University of Southern California Dissertations and Theses

The cost of opioid use in high-risk hospitalized infants

(USC Thesis Other)

The cost of opioid use in high-risk hospitalized infants

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content The Cost of Opioid Use in High-Risk Hospitalized Infants
By
Olivia A. Keane
A Thesis Presented to the
FACULTY OF THE UNIVERSITY OF SOUTHERN CALIFORNIA
KECK SCHOOL OF MEDICINE
In Partial Fulfillment of the
Requirements for the Degree
MASTER OF SCIENCE
(CLINICAL, BIOMEDICAL, AND TRANSLATIONAL INVESTIGATIONS)
May 2024

ii
DEDICATION
To my family, co-residents, mentors, patients, and partner-in-crime Christopher.

iii
ACKNOWLEDGEMENTS
I would like to thank my mentor and thesis committee member Dr. Lorraine Kelley-Quon
for allowing me to join her research, for being a fantastic and supportive primary investigator,
and for introducing me to the Master of Science of Clinical, Biomedical, and Translational
Investigations program.
I would like to thank my committee member Dr. Cameron Kaplan for lending his
expertise and never-ending patience especially when advising on the statistical analysis for this
project. In addition, I would like to thank my final committee member Dr. Cecilia Patino-Sutton
for her kind help and support and for bringing so much fun and joy to our clinical translational
research courses.
Finally, I would like to thank the remaining members of the Health Outcomes and Policy
Effects (HOPE) Lab at Children’s Hospital Los Angeles, with special mention to those that
provided significant assistance with this thesis manuscript including: Shadassa Ourshalimian, Dr.
Cynthia Gong, Dr. Ashwini Lakshmanan, Dr. Susan Hintz, Dr. Henry C. Lee, Madeleine Ing, Dr.
Rabab Barq, and Dr. Nam Nguyen, MD.

iv
TABLE OF CONTENTS
DEDICATION.................................................................................................................................ii
ACKNOWLEDGEMENTS............................................................................................................iii
LIST OF TABLES...........................................................................................................................v
ABSTRACT...................................................................................................................................vi
CHAPTER ONE: INTRODUCTION..............................................................................................1
CHAPTER TWO: METHODS........................................................................................................2
CHAPTER THREE: STATISTICAL ANALYSIS.........................................................................6
CHAPTER FOUR: RESULTS........................................................................................................8
CHAPTER FIVE: DISCUSSION..................................................................................................11
CHAPTER SIX: CONCLUSION..................................................................................................15
REFERENCES..............................................................................................................................16
TABLES........................................................................................................................................21
FIGURES.......................................................................................................................................25

v
LIST OF TABLES
Table 1: Demographic and clinical factors of the total cohort and stratified by cumulative
days of opioid use quartiles…….....………………………………………………………….21
Table 2: Gestational age, birthweight, length of stay, days on mechanical ventilation,
and days on TPN of the total cohort and stratified by cumulative days of opioid use
quartiles………………………………………………………………………………………22
Table 3: Standardized unit cost (SUC) of the total cohort and stratified by cumulative
days of opioid use quartiles………………………………………………….……………….23
Table 4: 2SRI analysis of the average marginal effects (AMEs) for categorical variables
represent the absolute change in total standardized unit costs……...……………………….24

vi
ABSTRACT
Background: Hospitalizations of high-risk infants are among the most expensive in the United
States, with many requiring surgery and prolonged intensive care. Healthcare costs and resource
use associated with hospitalized infant opioid exposure are less well known.
Methods: A retrospective cohort of high-risk infants <1yr admitted from 47 children’s hospitals
from 2010-2020 was identified from Pediatric Healthcare Information System (PHIS). High-risk
infants were identified by ICD-9/10 codes for congenital heart disease procedures, medical and
surgical necrotizing enterocolitis, extremely low birth weight, very low birth weight, hypoxemic
ischemic encephalopathy, extracorporeal membrane oxygenation, and gastrointestinal tract
malformations. Healthcare resource utilization was estimated using standardized unit costs
(SUC). Impact of opioid use on SUC was examined using general linear models and an
instrumental variable (IV).
Results: Overall, 126,897 high-risk infants were identified. The cohort was majority white
(57.1%), non-Hispanic (72.0%), and male (55.4%). Prematurity occurred in 26.4% and a
majority underwent surgery (77.9%). Median SUC was $120,585/infant (IQR:$57,602-
$276,562). On IV analysis, each day of opioid use was associated with an increase of $4,406 in
SUC. When adjusting for biologic sex, race, ethnicity, insurance type, diagnosis category,
number of comorbidities, mechanical ventilation and TPN use, each day of opioid use was
associated with an increase of $2,177 per infant.
Conclusion: Prolonged opioid use is significantly associated with healthcare utilization and
costs for high-risk infants, even when accounting for comorbidities, intensive care, ventilation,
and TPN use. Future studies are needed to estimate the long-term complications and additional
costs resulting from prolonged opioid exposures in high-risk infants.

1
CHAPTER ONE: INTRODUCTION
Neonatal intensive care unit (NICU) stays are among the most expensive hospitalizations
in the United States with costs upwards of $300,000.1-4 Many high-risk infants require months of
care and undergo surgical procedures, making up the majority of the cost.3,4 High-risk infants are
defined as newborns with neonatal-perinatal morbidities that require additional developmental
care through high-risk follow-up clinics after hospital discharge.5,6 High-risk infants that require
surgery are often exposed to opioids postoperatively.7-9 Today, more than 200,000 hospitalized
infants receive opioids in the US annually.7 Although opioids are effective and necessary for the
treatment of pain, prolonged opioid exposure in infants is associated with increased ventilator
requirements and prolonged parental nutrition (TPN) use, in addition to longer length of stay and
increased healthcare costs.8,10-13 In addition, opioid exposure in healthy infants undergoing
surgery has been shown to lead to prolonged hospitalization and increased healthcare costs.14,15
Existing literature surrounding neonatal opioid exposure, healthcare utilization, and costeffectiveness has shown increased healthcare utilization in infants with neonatal opioid
withdrawal syndrome (NOWS) or neonatal abstinence syndrome (NAS) compared to infants
without in-utero opioid exposure.16-18 Previous observational studies of the Kids’ Inpatient
Database (KIDS) have shown that a diagnosis of NOWS/NAS has a substantial burden in terms
of days in the hospital and costs.16 In this retrospective cohort study, we examine the healthcare
costs of various levels of opioid exposure in high-risk infant populations, excluding infants with
NOWS/NAS, admitted to institutions affiliated with the Children’s Hospital Association (CHA).
We hypothesized that increased days of opioid exposure is an independent risk factor for
increased healthcare costs for hospitalized high-risk infants.

