
Methods
Study Population:
The analysis population for this study are the patients who completed randomization and
enrollment in the TExT-MED+FANS RCT (n=166), those who completed 6 month follow up
(n=100) and those who completed 12 month follow up (n=106). To develop the Latent Profile
Analysis model, all 173 patients who initially enrolled in the study were included, even if their
supporter did not complete enrollment to continue participation in the trial. See Figure 9 for
study flow diagram.

130
Figure 9: Screening and Enrollment of patients into TExT-MED+FANS and study flow diagram

Assessed for eligibility (n= 2,004)
Not screened (n=1,959)
¨ Critically Ill (n=449)
¨ Discharged (n=414)
¨ Not approached (n=526)
¨ Not alert and oriented (n=267)
¨ Language barrier (n=116)
¨ Decline, did not want to hear
about study (n=92)
¨ Other (n=95)

Analyzed (n=51)
¨ Lost to follow-up (n=29)
¨ Withdrew from study (n=0)
Allocated to intervention (n=86)
¨ Received allocated intervention (n=80)
¨ Did not receive allocated intervention (FANS
did not complete enrollment) (n=6)
Analyzed (n=49)
¨ Lost to follow-up (n=36)
¨ Discontinued intervention (n=1)

Allocated to control (n=87)
¨ Received allocated intervention (n=86)
¨ Did not receive allocated intervention (FANS
did not complete enrollment) (n=1)
Allocation
3 Month Follow-Up
Randomized (n=173)
Enrollment
Screening
Patients with Diabetes identified by EHR (n=3,963)
Not meeting inclusion criteria (n=1,739)
¨ Unable to use text message (n=613)
¨ Does not have a stable mobile phone (n=427)
¨ Hba1c<8.5 (n=394)
¨ No supporter identified (n=122)
¨ In denial of DM (n=65)
¨ Refused Hba1c test (n=57)
¨ Supporter could not be reached (n=36)
¨ Decline, eligible but did not want to participate
(n=25)

Analyzed (n=45)
¨ Lost to follow-up (n=27)
¨ Discontinued intervention (n=8)
Analyzed (n=52)
¨ Lost to follow-up (n=29)
¨ Discontinued intervention (n=4)
Analyzed (n=36)
¨ Lost to follow-up (n=32)
¨ Discontinued intervention (n=4)
Analyzed (n=51)
¨ Lost to follow-up (n=27)
¨ Discontinued intervention (n=3)
6 Month Follow-Up
9 Month Follow-Up
Analyzed (n=56)
¨ Lost to follow-up (n=20)
¨ Discontinued intervention (n=2)
Analyzed (n=50)
¨ Lost to follow-up (n=15)
¨ Discontinued intervention (n=3)
12 Month Follow-Up
Patients included in LPA model development
Patients included in subgroup baseline description
Patients included 6 month estimate of efficacy
Patients included 12 month estimate of efficacy
131
Study design:
Mixed effects subgroup analysis of a randomized controlled trial of an mHealth based social
support intervention vs a pamphlet based support intervention for Emergency Department
patients with diabetes (Study 1) based on groups identified by:
1) A Latent Profile Analysis of baseline psychosocial factors for all initially enrolled patients
and
2) Themes identified in semi-structured interviews with patients and supporters after the
conclusion of the trial (Study 2) (lack of access to primary care, patient gender and
supporter relationship)

Latent Profile Analysis:
We conducted a Latent Profile Analysis of baseline psycho-social state of a cohort of 173 ED
patients randomized in an mHealth trial with augmented with social support.
Data preparation:
We started by ensuring all variables were complete, with minimal missing data

Variable Selection and screening:
We initially selected variables from the RCT initial baseline data collection that directly measure
a patient’s disease-specific psychologic distress as well as measures of baseline social support
(both disease specific and general social support). (see Figure 10) Measures included:
• Distress from Diabetes Management: Diabetes Distress Scale (Polonsky et al., 2005)
• Self-efficacy: Diabetes Empowerment Scale Short-Form (R. M. Anderson et al., 2003)
132
• Depression: PHQ-9 (Kroenke et al., 2010)
• Fatalism: Diabetes Fatalism Scale (Egede & Ellis, 2010)
• Family behaviors: supportive and non-supportive sub-scores of the Diabetes Family
Behavior Checklist (Lewin et al., 2005)
• Diabetes Care Profile support questions: Support wanted, and support received.
(Fitzgerald et al., 1996)
• General social support: Norbeck Social Support Questionnaire, emotional sub score and
tangible sub score. (Norbeck et al., 1981)

Figure 10: Theoretical model of LPA analysis

Rather than an automated selection of the best model, we conducted a manual review model
of models. Using the Mplus software,(Muthén, 2019) we elected for a manual serial review of

Latent
Class
Non-
Supportive
Family
Behaviors
Support
Needs Support
Recieved
Latent
Class
Depression
Self
Efficacy
Distress
Fatalism
General
Emotional
Social
Support
General
Tangible
Social
Support
Supportive
Family
Behaviors
Latent
Class
Medication
Adherence
Sumamry
of Diabetes
Self Care
Acitivites
BMI/weight
HbA1c
(Glycemic
Control)
133
LPA models incorporating the selected variables with sequentially increasing number of profiles
until the optimum model with the smallest AIC & BIC and with the highest entropy was
obtained.(Spurk, Hirschi, Wang, Valero, & Kauffeld, 2020) We also examined the Lo-Mendell-
Rubin likelihood ratio test as we increased the number of profiles to ensure the model with K+1
profiles fit the data better than the previous model with K profiles.(Lo, Mendell, & Rubin, 2001).
We assigned each patient to the profile they had the highest probability of belonging to. We
then examined the number of participants predicted to be in each latent profile, to ensure that
we did not select profiles that would be too small to use in a sub-analysis of the whole RCT.
(See Table 1) Using these techniques, we identified a 4-profile model as the optimal LPA model.

Table 10: Model Fit and profile size characteristics of proposed LPA models

# profiles AIC BIC Entropy LMR LRT p value Size of smallest profile
2 10828.793 10931.196 0.84 0.0006 88
3 10659.01 10797.749 0.888 0.0508 56
4 10575.288 10750.364 0.925 0.033 17
5 10471.266 10682.677 0.909 0.3773 12
6 10418.071 10665.819 0.92 0.1337 12

With the optimal 4 latent profile model, we categorized the participants in the baseline data set
based on their predicted latent profile. Using descriptive statistics, we examined the profile of
psychosocial variables for each of the 4 profiles and ranked the average score of each predicted
profile to explore each profile. We then developed descriptive names for the 4 latent profiles.

Using subgroups based on predicted latent profile, we examined subgroup differences in
demographic characteristics, clinical measures and baseline health behaviors with ANOVA or
134
Fischer’s exact test as appropriate for variable type. We repeated the mixed-effects analysis in
Study 1, the RCT, using predicted latent profile as an interaction term to randomization group
to test if intervention effects by mHealth support mode intervention vs pamphlet support mode
intervention differed by profile subgroup. We estimated predicted changes in A1C across
intervention groups at 6 and 12 months after enrollment for each predicted latent profile, and
also estimated predicted changes in A1C between intervention groups at 6 and 12 months after
enrollment for each predicted latent profile. Mixed-effects analysis was conducted in Stata
version 17.0 (StataCorp, 2021).

Mixed methods analysis of RCT with semi-structured interview data
In a series of semi-structured interview with patients and supporters from the mHealth
augmented FANS arms analyzed with a grounded theory framework detailed in Study 2, two
themes relevant to social context emerged: (1) persistent barriers in accessing primary care and
(2) a gendered relationship between male and female patients selecting a spouse or a non-
spouse supporter.

Access to Primary Care
During the interviews, or in the time preceding the interviews when the participants were
completing self-report measures at the conclusion of the trial, participants noted persistent
barriers to accessing continuity care and establishing therapeutic relationships with healthcare
providers.
135
“It’s good, the program is good, it helps us, but like I mentioned earlier some of us don’t
have a primary care doctor…hopefully I can get a primary care doctor right now so that
my diabetes can be better. “
104P, Patient, Spanish Language Preference

From Interviewer notes on small-talk prior to recorded interview:
Patient does not have primary care provider and has not been taking her insulin for more
than 5 months. She noted that she was tired of having a new doctor every 3 months.
293P, Patient, Spanish Language Preference

Gender and supporter relationship
Specifically, male patients recurrently described their selection of their wife or life-partner as a
supporter due to the emotional or instrumental support they provided the patient at baseline,
before the additional activation of support. These male patients identified their wife or life-
partner as the person in their life that actually drives many of the self-care behaviors:

“Because she's the woman who’s with me in the house and she is the person who helps
me with everything. I didn’t choose her; it's how we live.”
152P, Male, Patient, Spanish Preference

“Since the first time that she (patient’s wife) came, she was taught to make food a little
more healthier.”
136
125P, Male, Patient, Spanish Preference

After examining the excerpts from interviews and detailed notes from the interviewers to
understand the barriers to care, we examined the variables collected in the RCT and
determined that utilizing primary care in the 12 months prior to study enrollment captured
access to primary care as insurance alone did not address the issues of fragmented care and
frequent provider changes. We also determined that gender and relationship to supporter
(spouse or other) were relevant variables that represented the theme of gender differences in
supporter selection.

Using these variables to represent the qualitative themes to identify subgroups (baseline access
to primary care in the 12 months prior to the trial, and the patient gender & supporter
relationship), we examined subgroup differences in demographic characteristics, clinical
measures and baseline health behaviors with ANOVA or Fischer’s exact test as appropriate for
variable type. We repeated the mixed-effects analysis in Study 1, the RCT, using these variables
as interaction terms to randomization group. We estimated predicted A1C in combined
intervention groups at 6 and 12 months after enrollment, and estimated predicted diffrences in
A1C between intervention groups at 6 and 12 months after enrollment for each subgroup
variable.

137
Results:
Latent Profile Analysis:
The population to determine the Latent Profile characteristics consisted of 173 patients who
were initially enrolled in the TExT-MED+FANS trial. The four latent profiles identified consisted
of between 17 and 79 participants, and varied most in depression, fatalism and level of support
(see Table 11). The four descriptive names were:
(1) Alone, depressed & fatalistic
(2) Highly Supported & fatalistic, not depressed
(3) Supported, depressed & fatalistic
(4) Most supported, most depressed, not fatalistic
Table 11: Profile Descriptive Name, Size and Rank*** of Profile Average of Each Scale
Descriptive Name
Participants Predicted in Profile, n=
PHQ9 (depression)
DESF (self-efficacy)
DDS (Diabetes related distress)
Diabetes Supportive Family Behaviors
Diabetes Non-Supportive Family Behaviors
General Emotional Support
General Tangible Support
Support Needs (Diabetes Care Profile)
Support Received (Diabetes Care Profile)
Diabetes Specific Fatalism
Alone, depressed &
fatalistic
46 2 1 4 4 1 4 4 4 4 3
Highly Supported &
fatalistic, not depressed
59 1 4 2 2 3 2 2 2 2 4
Supported, depressed &
fatalistic
79 3 2 3 3 2 3 3 3 3 4
Most support, most
depressed, not fatalistic
17 4 3 1 1 4 1 1 1 1 1
*** Rank 1 = best average of the profiles 4 for that variable, Rank 2 is next best, etc
138

The patients in these 4 latent profiles were similar in terms of race/ethnicity, age, Spanish
language preference and nativity, with no differences in proportions between groups (see Table
12, top panel, for patient characteristics by predicted latent profile). Patients in the first latent
profile, who reported the least baseline social support by the Diabetes Family Behavior
Checklist – supportive (ANOVA between group p<0.0001), Diabetes Family Behavior Checklist –
nonsupportive (ANOVA between group p<0.0001), Diabetes Care Profile- Support received
score (ANOVA between group p<0.0001), Norbeck Social Support Questionaire – Emotional
(ANOVA between group p=0.002) and Norbeck Social Support Questionaire – Tangible (ANOVA
between group p=0.0001), also reported the worst self-care behaviors (SDSCA – self-monitoring
glucose ANOVA between group p=0.002; Wilson medication adherence scale ANOVA between
group p= 0.025). Patients in third latent profile had the lowest level of self-reported baseline
exercise (SDSCA – exercise ANOVA between group p=0.039), and lower levels of social support:
Diabetes Family Behavior Checklist – supportive (ANOVA between group p<0.0001), Diabetes
Family Behavior Checklist – nonsupportive (ANOVA between group p<0.0001), Diabetes Care
Profile- Support received score (ANOVA between group p<0.0001), Norbeck Social Support
Questionaire – Emotional (ANOVA between group p=0.002). Patients in the fourth latent profile
reported the highest levels of social support and reported the best exercise days, self-
monitoring of glucose and medication adherence (SDSCA – self-monitoring glucose ANOVA
between group p=0.002; SDSCA – exercise ANOVA between group p=0.039; Wilson medication
adherence scale ANOVA between group p= 0.025).

139
We examined overall TExT-MED+FANS intervention efficacy by subgroup identified by the
predicted latent profiles (see Table 12, bottom panel). Analysis by subgroup was
underpowered for clinically significant differences in A1C; no statistical differences between
latent profiles or between the FANS intervention and unaugmented social support arms were
found. With mixed effects analysis controlling for baseline A1C, across intervention group mean
A1C improvement from 0 to 6 months was twice as large for patients in the “Alone, depressed
and fatalistic” Profile #1 compared to the “Most support, most depressed, not fatalistic” Profile
#4. In the 6 to 12 month maintenance phase, the across intervention group mean A1C change
from 6 to 12 months was of similar magnitude across latent profiles.

