Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Factors associated with agreement between childhood cancer survivors (CCS) and their parents on knowledge and attitudes about health-related quality of life, treatment and follow-up care
(USC Thesis Other)
Factors associated with agreement between childhood cancer survivors (CCS) and their parents on knowledge and attitudes about health-related quality of life, treatment and follow-up care
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
i Factors associated with agreement between childhood cancer survivors (CCS) and their parents on knowledge and attitudes about Health Related Quality of Life, treatment and follow-up care By Xueyan Zhuang A Thesis Presented to the Faculty of Graduate School University of Southern California In Partial Fulfillment of the Requirement for the Degree of Master of Science In Applied Biostatistics and Epidemiology Date August 2016 Copyright 2016 Xueyan Zhuang ii TABLES OF CONTENTS ACKNOWLEDGEMENT ................................................................................................. iii ABSTRACT ....................................................................................................................... iv INTRODUCTION .............................................................................................................. 1 METHODS ......................................................................................................................... 4 RESULTS ......................................................................................................................... 14 DISCUSSION ................................................................................................................... 20 CONCLUSION ................................................................................................................. 23 REFERENCES ................................................................................................................. 24 TABLES ........................................................................................................................... 28 FIGURES .......................................................................................................................... 39 iii ACKNOWLEDGEMENT I would like to thank all the people who helped and supported me in the completion of my master thesis. I would like to express my deep gratitude to my thesis mentor Dr. Ann Hamilton, for her patient guidance, enthusiastic encouragement and useful critiques of this research project. Without her guidance, my master thesis could not have been successfully conducted. I would also like to thank my committee chair Dr. Wendy Mack, for her advice in using statistical methods and insightful comments. I am gratefully indebted to her for her valuable comments on this thesis. My grateful thanks are also extended to Dr. Joel Milam for his help in offering me the resources of dataset and providing me professional guidance. Finally, I wish to thank my parents and my husband for providing me with unfailing support and continuous encouragement through the process of researching and writing this thesis. iv ABSTRACT Background: The number of childhood cancer survivors (CCS) has been increasing greatly over recent decades due to longer term survival. Both CCS and their parents experience health problems caused by the child’s cancer and cancer treatment. Parents often make health decisions as a proxy for their children when children are too ill or young. Knowing how well the parent’s view of the child’s health compares to the child’s perspective is important, however previous studies have not provided consistent conclusions about the agreement between child-parent dyads on ratings of child’s health-related quality of life (HRQoL) and have been based on small numbers of dyads. The objectives of this study were: (1) to assess the consistency of responses between child-parent dyads to questions regarding use of health care and health assessment of the child; and (2) to test whether the magnitude of agreement within dyads is associated with socio-demographic characteristics of the parent and child. Methods: CCS seen at two large children’s hospitals in Los Angeles were selected from a population based cancer registry in 2009 and mailed a survey at 2-10 years after diagnosis. A survey with similar questions concerning the child’s cancer was mailed to their parents. 