2
CHAPTER TWO: METHODS
Study Design
A cohort of high-risk infants <1 year of age at time of admission from January 1st, 2010
to December 31st, 2020 was built using the Pediatric Health Information Systems (PHIS)
database. The PHIS database is maintained by Children’s Hospital Association (CHA; Lenexa,
KS), which includes clinical and resource utilization data for both inpatient and outpatient
encounters for 47 children’s hospitals across the United States. At the time evaluated for the
study cohort, 47 hospitals contributed data to the PHIS database. All data are de-identified and
data integrity is checked by the CHA data quality program, which issues quarterly reports to
participating hospitals that detail any quality concerns. The study was approved by the
Institutional Review Board (#CHLA-21-00367).
High-risk infants were selected as the population of interest because these patients often
require prolonged hospitalizations and are likely to receive opioids during admission.3,5 We
defined high-risk infants using International Classification of Diseases, Ninth Revision, Clinical
Modification (ICD-9-CM) and International Classification of Diseases, Tenth Revision, Clinical
Modification (ICD-10-CM) codes (Supplement Table 1) that included the following diagnoses:
congenital heart disease (CHD) procedure, medical necrotizing enterocolitis (NEC), surgical
NEC, extremely low birth weight (ELBW), very low birth weight (VLBW), hypoxemic ischemic
encephalopathy (HIE), extracorporeal membrane oxygenation (ECMO), and gastrointestinal tract
malformations (GI malformations). The category “gastrointestinal tract malformations” included
codes for the following diagnoses: biliary atresia, gastroschisis, omphalocele, malrotation,
volvulus, Hirschsprung’s disease, intestinal atresias, annular pancreas, gastric ulcer perforation,
intestinal perforation, peritonitis, meconium ileus, intestinal ischemia or mesenteric ischemia,

3
congenital diaphragmatic hernia (CDH), and tracheoesophageal fistula (TEF). These categories
were based on ICD-9/10 codes for a patient encounter and are not mutually exclusive, thus
patients may fall into multiple categories.
Patients with the following ICD-9 or ICD-10 codes were excluded: 779.5 (drug
withdrawal syndrome in a newborn), P96.1 (neonatal withdrawal symptoms from maternal use of
drugs of addiction), P96.2 (withdrawal symptoms from therapeutic use of drugs in newborn),
760.7x (noxious influences affecting fetus or newborn via placenta or breastmilk) and P90.49
(newborn affected by maternal use of other drugs of addiction) (Figure 1). Patients with a flag
for malignancy in the PHIS database were excluded as pain management in these patients can be
more complex and is often chronic or palliative in nature.19 Patients with methadone use alone
and no other opioid exposure were excluded as this was likely a surrogate for neonatal
abstinence syndrome or withdrawal from an in-utero exposure. Within the congenital heart
disease group, patients who did not undergo a procedural/surgical intervention and possessed a
diagnosis code alone were excluded from the study as patients with a congenital heart defect (i.e.
atrial septal defect, ventricular septal defect, patent ductal arteriosus, etc.) not requiring
procedural intervention within the first year of life are not considered high-risk. Finally, patients
who had no opioid exposure during hospitalization (exposure of interest) were excluded.
Patient Characteristics and Clinical Factors
Demographic factors included for analysis were gestational age at admission (weeks),
sex, birthweight (grams), race, ethnicity, insurance type, and region of hospital location.
Race/ethnicity were included in our study as race and ethnicity have been shown to influence
prescribing decisions in pediatric populations.20-23 Additionally, literature suggests sex-based

4
differences in behavior and gene expression associated with neonatal opioid use.24,25 The cohort
was divided into United States census and PHIS regions based on hospital location. The regional
distribution of the 47 included hospitals was: Northeast-6, South-17, Midwest-14, West-10.
Clinical factors examined included: presence of prematurity diagnosis code (yes/no),
high-risk diagnosis category, surgical intervention during encounter of interest, number of
complex chronic conditions (CCCs), ICU stay (yes/no), length of stay (LOS), mechanical
ventilation (yes/no), total parental nutrition (TPN, yes/no), and cumulative days of ventilation or
on TPN. Notably, the PHIS database only includes NICU and ICU stays, but not specifically
other specialty-designated ICUs such as a cardiothoracic ICU. Additionally, mechanical
ventilation is defined in PHIS as the requirement of mechanical ventilation for a reason apart
from procedural or surgical intervention.
Opioid Exposure
Opioids were identified based on PHIS pharmacy billing codes for any opioid analgesic
and included: alfentanil, fentanyl citrate, hydromorphone, meperidine, oxycodone, remifentanil,
sufentanil citrate, nalbuphine, and other narcotic analgesic combinations. Using billing codes,
cumulative number of opioid exposure days during hospitalization was determined. Cumulative
days of opioid exposure was divided into quartiles: 1-2 days, 3-5 days, 6-11 days, and 12+ days.
Healthcare Utilization and Cost
The PHIS database generates adjusted costs for all billed items at PHIS-participating
hospitals. PHIS standardized unit costs (SUC) are created from a Cost Master Index for each
cost-to-charge (CTC) code and discharge year. The SUC is derived by multiplying the adjusted

5
charges for the CTC code by the corresponding ratio of costs-to-charges (RCC) and dividing by
the number of units. For every discharge year, the median of these costs is determined for each
hospital and CTC code. A median cost for a CTC code is determined using the median cost of
each hospital within that CTC code and discharge year. By using a standard unit cost for an item
across all PHIS hospitals, differences in cost can be attributed to variation in volume. SUC is not
representative of the true hospital cost, but rather a marker for resource utilization. Standardized
unit costs include six main billing categories: pharmacy, supplies, laboratory, imaging, and
clinical are available. Data can be further broken down to evaluate cost associated with
mechanical ventilation and total parental nutrition (TPN) use specifically.
Primary outcome of interest was the impact of opioids on total standardized unit cost
among high-risk infants admitted in the first year of life. Secondary outcomes of healthcare
utilization assessed included: radiology, pharmacy, laboratory, mechanical ventilation, and TPN
costs.

6
CHAPTER THREE: STATISTICAL ANALYSIS
Continuous variables were described using median and interquartile range (IQR) and
Wilcoxon-Mann-Whitney tests were used to compare between groups. Categorical variables
were described using frequencies and percentages and Chi-square or Fischer’s exact tests were
used to compare between groups. Non-parametric tests and median/IQR were used due to our
data not following a normal distribution.
We used a generalized linear model (GLM) using a log-link and gamma distribution to
evaluate standardized unit cost differences. An instrumental variable (IV) model was then
implemented to minimize unmeasured confounding and reduce bias in the estimate due to the
possibility that opioid use may be endogenous. The IV was cumulative opioid use in the
preceding 365 days (2009) and was stratified by whether a patient was surgically treated or not
and how many complex chronic conditions were present which will allow for the incorporation
of variation at both the hospital and patient level.26-30 We operationalized the IV model using a
two-stage residual inclusion (2SRI) model, which allows the use of non-linear regression
models.29 The first stage of the model regresses the potentially endogenous cumulative days of
opioid use variable on the exogenous variables and the IV. In the second stage, the residuals
from the first stage are entered together with the original cumulative days of opioid use variable
along with relevant covariates, clustering on hospital and implementing bootstrapping to
approximate the asymptotically correct standard errors.31 Covariate selection for model inclusion
was based on clinical judgement, availability from PHIS, and variables with a significant
bivariate association with SUC where p<0.05. Covariates included sex, ethnicity, race, insurance
status, prematurity, high-risk diagnosis, number of CCCs, ICU stay, mechanical ventilation, and
TPN use.