In examining the FANS intervention vs unaugmented social support groups, the latent profiles
showed similar efficacy with similar between group A1C changes at 6 months (See Table 12, last
2 rows). In examining the maintenance of the intervention effects, the highly supported but
most depressed Profile #4 showed the largest between intervention group difference in
predicted A1C, with the FANS augmented social support with a rebound in A1C 1.49% mg/dL
higher in the FANS group compared to the unaugmented social support group.

140
Table 12: Characteristics and efficacy outcomes by predicted latent profile
Alone,
depressed &
fatalistic
n= 38
High support,
fatalistic, less
depressed
n= 51
Med support,
fatalism, more
depressed n=
63
Most support,
depressed,
low fatalism
n= 14
Baseline Characteristics and Measures (mean, sd)
Insurance type (%, n) None 0% (0) 4% (2) 5% (3) 0% (0)
Means tested 95% (36) 92% (47) 90% (57) 100% (14)
Non-means tested 5% (2) 4% (2) 5% (3) 0% (0)
Race/Ethnicity Latino 95% (36) 92% (47) 90% (57) 93% (13)
Non-Hispanic White 0% (0) 2% (1) 2% (1) 0% (0)
Asian 3% (1) 2% (1) 0% (0) 0% (0)
Black 3% (1) 4% (2) 8% (5) 7% (1)
Spanish Language Preference 68% (26) 71% (36) 70% (44) 71% (10)
Nativity – Born in US 22% (8) 24% (12) 17% (11) 36% (5)
Age 48.7 (10.4) 45.5 (10.6) 49.0 (9.8) 46.2 (12.3)
A1C 10.4 (1.7) 10.8 (1.7) 10.8 (1.7) 11.2 (1.6)
BMI 29.1 (7.0) 30.0 (8.0) 30.9 (7.9) 28.9 (6.2)
Systolic BP 132.0 (21.1) 136.1 (21.1) 135.4 (26.2) 132.9 (26.2)
Medication Adherence 56.6 (34.5) 73.1 (25.1) 64.6 (29.4) 78.3 (23.6)
SDSCA: general diet 2.9 (2.9) 3.4 (2.4) 3.2 (2.4) 3.2 (2.4)
SDSCA: specific diet 3.5 (2.1) 4.0 (1.7) 3.8 (1.9) 4.3 (1.8)
SDSCA: glucose monitoring 1.5 (2.4) 3.5 (3.0) 2.3 (2.9) 4 (2.9)
SDSCA: foot care 3.4 (3.1) 4.4 (2.9) 4.0 (2.8) 4.0 (3.3)
SDSCA: carb spacing 2.5 (2.7) 3.1 (2.3) 2.7 (2.7) 3.6 (2.6)
SDSCA: exercise 2.7 (2.7) 2.4 (2.5) 2.0 (2.2) 4.1 (3.3)
Self-efficacy 3.4 (0.7) 4.1 (0.6) 3.9 (0.6) 4.0 (0.5)
DM distress score 2.5 (1.1) 2.4 (1.0) 2.6 (1.1) 2.4 (0.8)
Depression (PHQ9) 8.9 (7.5) 8.2 (5.4) 9.7 (7.0) 11.1 (6.8)
Quality of life 55.5 (31.1) 62.0 (26.1) 60.8 (27.6) 61.4 (29.6)
Fatalism score 33.2 (12.0) 35.8 (9.5) 36.0 (8.9) 31.4 (8.4)
Supportive Family Behaviors 11.9 (2.5) 31 (2.6) 21.8 (3.0) 39.7 (2.4)
Nonsupportive Family Behaviors 9.4 (1.9) 23.5 (1.7) 16.6 (2.2) 30.2 (2.2)
DCP support needs 19.76 (8.6) 25.4 (6.9) 23.6 (7.0) 27.9 (3.9)
DCP support received 9.6 (6.1) 22.5 (8.0) 18.9 (7.4) 27.1 (4.3)
DCP support attitudes 4.1 (5.4) 7.9 (4.8) 6.4 (5.1) 7.6 (4.3)
Emotional Support 12.6 (3.8) 14.9 (2.7) 13.5 (3.2) 15.5 (1.0)
Tangible support 6.0 (2.1) 7.5 (1.2) 7 (1.7) 7.9 (0.3)
Mixed Effects Estimates of Intervention Efficacy by Subgroup: mean(UL95% CI, LL95%CI)
Combined intervention group
mean 0 to 6 month A1C change
-1.6 (-2.4, -0.8) -1.3 (-2.1, -0.6) -1.4 (-2.0, -0.7) -0.8 (-2.1, 0.4)
Combined intervention group
mean 6 to 12 month A1C change
-0.07(-1.0, 0.9) 0.4 (-0.4,1.2) -0.1 (-0.6, 0.8) -0.2 (-1.6, 1.2)
FANS vs unaugmented social
support 6 month A1C change
-0.04(-1.7, 1.7) 0.5 (-1.0, 2.1) 0.05(-1.2, 1.3) -0.4(-2.9, 2.1)
FANS VS unaugmented social
support at 12 months
-0.8 (-2.7, 1.1) -0.9(-2.5, 0.7) -0.4(-1.8, 1.0) 1.49 (-1.3, 4.3)
141
Qualitative Findings Mixed-Methods Analysis

Analysis by Access to Primary Care:
There were 39 patients who had not had a primary care visit in the year prior to enrollment and
127 who reported a primary care visit in the year prior to enrollment. The patients with access
to primary care vs those who did not have access had similar insurance types, racial and ethnic
backgrounds, language preference and likelihood of being born in the United States. See Table
4, top panel for all baseline characteristics. Those without a recent primary care visit reported
worse diabetes self-care behaviors: SDSCA – self-monitoring glucose ANOVA between group
p=0.036; SDSCA – disease specific diet ANOVA between group p=0.0105; SDSCA – footcare
ANOVA between group p=0.0706; Wilson medication adherence scale ANOVA between group
p= 0.0003. Those with primary care access reported an average lower quality of life: WHO QOL
ANOVA between group p= 0.0714; they also reported higher levels of diabetes specific social
support Diabetes Family Behavior Checklist – supportive (ANOVA between group p<0.0943),
Diabetes Family Behavior Checklist – nonsupportive (ANOVA between group p<0.0624). (See
Table 13)

We examined overall TExT-MED+FANS intervention efficacy by subgroups identified by having
had or not had a primary care visit in the year prior to the TExT-MED+FANS intervention.
Analysis by subgroup was underpowered for clinically significant differences in A1C; no
statistical differences between those with and without primary care was found between the
FANS intervention and unaugmented social support arms. With mixed effects analysis
142
controlling for baseline A1C, across intervention group mean A1C change from 0 to 6 months
was -2.1 (95%CI -3.0 to -1.2) %mg/dL for patients without access to primary care and -1.2
(95%CI -1.6 to -0.7) %mg/dL; across intervention group mean A1C change from the post
intervention maintenance phase of 6 to 12 months trended to a larger rebound among patients
without existing primary care access (0.8 (95%CI -0.1 to 1.9) %mg/dL compared to 0.04 (95%CI -
0.5 to 0.4) %mg/dL).

Those with and without primary care access showed similar efficacy between the FANS
intervention and the unaugmented social support arm at 6 months, and similar small rebound
of A1C at 12 months (see Table 13, last 2 rows).

143
Table 13: Characteristics and efficacy outcomes of patients with and without primary care

No Primary Care
n=39
Has Primary Care
n=127
Baseline Measures and Characteristics (mean, sd)
Insurance type
(%, n) None 5% (2) 2% (3)
Means tested 92% (36) 93% (118)
Non-means tested 3% (1) 5% (6)
Race/Ethnicity Latino 92% (36) 92% (117)
Non-Hispanic White 0% (0) 2% (2)
Asian 3% (1) 1% (1)
Black 5% (2) 6% (7)
Spanish Language Preference 64% (25) 72% (91)
Nativity – Born in US 23% (9) 21% (27)
Age 47.5 (9.7) 47.6 (10.7)
A1C - baseline 10.8 (1.4) 10.8 (1.8)
BMI - baseline 32.0 (9.9) 29.5 (6.7)
Systolic BP - baseline 136.0 (27.0) 134.2 (24.0)
Medication Adherence 51.8 (36.7) 71.2 (25.4)
SDSCA: general diet 2.9 (2.6) 3.3 (2.4)
SDSCA: specific diet 3.2 (1.8) 4.1 (1.9)
SDSCA: glucose monitoring 1.8 (2.8) 2.9 (2.9)
SDSCA: foot care 3.3 (3.2) 4.3 (2.8)
SDSCA: carb spacing 2.3 (2.7) 3.1 (2.5)
SDSCA: exercise 2.5 (2.6) 2.5 (2.6)
Self-efficacy 3.8 (0.8) 3.9 (0.6)
DM distress score 2.5 (1.1) 2.5 (1.0)
Depression (PHQ9) 8.7 (6.2) 9.3 (6.8)
Quality of life 67.1 (28.0) 57.8 (278)
Fatalism score 35.9 (8.4) 35.7 (10.3)
Supportive Family Behavior 21.8 (7.4) 24.5 (9.2)
Nonsupportive Family Behavior 16.5 (5.5) 18.8 (6.9)
DCP support needs 22.4 (9.0) 24.0 (7.0)
DCP support received 17.4 (8.3) 18.9 (9.1)
DCP support attitudes 6.4 (4.4) 6.5 (5.4)
Emotional Support 13.7 (3.2) 13.9 (3.3)
Tangible support 6.8 (1.8) 7.0 (1.6)
Mixed Effects Estimates of Intervention Efficacy by Subgroup: mean(UL95% CI, LL95%CI)
Combined intervention group
mean 0 to 6 month A1C change
-2.1 (-3.0, -1.2) -1.2 (-1.6, -0.7)
Combined intervention group
mean 6 to 12 month A1C change
0.8 (-0.1, 1.9) 0.04 (-0.5, 0.4)
FANS VS unaugmented social
support at 6 months
0.6 (-1.2, 2.3) 0.08 (-0.8, 1.0)
FANS VS unaugmented social
support at 12 months
-0.3 (-2.3, 1.7) -0.5 (-1.4, 0.5)
144
Analysis by Gender and Supporter relationship subgroups:
There were 4 combinations of gender and supporter relationships, as no patients identified as
non-binary or non-gender-conforming: (1) female patients with non-spouse supporter n=59 (2)
female patients with spouse supporter n=25; (3) male patients with non-spouse supporter
n=42; and (4) male patients with spouse supporter n=40. The patients in these subgroups had
similar proportions of insurance types, racial and ethnic backgrounds, nativity and language
preference, except for male patients with non-spouse supporters. Male patients with non-
spouse supporters tended to be more likely to have an English language preference than other
patients (45% English preference vs 22%, 24% and 30% in other groups, p=0.08) (see Table 14,
top panel). Male with non-spouse supporters generally reported the lowest medication
adherence and self-efficacy (Wilson medication adherence scale ANOVA between group p=
0.0012, Diabetes Empowerment - Short Form ANOVA between group p= 0.0482
frequency of self-care activities) and lower levels of disease specific social support: Diabetes
Care Profile- Support received score (ANOVA between group p=0.0828), Diabetes Care Profile-
Support attitudes score (ANOVA between group p=0.0324). Males with spouse supporters
reported the highest level of non-supportive family behaviors (i.e., nagging): Diabetes Family
Behavior Checklist – nonsupportive (ANOVA between group p=0.0919) and also general
instrumental support Norbeck Social Support Questionaire – Tangible (ANOVA between group
p=0.0845). Females with non-spouse supporters reported the lowest quality of life (WHO QOL
ANOVA between group p=0.0393). Both males and females with non-spouse supporters had
higher levels of depression than patients who selected a spouse supporter (PHQ-9 ANOVA
between group p=0.0346)
145
We examined overall TExT-MED+FANS intervention efficacy by subgroups identified by patient
gender and supporter relationship. Analysis by subgroup was underpowered for clinically
significant differences in A1C; no statistical differences between the FANS intervention and
unaugmented social support arms were found (see Table 14, bottom panel). With mixed effects
analysis controlling for baseline A1C, across intervention group mean A1C change from 0 to 6
months was largest among males with non-spouse supporters (-1.9 (95%CI -2.8 to -1.1)
%mg/dL), followed by female with spouse supporters (-1.7 (95%CI -2.7 to -0.7) %mg/dL),
females with non-spouse supporters (-1.1 (95%CI -1.8 to -0.5) %mg/dL) and then males with
spouse supporters(-0.8 (95%CI -1.7 to 0.1) %mg/dL). In the 6-12 month post intervention
maintenance phase, male patients with spouse and non-spouse supporters trended toward a
continued improvement in A1C (-0.2 (95%CI -1.1 to 0.7) %mg/dL and -0.6 (95%CI -1.6 to 0.3)
%mg/dL, respectively.