160 child-parent dyads completed child and parent surveys (CCS mean± standard deviation age at survey response,19.36 ± 2.87 years; age at diagnosis,11.95 ± 2.91 years; 51.3% Hispanic). Agreement between dyads on 8 categories of questions that requested a categorical response was assessed by Kappa statistics. Chi-Square tests and multivariable logistic regression were used to identify demographic and clinical factors associated with agreement on 3 questions selected as examples. All regression models were adjusted for CCS age and race/ethnicity. (Question 1: When did the child last see a doctor for cancer related care? Question 2: How long should the v child continue having cancer follow-up care? Question 3: During the next 2 years, what are the chances that the child will go for a cancer follow-up visit?) Agreement on child’s health-related quality of life (HRQoL) measure (the PedsQL™) in various domains was analyzed via mean difference, intra-class correlation coefficient (ICC), and Bland-Altman plots. Results: Kappa statistics indicated moderate agreement (0.25≤Kappa<0.5) for most categories of common questions. Excellent agreement (Kappa≥0.75) was found for questions about cancer treatments and diagnosis of cancer, while poor agreement (Kappa<0.25) was found for questions about plans for future follow-up care and family influence on health decisions. Agreement on Question 1 regarding when the child last saw a doctor for cancer-related care was significantly higher in dyads where children had lymphoma (vs. leukemia) (OR=4.88,95%CI=1.23-19.36). Agreement on the preferred answer (‘child’s whole lifetime’) for Question 2 about how long should the child continue cancer follow up care was significantly higher for parents who spoke English than parents who only spoke Spanish (p=0.01); this association was not significant after adjusting for child’s age at survey completion, child’s ethnicity, and parent’s educational attainment. Agreement on the preferred answer (‘very likely’) for Question 3 regarding the chances that the child will go for cancer follow-up care was significantly lower for dyads including Hispanic CCS than for dyads including non-Hispanic CCS (p=0.003) in univariate analysis; this association was not significant after adjusting for child’s age at survey completion, parent’s educational attainment and parent’s language. The 3 questions all had significant trends of increasing agreement within dyads as the parent’s education level increased (p<0.05). There was no statistically significant difference between the child-rated and the parent rated PedsQL scores, except for the Social Functioning Score (mean difference of child rated-parent rated =5.02, p=0.005). Most domains of the PedsQL showed moderate agreement (0.41<ICC<0.60). vi The Physical Function Score showed good agreement (ICC=0.62) while the Social Function Score had fair agreement (ICC=0.387). ICC’s were almost all lower for the middle 18-20 age group than the other two age groups (15-17 and 21-25 age groups) for each subscale of the PedsQL. The Bland-Altman plots for the PedsQL™ Social Function Score showed relatively poor agreement between dyads (95% limits of agreement=-48.3-38.3). Conclusion: Our study demonstrated better agreement between child-parent dyads for objective questions and rating of child’s quality of life in the physical functioning domain. Relatively poor agreement was found for subjective questions and assessment of child’s social functioning. Higher agreement was related to parents who spoke English (vs. parents who only spoke Spanish) and higher parent education level. Further studies are needed to establish the optimal predictors of levels of parent-child agreement. Keywords: Childhood, Cancer, Survivorship, Follow-up care, Health-related quality of life, Parent agreement 1 INTRODUCTION An estimated 15,780 new cases of child and adolescent cancer were diagnosed in the United States in 2014 [1]. From the SEER Cancer Statistics Review of 1975-2012, the childhood (age 0- 19) cancer rate increased during this period (from 13.0 to 17.7 per 100,000); the overall 5-year survival rate for childhood cancer has dramatically increased from 62% to about 84% [2]. As a result, the number of CCS has increased from 70, 868 in 1975 to 337,415 in 2012 [2]. Childhood cancer survivors (CCS) experience adverse effects from cancer and cancer treatment such as worse overall health, more mental problems, more activity limitations and poorer academic performance than their peers and healthy siblings [3-4]. Parents also suffer from depression, psychological stress and sleep problems related to their child’s cancer experience [5]. Parents often make health decisions as a proxy for their children (e.g., when pediatric patients are too young or too ill) [6]. Therefore, understanding the agreement between CCS and their parents on the child’s health status and knowledge on cancer and related follow-up care is important. Previous studies have compared parent’s and child’s rating of health-related quality of life (HRQoL) for children with cancer. Most studies showed CCS rated themselves higher on child’s HRQoL than their parent proxy reports [7-10]. A few studies indicated parents rated their child’s HRQoL lower than healthy children [8, 11]. Agreement for HRQoL between parent and child varies by domain [12] and is sometimes greater for physical aspects of health compared to 2 emotional or social aspects in some studies [13-14], with other studies reporting contrary conclusions [11,15]. Disagreement may occur when parents know more about their child’s health information than the child does. There might also be discrepancies in ratings of child’s health, since parents