7
Before implementing our 2-stage residual inclusion model, we evaluated the robustness
of our instrumental variable (IV) by verifying that the IV is correlated with the treatment variable
(i.e. opioid exposure), then determining the strength of that relationship using two statistical tests
from a 2-stage least squares model. From the first-stage equation, we calculated our F-statistic =
22492.7, p-value <0.0001 (an F-statistic >10 indicated a strong IV).32 Second, we assessed our
partial R2 (0.15); a large value indicates that a significant proportion of the variation found in the
treatment variable can be explained by the IV and not by the measured covariates.33,34 Based on
these criteria, we determined previous drug use within 365 days by hospital and patient type to
be a good candidate instrument based given the strong F-statistic and moderate partial R2
.
Additionally, previous opioid prescribing within 365 days by hospital and patient type per se
should theoretically have no effect on hospital resource use at the observed unit except through
the treatment variable (cumulative days of opioid use).
All analyses were conducted with an α equal to 0.05. Bivariate analyses were conducted
using SAS® software 9.4 (copyright © 2016 SAS Institute Inc., Cary, NC) and 2-stage residual
inclusion methods were performed using Stata version 15, StataCorp LP.

8
CHAPTER FOUR: RESULTS
Overall, 126,897 high-risk infants were identified. The cohort was majority white
(57.1%), non-Hispanic (72.0%), male (55.4%), and prematurity occurred in 26.4% (Table 1).
The largest proportion of the cohort were admitted to a hospital within the Southern region of the
United States (36.7%), and the majority of the cohort was publicly insured (55.2%). High-risk
diagnoses included: 74.5% CHD surgery, 4.6% medical NEC, 3.1% surgical NEC, 25.1%
gastrointestinal tract malformations, 8.3% VLBW, 13.4% ELBW, 1.9% ECMO, and 7.2% HIE.
The majority underwent surgery during the encounter (77.9%).
There were significant differences in demographic and clinical factors, amongst the
different quartiles of days of opioid exposure (p<0.001). Amongst the cohort overall, 13,387
patients (10.5%) were exposed to methadone (Table 1). Methadone exposure stratified by
cumulative days of opioid use quartiles was 0.4% in the first quartile (1-2 days of opioids), 1.9%
in the second quartile (3-5 days), 3.7% in the third quartile (6-11 days), and 35.8% in the fourth
quartile (12+ days) (p<0.001). There was a progressive increase in the percent of infants
requiring mechanical ventilation: 55.5% in the first quartile, 75.1% in the second quartile, 88.7%
in the third quartile, and 96.3% in the fourth quartile. A similar pattern in the proportion of TPN
use was observed in the quartiles of days of opioid exposure: 60.5% in the first quartile, 46.9%
of the second quartile, 65.3% of the third quartile, and 91.4% of those in the fourth quartile.
Median length of stay for the entire cohort was 23 days (IQR 9-66). When stratified by
days of opioid exposure, those with 1-2 days of opioid exposure had a median LOS of 18 days
(IQR 7-56), 3-5 days had 10 days median LOS (IQR 6-23), 6-11 days had 20 days median LOS
(IQR 9-66), and infants with 12+ days of opioid exposure had a median LOS of 75 days (IQR
38-127) (Table 2). Cumulative days of mechanical ventilation gradually increased with more

9
days of opioid exposure with a median of 18 days of ventilation for infants in the fourth quartile.
Median cumulative days of TPN use was highest in the fourth quartile (46 days, IQR 22-94) than
in the other 3 quartiles.
The estimated median total standardized unit cost for the hospitalization of a high-risk
infant was $120,585 (IQR $57,602-$276,562). The estimated median total standardized unit cost
for the hospitalization of a high-risk infant differed across the days of opioid exposure quartiles:
$77,807 (IQR $38,341-$189,033) in the first quartile, $63,521 (IQR $43,211-$113,820) in the
second quartile, $118,159 (IQR $74,900-$199,403) in the third quartile, and $363,942 (IQR
$216,106-$584,713) in the fourth quartile of days of exposure (Table 3). Estimated median total
cost per high-risk infant was then broken down by clinical, imaging, lab, pharmacy, supply, and
other costs with significant variation across the days of opioid quartiles (p<0.001, Table 3). The
cohort was then broken into the mutually exclusive diagnoses to show the estimated median total
cost for each high-risk infant category (Supplemental Tables 2-8).
On 2SRI analysis, the average marginal effects (AMEs) for categorical variables
represent the change in total cost if patients were in the category of interest compared to if
patients were in the reference category, holding other covariates at their mean values. For
continuous variables, AMEs represent the absolute change in total cost associated with a 1-unit
increase, holding other covariates at their mean values. In the simple instrumental variable
analysis not adjusted for potential covariates of interest, each day of opioid use was associated
with an increase of $4,406 in cost per infant (Table 4). When adjusting for biologic sex, race,
ethnicity, insurance type, HRIF diagnosis category, number of CCCs, mechanical ventilation,
and TPN use, each day of opioid use was associated with an average marginal effect of $2,177 in
resource utilization (SUC) per infant. In adjusted analyses, the factors associated with higher

10
total standardized unit costs included Black race, public insurance, prematurity, increased
number of CCCs, ICU stay, mechanical ventilation, and TPN use. With the exception of hypoxic
ischemic encephalopathy, all high-risk diagnoses were associated with higher total SUC.

11
CHAPTER FIVE: DISCUSSION
In this retrospective study of 126,897 high-risk hospitalized infants admitted to 47 U.S.
children’s hospitals, we found that opioid use was independently a significant contributor to
healthcare resource utilization. When adjusting for demographics, clinical comorbidities, and
unmeasured confounding, each day of opioid use was associated with $2,177 in resource
utilization per infant. This is the first study to our knowledge that quantifies the amount of
resource utilization of each additional day of opioid exposure within high-risk hospitalized
infants with a diverse array of diagnoses. Our findings highlight the impact that even one
additional day of opioid prescribing can have on healthcare resources and underscore the need
for thoughtful and judicious opioid prescribing in hospitalized infants.
Neonatal intensive care unit (NICU) stays are among the most expensive hospitalizations
in the United States.1,2 Many high-risk infants require months of care resulting in costs upwards
of $300,000, and infants undergoing surgical procedures make up the highest costs.3,4 Indeed,
our study reports a median total cost for a hospitalized infant of $120,585 with IQR $57,602-
$276,562. Additionally, a majority of our cohort (77.9%) required at least one surgical
procedure. High-risk infants that require pediatric surgical intervention and are often exposed to
opioids postoperatively, in addition to other pain management requirements.1-4 We reported an
increase in resource utilization of $2,177 with each day of opioid use per infant when adjusting
for other clinical factors. With more than 200,000 hospitalized infants receiving opioids in the
US annually, our results highlight a substantial burden on healthcare resource utilization in the
U.S. and a target for future quality efforts.1