The FANS intervention vs unaugmented social support showed the largest difference in A1C
between intervention arms at 6 months for female patients with spouse supporters (-0.8
(95%CI -2.7, 1.1) %mg/dL compared to the unaugmented social support arm). Males with non-
spouse supporters in the FANS arm showed the largest continued improvement in the 6-12
month post intervention maintenance phase (-1.5 (95%CI -3.3, 0.2) %mg/dL) when compared to
males with non-spouse supporters in the unaugmented social support arm, while males with
spouse supporters in the FANS arm showed the largest rebound compared to males with
spouse supporters in the unaugmented social support arm with an increase in A1C of 1.6
(95%CI -3.3, 0.2) %mg/dL. (see Table 14, last two rows)
146
Table 14: Characteristics and efficacy outcomes by patient gender & supporter relationship type

Female with
non-spouse
supporter
n=59
Female with
spouse
supporter
n=25
Male with
non-spouse
supporter
n=42
Male with
spouse
supporter
n=40
Baseline Characteristics & Measures
Insurance type (%, n) None 3% (2) 4% (1) 2% (1) 3% (1)
Means tested 93% (55) 88% (22) 98% (41) 90% (36)
Non-means tested 3% (2) 8% (2) 0% (0) 8% (3)
Race/Ethnicity Latino 90% (53) 92% (23) 98% (41) 90% (36)
Non-Hispanic White 2% (1) 0% (0) 0% (0) 3% (1)
Asian 3% (1) 0% (0) 0% (0) 3% (1)
Black 7% (5) 8% (2) 2% (1) 5% (2)
Spanish Language Preference 78% (46) 76% (19) 55% (23) 70% (28)
Nativity – Born in US 16% (9) 20% (5) 36% (15) 18% (7)
Age 48.6 (10.5) 47.6 (12.3) 46.2 (10.8) 47.6 (8.7)
A1C - baseline 10.4 (1.4) 10.5 (1.6) 11.3 (1.7) 11.2 (1.9)
BMI - baseline 30.6 (6.6) 32.5 (10.0) 29.5 (8.4) 28.4 (6.0)
Systolic BP 130.3 (23.3) 133.6 (22.7) 136.2 (24.9) 140.0 (27.1)
Medication Adherence 73.1 (25.5) 68.8 (27.0) 51.3 (32.8) 71.7 (28.2)
SDSCA: general diet 3.5 (2.5) 3.2 (2.7) 3.2 (2.4) 2.9 (2.3)
SDSCA: specific diet 4.3 (1.7) 4.0 (2.0) 3.7 (2.0) 3.3 (1.9)
SDSCA: glucose monitoring 3.1 (3.0) 2.6 (2.9) 2.1 (2.9) 2.6 (3.0)
SDSCA: foot care 3.9 (3.0) 5.6 (2.7) 3.6 (2.9) 4.4 (3.0)
SDSCA: carb spacing 2.9 (2.6) 2.2 (2.3) 2.9 (2.6) 3.3 (2.6)
SDSCA: exercise 2.3 (2.5) 2.5 (2.7) 2.7 (2.7) 2.5 (2.5)
Self-efficacy 3.8 (0.5) 3.9 (0.8) 3.7 (0.8) 4.1 (0.6)
DM distress score 2.6 (1.0) 2.8 (1.0) 2.2 (0.8) 2.4 (1.2)
Depression (PHQ9) 10.6 (6.5) 7.6 (6.3) 10.0 (6.7) 7.2 (6.4)
Quality of life 51.8 (26.8) 63.0 (28.8) 63.6 (27.6) 66.5 (27.6)
Fatalism score 37.1 (9.9) 32.0 (9.5) 34.1 (10.2) 34.5 (9.5)
Supportive Family Behavior 23.5 (9.0) 23.6 (8.8) 22.1 (9.4) 18.0 (6.7)
Nonsupportive Family Behavior 18.0 (6.7) 18.0 (6.5) 16.7 (7.2) 20.3 (5.7)
DCP support needs 24.8 (6.6) 23.4 (8.5) 21.8 (7.8) 24 (7.7)
DCP support received 20.1 (8.2) 17.4 (9.7) 15.9 (8.2) 19.75 (9.6)
DCP support attitudes 5.9 (5.2) 9.1 (3.6) 5.5 (4.6) 6.6 (6.0)
Emotional Support 14.1 (3.1) 13.1 (4.4) 13.1 (3.2) 14.7 (2.2)
Tangible support 6.9 (1.7) 6.8 (2.1) 6.6 (1.7) 7.6 (1.0)
Mixed Effects Estimates of Intervention Efficacy by Subgroup: mean(UL95% CI, LL95%CI)
Combined intervention group
mean 0 to 6 month A1C change
-1.1 (-1.8, -0.5) -1.7 (-2.7, -0.7) -1.9 (-2.8, -1.1) -0.8 (-1.7, 0.1)
Combined intervention group
mean 6 to 12 month A1C change
0.4 (-0.3, 1.2) 0.6 (-0.4, 1.7) -0.2 (-1.1, 0.7) -0.6 (-1.6, 0.3)
FANS VS unaugmented social
support at 6 months
0.1 (-1.2, 1.4) -0.8 (-2.7, 1.1) 0.5 (-1.1, 2.1) 0.3 (-1.5, 2.0)
FANS VS unaugmented social
support at 12 months
-0.03(-1.5, 1.4) -0.9 (-2.9, 1.2) -1.5 (-3.3, 0.2) 1.6 (-0.3, 3.5)
147
Discussion:
In this study, we conducted subgroup analyses of the trial of the TExT-MED+FANS intervention,
designed to activate diabetes specific social support through mHealth modalities. We utilized
Latent Profile Analysis to identify subgroups that entered the study with unique psycho-social
profiles and employed mixed method qualitative methodology to identify subgroups of
participants who experienced differences in the effectiveness of the intervention based on
qualitative findings: those with lack of access to primary care, and differences in the existing
patient-supporter relationship by patient gender and supporter relationship type. In the overall
analysis (Study 1) patients whose supporter received support curriculum via the mHealth mode
and those whose supporter received the curriculum via the pamphlet mode improved equally
in A1C. In this study (Study 3), we examined subgroup differences in support mode efficacy in
patient A1C improvement for social support augmented via the mHealth mode from the FANS
intervention added to the patient focused TExT-MED compared to the pamphlet mode support.
While the subgroups were underpowered to detect differences in the support intervention
mode efficacy, in this exploratory analysis, we identified gender based patterns in social
support which may impact the efficacy of social support mediated behavior change
interventions and persistence of improved outcomes. We found promising results in across
intervention arm improvements in glycemic control, especially for patients without primary
care access and those with low baseline social support and higher levels of comorbid
depression.

148
We identified patterns of social support and self-care differences by gender and supporter
spousal relationship, with trends to differences in intervention efficacy by support mode.
Combined across support intervention groups, male patients with spouse supporters enrolled in
the trial did not have as rapid of an improvement in A1C, but has persistent improvement in the
post intervention phase, while male patients with non-spouse supporters had a rapid
improvement in A1C, but a more modest continued improvement in the post intervention
phase. Female patients with spouse and non-spouse supporters had moderate improvements
in A1C, but tended toward regression towards poorer glycemic control. Female patients with
non-spouse supporters and male patients with spouse supports showed the largest continued
benefit in the post intervention maintenance phase between modes of support, with the
mHealth mode support arm showing predicted A1C -0.9%mg/dL (95% CI -2.9, 1.2) and -
1.5%mg/dL (95% CI -3.3, 0.2) lower than the pamphlet mode support in those subgroups,
respectively. However, male with non-spouse supporters with mHealth mode support fared
worse in the post intervention maintenance phase, with predicted mean A1C 1.6 %mg/dL
higher (95% CI -0.3, 3.5) than those with pamphlet mode support. Observational studies of the
role of marital relationships in diabetes self-care have explored the tension of providing support
to a spouse with diabetes versus controlling the spouse’s disease adherence
behavior.(Vahedparast, Mohammadi, Ahmadi, & Farhadi, 2018) Existing support from friends is
more directly related to glycemic control than support from spouses or other family members,
likely mediated by internal locus of control of disease related selfcare.(Ford & Robitaille, 2023)
The quality of the support offered by family and spouses likely plays a large role in how social
support mediated disease management and glycemic control; non supportive behaviors such as
149
nagging and allowing patients to sleep through their scheduled medication times have been
associated with less glycemic control.(L. S. Mayberry & Osborn, 2012). Women have reported a
diverse group of support providers for their diabetes self-management and less frequently find
support from spouses at baseline, while men tend to report a spouse or one closely connected
person as their support. (Albanese et al., 2019; A. Cherrington et al., 2011; Koch, Kralik, &
Taylor, 2000; Nagelkerk, Reick, & Meengs, 2006; Rosland, Heisler, Choi, Silveira, & Piette, 2010)
Understanding how to best inspire support from family members and spouses rather than
nagging or enabling support behaviors, and how gender roles may reflect different patient and
supporter needs will inform future interventions.

Primary care access was a critical barrier from our qualitative analysis, and was an important
driver in intervention efficacy in this mixed methods analysis. We found that in the combined
intervention group analysis, patients with low access to primary care prior to the TExT-
MED+FANS trial were rapid responders to the TExT-MED intervention, with nearly twice the
effect size of patient who had existing access to primary care at the end of the 6 month
intervention (decrease in A1C at 6 months of -2.1%mg/dL (95% CI -3.0, -1.2) vs -1.2%mg/dL
(95% CI -1.6, -0.7). However, the combined support mode patients who lacked primary care
lost about half of their gains in the post-intervention maintenance phase, compared to patients
who had existing primary care access and maintained their more modest gains in the post
intervention phase (post intervention phase A1C increase of in 0.8%mg/dL (95% CI -0.1, 1.9) vs
0.04%mg/dL (95% CI -0.5, 0.4)). The support mode arms had no difference in A1C when
examined by primary care access subgroups in either phase. The persistent barrier continuity
150
care for disease management is a unique barrier in the diabetes selfcare literature, as TExT-
MED+FANS is based out of the ED and patients were enrolled during unscheduled care visits.
Nearly all studies of mHealth and ICT based are based out of continuity care clinics or
endocrinology clinics. Primary care is an important driver of diabetes selfcare, recognized by
the American Diabetes Association as the cornerstone of patient education, medication
adjustments and referrals to specialists as necessary.(American Diabetes, 2022) Those patient
lacking access to quality primary care relationships are more likely to be hospitalized, have
poorer outcomes and to seek care in emergency departments.(van Loenen, van den Berg,
Westert, & Faber, 2014), (Rosano et al., 2013) Natural experiments in which patients were
switched to high deductible insurance plans, or insurance coverage decreased for routine
outpatient care such as podiatry have resulted in increased costs from unscheduled Emergency
Department care and hospitalizations. (Wharam et al., 2018), (Skrepnek, Mills, & Armstrong,
2014) Simply having insurance and potential access to care is not sufficient; patients must have
realized access to care.

When examining the latent profiles of baseline psychosocial characteristics, the most
pronounced differences in selfcare activities at baseline, and in improvements in A1C in the
combined arms of the patient intervention was between those with low perceived social
support at baseline, and those with the highest social support. The mode of social support
intervention may have differential impact based on baseline support. The group with the
highest baseline support showed the largest relapse in glycemic control, with a rise in predicted
A1C 1.49%mg/dL (95% CI -1.3, 4.3) higher in the FANS mHealth support mode arm than the
151
pamphlet support arm in the post intervention maintenance phase. In contrast, in the subgroup
with the lowest baseline support, patients in the FANS mHealth support mode arm had a
decrease in predicted mean A1C than the patients in the pamphlet support mode arm during
the post intervention maintenance phase (-0.8%mg/dL (-95% CI 2.7, 1.1)). Perceived social
support is known to be associated with improved medication adherence, dietary practices,
primary care adherence and self-monitoring of blood glucose. (Gray, Hoerster, Reiber, Bastian,
& Nelson, 2019), (King et al., 2010) Lower social support is associated with downward disease
trajectories and worsening functional status. (Levy, Deschenes, Burns, Elgendy, & Schmitz,
2019). Ethnic minorities find social support for diabetes to be more helpful than non-Hispanic
whites, highlighting how social support may be an important strategy to reduce disease
disparities.(Peyrot et al., 2018) Additionally, prior work in vulnerable populations has also
identified baseline level of social support as key components to the success of selfcare
interventions, with higher levels of social support generally resulting in more successful
behavior change, however translation to better glycemic control has been rare.(Brew-Sam,
Chib, & Rossmann, 2020; L. S. Mayberry et al., 2021; Ramkisson, Pillay, & Sibanda, 2017; Roddy,
Nelson, Greevy, & Mayberry, 2022) Further studies of the role of mHealth augmented social
support vs simulated social support may identify patients that are more likely to benefit from
each strategy. Computer simulated social support would also allow for those patients who lack
the baseline social support to enroll a supporter to be able to benefit from social support based
mHealth interventions for diabetes self-care.

152
While this study provides important insights into subpopulations who may benefit most from
mHealth and social support based interventions, there are several limitations to the findings.
The study sample size was small for the latent profile method. Typically, sample sizes of 500 or
more participants are ideal,(Spurk et al., 2020) although prior work has indicated success with
small sample sizes when more variables are included if large degrees of separation exist.(Tein,
Coxe, & Cham, 2013) However, the differences in intervention efficacy may not have been large
enough to find a statistical difference between groups. In addition to the limitations of the
latent profile analysis, the qualitative mixed-methods approach is context based, and was
drawn only from the FANS intervention arm. The themes identified from it may not reveal
differences between the intervention and unaugmented support arm, but only of overall
intervention efficacy. The subgroup analyses statistical power were exploratory and planned
post hoc. They are also limited by small subgroups, however the effect sizes are large and
warrant confirmation in larger studies.

In this study, we found that there are specific groups of patients who may benefit most from
TExT-MED+FANS, an mHealth based diabetes self-care interventions that include a social
support component (patients with depression and low baseline support or without existing
access to primary care), as well as patient groups that may benefit from the additional
augmentation of social support intervention from the FANS support intervention (men without
spousal social support). By identifying these groups, future interventions can be designed to
reach the groups that would have the great potential for benefit.

153
Chapter 5: Conclusions

With the persistent health disparities faced by safety-net Emergency Department patients, this
dissertation provides new insight on the potential impact of mHealth augmented social support
on psychological measures, health behaviors and subsequent glycemic control. Social support
based interventions are more positively received among ethnic minority and vulnerable
populations, such as the patient population served at LACUSC Emergency Department.
mHealth interventions can break through the transportation and inflexible schedule barriers
faced by these populations as well. In Study 1, the overall sample of emergency department
patients receiving the TExT-MED intervention along with identifying a support person, all
patients regardless of study intervention arm showed a substantial improvement in health
behaviors and glycemic control. In Study 2, patients reported changes in intrinsic and extrinsic
motivators to good selfcare, as well as gendered choices in supporter selection. Supporters and
patients noted persistent barriers to diabetes selfcare, such as medication affordability and
access to primary care. In Study 3, a post hoc mixed-methods exploratory analysis of the
intervention highlighted groups that may benefit the most from the additional support of the
FANS mHealth activated social support. Within the limitations of these studies and the potential
implications for secondary prevention interventions such as TExT-MED and TExT-MED+FANS,
there are many opportunities for future research to investigate the ideal mechanisms to
activate social support and the ideal nodes to do so from.

154
The evidence in Study 1 strengthens prior literature that mHealth based strategies for
vulnerable, especially those that incorporate building social support resources, are powerful
tools to improve diabetes self-management and glycemic control.(Arora et al., 2014; Presley,
Agne, Shelton, Oster, & Cherrington, 2020; Steinberg et al., 2022) Overall, healthy behaviors
and glycemic improved for all patients in the study during the intervention period, with non-
significant regressions to prior behavior measures and glycemic control during the post-
intervention maintenance phase. However, there was no between group difference for the
mHealth augmented social support arm (the FANS mHealth intervention sent to supporters’
mobile phones) versus the pamphlet social support intervention, even after controlling for
uneven randomization with baseline differences in gender and language preference between
the support intervention groups. In planned subgroup analysis, patients with newly diagnosed
diabetes and patients who received texts in English showed promising trends of improvement
in glycemic control when adding the mHealth augmented social support of FANS to TExT-MED
compared to the pamphlet version. In structural equation mediation modeling, there was no
evidence of mediation of the effect of the FANS intervention on glycemic control by healthy
behaviors. While augmenting support via mHealth may have benefited those with newly
diagnosed diabetes and those who spoke English, for most patients, the additional cost and
time of enrolling supporters in an mHealth intervention may not result in improved glycemic
control over the benefit of simply identifying a specific person as the supporter and providing
some basic training in offering disease specific support.

155
In Study 2, a post intervention qualitative analysis of FANS intervention arm patients and
supporters, we identified changes in intrinsic and extrinsic motivators to good selfcare.
Patients identified a desire to fulfill ideal family roles and to feel healthy for themselves as
internal motivator. Fear of diabetic complications (especially witnessed previously in others)
and having family members providing support and encouragement were perceived as strong
extrinsic motivators. These intrinsic and extrinsic motivators were more present in patients’
health behavior decisions after the intervention. Patients also identified reductions in negative
coping strategies as a result of their participation in the trial. Patients of each gender, as well as
patients who selected male versus female supporters, reported different reasons for selecting
their supporter. Female supporters where often chosen for their perceived emotional strength,
or current role as the caregiver or food preparer in a patient’s life, especially among male
patients. This gendered difference in supporter selection and role may impact the efficacy of
the intervention. In order to best support their loved one with diabetes, wives or caregivers
may need a more specialized support curriculum that focused more on the tangible support
described in the interview. Less proximal family members or friends may need a curriculum
focused on emotional, informational and appraisal support to offer the most benefit to
patients. While the qualitative findings informed the mechanism of action and future directions
for interventions, we also identified potential barriers to diabetes selfcare that were not
addressed by the TExT-MED+FANS intervention design, including medication affordability and
access to primary care. These barriers to care are difficult to address on the individual level, and
specialized curriculum may be needed to activate patients to successfully build a continuity
relationship and a means to pay for that medical care.
156

In Study 3, we conducted a post hoc exploratory analysis of the FANS intervention by subgroup
identified by the qualitative findings (gender and supporter relationship types, access to
primary care) and a Latent Profile Analysis of the baseline psychosocial states of the
participants. The FANS intervention (compared to the pamphlet support intervention) may be
most efficacious among those without primary care access at baseline, and males who selected
a non-spouse supporter. In the post intervention maintenance phase, those without primary
care rebounded towards prior levels of poor glycemic control, while male patients with newly
augmented non-spousal support continued to show improvements in glycemic control. This is
consistent with prior experiences with the TExT-MED patient-only intervention; patients in the
initial trial reported reviewing their patient focused intervention with spouse or other family
members who lived with them before the existence of the supporter focused curriculum. The
additional support focused curriculum may not add to the existing support provided by wives.
The TExT-MED+FANS intervention was designed to be scalable and widely deployed when a
patient with poorly controlled diabetes presents to the emergency department; however this
trade-off between a widely generalizable intervention and the specific context of individual
patients may result in less overall efficacy.

The Transactional Model of Stress and Coping (TMSC)(Lazarus, 1966; Lazarus & Folkman, 1984)
provides a lens to understand the how changes in psychosocial factors (disease related distress,
self-efficacy, fatalism, depression) and coping behaviors (selfcare activities, medication
adherence) that are key in improved disease management and glycemic control, reducing
157
complications of diabetes. In Study 1, the primary appraisal of the threat of diabetes was
reduced in both intervention groups, as evidenced by a decrease in diabetes specific distress.
Intrinsic resources such as self-efficacy and fatalism were also similarly improved between the
groups. Similarly, both the mHealth augmented social support group and the standard social
support group showed similar changes in measures of extrinsic support, measured by general
and disease specific social support. While the FANS intervention was intended to increase the
perceptions of extrinsic capacity to manage diabetes through activating a support person on a
daily basis, both the mHealth augmented group and standard support group improved this
perception of increased external capacity. As the primary and secondary appraisals were
similarly changed between the groups, it is unsurprising that the groups displayed similar
improvements in coping, with similar changes in diabetes self-care activities, medication
adherence, depression and overall glycemic control. Study 2’s qualitative findings elucidate
how these changes in secondary appraisal and coping strategies are related to the TExT-
MED+FANS intervention. The secondary appraisal of a patient’s capacity to meet the challenge
of diabetes management was primarily viewed through an intrinsic lens; the patients perceived
they had increased capacity to make healthy decisions, and through more active coping
strategies, reinforced and increased their appraisal of their capacity to continue doing so. Study
3’s mixed methods sub-group analysis indicates that there are differences in the appraisal and
coping mechanism by patient gender and support relationship type, as well as extrinsic capacity
that was not impacted by the TExT-MED+FANS intervention (access to primary care).

158
While the TMSC encapsulates the internal changes in TExT-MED+FANS participants, it does not
well represent many of the challenges in extrinsic capacity faced by the participants in the
LAC+USC Emergency Department, or why patients may be seeking continuity care in the
Emergency Department or succumbing to ambulatory care sensitive conditions. The most
salient challenge to extrinsic capacity expressed in the qualitative analysis was access to
primary care. Additional models of health behavior must be incorporated into the TMSC to
understand how that access impacts extrinsic capacity and disease management. One model
that has consistently described access to care in multiple iterations since the 1970s is the
Andersen Aday Behavioral Model and Access to Medical Care (Figure 11) (Access to Care
Model).(Andersen, 1995; Andersen, Davidson, & Baumeister, 2007) This Access to Care Model
and more recent emphasis on contextual factors allows for a more nuanced understanding of a
patient’s perceived and actual capacity to realize potential access to care and can clarify the
findings of the TExT-MED+FANS trial that are not explained fully with the TMSC.
Figures 11 & 12: Andersen Aday Behavioral Model and Access to Medical Care from Andersen
and Davidson, Improving Access to Care in America

159

Anderson has elaborated this model to better reflect the different levels of access to care, and
how health policy is often required to improve this access (Figure 12) (Andersen et al., 2007)).
Realized access to care was a persistent barriers to self-management found in the qualitative
analysis of the individual interviews in Study 2, and a strong driving factor in the efficacy of the
trial in the subgroup analysis in Study 3. While less than 5% of patients in the TExT-MED+FANS
intervention lacked medical insurance or any potential access to primary care. Nearly all
patients had potential access to care due to medical insurance that was available to them
because of low-income status, either Medicaid/Medical or LA County funded insurance such as
Healthy Way LA. Despite this near universal potential access, there was a difference in baseline
self-care and intervention efficacy based on realized access to continuity medical care. 23% of
patients in the trial had not had a clinic visit with primary care in the 12 months preceding the
trial; these patients who lacked realized access benefited the most in the intervention period,
but also had the largest rebound when the additional prompting from the TExT-MED
intervention was removed. By incorporating the Access to Care Model concepts of Potential
Access and Realized Access with the TMSC component of secondary appraisal of extrinsic
resources, we can add to the modeling of extrinsic capacity within the TMSC. While the TMSC
models the intervention related changes in primary appraisal, intrinsic capacity and coping, as
well as the relationships between these factors, the additional information processed by the
Access to Care Model allows for a better representation of the experiences of the TExT-MED,
who experience inequitable access to primary care and therefore lack effective access.
Increasing this access to care is a policy level intervention in most situations, and cannot be
160
efficiently managed through an individual patient level intervention in a safety-net emergency
department. Future interventions to improve self-care of chronic diseases out of an emergency
department setting most use this lens of potential vs realized vs effective care to overcome the
challenges faced by these vulnerable patients.

FIGURE 13: Integrated Transactional Model of Stress and Coping and Access to Care Model

Limitations and Strengths:
While this series of studies provides important insights into ED populations who may benefit
most from mHealth and social support based interventions for diabetes self-management,
there are several limitations to the findings which limit both generalizability and potentially
represent biased estimates of effects. The sparse coding structure of the mHealth delivery
platform made it impossible to try measure the dose of the intervention delivered, but did
Stressor
•Disease
Management
Choices
1
o
Appraisal
Psychological
• Distress
(Diabetes
Distress Scale)
Outcomes
Physiologic
•Glycemic
Control (A1C)
•Blood
Pressure
•BMI
2
o
Appraisal
Intrinsic:
•Self-efficacy
•Fatalism
Extrinsic:
•General
Support
•Disease
Specific Social
support
Extrinsic:
•Potential
Access to Care
•Realized
Access to Care
Coping
Problem Solving
Mechanisms
•Diabetes Self
Care Activities
(SDSCA)
•Medication
Adherence
(Wilson 3-
item scale)
•Healthcare
utilization
Emotional
Mechanisms
•Depression
(PHQ-9)
•Fatalism
Psychological
•Quality of Life
(WHO-5)
•PHQ-9
161
allow for the intervention to be delivered over SMS-text messaging, the format most patients
were familiar with. There was significant loss to follow up, even for this highly transient patient
population, reducing the power to detect differences in the FANS arm and the pamphlet social
support arm. However, the estimates of effects of the primary clinical outcome (A1C) and
nearly all mediators were similar, making it unlikely that a true difference for the average
patient was missed. In the qualitative analysis, only intervention patients and supporters were
interviewed due to financial and logistical constraints. The themes identified in Study 2 and the
resultant mixed methods analysis in Study 3 are somewhat limited by not having the
perspective of patients and supporters who only received the paper pamphlets to reinforce
diabetes specific social support. Study 3 is also limited by small sample size for the
methodology; Latent Profile Analysis is typically performed with at least twice the sample size.
However, the inclusion of additional indicators allowed for identification of latent profiles. The
subgroup analysis based on the Latent Profile Analysis and mixed-methods approach are
underpower, but as they are post hoc and exploratory, these findings will need to be repeated
in studies designed for these outcomes.

There are several strengths of this series of studies that add value to the knowledge of mHealth
interventions for chronic disease self-management. The patients and supporters used their own
phones and the platform was designed on a universally compatible SMS text-messaging; these
design decisions maintain the pragmatic nature of this trial and decrease the barrier to
translation to a larger population. The comparison group in the study was not a “control” arm,
but received an active intervention; this was for both ethical reasons, as the TExT-MED patient
162
only curriculum is proven to improve chronic disease management, but also to isolate the
impact of an mHealth activated supporter versus lower cost alternatives. All of these strengths
contribute to the strength of the findings of the study and add to the evidence supporting “low-
tech” mHealth interventions that focus on SMS technology and activating existing patient
resources over complex designs that require a substantial learning curve.