Hospitalized infants are often exposed to opioids for both procedural/postoperative
analgesia and for sedation during mechanical ventilation.35 Indeed, use of opioids as sedation in

12
mechanically ventilated infants has been increasing despite illness severity remaining stable.33
Prolonged opioid exposure in infants has been shown to lead to increased likelihood of
methadone exposure.4 Similarly, we report a significantly larger proportion of infants exposed to
methadone in those with greater than 12 days of opioid use. Prolonged opioid exposure is also
associated with increased ventilator requirements, prolonged TPN use, and longer length of stay
in infants.10-13 We report that the number of cumulative days of both mechanical ventilation and
TPN use were substantially increased in infants with greater than 12 days of opioid use. The side
effect profile of opioids includes decreased respiratory drive and gut motility, which can lead to
prolonged ventilator days and increased TPN dependence.14,15,36-41 Our study also found that both
TPN use and mechanical ventilation contribute substantially to costs in hospitalized infants with
an average marginal effect of $70,644 and $139,487 respectively. Therefore, the adjusted effect
of opioid exposure on cost reported may be underestimating the impact that prolonged opioid
exposure has on the requirement of even more costly healthcare resources such as TPN and
mechanical ventilation.
In addition to the short-term effects of prolonged opioid exposure such as increased
ventilator requirements, TPN use, need for weaning agents, and longer length of stay, there are
also potential long-term effects of prolonged opioid exposure.8,10-13 Higher cumulative opioid
exposure has been associated with impaired neurodevelopment in early childhood.41-45 Thus,
there may be long-term healthcare resource utilization resulting from early opioid exposure
during hospitalization. Future studies are needed to investigate costs associated with prolonged
opioid exposure and neurodevelopmental impairment in early childhood.
The healthcare burden of prolonged prescription opioid use has been extensively studied
in the adult population and has also been investigated in healthy children receiving outpatient

13
opioid prescriptions following surgery.46,47 Cummings Joyner et al. reported that healthcare costs
are $4,604 higher in adolescents with prolonged opioid use following elective surgery than those
without prolonged use.46 Despite the widely accepted understanding that opioid use can lead to
substantial healthcare resource utilization, the cost of opioid exposure in infants is less studied.
Studies of healthy infants undergoing pyloromotomy found that opioid exposure led to prolonged
hospitalization and increased healthcare costs.14,15 However, a majority of the existing literature
examining infant inpatient opioid exposure on resource utilization focuses on infants with
NOWS/NAS.16-18,44 Studies have shown substantial burden from increased hospital length of
stay, healthcare utilization, and costs in infants with NOWS/NAS.16-18,48 Opioid exposure among
infants with NAS/NOWS is associated with increased first-year healthcare utilization and in the
subsequent years.17,18 The present study examined a cohort of high-risk hospitalized infants who
received opioids for pain management and not prevention of withdrawal from in utero exposure,
and who, therefore, are an ideal target for future interventions.
There are limitations present and our findings should be interpreted with care. First, this
study is a retrospective analysis of a large administrative claims database, which is subject to
possible coding errors, missing data, leading to potential for bias. However, if those errors are
equally distributed among the sample, we anticipate non-differential misclassification resulting
in estimates towards the null. Utilization of an administrative database (PHIS) leads to a nonrandom sample of hospitals, which although large in number, may have different practices than
hospitals not included in PHIS. The PHIS database represents patients cared for at tertiary
referral children’s hospitals, which limits generalizability of our study’s findings to nonchildren’s hospitals. Furthermore, we did not examine inter-hospital variability in opioid
prescribing or individual hospital charges and, therefore, did not quantify inter-hospital

14
variability, which may significantly contribute to differences in opioid use healthcare pricing.8
Finally, use of multimodal pain management strategies and non-pharmacologic interventions,
such as therapies and skin-to-skin time, are difficult to examine in claims data and were not
examined within this study which may have influenced the results reported. There is promising
data surrounding the use of opioid-alternatives for sedation in infants including
dexmedetomidine and the potential improved outcomes including reduced ventilator days and
neuroprotective affects in CHD and HIE populations.49-55 However, further study is needed to
allow for standardized recommendations on the management of neonatal pain and sedation
management incorporating these adjuncts.

15
CHAPTER SIX: CONCLUSION
Prolonged opioid use significantly contributes to healthcare utilization and costs for highrisk hospitalized infants, even when accounting for clinical comorbidities, intensive care,
mechanical ventilation, and TPN use. These findings underscore the need for quality
improvement efforts centered around thoughtful and judicious opioid prescribing in hospitalized
infants. Future studies are needed to estimate the possible long-term complications and their
associated costs resulting from prolonged opioid exposures in high-risk infants.

16
REFERENCES
1. Rogowski J. Measuring the cost of neonatal and perinatal care. Pediatrics. 1999
Jan;103(1 Suppl E):329-35. PMID: 9917475
2. Phibbs CS, Schmitt SK, Cooper M, Gould JB, Lee HC, Profit J, Lorch SA. Birth
Hospitalization Costs and Days of Care for Mothers and Neonates in California, 2009-
2011. J Pediatr. 2019 Jan;204:118-125.e14. doi: 10.1016/j.jpeds.2018.08.041.
3. Apfeld JC, Kastenberg ZJ, Gibbons AT, Phibbs CS, Lee HC, Sylvester KG. The
disproportionate cost of operation and congenital anomalies in infancy. Surgery. 2019
Jun;165(6):1234-1242. doi: 10.1016/j.surg.2018.12.022
4. Kirkby S, Greenspan JS, Kornhauser M, Schneiderman R. Clinical outcomes and cost of
the moderately preterm infant. Adv Neonatal Care. 2007 Apr;7(2):80-7. doi:
10.1097/01.anc.0000267913.58726.f3
5. Tang BG, Feldman HM, Huffman LC, Kagawa KJ, Gould JB. Missed opportunities in
the referral of high-risk infants to early intervention. Pediatrics. 2012 Jun;129(6):1027-
34. doi: 10.1542/peds.2011-2720.
6. California Perinatal Quality Care Collaborative (CPQCC). What is HRIF? CPQCC High
Risk Infant Follow-Up. 2023 Jan. https://www.cpqcc.org/follow/what-hrif
7. Womer J, Zhong W, Kraemer FW, et al. Variation of opioid use in pediatric inpatients
across hospitals in the U.S. J Pain Symptom Manage. 2014;48(5):903-914.
doi:10.1016/j.jpainsymman.2013.12.241
8. Dervan LA, Yaghmai B, Watson RS, Wolf FM. The use of methadone to facilitate opioid
weaning in pediatric critical care patients: a systematic review of the literature and metaanalysis. Paediatr Anaesth. 2017;27(3):228-239. doi:10.1111/pan.13056
9. Borenstein-Levin L, Synnes A, Grunau RE, Miller SP, Yoon EW, Shah SP; Canadian
Neonatal Network Investigators. Narcotics and Sedative Use in Preterm Neonates. J
Pediatr. 2017;180:92-98.e1. doi: 10.1016/j.jpeds.2016.08.031. PMID: 27614931.
10. Keane OA, Zamora AK, Ourshalimian S, Mahdi EM, Song AY, Kim E, Lakshmanan A,
Kim ES, Kelley-Quon LI. Opioid and Methadone Use for Infants With Surgically Treated
Necrotizing Enterocolitis. JAMA Netw Open. 2023 Jun 1;6(6):e2318910. doi:
10.1001/jamanetworkopen.2023.18910.
11. Menon G, McIntosh N. How should we manage pain in ventilated neonates?
Neonatology. 2008;93(4):316-23. doi: 10.1159/000121458
12. Kelley-Quon LI, Zamora AK, Ourshalimian S, Kim E, Leventhal AM, Kaplan C,
Lakshmanan A. Iatrogenic opioid withdrawal in hospitalized infants. J Perinatol. 2022
Mar;42(3):399-400. doi: 10.1038/s41372-022-01332-6

17
13. Lewis T, Erfe BL, Ezell T, Gauda E. Pharmacoepidemiology of opiate use in the neonatal
ICU: Increasing cumulative doses and iatrogenic opiate withdrawal. J Opioid Manag.
2015 Jul-Aug;11(4):305-12. doi: 10.5055/jom.2015.0279
14. McLaughlin C, Squillaro AI, Ourshaliman S, Song A, Lakshmanan A, Cucchiaro G, Hall
M, Burke RV, Kelley-Quon LI. The Association Between Opioid Use and Outcomes in
Infants Undergoing Pyloromyotomy. Clin Ther. 2019 Sep;41(9):1690-1700
15. Squillaro AI, Ourshalimian S, McLaughlin CM, Lakshmanan A, Friedlich P, Gong C,
Song A, Kelley-Quon LI. Postoperative Opioid Analgesia Impacts Resource Utilization
in Infants Undergoing Pyloromyotomy. J Surg Res. 2020 Nov;255:594-601.
16. Corr TE, Hollenbeak CS. The economic burden of neonatal abstinence syndrome in the
United States. Addiction. 2017 Sep;112(9):1590-1599.
17. Percy Z, Brokamp C, McAllister JM, Ryan P, Wexelblatt SL, Hall ES. Subclinical and
Overt Newborn Opioid Exposure: Prevalence and First-Year Healthcare Utilization. J
Pediatr. 2020 Jul;222:52-58.e1.
18. Liu G, Kong L, Leslie DL, Corr TE. A Longitudinal Healthcare Use Profile of Children
with a History of Neonatal Abstinence Syndrome. J Pediatr. 2019 Jan;204:111-117.e1.
19. Feudtner et al. Pediatric complex chronic conditions classification system version
updated for ICD-10 and complex medical technology dependence and transplantation.
BMC Pediatrics 2014 14:199
20. Ehwerhemuepha L, Donaldson CD, Kain ZN, et al. Race, Ethnicity, and Insurance: the
Association with Opioid Use in a Pediatric Hospital Setting. J Racial Ethn Health
Disparities. 2021 Oct;8(5):1232-1241. doi: 10.1007/s40615-020-00882-9
21. Stake CE, Manworren RCB, Rizeq YK, Minhas S, Quan H, Barsness KA. Use of Opioids
and Nonopioid Analgesics to Treat Pediatric Postoperative Pain in the Emergency
Department. Pediatr Emerg Care. 2022 Jan 1;38(1):e234-e239. doi:
10.1097/PEC.0000000000002227.
22. Miller MM, Williams AE, Zapolski TCB, Rand KL, Hirsh AT. Assessment and
Treatment Recommendations for Pediatric Pain: The Influence of Patient Race, Patient
Gender, and Provider Pain-Related Attitudes. J Pain. 2020 Jan-Feb;21(1-2):225-237. doi:
10.1016/j.jpain.2019.07.002.
23. Hooten WM, Brummett CM, Sullivan MD, et al. A Conceptual Framework for
Understanding Unintended Prolonged Opioid Use. Mayo Clin Proc. 2017
Dec;92(12):1822-1830. doi: 10.1016/j.mayocp.2017.10.010.
24. Borrelli KN, Yao EJ, Yen WW, et al. Sex Differences in Behavioral and Brainstem
Transcriptomic Neuroadaptations following Neonatal Opioid Exposure in Outbred Mice.
eNeuro. 2021 Sep 20;8(5):ENEURO.0143-21.2021. doi: 10.1523/ENEURO.0143-
21.2021

18
25. Yen E, Kaneko-Tarui T, Ruthazer R, Harvey-Wilkes K, Hassaneen M, Maron JL. SexDependent Gene Expression in Infants with Neonatal Opioid Withdrawal Syndrome. J
Pediatr. 2019 Nov;214:60-65.e2. doi: 10.1016/j.jpeds.2019.07.032.
26. Stukel TA, Fisher ES, Wennberg DE. Using observational data to estimate treatment
effects. JAMA. 2007 May 16;297(19):2078-9; author reply 2079. doi:
10.1001/jama.297.19.2078-b.
27. Malik S, Bird TM, Jaquiss RD, Morrow WR, Robbins JM. Comparison of in-hospital and
longer-term outcomes of hybrid and Norwood stage 1 palliation of hypoplastic left heart
syndrome. J Thorac Cardiovasc Surg. 2015 Sep;150(3):474-80.e2. doi:
10.1016/j.jtcvs.2015.06.071.
28. Rassen JA, Brookhart MA, Glynn RJ, Mittleman MA, Schneeweiss S. Instrumental
variables I: instrumental variables exploit natural variation in nonexperimental data to
estimate causal relationships. J Clin Epidemiol. 2009 Dec;62(12):1226-32. doi:
10.1016/j.jclinepi.2008.12.005.
29. Lousdal ML. An introduction to instrumental variable assumptions, validation and
estimation. Emerg Themes Epidemiol. 2018 Jan 22;15:1. doi: 10.1186/s12982-018-0069-
7.
30. Slaughter JL, Reagan PB, Newman TB, Klebanoff MA. Comparative Effectiveness of
Nonsteroidal Anti-inflammatory Drug Treatment vs No Treatment for Patent Ductus
Arteriosus in Preterm Infants. JAMA Pediatr. 2017 Mar 6;171(3):e164354. doi:
10.1001/jamapediatrics.2016.4354.
31. Terza, JV. 2017. Two-stage Residual Inclusion Estimation: A Practitioners Guide to Stata
Implementation. The Stata Journal, 17(4), 916-938.
https://doi.org/10.1177/1536867X1801700409
32. Stock JH, Wright JH, Yogo M. (2002) A Survey of Weak Instruments and Weak
Identification in Generalized Method of Moments, Journal of Business & Economic
Statistics, 20:4, 518-529, DOI: 10.1198/073500102288618658
33. Ertefaie A, Small DS, Flory JH, Hennessy S. A tutorial on the use of instrumental
variables in pharmacoepidemiology. Pharmacoepidemiol Drug Saf. 2017 Apr;26(4):357-
367. doi: 10.1002/pds.4158.
34. Shea, J. (1997). Instrument Relevance in Multivariate Linear Models: A Simple
Measure. The Review of Economics and Statistics, 79(2), 348–352.
http://www.jstor.org/stable/2951471
35. Zimmerman KO, Smith PB, Benjamin DK, et al. Sedation, Analgesia, and Paralysis
during Mechanical Ventilation of Premature Infants. J Pediatr. 2017 Jan;180:99-104.e1.
doi:10.1016/j.jpeds.2016.07.001.