Implications for Future Interventions and Research
This series of studies on an mHealth-based intervention for emergency department patients
with poorly controlled diabetes presents several lessons for future interventions and directions
for future research. In comparing mHealth augmented social support with a traditional paper-
based curriculum, we were able to isolation groups that may benefit from the activation of a
support person multiple times a day. Patients with newly diagnosed diabetes were a group that
particularly benefited from the FANS activation; the massive shift in health behaviors and
support required to manage a new diagnosis of diabetes may override the potential of non-
supportive behaviors such as nagging. The post hoc analysis also identified male patients with
non-spouse supporters as benefiting from the extra support activated by FANS; for those males
with out spousal support, FANS may activate the close connections needed for lifestyle change
and improved glycemic control. There is also an important need to better understand how
support changed throughout the intervention. A mediation analysis examining how the
changes in social support relate to behavioral and clinical outcomes would also inform future
intervention design. A mixed methods study examining the differences in themes between
163
patient supporter dyads who increased support and those who did not increase support would
also inform social support interventions for chronic diseases such as diabetes.
The findings of these studies also inform future directions in mHealth social support trials for
medically vulnerable patients, such as adaptive trial principals and adding lay health worker
support for navigating the health care system. The importance of scaling up or down
intervention intensity to meet the needs of a patient with diabetes, especially those with
depression or other barriers to self-management, could be address in an adaptive trial designs
such as a Sequential Multiple Assignment Randomized Trial (SMART) design.(Almirall, Nahum-
Shani, Sherwood, & Murphy, 2014) SMART trials are advocated for in diseases that require
continuous adaptions to maintain a long-term behavior change and clinical outcome. Applying
adaptive trial designs to a social support intervention could accommodate the need to activate
a supporter without instigating over-support or non-supportive behaviors such as nagging.
Adaptive trial designs could also allow for the integration of lay health workers to connect
those patients who have not transitioned from potential to realized access to care. ED based
navigators and other lay health workers have improved access to care for patients with chronic
disease, and may be a necessary component to develop meaningful interventions in safety-net
populations. (Atzema & Maclagan, 2017; Fitzpatrick et al., 2022; Horny et al., 2017)