19
36. Chua KP, Brummett CM, Conti RM, Bohnert AS. Opioid Prescribing to US Children and
Young Adults in 2019. Pediatrics. 2021 Sep;148(3):e2021051539.
37. Young LW, Hu Z, Annett RD, Das A, Fuller JF, Higgins RD, Lester BM, Merhar SL,
Simon AE, Ounpraseuth S, Smith PB, Crawford MM, Atz AM, Cottrell LE, Czynski AJ,
Newman S, Paul DA, Sánchez PJ, Semmens EO, Smith MC, Turley CB, Whalen BL,
Poindexter BB, Snowden JN, Devlin LA; Eunice Kennedy Shriver National Institute of
Child Health and Human Development Neonatal Research Network and the NIH
Environmental Influences on Child Health Outcomes (ECHO) Program Institutional
Development Awards States Pediatric Clinical Trials Network. Site-Level Variation in
the Characteristics and Care of Infants with Neonatal Opioid Withdrawal. Pediatrics.
2021 Jan;147(1):e2020008839. doi: 10.1542/peds.2020-008839
38. Liu G, Kong L, Leslie DL, Corr TE. A Longitudinal Healthcare Use Profile of Children
with a History of Neonatal Abstinence Syndrome. J Pediatr. 2019 Jan;204:111-117.e1.
39. Patrick SW, Kaplan HC, Passarella M, Davis MM, Lorch SA. Variation in treatment of
neonatal abstinence syndrome in US children's hospitals, 2004-2011. J Perinatol. 2014
Nov;34(11):867-72. doi: 10.1038/jp.2014.114
40. Milliren CE, Gupta M, Graham DA, Melvin P, Jorina M, Ozonoff A. Hospital Variation
in Neonatal Abstinence Syndrome Incidence, Treatment Modalities, Resource Use, and
Costs Across Pediatric Hospitals in the United States, 2013 to 2016. Hosp Pediatr. 2018
Jan;8(1):15-20. doi: 10.1542/hpeds.2017-0077
41. Kocek M, Wilcox R, Crank C, Patra K. Evaluation of the relationship between opioid
exposure in extremely low birth weight infants in the neonatal intensive care unit and
neurodevelopmental outcome at 2 years. Early Hum Dev. 2016 Jan;92:29-32. doi:
10.1016/j.earlhumdev.2015.11.001
42. Zwicker JG, Miller SP, Grunau RE, Chau V, Brant R, Studholme C, Liu M, Synnes A,
Poskitt KJ, Stiver ML, Tam EW. Smaller Cerebellar Growth and Poorer
Neurodevelopmental Outcomes in Very Preterm Infants Exposed to Neonatal Morphine.
J Pediatr. 2016 May;172:81-87.e2. doi: 10.1016/j.jpeds.2015.12.024
43. Anand KJ, Anderson BJ, Holford NH, Hall RW, Young T, Shephard B, Desai NS, Barton
BA; NEOPAIN Trial Investigators Group. Morphine pharmacokinetics and
pharmacodynamics in preterm and term neonates: secondary results from the NEOPAIN
trial. Br J Anaesth. 2008 Nov;101(5):680-9. doi: 10.1093/bja/aen248
44. Ferguson SA, Ward WL, Paule MG, Hall RW, Anand KJ. A pilot study of preemptive
morphine analgesia in preterm neonates: effects on head circumference, social behavior,
and response latencies in early childhood. Neurotoxicol Teratol. 2012 Jan-Feb;34(1):47-
55.
45. Puia-Dumitrescu M, Comstock BA, Li S, Heagerty PJ, Perez KM, Law JB, Wood TR,
Gogcu S, Mayock DE, Juul SE; PENUT Consortium. Assessment of 2-Year
Neurodevelopmental Outcomes in Extremely Preterm Infants Receiving Opioids and

20
Benzodiazepines. JAMA Netw Open. 2021 Jul 1;4(7):e2115998. doi:
10.1001/jamanetworkopen.2021.15998
46. Cummings Joyner AK, King MR, Safranek C, Parvathinathan G, De Souza E, Anderson
TA. Health Care Burden Associated With Adolescent Prolonged Opioid Use After
Surgery. Anesth Analg. 2023 Feb 1;136(2):317-326. doi:
10.1213/ANE.0000000000006111.
47. Ruhe AK, Frosch M, Wager J, Linder R, Pfenning I, Sauerland D, Zernikow B. Health
Care Utilization and Cost in Children and Adolescents With Chronic Pain: Analysis of
Health Care Claims Data 1 Year Before and After Intensive Interdisciplinary Pain
Treatment. Clin J Pain. 2017 Sep;33(9):767-776. doi: 10.1097/AJP.0000000000000460.
48. Milliren CE, Gupta M, Graham DA, Melvin P, Jorina M, Ozonoff A. Hospital Variation
in Neonatal Abstinence Syndrome Incidence, Treatment Modalities, Resource Use, and
Costs Across Pediatric Hospitals in the United States, 2013 to 2016. Hosp Pediatr. 2018
Jan;8(1):15-20. doi: 10.1542/hpeds.2017-0077.
49. Oschman A, McCabe T, Kuhn RJ. Dexmedetomidine for opioid and benzodiazepine
withdrawal in pediatric patients. Am J Health Syst Pharm. 2011 Jul 1;68(13):1233-8. doi:
10.2146/ajhp100257.
50. Ojha S, Abramson J, Dorling J. Sedation and analgesia from prolonged pain and stress
during mechanical ventilation in preterm infants: is dexmedetomidine an alternative to
current practice? BMJ Paediatr Open. 2022 May;6(1):e001460. doi: 10.1136/bmjpo2022-001460
51. Squillaro A, Mahdi EM, Tran N, Lakshmanan A, Kim E, Kelley-Quon LI. Managing
Procedural Pain in the Neonate Using an Opioid-sparing Approach. Clin Ther. 2019
Sep;41(9):1701-1713. doi: 10.1016/j.clinthera.2019.07.014.
52. Whalen LD, Di Gennaro JL, Irby GA, Yanay O, Zimmerman JJ. Long-term
dexmedetomidine use and safety profile among critically ill children and neonates.
Pediatr Crit Care Med. 2014 Oct;15(8):706-14. doi: 10.1097/PCC.0000000000000200.
53. Curtis S, Kilpatrick R, Billimoria ZC, Zimmerman K, Tolia V, Clark R, Greenberg RG,
Puia-Dumitrescu M. Use of Dexmedetomidine and Opioids in Hospitalized Preterm
Infants. JAMA Netw Open. 2023 Nov 1;6(11):e2341033. doi:
10.1001/jamanetworkopen.2023.41033.
54. O'Mara K, McPherson C. Neuroprotective Agents for Neonates with Hypoxic-Ischemic
Encephalopathy. Neonatal Netw. 2021 Nov 1;40(6):406-413. doi: 10.1891/11-T-755
55. Achuff BJ, Nicolson SC, Elci OU, Zuppa AF. Intraoperative dexmedetomidine reduces
postoperative mechanical ventilation in infants after open heart surgery. Pediatr Crit Care
Med. 2015 Jun;16(5):440-7. doi: 10.1097/PCC.0000000000000382

21
Table 1. Demographic and clinical factors of the total cohort and stratified by cumulative days of
opioid use quartiles
Total Cohort 1st Quartile
(1-2 days)
2nd Quartile
(3-5 days)
3rd Quartile
(6-11 days)
4th Quartile
(12+ days)
p-value
n = 126897 n = 27344 n = 38504 n = 28931 n= 32118
N % N % N % N % N %
Biologic Sex <.001
Female 56530 44.5 12868 47.1 17204 44.7 12635 43.7 13823 43
Male 70318 55.4 14471 52.9 21283 55.3 16284 56.3 18280 56.9
Unknown 49 0 5 0 17 0 12 0 15 0
Race <.001
Black 22012 17.3 5252 19.2 6121 15.9 4294 14.8 6345 19.8
White 72463 57.1 15175 55.5 22162 57.6 17183 59.4 17943 55.9
Asian 3692 2.9 850 3.1 1210 3.1 877 3 755 2.4
Other 21034 16.6 4283 15.7 6579 17.1 4921 17 5251 16.3
Unknown 7696 6.1 1784 6.5 2432 6.3 1656 5.7 1824 5.7
Ethnicity <.001
Hispanic 22708 17.9 4701 17.2 6680 17.3 5389 18.6 5938 18.5
Non-Hispanic 91395 72 19917 72.8 27678 71.9 20587 71.2 23213 72.3
Unknown 12794 10.1 2726 10 4146 10.8 2955 10.2 2967 9.2
Insurance Type <.001
Private 52776 41.6 11411 41.7 16866 43.8 12232 42.3 12267 38.2
Public 70057 55.2 15103 55.2 20227 52.5 15797 54.6 18930 58.9
Other 4064 3.2 830 3 1411 3.7 902 3.1 921 2.9
U.S. Region <.001
Midwest 35168 27.7 9026 33 10111 26.3 7276 25.1 8755 27.3
Northeast 17532 13.8 3444 12.6 6040 15.7 3929 13.6 4119 12.8
South 46607 36.7 9411 34.4 13906 36.1 10898 37.7 12392 38.6
West 27590 21.7 5463 20 8447 21.9 6828 23.6 6852 21.3
Prematurity 33526 26.4 11014 40.3 6159 16 5417 18.7 10936 34 <.001
High-risk
Diagnosis
CHD Surgery 69914 55.1 8618 31.5 24039 62.4 19551 67.6 17706 55.1 <.001
GI Malformations 31887 25.1 5406 19.8 7484 19.4 6772 23.4 12225 38.1 <.001
Medical NEC 5811 4.6 1669 6.1 937 2.4 975 3.4 2230 6.9 <.001
Surgical NEC 3897 3.1 109 0.4 318 0.8 702 2.4 2768 8.6 <.001
VLBW 10550 8.3 5523 20.2 1777 4.6 1344 4.6 1906 5.9 <.001
ELBW 16952 13.4 5027 18.4 2836 7.4 2487 8.6 6602 20.6 <.001
HIE 9125 7.2 2875 10.5 3745 9.7 1224 4.2 1281 4 <.001
ECMO 2383 1.9 160 0.6 121 0.3 284 1 1818 5.7 <.001
Surgically treated 98792 77.9 14018 51.3 31336 81.4 25649 88.7 27789 86.5 <.001
ICU Stay 68847 54.3 7416 27.1 23803 61.8 19389 67 18239 56.8 <.001
Methadone
Exposure 13387 10.5 113 0.4 719 1.9 1063 3.7 11492 35.8 <.001
Mechanical
Ventilation 100667 79.3 15173 55.5 28907 75.1 25668 88.7 30919 96.3 <.001
TPN use 82843 65.3 16537 60.5 18050 46.9 18886 65.3 29370 91.4 <.001

22
Table 2. Gestational age, birthweight, length of stay, days on mechanical ventilation, and days
on TPN of the total cohort and stratified by cumulative days of opioid use quartiles
Total Cohort 1st Quartile
(1-2 days)
2nd Quartile
(3-5 days)
3rd Quartile
(6-11 days)
4th Quartile
(12+ days) pvalue
n = 126897 n = 27344 n = 38504 n = 28931 n= 32118
Median IQR N Median IQR N Median IQR N Median IQR N Median IQR N
Gestational Age
(weeks) 37 30-39 79854 33 28-38 18729 38 35-39 21518 37 30-39 17732 36 27-38 21875 <.001
Birthweight
(grams) 2640 1230-
3260 82446 1720 1036-
3070 19601 2920 2070-
3400 21359 2843 1230-
3260 17903 2470 960-
3160 23583 <.001
Length of Stay 23 9-66 126897 18 7-56 27344 10 6-23 38504 20 9-66 28931 75 38-127 32118 <.001
Days on
Ventilation 5 2-14 100667 2 1-6 15173 2 1-4 28907 5 2-8 25668 18 10-42 30919 <.001
Days on TPN 24 12-50 82843 17 10-30 16537 15 8-30 18050 20 10-40 18886 46 22-94 29370 <.001