164
References
Aceves, B., Ezekiel-Herrera, D., Marino, M., Datta, R., Lucas, J., Giebultowicz, S., & Heintzman, J.
(2022). Disparities in HbA1c testing between aging US Latino and non-Latino white
primary care patients. Prev Med Rep, 26, 101739. doi:10.1016/j.pmedr.2022.101739
Ahmed, A. T., Quinn, V. P., Caan, B., Sternfeld, B., Haque, R., & Van Den Eeden, S. K. (2009).
Generational status and duration of residence predict diabetes prevalence among
Latinos: the California Men's Health Study. BMC Public Health, 9, 392.
doi:10.1186/1471-2458-9-392
Aikens, J. E., Rosland, A. M., & Piette, J. D. (2015). Improvements in illness self-management
and psychological distress associated with telemonitoring support for adults with
diabetes. Prim Care Diabetes, 9(2), 127-134. doi:10.1016/j.pcd.2014.06.003
Aikens, J. E., Trivedi, R., Aron, D. C., & Piette, J. D. (2015). Integrating support persons into
diabetes telemonitoring to improve self-management and medication adherence. J Gen
Intern Med, 30(3), 319-326. doi:10.1007/s11606-014-3101-9
Aikens, J. E., Zivin, K., Trivedi, R., & Piette, J. D. (2014). Diabetes self-management support using
mHealth and enhanced informal caregiving. J Diabetes Complications, 28(2), 171-176.
doi:10.1016/j.jdiacomp.2013.11.008
Albanese, A. M., Huffman, J. C., Celano, C. M., Malloy, L. M., Wexler, D. J., Freedman, M. E., &
Millstein, R. A. (2019). The role of spousal support for dietary adherence among type 2
diabetes patients: a narrative review. Soc Work Health Care, 58(3), 304-323.
doi:10.1080/00981389.2018.1563846
Almirall, D., Nahum-Shani, I., Sherwood, N. E., & Murphy, S. A. (2014). Introduction to SMART
designs for the development of adaptive interventions: with application to weight loss
research. Transl Behav Med, 4(3), 260-274.
American Diabetes, A. (2022). Standards of Care in Diabetes-2023 Abridged for Primary Care
Providers. Clin Diabetes, 41(1), 4-31. doi:10.2337/cd23-as01
Andersen, R. M. (1995). Revisiting the behavioral model and access to medical care: does it
matter? Journal of health and social behavior, 1-10.
Andersen, R. M., Davidson, P. L., & Baumeister, S. E. (2007). Improving access to care in
America. Changing the US health care system: key issues in health services policy and
management. 3a. edición. San Francisco: Jossey-Bass, 3-31.
Anderson, R. J., Freedland, K. E., Clouse, R. E., & Lustman, P. J. (2001). The prevalence of
comorbid depression in adults with diabetes: a meta-analysis. Diabetes Care, 24(6),
1069-1078.
165
Anderson, R. M., Fitzgerald, J. T., Gruppen, L. D., Funnell, M. M., & Oh, M. S. (2003). The
Diabetes Empowerment Scale-Short Form (DES-SF). Diabetes Care, 26(5), 1641-1642.
Arora, S., Abramson, T., Ruiz, R., Coyne, C., DeSantos, R., & Menchine, M. (2013). Describing the
Mobile Health Capacity of Inner City Latino Emergency Department Patients: Are
National Estimates Accurate? Academic Emergency Medicine, 20(S), 187.
Arora, S., Peters, A. L., Burner, E., Lam, C. N., & Menchine, M. (2014). Trial to examine text
message-based mHealth in emergency department patients with diabetes (TExT-MED):
a randomized controlled trial. Ann Emerg Med, 63(6), 745-754.e746.
doi:10.1016/j.annemergmed.2013.10.012
Aschbrenner, K. A., Kruse, G., Gallo, J. J., & Plano Clark, V. L. (2022). Applying mixed methods to
pilot feasibility studies to inform intervention trials. Pilot and Feasibility Studies, 8(1),
217.
Atzema, C. L., & Maclagan, L. C. (2017). The Transition of Care Between Emergency Department
and Primary Care: A Scoping Study. Acad Emerg Med, 24(2), 201-215.
doi:10.1111/acem.13125
Beck, J., Greenwood, D. A., Blanton, L., Bollinger, S. T., Butcher, M. K., Condon, J. E., . . . Wang, J.
(2017). 2017 National Standards for Diabetes Self-Management Education and Support.
Diabetes Spectr, 30(4), 301-314. doi:10.2337/ds17-0067
Beckles, G. L., & Chou, C. F. (2016). Disparities in the Prevalence of Diagnosed Diabetes - United
States, 1999-2002 and 2011-2014. MMWR Morb Mortal Wkly Rep, 65(45), 1265-1269.
doi:10.15585/mmwr.mm6545a4
Bell, A. M., Fonda, S. J., Walker, M. S., Schmidt, V., & Vigersky, R. A. (2012). Mobile phone-based
video messages for diabetes self-care support. J Diabetes Sci Technol, 6(2), 310-319.
Bentler, P. M., & Chou, C.-P. (1987). Practical issues in structural modeling. Sociological
Methods & Research, 16(1), 78-117.
Berg, M., Linden, K., Adolfsson, A., Sparud Lundin, C., & Ranerup, A. (2018). Web-Based
Intervention for Women With Type 1 Diabetes in Pregnancy and Early Motherhood:
Critical Analysis of Adherence to Technological Elements and Study Design. J Med
Internet Res, 20(5), e160. doi:10.2196/jmir.9665
Birtwhistle, R., Green, M. E., Frymire, E., Dahrouge, S., Whitehead, M., Khan, S., . . . Glazier, R.
H. (2017). Hospital admission rates and emergency department use in relation to
glycated hemoglobin in people with diabetes mellitus: a linkage study using electronic
medical record and administrative data in Ontario. CMAJ Open, 5(3), E557-E564.
doi:10.9778/cmajo.20170017
166
Boudreaux, E. D., Bock, B., & O'Hea, E. (2012). When an event sparks behavior change: an
introduction to the sentinel event method of dynamic model building and its application
to emergency medicine. Acad Emerg Med, 19(3), 329-335. doi:10.1111/j.1553-
2712.2012.01291.x
Brew-Sam, N., Chib, A., & Rossmann, C. (2020). Differential influences of social support on app
use for diabetes self-management–a mixed methods approach. BMC Medical
Informatics and Decision Making, 20, 1-13.
Brislin, R. W., & Freimanis, C. (2001). Back-translation. An Encyclopaedia of Translation:
Chinese-English, English-Chinese, 22.
Burner, E., Lam, C. N., DeRoss, R., Kagawa-Singer, M., Menchine, M., & Arora, S. (2018). Using
Mobile Health to Improve Social Support for Low-Income Latino Patients with Diabetes:
A Mixed-Methods Analysis of the Feasibility Trial of TExT-MED + FANS. Diabetes Technol
Ther, 20(1), 39-48. doi:10.1089/dia.2017.0198
Burner, E., Menchine, M. D., Kubicek, K., Robles, M., Kagawa Singer, M., & Arora, S. (2018).
Perceptions of Family Among Low-Income Patients With Diabetes in a Text-Based
Mobile Health Intervention. J Diabetes Sci Technol, 12(6), 1203-1210.
doi:10.1177/1932296818770702
Burner, E. R., Menchine, M. D., Kubicek, K., Robles, M., & Arora, S. (2014). Perceptions of
successful cues to action and opportunities to augment behavioral triggers in diabetes
self-management: qualitative analysis of a mobile intervention for low-income Latinos
with diabetes. J Med Internet Res, 16(1), e25. doi:10.2196/jmir.2881
Carpenter, R. D., Theeke, L. A., Mallow, J. A., Theeke, E., & Gilleland, D. (2017). Relationships
among Distress, Appraisal, Self-Management Behaviors, and Psychosocial Factors in a
Sample of Rural Appalachian Adults with Type 2 Diabetes. Online Journal of Rural
Nursing and Health Care, 17(2), 31.
Charfen, M. A., Ipp, E., Kaji, A. H., Saleh, T., Qazi, M. F., & Lewis, R. J. (2009). Detection of
undiagnosed diabetes and prediabetic states in high-risk emergency department
patients. Acad Emerg Med, 16(5), 394-402. doi:10.1111/j.1553-2712.2009.00374.x
Charmaz, K. (2006). Constructing grounded theory: A practical guide through qualitative
analysis: Sage Publications Limited.
Charmaz, K. (2014). Constructing grounded theory: Sage.
Cherrington, A., Ayala, G. X., Amick, H., Scarinci, I., Allison, J., & Corbie-Smith, G. (2008).
Applying the community health worker model to diabetes management: using mixed
methods to assess implementation and effectiveness. J Health Care Poor Underserved,
19(4), 1044-1059.
167
Cherrington, A., Ayala, G. X., Scarinci, I., & Corbie-Smith, G. (2011). Developing a family-based
diabetes program for Latino immigrants: do men and women face the same barriers?
Fam Community Health, 34(4), 280-290. doi:10.1097/FCH.0b013e31822b5359
Chew, L. D., Griffin, J. M., Partin, M. R., Noorbaloochi, S., Grill, J. P., Snyder, A., . . . Vanryn, M.
(2008). Validation of screening questions for limited health literacy in a large VA
outpatient population. J Gen Intern Med, 23(5), 561-566. doi:10.1007/s11606-008-0520-
5
Chou, C.-P., & Bentler, P. M. (1995). Estimates and tests in structural equation modeling.
Cohen, S. (1988). Psychosocial models of the role of social support in the etiology of physical
disease. Health psychology, 7(3), 269.
Creswell, J. W., Fetters, M. D., Plano Clark, V. L., & Morales, A. (2009). Mixed methods
intervention trials. Mixed methods research for nursing and the health sciences, 161-
180.
Danielson, K. K., Rydzon, B., Nicosia, M., Maheswaren, A., Eisenberg, Y., Lin, J., & Layden, B. T.
(2023). Prevalence of Undiagnosed Diabetes Identified by a Novel Electronic Medical
Record Diabetes Screening Program in an Urban Emergency Department in the US.
JAMA Network Open, 6(1), e2253275-e2253275.
doi:10.1001/jamanetworkopen.2022.53275
Daniulaityte, R. (2004). Making sense of diabetes: cultural models, gender and individual
adjustment to Type 2 diabetes in a Mexican community. Soc Sci Med, 59(9), 1899-1912.
doi:10.1016/j.socscimed.2004.03.001
De Vries, H., Elliott, M. N., Kanouse, D. E., & Teleki, S. S. (2008). Using pooled kappa to
summarize interrater agreement across many items. Field Methods, 20(3), 272-282.
. Dedoose (Version 4.5.91). (2012). Los Angeles, CA: SocioCultural Resources Counsultants, LLC.
Denham, S. A., Ware, L. J., Raffle, H., & Leach, K. (2011). Family inclusion in diabetes education:
a nationwide survey of diabetes educators. Diabetes Educ, 37(4), 528-535.
doi:10.1177/0145721711411312
Diabetes in LAC Adults. (2007). Retrieved from
Diseases, N. I. o. D. a. D. a. K. About NDEP. Retrieved from http://ndep.nih.gov/about-
ndep/index.aspx
Dominguez, K., Penman-Aguilar, A., Chang, M. H., Moonesinghe, R., Castellanos, T., Rodriguez-
Lainz, A., . . . Prevention. (2015). Vital signs: leading causes of death, prevalence of
diseases and risk factors, and use of health services among Hispanics in the United
168
States - 2009-2013. MMWR Morb Mortal Wkly Rep, 64(17), 469-478. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/25950254
Edwards, C., Orellana, E., Rawlings, K., Rodriguez-Pla, M., & Venkatesan, A. (2022). Changes in
Glycemic Control Following Use of a Spanish-Language, Culturally Adapted Diabetes
Program: Retrospective Study. JMIR Form Res, 6(12), e40278. doi:10.2196/40278
Egede, L. E., & Ellis, C. (2010). Development and psychometric properties of the 12-item
diabetes fatalism scale. J Gen Intern Med, 25(1), 61-66. doi:10.1007/s11606-009-1168-5
Fernandez, A., Schillinger, D., Warton, E. M., Adler, N., Moffet, H. H., Schenker, Y., . . . Karter, A.
J. (2011). Language barriers, physician-patient language concordance, and glycemic
control among insured Latinos with diabetes: the Diabetes Study of Northern California
(DISTANCE). J Gen Intern Med, 26(2), 170-176. doi:10.1007/s11606-010-1507-6
Fisher, L., Chesla, C. A., Bartz, R. J., Gilliss, C., Skaff, M. A., Sabogal, F., . . . Lutz, C. P. (1998). The
family and type 2 diabetes: a framework for intervention. Diabetes Educ, 24(5), 599-607.
Fitzgerald, J. T., Davis, W. K., Connell, C. M., Hess, G. E., Funnell, M. M., & Hiss, R. G. (1996).
Development and validation of the Diabetes Care Profile. Eval Health Prof, 19(2), 208-
230. doi:10.1177/016327879601900205
Fitzpatrick, S. L., Papajorgji-Taylor, D., Schneider, J. L., Lindberg, N., Francisco, M., Smith, N., . . .
Hill-Briggs, F. (2022). Bridge to Health/Puente a la Salud: a pilot randomized trial to
address diabetes self-management and social needs among high-risk patients. Transl
Behav Med, 12(7), 783-792. doi:10.1093/tbm/ibac016
Fogg, B. (2009). A behavior model for persuasive design. Paper presented at the Proceedings of
the 4th International Conference on Persuasive Technology, Claremont, California.
Ford, K. J., & Robitaille, A. (2023). How sweet is your love? Disentangling the role of marital
status and quality on average glycemic levels among adults 50 years and older in the
English Longitudinal Study of Ageing. BMJ Open Diabetes Research and Care, 11(1),
e003080.
Galarraga, J. E., Mutter, R., & Pines, J. M. (2015). Costs associated with ambulatory care
sensitive conditions across hospital-based settings. Acad Emerg Med, 22(2), 172-181.
doi:10.1111/acem.12579
Gallegos-Macias, A. R., Macias, S. R., Kaufman, E., Skipper, B., & Kalishman, N. (2003).
Relationship between glycemic control, ethnicity and socioeconomic status in Hispanic
and white non-Hispanic youths with type 1 diabetes mellitus. Pediatr Diabetes, 4(1), 19-
23. doi:10.1034/j.1399-5448.2003.00020.x
Gao, F., Chen, J., Liu, X., Wang, X., Zhao, H., Han, D., . . . Ma, J. (2017). Latent class analysis
suggests four classes of persons with type 2 diabetes mellitus based on complications
169
and comorbidities in Tianjin, China: a cross-sectional analysis. Endocr J, 64(10), 1007-
1016. doi:10.1507/endocrj.EJ17-0199
Gariepy, G., Malla, A., Wang, J., Messier, L., Strychar, I., Lesage, A., & Schmitz, N. (2012). Types
of smokers in a community sample of individuals with Type 2 diabetes: a latent class
analysis. Diabet Med, 29(5), 586-592. doi:10.1111/j.1464-5491.2011.03493.x
Geiss, L. S., Wang, J., Cheng, Y. J., Thompson, T. J., Barker, L., Li, Y., . . . Gregg, E. W. (2014).
Prevalence and incidence trends for diagnosed diabetes among adults aged 20 to 79
years, United States, 1980-2012. JAMA, 312(12), 1218-1226.
doi:10.1001/jama.2014.11494
Glasgow, R. E., Barrera, M., Jr., McKay, H. G., & Boles, S. M. (1999). Social support, self-
management, and quality of life among participants in an internet-based diabetes
support program: a multi-dimensional investigation. Cyberpsychol Behav, 2(4), 271-281.
doi:10.1089/cpb.1999.2.271
Glasgow, R. E., Boles, S. M., McKay, H. G., Feil, E. G., & Barrera, M., Jr. (2003). The D-Net
diabetes self-management program: long-term implementation, outcomes, and
generalization results. Prev Med, 36(4), 410-419.
Gray, K. E., Hoerster, K. D., Reiber, G. E., Bastian, L. A., & Nelson, K. M. (2019). Multiple domains
of social support are associated with diabetes self-management among Veterans.
Chronic Illn, 15(4), 264-275. doi:10.1177/1742395318763489
Gregg, E. W., Zhuo, X., Cheng, Y. J., Albright, A. L., Narayan, K. M., & Thompson, T. J. (2014).
Trends in lifetime risk and years of life lost due to diabetes in the USA, 1985-2011: a
modelling study. Lancet Diabetes Endocrinol, 2(11), 867-874. doi:10.1016/S2213-
8587(14)70161-5
Hales, C. M., Carroll, M. D., Kuo, T., & Simon, P. A. (2019). Diabetes and prediabetes among
adults in Los Angeles County and the United States, 1999–2006 and 2007–2014.
Hansen, U. M., Jones, A., Zander, M., & Willaing, I. (2015). Denmark's comparative position
regarding health status, healthcare provision, self-management and social support:
Diabetes Attitudes, Wishes and Needs second study (DAWN2). Scand J Public Health,
43(2), 111-116. doi:10.1177/1403494814563372
Harkness, E., Macdonald, W., Valderas, J., Coventry, P., Gask, L., & Bower, P. (2010). Identifying
psychosocial interventions that improve both physical and mental health in patients
with diabetes: a systematic review and meta-analysis. Diabetes Care, 33(4), 926-930.
Hay, J. W., Lee, P. J., Jin, H., Guterman, J. J., Gross-Schulman, S., Ell, K., & Wu, S. (2018). Cost-
Effectiveness of a Technology-Facilitated Depression Care Management Adoption Model
in Safety-Net Primary Care Patients with Type 2 Diabetes. Value Health, 21(5), 561-568.
doi:10.1016/j.jval.2017.11.005
170
Heisler, M. (2007). Overview of peer support models to improve diabetes self-management and
clinical outcomes. Diabetes Spectrum, 20(4), 214-221.
Heisler, M., & Piette, J. D. (2005). "I help you, and you help me": facilitated telephone peer
support among patients with diabetes. Diabetes Educ, 31(6), 869-879.
doi:10.1177/0145721705283247
Heisler, M., Vijan, S., Makki, F., & Piette, J. D. (2010). Diabetes control with reciprocal peer
support versus nurse care management: a randomized trial. Ann Intern Med, 153(8),
507-515. doi:10.7326/0003-4819-153-8-201010190-00007
Ho, K., Newton, L., Boothe, A., & Novak-Lauscher, H. (2015). mobile Digital Access to a Web-
enhanced Network (mDAWN): Assessing the Feasibility of Mobile Health Tools for Self-
Management of Type-2 Diabetes. AMIA Annu Symp Proc, 2015, 621-629.
Horigan, G., Davies, M., Findlay-White, F., Chaney, D., & Coates, V. (2017). Reasons why
patients referred to diabetes education programmes choose not to attend: a systematic
review. Diabet Med, 34(1), 14-26. doi:10.1111/dme.13120
Horny, M., Glover, W., Gupte, G., Saraswat, A., Vimalananda, V., & Rosenzweig, J. (2017).
Patient navigation to improve diabetes outpatient care at a safety-net hospital: a
retrospective cohort study. BMC Health Serv Res, 17(1), 759. doi:10.1186/s12913-017-
2700-7
Hou, C., Carter, B., Hewitt, J., Francisa, T., & Mayor, S. (2016). Do Mobile Phone Applications
Improve Glycemic Control (HbA1c) in the Self-management of Diabetes? A Systematic
Review, Meta-analysis, and GRADE of 14 Randomized Trials. Diabetes Care, 39(11),
2089-2095. doi:10.2337/dc16-0346
House, J. S., Umberson, D., & Landis, K. R. (1988). Structures and processes of social support.
Annual review of sociology, 14(1), 293-318.
Hu, J., Wallace, D. C., McCoy, T. P., & Amirehsani, K. A. (2014). A family-based diabetes
intervention for Hispanic adults and their family members. Diabetes Educ, 40(1), 48-59.
doi:10.1177/0145721713512682
Huang, E. S., Liu, J. Y., Moffet, H. H., John, P. M., & Karter, A. J. (2011). Glycemic control,
complications, and death in older diabetic patients: the diabetes and aging study.
Diabetes Care, 34(6), 1329-1336. doi:10.2337/dc10-2377
Igarashi, T., Takai, J., & Yoshida, T. (2005). Gender differences in social network development
via mobile phone text messages: A longitudinal study. Journal of Social and Personal
Relationships, 22(5), 691-713.
Janz, N. K., & Becker, M. H. (1984). The Health Belief Model: a decade later. Health Educ Q,
11(1), 1-47.
171
Janz, N. K., & Becker, M. H. (1985). The Health Belief Model: Applied to Understanding Diabetes
Regimen Compliance. Diabetes Educ, 11(1), 41-47.
Jennings, A., Powell, J., Armstrong, N., Sturt, J., & Dale, J. (2009). A virtual clinic for diabetes
self-management: pilot study. J Med Internet Res, 11(1), e10. doi:10.2196/jmir.1111
Jiang, L., Beals, J., Zhang, L., Mitchell, C. M., Manson, S. M., Acton, K. J., . . . Special Diabetes
Program for Indians Demonstration, P. (2012). Latent class analysis of stages of change
for multiple health behaviors: results from the Special Diabetes Program for Indians
Diabetes Prevention Program. Prev Sci, 13(5), 449-461. doi:10.1007/s11121-011-0272-z
Khosravizade Tabasi, H., Madarshahian, F., Khoshniat Nikoo, M., Hassanabadi, M., &
Mahmoudirad, G. (2014). Impact of family support improvement behaviors on anti
diabetic medication adherence and cognition in type 2 diabetic patients. J Diabetes
Metab Disord, 13(1), 113. doi:10.1186/s40200-014-0113-2
King, D. K., Glasgow, R. E., Toobert, D. J., Strycker, L. A., Estabrooks, P. A., Osuna, D., & Faber, A.
J. (2010). Self-efficacy, problem solving, and social-environmental support are
associated with diabetes self-management behaviors. Diabetes Care, 33(4), 751-753.
doi:10.2337/dc09-1746
Kirk, J. K., Ebert, C. N., Gamble, G. P., & Ebert, C. E. (2013). Social support strategies in adult
patients with diabetes: a review of strategies in the USA and Europe. Expert Review of
Endocrinology & Metabolism, 8(4), 379-389. doi:10.1586/17446651.2013.811895
Koch, T., Kralik, D., & Taylor, J. (2000). Men living with diabetes: minimizing the intrusiveness of
the disease. J Clin Nurs, 9(2), 247-254. doi:10.1046/j.1365-2702.2000.00341.x
Koetsenruijter, J., van Eikelenboom, N., van Lieshout, J., Vassilev, I., Lionis, C., Todorova, E., . . .
Wensing, M. (2016). Social support and self-management capabilities in diabetes
patients: An international observational study. Patient Educ Couns, 99(4), 638-643.
doi:10.1016/j.pec.2015.10.029
Kovacs Burns, K., Nicolucci, A., Holt, R. I., Willaing, I., Hermanns, N., Kalra, S., . . . Group, D. S.
(2013). Diabetes Attitudes, Wishes and Needs second study (DAWN2): cross-national
benchmarking indicators for family members living with people with diabetes. Diabet
Med, 30(7), 778-788. doi:10.1111/dme.12239
Kroenke, K., Spitzer, R. L., Williams, J. B., & Lowe, B. (2010). The Patient Health Questionnaire
Somatic, Anxiety, and Depressive Symptom Scales: a systematic review. Gen Hosp
Psychiatry, 32(4), 345-359. doi:10.1016/j.genhosppsych.2010.03.006
Langford, C. P., Bowsher, J., Maloney, J. P., & Lillis, P. P. (1997). Social support: a conceptual
analysis. J Adv Nurs, 25(1), 95-100.
172
Lari, H., Tahmasebi, R., & Noroozi, A. (2018). Effect of electronic education based on health
promotion model on physical activity in diabetic patients. Diabetes Metab Syndr, 12(1),
45-50. doi:10.1016/j.dsx.2017.08.013
Lasater, L. M., Davidson, A. J., Steiner, J. F., & Mehler, P. S. (2001). Glycemic control in English-
vs Spanish-speaking Hispanic patients with type 2 diabetes mellitus. Arch Intern Med,
161(1), 77-82. doi:10.1001/archinte.161.1.77
Lawrence, J. M., Mayer-Davis, E. J., Reynolds, K., Beyer, J., Pettitt, D. J., D'Agostino, R. B., Jr., . . .
Hamman, R. F. (2009). Diabetes in Hispanic American youth: prevalence, incidence,
demographics, and clinical characteristics: the SEARCH for Diabetes in Youth Study.
Diabetes Care, 32 Suppl 2, S123-132. doi:10.2337/dc09-S204
Lazarus, R. S. (1966). Psychological stress and the coping process.
Lazarus, R. S., & Folkman, S. (1984). Stress, appraisal, and coping: Springer publishing company.
Levy, M., Deschenes, S. S., Burns, R. J., Elgendy, R., & Schmitz, N. (2019). Trajectories of social
support in adults with type 2 diabetes: Associations with depressive symptoms and
functional disability. Int J Geriatr Psychiatry, 34(3), 480-487. doi:10.1002/gps.5042
Lewin, A. B., Geffken, G. R., Heidgerken, A. D., Duke, D. C., Novoa, W., Williams, L. B., & Storch,
E. A. (2005). The diabetes family behavior checklist: A psychometric evaluation. Journal
of Clinical Psychology in Medical Settings, 12(4), 315-322.
Li, R., Shrestha, S. S., Lipman, R., Burrows, N. R., Kolb, L. E., Rutledge, S., . . . Prevention. (2014).
Diabetes self-management education and training among privately insured persons with
newly diagnosed diabetes--United States, 2011-2012. MMWR Morb Mortal Wkly Rep,
63(46), 1045-1049. Retrieved from https://www.ncbi.nlm.nih.gov/pubmed/25412060
Linden, K., Berg, M., Adolfsson, A., & Sparud-Lundin, C. (2018). Person-centred, web-based
support in pregnancy and early motherhood for women with Type 1 diabetes mellitus: a
randomized controlled trial. Diabet Med, 35(2), 232-241. doi:10.1111/dme.13552
Livingston, G. (2011). Latinos and digital technology, 2010. Pew Hispanic Center, web page,
February. Retrieved from https://www.pewresearch.org/wp-
content/uploads/sites/5/reports/134.pdf
Lo, Y., Mendell, N. R., & Rubin, D. B. (2001). Testing the number of components in a normal
mixture. Biometrika, 88(3), 767-778.
Lopez, M. H., Gonzalez-Barrera, A., & Patten, E. (2013). Closing the digital divide: Latinos and
technology adoption.
Magill, R. S. (1993). Focus groups, program evaluation, and the poor. J. Soc. & Soc. Welfare, 20,
103.
173
Mayberry, L. S., Berg, C. A., Greevy, R. A., Nelson, L. A., Bergner, E. M., Wallston, K. A., . . . Elasy,
T. A. (2021). Mixed-Methods Randomized Evaluation of FAMS: A Mobile Phone-
Delivered Intervention to Improve Family/Friend Involvement in Adults' Type 2 Diabetes
Self-Care. Ann Behav Med, 55(2), 165-178. doi:10.1093/abm/kaaa041
Mayberry, L. S., Berg, C. A., Harper, K. J., & Osborn, C. Y. (2016). The Design, Usability, and
Feasibility of a Family-Focused Diabetes Self-Care Support mHealth Intervention for
Diverse, Low-Income Adults with Type 2 Diabetes. J Diabetes Res, 2016, 7586385.
doi:10.1155/2016/7586385
Mayberry, L. S., & Osborn, C. Y. (2012). Family support, medication adherence, and glycemic
control among adults with type 2 diabetes. Diabetes Care, 35(6), 1239-1245.
doi:10.2337/dc11-2103
Mayberry, L. S., & Osborn, C. Y. (2014). Family involvement is helpful and harmful to patients'
self-care and glycemic control. Patient Educ Couns, 97(3), 418-425.
doi:10.1016/j.pec.2014.09.011
Mayberry, L. S., Rothman, R. L., & Osborn, C. Y. (2014). Family Members' Obstructive Behaviors
Appear to Be More Harmful Among Adults With Type 2 Diabetes and Limited Health
Literacy. J Health Commun, 19(sup2), 132-143. doi:10.1080/10810730.2014.938840
McKay, H. G., Glasgow, R. E., Feil, E. G., Boles, S. M., & Barrera Jr, M. (2002). Internet-based
diabetes self-management and support: Initial outcomes from the diabetes network
project. Rehabilitation Psychology, 47(1), 31.
McManus, R., Miller, D., Mottola, M., Giroux, I., & Donovan, L. (2018). Translating Healthy Living
Messages to Postpartum Women and Their Partners After Gestational Diabetes (GDM):
Body Habitus, A1C, Lifestyle Habits, and Program Engagement Results From the Families
Defeating Diabetes (FDD) Randomized Trial. Am J Health Promot, 32(6), 1438-1446.
doi:10.1177/0890117117738210
Menchine, M., Marzec, K., Solomon, T., & Arora, S. (2013). Fragile health status of Latino
patients with diabetes seen in the emergency department of an urban, safety-net
hospital. Ethn Dis, 23(1), 49-55.
Moghani Lankarani, M., & Assari, S. (2017). Diabetes, hypertension, obesity, and long-term risk
of renal disease mortality: Racial and socioeconomic differences. Journal of diabetes
investigation, 8(4), 590-599.
Mohebi, S., Parham, M., Sharifirad, G., Gharlipour, Z., Mohammadbeigi, A., & Rajati, F. (2018).
Relationship between perceived social support and self-care behavior in type 2
diabetics: A cross-sectional study. J Educ Health Promot, 7, 48.
doi:10.4103/jehp.jehp_73_17
Muthén, L. K., & Muthén, B. O. (2019). Mplus (Version 8.3). Los Angeles, CA: Muthen & Muthen.
174
Nagelkerk, J., Reick, K., & Meengs, L. (2006). Perceived barriers and effective strategies to
diabetes self-management. J Adv Nurs, 54(2), 151-158. doi:10.1111/j.1365-
2648.2006.03799.x
Narayan, K. M., Boyle, J. P., Thompson, T. J., Sorensen, S. W., & Williamson, D. F. (2003).
Lifetime risk for diabetes mellitus in the United States. JAMA, 290(14), 1884-1890.
doi:10.1001/jama.290.14.1884
Nathan, D. M., McGee, P., Steffes, M. W., Lachin, J. M., & Group, D. E. R. (2014). Relationship of
glycated albumin to blood glucose and HbA1c values and to retinopathy, nephropathy,
and cardiovascular outcomes in the DCCT/EDIC study. Diabetes, 63(1), 282-290.
doi:10.2337/db13-0782
Norbeck, J. S., Lindsey, A. M., & Carrieri, V. L. (1981). The development of an instrument to
measure social support. Nursing research, 30(5), 264-269.
NS Endler, J. P. (1990). Multidimensional assessment of coping: A critical evaluation. Journal of
Personality and Social Psychology, 58(5), 11.
Oberski, D. (2016). Mixture models: Latent profile and latent class analysis. In Modern statistical
methods for HCI (pp. 275-287): Springer.
Pal, K., Dack, C., Ross, J., Michie, S., May, C., Stevenson, F., . . . Murray, E. (2018). Digital Health
Interventions for Adults With Type 2 Diabetes: Qualitative Study of Patient Perspectives
on Diabetes Self-Management Education and Support. J Med Internet Res, 20(2), e40.
doi:10.2196/jmir.8439
Peyrot, M., Egede, L., Funnell, M., Hsu, W., Ruggiero, L., & Stuckey, H. (2018). US Ethnic Group
Differences in Family Member Support for People With Diabetes in the 2nd Diabetes
Attitudes, Wishes and Needs (DAWN2) Study. Diabetes Educ, 44(3), 249-259.
doi:10.1177/0145721718770767
Peyrot, M., Rubin, R. R., Funnell, M. M., & Siminerio, L. M. (2009). Access to diabetes self-
management education: results of national surveys of patients, educators, and
physicians. Diabetes Educ, 35(2), 246-248, 252-246, 258-263.
doi:10.1177/0145721708329546
Piette, J. D., Kerr, E., Richardson, C., & Heisler, M. (2008). Veterans Affairs research on health
information technologies for diabetes self-management support. In: SAGE Publications.
Piette, J. D., Marinec, N., Gallegos-Cabriales, E. C., Gutierrez-Valverde, J. M., Rodriguez-Saldana,
J., Mendoz-Alevares, M., & Silveira, M. J. (2013). Spanish-speaking patients' engagement
in interactive voice response (IVR) support calls for chronic disease self-management:
data from three countries. J Telemed Telecare, 19(2), 89-94.
doi:10.1177/1357633x13476234
175
Piette, J. D., Marinec, N., Janda, K., Morgan, E., Schantz, K., Yujra, A. C., . . . Aikens, J. E. (2016).
Structured Caregiver Feedback Enhances Engagement and Impact of Mobile Health
Support: A Randomized Trial in a Lower-Middle-Income Country. Telemed J E Health,
22(4), 261-268. doi:10.1089/tmj.2015.0099
Piette, J. D., Mendoza-Avelares, M. O., Ganser, M., Mohamed, M., Marinec, N., & Krishnan, S.
(2011). A preliminary study of a cloud-computing model for chronic illness self-care
support in an underdeveloped country. Am J Prev Med, 40(6), 629-632.
doi:10.1016/j.amepre.2011.02.014
Piette, J. D., Rosland, A. M., Marinec, N. S., Striplin, D., Bernstein, S. J., & Silveira, M. J. (2013).
Engagement with automated patient monitoring and self-management support calls:
experience with a thousand chronically ill patients. Med Care, 51(3), 216-223.
doi:10.1097/MLR.0b013e318277ebf8
Polonsky, W. H., Fisher, L., Earles, J., Dudl, R. J., Lees, J., Mullan, J., & Jackson, R. A. (2005).
Assessing psychosocial distress in diabetes: development of the diabetes distress scale.
Diabetes Care, 28(3), 626-631. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/15735199
Polonsky, W. H., Fisher, L., Hessler, D., & Johnson, N. (2016). Emotional Distress in the Partners
of Type 1 Diabetes Adults: Worries About Hypoglycemia and Other Key Concerns.
Diabetes Technol Ther, 18(5), 292-297. doi:10.1089/dia.2015.0451
Powell, J., Jennings, A., Armstrong, N., Sturt, J., & Dale, J. (2009). Pilot study of a virtual diabetes
clinic: satisfaction and usability. J Telemed Telecare, 15(3), 150-152.
doi:10.1258/jtt.2009.003014
Powers, M. A., Bardsley, J., Cypress, M., Duker, P., Funnell, M. M., Fischl, A. H., . . . Vivian, E.
(2015). Diabetes self-management education and support in type 2 diabetes: a joint
position statement of the American Diabetes Association, the American Association of
Diabetes Educators, and the Academy of Nutrition and Dietetics. Diabetes Educ, 41(4),
417. doi:10.1177/0145721715588904
Powers, M. A., Bardsley, J. K., Cypress, M., Funnell, M. M., Harms, D., Hess-Fischl, A., . . .
Uelmen, S. (2020). Diabetes self-management education and support in adults with type
2 diabetes: A consensus report of the American Diabetes Association, the Association of
Diabetes Care and Education Specialists, the Academy of Nutrition and Dietetics, the
American Academy of Family Physicians, the American Academy of PAs, the American
Association of Nurse Practitioners, and the American Pharmacists Association. JAAPA,
33(7), 1-20. doi:10.1097/01.JAA.0000668828.47294.2a
Presley, C., Agne, A., Shelton, T., Oster, R., & Cherrington, A. (2020). Mobile-enhanced peer
support for African Americans with type 2 diabetes: a randomized controlled trial. J Gen
Intern Med, 35, 2889-2896.
176
QuickFacts: Los Angeles County, California. (2022, 2022). United States Census Bureau.
Retrieved from
https://www.census.gov/quickfacts/fact/table/losangelescountycalifornia/RHI625221#R
HI625221
Quinn, C. C., Clough, S. S., Minor, J. M., Lender, D., Okafor, M. C., & Gruber-Baldini, A. (2008).
WellDoc™ mobile diabetes management randomized controlled trial: change in clinical
and behavioral outcomes and patient and physician satisfaction. Diabetes Technol Ther,
10(3), 160-168.
Quinn, C. C., Shardell, M. D., Terrin, M. L., Barr, E. A., Ballew, S. H., & Gruber-Baldini, A. L.
(2011). Cluster-randomized trial of a mobile phone personalized behavioral intervention
for blood glucose control. Diabetes Care, 34(9), 1934-1942.
Ramirez, M., & Wu, S. (2017). Phone Messaging to Prompt Physical Activity and Social Support
Among Low-Income Latino Patients With Type 2 Diabetes: A Randomized Pilot Study.
JMIR Diabetes, 2(1), e8. doi:10.2196/diabetes.7063
Ramkisson, S., Pillay, B. J., & Sibanda, W. (2017). Social support and coping in adults with type 2
diabetes. Afr J Prim Health Care Fam Med, 9(1), e1-e8. doi:10.4102/phcfm.v9i1.1405
Reese, P. P., Kessler, J. B., Doshi, J. A., Friedman, J., Mussell, A. S., Carney, C., . . . Volpp, K.
(2016). Two Randomized Controlled Pilot Trials of Social Forces to Improve Statin
Adherence among Patients with Diabetes. J Gen Intern Med, 31(4), 402-410.
doi:10.1007/s11606-015-3540-y
Rinker, J., Dickinson, J. K., Litchman, M. L., Williams, A. S., Kolb, L. E., Cox, C., & Lipman, R. D.
(2018). The 2017 Diabetes Educator and the Diabetes Self-Management Education
National Practice Survey. Diabetes Educ, 44(3), 260-268.
doi:10.1177/0145721718765446
Roddy, M. K., Nelson, L. A., Greevy, R. A., & Mayberry, L. S. (2022). Changes in family
involvement occasioned by FAMS mobile health intervention mediate changes in
glycemic control over 12 months. J Behav Med, 45(1), 28-37. doi:10.1007/s10865-021-
00250-w
Rosano, A., Loha, C. A., Falvo, R., van der Zee, J., Ricciardi, W., Guasticchi, G., & de Belvis, A. G.
(2013). The relationship between avoidable hospitalization and accessibility to primary
care: a systematic review. Eur J Public Health, 23(3), 356-360.
doi:10.1093/eurpub/cks053
Rosland, A. M., Heisler, M., Choi, H. J., Silveira, M. J., & Piette, J. D. (2010). Family influences on
self-management among functionally independent adults with diabetes or heart failure:
do family members hinder as much as they help? Chronic Illn, 6(1), 22-33.
doi:10.1177/1742395309354608
177
Sahin, C., & Naylor, P.-J. (2017). Mixed-methods research in diabetes management via mobile
health technologies: a scoping review. JMIR Diabetes, 2(1), e6667.
Samuel-Hodge, C. D., Watkins, D. C., Rowell, K. L., & Hooten, E. G. (2008). Coping styles, well-
being, and self-care behaviors among African Americans with type 2 diabetes. Diabetes
Educ, 34(3), 501-510. doi:10.1177/0145721708316946
Sarkar, U., Fisher, L., & Schillinger, D. (2006). Is self-efficacy associated with diabetes self-
management across race/ethnicity and health literacy? Diabetes Care, 29(4), 823-829.
Scollan-Koliopoulos, M., Schechter, C. B., Caban, A., & Walker, E. A. (2012). Hispanic
acculturation, psychosocial functioning, and routine support for diabetes self-
management. Diabetes Educ, 38(5), 715-722. doi:10.1177/0145721712450924
Seto, E., Leonard, K. J., Cafazzo, J. A., Barnsley, J., Masino, C., & Ross, H. J. (2012). Mobile
phone-based telemonitoring for heart failure management: a randomized controlled
trial. J Med Internet Res, 14(1), e31. doi:10.2196/jmir.1909
Shah, B. M., Gupchup, G. V., Borrego, M. E., Raisch, D. W., & Knapp, K. K. (2012). Depressive
symptoms in patients with type 2 diabetes mellitus: do stress and coping matter? Stress
Health, 28(2), 111-122. doi:10.1002/smi.1410
Skrepnek, G. H., Mills, J. L., & Armstrong, D. G. (2014). Foot-in-wallet disease: tripped up by
"cost-saving" reductions? Diabetes Care, 37(9), e196-197. doi:10.2337/dc14-0079
Spanakis, E. K., & Golden, S. H. (2013). Race/ethnic difference in diabetes and diabetic
complications. Curr Diab Rep, 13(6), 814-823. doi:10.1007/s11892-013-0421-9
Spencer, M. S., Kieffer, E. C., Sinco, B., Piatt, G., Palmisano, G., Hawkins, J., . . . Heisler, M.
(2018). Outcomes at 18 Months From a Community Health Worker and Peer Leader
Diabetes Self-Management Program for Latino Adults. Diabetes Care, 41(7), 1414-1422.
doi:10.2337/dc17-0978
Spencer-Bonilla, G., Ponce, O. J., Rodriguez-Gutierrez, R., Alvarez-Villalobos, N., Erwin, P. J.,
Larrea-Mantilla, L., . . . Montori, V. M. (2017). A systematic review and meta-analysis of
trials of social network interventions in type 2 diabetes. BMJ Open, 7(8), e016506.
doi:10.1136/bmjopen-2017-016506
Spurk, D., Hirschi, A., Wang, M., Valero, D., & Kauffeld, S. (2020). Latent profile analysis: A
review and “how to” guide of its application within vocational behavior research.
Journal of Vocational Behavior, 120, 103445.
doi:https://doi.org/10.1016/j.jvb.2020.103445
StataCorp. (2021). Stata Statistical Software: Release 17. College Station, TX: StataCorp.
178
Steed, L., Cooke, D., & Newman, S. (2003). A systematic review of psychosocial outcomes
following education, self-management and psychological interventions in diabetes
mellitus. Patient Educ Couns, 51(1), 5-15. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/12915275
Steinberg, J. R., Yeh, C., Jackson, J., Saber, R., Niznik, C. M., Leziak, K., & Yee, L. M. (2022).
Optimizing Engagement in an mHealth Intervention for Diabetes Support During
Pregnancy: the Role of Baseline Patient Health and Behavioral Characteristics. J Diabetes
Sci Technol, 16(6), 1466-1472. doi:10.1177/19322968211035441
Stevens, C. D., Schriger, D. L., Raffetto, B., Davis, A. C., Zingmond, D., & Roby, D. H. (2014).
Geographic clustering of diabetic lower-extremity amputations in low-income regions of
California. Health Aff (Millwood), 33(8), 1383-1390. doi:10.1377/hlthaff.2014.0148
Strawbridge, L. M., Lloyd, J. T., Meadow, A., Riley, G. F., & Howell, B. L. (2015). Use of
Medicare's Diabetes Self-Management Training Benefit. Health Educ Behav, 42(4), 530-
538. doi:10.1177/1090198114566271
Strom, J. L., & Egede, L. E. (2012). The impact of social support on outcomes in adult patients
with type 2 diabetes: a systematic review. Curr Diab Rep, 12(6), 769-781.
doi:10.1007/s11892-012-0317-0
Strom, J. L., & Egede, L. E. (2012). The Impact of Social Support on Outcomes in Adult Patients
with Type 2 Diabetes: A Systematic Review. Curr Diab Rep, 12(6), 769-781.
doi:10.1007/s11892-012-0317-0
Suh, S., Jean, C., Koo, M., Lee, S. Y., Cho, M. J., Sim, K. H., . . . Kim, J. H. (2014). A randomized
controlled trial of an internet-based mentoring program for type 1 diabetes patients
with inadequate glycemic control. Diabetes Metab J, 38(2), 134-142.
doi:10.4093/dmj.2014.38.2.134
Tanaka, R., Shibayama, T., Sugimoto, K., & Hidaka, K. (2020). Diabetes self-management
education and support for adults with newly diagnosed type 2 diabetes mellitus: A
systematic review and meta-analysis of randomized controlled trials. Diabetes Res Clin
Pract, 169, 108480. doi:10.1016/j.diabres.2020.108480
Tang, T. S., Ayala, G. X., Cherrington, A., & Rana, G. (2011). A Review of Volunteer-Based Peer
Support Interventions in Diabetes. Diabetes Spectrum, 24(2), 85-98.
doi:10.2337/diaspect.24.2.85
Tang, T. S., Funnell, M. M., Sinco, B., Spencer, M. S., & Heisler, M. (2015). Peer-Led,
Empowerment-Based Approach to Self-Management Efforts in Diabetes (PLEASED): A
Randomized Controlled Trial in an African American Community. Ann Fam Med, 13
Suppl 1, S27-35. doi:10.1370/afm.1819
179
Tein, J.-Y., Coxe, S., & Cham, H. (2013). Statistical Power to Detect the Correct Number of
Classes in Latent Profile Analysis. Structural Equation Modeling: A Multidisciplinary
Journal, 20(4), 640-657. doi:10.1080/10705511.2013.824781
Teufel-Shone, N. I., Drummond, R., & Rawiel, U. (2005). Developing and adapting a family-based
diabetes program at the U.S.-Mexico border. Prev Chronic Dis, 2(1), A20.
Thompson, J. R., Horton, C., & Flores, C. (2007). Advancing diabetes self-management in the
Mexican American population: a community health worker model in a primary care
setting. Diabetes Educ, 33 Suppl 6, 159s-165s. doi:10.1177/0145721707304077
Tong, A., Sainsbury, P., & Craig, J. (2007). Consolidated criteria for reporting qualitative research
(COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care,
19(6), 349-357. doi:10.1093/intqhc/mzm042
Toobert, D. J., & Glasgow, R. E. (1994). Assessing diabetes self-management: the summary of
diabetes self-care activities questionnaire. Handbook of psychology and diabetes: A
guide to psychological measurement in diabetes research and practice, 351, 75.
Toobert, D. J., Hampson, S. E., Glasgow, & RE. (2000). The summary of diabetes self-care
activities measure: results from 7 studies and a revised scale. Diabetes Care, 23(7), 943-
950.
Topp, C. W., Ostergaard, S. D., Sondergaard, S., & Bech, P. (2015). The WHO-5 Well-Being Index:
a systematic review of the literature. Psychother Psychosom, 84(3), 167-176.
doi:10.1159/000376585
Torres Mosst, J. K., Tony. (2017). Diabetes and Prediabetes: Screening Guidelines and the
Expansion of the National Diabetes Prevention Program in Los Angeles County. Rx for
Prevention, 7(2). Retrieved from http://rx.ph.lacounty.gov/RxDiabetes0717
Treacy-Abarca, S., Mercado, J., Serrano, J., Gonzalez, J., Menchine, M., Arora, S., . . . Burner, E.
(2022). Technological Proficiencies, Engagement, and Practical Considerations for
mHealth Programs at an Urban Safety-Net Hospital Emergency Departments: Data
Analysis. JMIR Diabetes, 7(2), e23641.
Two Feathers, J., Kieffer, E. C., Palmisano, G., Anderson, M., Sinco, B., Janz, N., . . . James, S. A.
(2005). Racial and Ethnic Approaches to Community Health (REACH) Detroit partnership:
improving diabetes-related outcomes among African American and Latino adults. Am J
Public Health, 95(9), 1552-1560. doi:10.2105/ajph.2005.066134
UKPDSGroup, U. U. P. D. S. G. (1998). Intensive blood-glucose control with sulphonylureas or
insulin compared with conventional treatment and risk of complications in patients with
type 2 diabetes (UKPDS 33). . Lancet, 352(9131), 837-853. Retrieved from
https://www.ncbi.nlm.nih.gov/pubmed/9742976
180
Vahedparast, H., Mohammadi, E., Ahmadi, F., & Farhadi, A. (2018). The role of social support in
adherence to treatment regimens: Experiences of patients with chronic diseases.
Medical-Surgical Nursing Journal, 7(1).
van Dam, H. A., van der Horst, F. G., Knoops, L., Ryckman, R. M., Crebolder, H. F., & van den
Borne, B. H. (2005). Social support in diabetes: a systematic review of controlled
intervention studies. Patient Educ Couns, 59(1), 1-12. doi:10.1016/j.pec.2004.11.001
van Loenen, T., van den Berg, M. J., Westert, G. P., & Faber, M. J. (2014). Organizational aspects
of primary care related to avoidable hospitalization: a systematic review. Fam Pract,
31(5), 502-516. doi:10.1093/fampra/cmu053
Venermo, M., Manderbacka, K., Ikonen, T., Keskimaki, I., Winell, K., & Sund, R. (2013).
Amputations and socioeconomic position among persons with diabetes mellitus, a
population-based register study. BMJ Open, 3(4). doi:10.1136/bmjopen-2012-002395
Vijayan, M., Radhakrishnan, S., Abraham, G., Mathew, M., Sampathkumar, K., & Mancha, N. P.
(2016). Diabetic kidney disease patients on hemodialysis: a retrospective survival
analysis across different socioeconomic groups. NDT Plus, 9(6), 833-838.
Vorderstrasse, A., Lewinski, A., Melkus, G. D., & Johnson, C. (2016). Social Support for Diabetes
Self-Management via eHealth Interventions. Curr Diab Rep, 16(7), 56.
doi:10.1007/s11892-016-0756-0
Walker, R. J., Gebregziabher, M., Martin-Harris, B., & Egede, L. E. (2014). Relationship between
social determinants of health and processes and outcomes in adults with type 2
diabetes: validation of a conceptual framework. BMC Endocr Disord, 14(1), 82.
doi:10.1186/1472-6823-14-82
Wang, X., Chen, J., Liu, X., Gao, F., Zhao, H., Han, D., . . . Ma, J. (2017). Identifying Patterns of
Lifestyle Behaviors among People with Type 2 Diabetes in Tianjin, China: A Latent Class
Analysis. Diabetes Ther, 8(6), 1379-1392. doi:10.1007/s13300-017-0327-7
Weiler, D. M., & Crist, J. D. (2009). Diabetes self-management in a Latino social environment.
Diabetes Educ, 35(2), 285-292. doi:10.1177/0145721708329545
Wharam, J. F., Zhang, F., Eggleston, E. M., Lu, C. Y., Soumerai, S. B., & Ross-Degnan, D. (2018).
Effect of High-Deductible Insurance on High-Acuity Outcomes in Diabetes: A Natural
Experiment for Translation in Diabetes (NEXT-D) Study. Diabetes Care, 41(5), 940-948.
doi:10.2337/dc17-1183
Wilson, I. B., Fowler, F. J., Jr., Cosenza, C. A., Michaud, J., Bentkover, J., Rana, A., . . . Rogers, W.
H. (2014). Cognitive and field testing of a new set of medication adherence self-report
items for HIV care. AIDS Behav, 18(12), 2349-2358. doi:10.1007/s10461-013-0610-1