23
Table 3. Standardized unit cost (SUC) of the total cohort and stratified by cumulative days of
opioid use quartiles
Standardized Unit
Costs
Total 1st Quartile
(1-2 days)
2nd Quartile
(3-5 days)
3rd Quartile
(6-11 days)
4th Quartile
(12+ days)
P-value
n = 126897 n = 27344 n = 38504 n = 28931 n = 32118
Median IQR Median IQR Median IQR Median IQR Median IQR
Total 120,585 57,602-276,562 77,807 38,341-189,033 63,521 43,211-113,820 118,159 74,900-199,403 363,942 216,106-584,713 <.001
Clinical 11,057 4,026.8-30,948 6150.7 2,446.2-15,106 5,371 2,371.1-12,910 10,806 5,088.5-24,345 44,095 21,474-84,892 <.001
Ventilator* 2,224.4 940.9-6,386.2 1071.7 537.3-2,749.1 979 534.6-1,957.9 2,040.2 979-3,550.4 7,977.1 4,205.1-18,698 <.001
Imaging 3,978.1 2,119.6-7,689.5 2180.3 1,098.6-3,890.1 2,844.7 1,726.1-4,401.9 4,539.2 2,858.1-7,029.4 10,076 6,312.2-15,897 <.001
Laboratory 6,734.6 3,528.9-14,624 3033.3 1,446.5-5,465.2 4,732.2 3,058-7,337.8 8,318.9 5,157.7-13,374 21,817 13,234-35,543 <.001
Other 73,401 33,743-188,257 52908 22,667-154,035 37,206 24,468-70,048 67,891 42,320-125,119 221,560 120,040-373,359 <.001
Total Pharmacy 6,733.4 2,808-16,197 3392.3 1,274.8-7,186.5 3,525.2 2,001.9-6,911.3 7,530.1 4,150.1-13,277 25,015 13,909-47,665 <.001
Methadone** 213.3 46-737.1 22.8 6.6-28.1 14 6.4-25.7 20.7 6.6-44.8 296 88.7-904.7 <.001
Other opioids 169.9 54.3-445.3 22.2 9.4-59.7 119.3 59.5-208.9 249.3 140.3-410.7 741.2 376.6-1,576.7 <.001
TPN*** 2,831.8 1,082.1-6,801.2 2215.6 914.3-4,321.5 1,727.9 692.5-4,029.9 2,241.6 875.2-5,316.3 5,504.9 2,194.7-12,553 <.001
Supplies 4,680.1 1,379.2-1,1523 1898 436.6-5,279 3,751 1,182-8,180.7 6,032.3 2,034.2-12,972 10,935 3,633.4-25,048 <.001
*Ventilator: n=100,667
**Methadone: n=13,387
***TPN: n=82,842

24
Table 4. 2SRI analysis of the average marginal effects (AMEs) for categorical variables
represent the absolute change in total standardized unit costs
dy/dx*
*
pvalue 95% CI dy/dx p-value 95% CI
Opioids 4406 <0.001 3089
572
3 2177 <0.001 1421 2933
Female sex 1303 0.186 -628 3233
Race
Black 5462 0.031 498 10425
Asian 2385 0.584 -6158 10928
Other -4267 0.247 -11495 2962
Unknown -2256 0.719 -14532 10019
Ethnicity
Hispanic -3266 0.385 -10639 4106
Unknown 8210 0.455 -13318 29738
Insurance Type
Public 7672 <0.001 4553 10792
Other -4908 0.463 -18008 8193
Prematurity 44232 <0.001 37725 50739
High-Risk Diagnosis
CHD Procedure 16369 0.015 3147 29592
GI Malformations 11966 0.037 720 23213
Medical NEC 36630 <0.001 29170 44090
Surgical NEC 33012 <0.001 16234 49790
VLBW 68486 <0.001 59318 77654
ELBW 117699 <0.001 100050 135348
HIE -29036 <0.001 -34738 -23334
ECMO 57505 0.002 21383 93628
ICU stay 43784 0.003 15343 72225
Number of CCCs
1 25980 <0.001 22657 29303
2 55288 <0.001 50240 60336
3 89658 <0.001 79613 99702
4 125059 <0.001 102545 147573
5 144065 <0.001 105279 182852
6 169561 <0.001 114791 224330
Mechanical Ventilation 42099 <0.001 32570 51627
TPN 91575 <0.001 81781 101368
*Residual u -0.750 0.999
-
1459
145
7 128 0.757 -684 941
*Residual u is a residual error term included in the model that incorporates information from our
instrumental variable.
**dy/dx = marginal effects for the change in standardized unit cost (SUC)

25
FIGURES
Figure 1. Study flow diagram
Inclusion: All high-risk infants <1 year of
age admitted to PHIS hospitals from
1/1/2010-12/31/2020
N = 575262
Exclusions
Diagnoses:
• Neonatal opioid withdrawal
syndrome/neonatal abstinence
syndrome (N = 15430)
• Prematurity alone (N = 94099)
• Congenital heart disease (CHD)
diagnosis without CHD procedure
(N = 181401)
• Abdominal diagnosis alone
without abdominal procedure (N
= 49291)
• Malignancy (N = 1720)
Data Quality:
• Missing pharmacy records (N =
45366)
• Missing charges data (N = 135)
• Suspected inaccurate charge
data from hospital to PHIS (N =
6223)
Patient/Encounter Factors:
• Patients who received methadone
alone (N = 142)
• Patients that did not receive any
opioids (N = 44428)
• Length of stay >365 days (N = 633)
• Expired or transferred to hospice
N = 126897

Abstract (if available)

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Medical comorbidities and multiple sclerosis in the Hispanic population: linking healthcare disparities

PDF

Risk factors for unanticipated hospitalizations in children and youth with spina bifida at an urban children’s hospital: a cross-sectional study

PDF

Relationships matter: contextualizing developmental outcomes of high risk infants

PDF

Predictors of thrombosis in hospitalized children with central venous catheters: a multi-center predictive study from the CHAT Consortium

PDF

Association of chronic obstructive pulmonary disease and mortality following thoracic and complex endovascular aortic repair: a retrospective cohort study

PDF

Simulation-based training is associated with lower risk-adjusted mortality in ACS Pediatric TQIP centers

PDF

Cardiac function in children and young adults treated with MEK inhibitors: a retrospective cohort study of routinely collected health data

PDF

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

PDF

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

PDF

Association between cardiovascular risk factors and coronary CT measures of coronary atherosclerosis in healthy postmenopausal women

PDF

Integrated management of atrial fibrillation in women in an underserved, safety-net health care system: a multicenter, single health system randomized control efficacy trial protocol

PDF

Defining the global burden of burn injuries: a 29 year review of data from 196 countries

PDF

A retrospective assessment of the safety of sildenafil use in pediatric pulmonary hypertension

PDF

Understanding anti-depressant treatment failure in an underserved vulnerable population

PDF

Validation of the pancreatitis activity scoring system in a large prospective cohort

PDF

Improving health equity after discharge from the neonatal intensive care unit: a qualitative study

PDF

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

PDF

Potential of aqueous humor as a liquid biopsy for uveal melanoma

PDF

The effect of cytomegalovirus on gene expression of pediatric acute lymphoblastic leukemia

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

Asset Metadata

Creator Keane, Olivia A. (author)

Core Title The cost of opioid use in high-risk hospitalized infants

School Keck School of Medicine

Degree Master of Science

Degree Program Biomedical and Translational Investigations,Clinical

Degree Conferral Date 2024-05

Publication Date 04/10/2024

Defense Date 04/09/2024

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

Tag Cost,healthcare resource utilization,Infants,OAI-PMH Harvest,opioids

Format theses (aat)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Patino-Sutton, Cecilia (committee chair), Kaplan, Cameron (committee member), Kelley-Quon, Lorraine (committee member)

Creator Email okeane@chla.usc.edu,okeane@emory.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-oUC113871419

Unique identifier UC113871419

Identifier etd-KeaneOlivi-12793.pdf (filename)

Legacy Identifier etd-KeaneOlivi-12793

Document Type Thesis

Format theses (aat)

Rights Keane, Olivia A.

Internet Media Type application/pdf

Type texts

Source 20240412-usctheses-batch-1139 (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