181

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Using mobile health to improve social support for low-income Latino patients with diabetes: a randomized mixed methods feasibility trial of TExT-MED FANS

PDF

The association between Echocardiography use in the Intensive Care Unit and 28-day mortality among patients with septic shock: a systematic review

PDF

A theoretical framework and mixed-methods investigation of document status as a social determinant of emergency department utilization…

PDF

Exploring geography and multi-level factors associated with disparities in HPV vaccination among Latino adolescents in Los Angeles County

PDF

Depression severity, self-care behaviors, and self-reported diabetes symptoms and daily functioning among low-income patients receiving depression care

PDF

Smoke-free housing policies and secondhand smoke exposure in low income multiunit housing in Los Angeles County

PDF

Planning care with the patient in the room: a patient-focused approach to reducing heart failure readmissions

PDF

Model development of breast cancer detection and staging via rare event enumeration from a liquid biopsy: a retrospective descriptive clinical research study

PDF

Role transitions, past life events, and their associations with multiple categories of substance use among emerging adults

PDF

Anxiety symptoms and nicotine use among adolescents and young adults

PDF

A cross-sectional study of the association of PTH on bone quality across levels of propionic acid among adult patients with uremia

PDF

Primary care physicians' experiences working within the patient-portal to improve the quality of patient care

PDF

Clinical outcomes of allogeneic hematopoietic stem cell transplant in acute lymphoblastic leukemia patients: a quality improvement project and systematic review meta-analysis

PDF

Examining tobacco regulation opinions and policy acceptance among key opinion leaders and tobacco retailers in low socioeconomic status African American, Hispanic, and non-Hispanic White communities

PDF

Using a human factors engineering perspective to design and evaluate communication and information technology tools to support depression care and physical activity behavior change among low-inco...

PDF

A series of longitudinal analyses of patient reported outcomes to further the understanding of care-management of comorbid diabetes and depression in a safety-net healthcare system

PDF

The effect of renal function on toxicity of E7389 (eribulin) among patients with bladder cancer

PDF

Barriers to surgery for in low- and middle-income countries: a cross-sectional study of cleft lip and cleft palate patients in Vietnam

PDF

Chronic eye disease epidemiology in the multiethnic ophthalmology cohorts of California study

Asset Metadata

Creator Burner, Elizabeth Rhea Erwin (author)

Core Title Addressing unmet needs and harnessing social support to improve diabetes self-care among low-income, urban emergency department patients with diabetes

School Keck School of Medicine

Degree Doctor of Philosophy

Degree Program Preventive Medicine

Degree Conferral Date 2023-05

Publication Date 05/23/2023

Defense Date 05/02/2023

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag continuity care,Diabetes,emergency care,mHealth,OAI-PMH Harvest,social support

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Baezconde-Garbanati, Lourdes (committee chair), Bluthenthal, Ricky (committee member), Mack, Wendy (committee member), Patino Sutton, Cecilia (committee member), Wu, Shinyi (committee member)

Creator Email eburner@usc.edu,elizabeth.Burner@med.usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113134822

Unique identifier UC113134822

Identifier etd-BurnerEliz-11884.pdf (filename)

Legacy Identifier etd-BurnerEliz-11884

Document Type Dissertation

Format theses (aat)

Rights Burner, Elizabeth Rhea Erwin

Internet Media Type application/pdf

Type texts

Source 20230524-usctheses-batch-1048 (batch), University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the author, as the original true and official version of the work, but does not grant the reader permission to use the work if the desired use is covered by copyright. It is the author, as rights holder, who must provide use permission if such use is covered by copyright.

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA

Repository Email cisadmin@lib.usc.edu