may tend to report more behavioral symptoms while children may report more subjective symptoms [16]. The disagreement may also be related to the child’s age or degree of relationship with the parent who provides the response; for example, father-son agreement was lower than father- daughter agreement on child’s HRQoL in one study [17]. In addition, among Hispanics, disagreement between parent and child may arise when Hispanic parents who were not born in America and have difficulty communicating with doctors because of language and cultural barriers [18], while their children who were born in America do not have such problems. This thesis will analyze agreement between CCS and their parents with the following objectives: (1) to assess the consistency of responses between parent-child dyads to questions regarding use of health care and health assessment of the child; and (2) to test whether the magnitude of agreement within dyads is associated with socio-demographic characteristics and discordance in acculturation between the parent and child. The study will determine categories of variables that have the best agreement between parents and children and which do not. Our study hypotheses were: (1) Agreement between CCS and parent reports is expected to be higher for objective questions 3 (e.g. cancer treatment) than subjective questions (e.g. plans for future follow-up care) (2) Agreement between CCS and parent’s reports is expected to be positively related to higher parent’s education level (vs. lower education level) and parents who speak English (vs. parents who only speak Spanish) (3) Children are expected to rate child’s HRQoL higher than do their parents 4 METHODS Source of subjects The ‘Project Forward Study’ evaluated factors associated with cancer-related follow-up care among Hispanic and non-Hispanic childhood cancer survivors (CCS) [19]. A self-administered survey was sent to selected CCS who were identified from the Los Angeles Cancer Surveillance Program (CSP). CCS were contacted if they: (1) were diagnosed with cancer at Children’s Hospital Los Angeles (CHLA) or Miller Children’s Hospital, Long Beach (LBMMC), (2) were diagnosed between 2000 and 2007, and (3) their age in 2009 was between 14 and 25 years. Another survey that included similar questions about the child’s experience was sent to their parents. Children (and their parents) with Hodgkin disease were excluded from the study because they were included in another registry study [19]. The parent and child surveys asked a number of questions in common about the child’s health care including: program attendance, diagnosis, treatment, doctors seen over the past 2 years, current doctors seen regularly, transportation problems, referrals to specialists, follow-up care, parental/family influence, overall health, child’s health-related quality of life (HRQoL), and health insurance. After the study was approved by the USC, CPHS, and Miller’s Hospital IRBs, CCS were mailed a survey along with a postage paid envelope for return. For CCS <18, the initial contact was with their parents, whose consent was required before asking the CCS to participate. For 5 CCS 18+, we requested that the CCS ask their parents if they would be willing to participate. Extensive follow-up, including telephone calls, second mailings, tracing of lost individuals, personal interviews, as well as offering the online option for survey completion, was used to increase response. Among 470 eligible CCS who were sent surveys, 235 CCS (50%) and 173 (37%) parents participated. Parents of younger children were more likely to participate (p<0.05), probably due to the initial contact with the parents for those <18 [20]. There were 160 dyads (34%) where both the parent and the child participated. Measurements 1. Categories of common questions There were 36 questions in common in the child and parent surveys. We divided these questions into 8 categories that included: (1) Program attendance; (2) Cancer diagnosis and treatment (3) Follow-up care received (4) Future follow-up care (5) Parental/Family influence (6) Problems in seeking care (7) Child health assessment, and; (8) Health insurance. The continuous variable -Pediatric Quality of Life Inventory (PedsQL) was also examined. (1) Program attendance: There were 3 questions about the child’s previous attendance in survivorship follow-up care programs (LIFE Program, Late Effects Program or Susan Shannon services, and cancer clinic at other facility). Detailed questions: Has the child ever gone to the LIFE Program? Has the child 6 ever gone to the Late Effects Program or Susan Shannon services? Has the child ever gone to cancer clinic at other facility? (2) Cancer diagnosis and treatment: This category included 6 questions regarding diagnosis of cancer and treatment. Questions included: the name of the child’s cancer (1 question); the type of treatment (chemotherapy, radiation therapy, surgery; 3 questions); and the time when child finished his/her cancer treatment (1 question). An additional question queried child and parent about sharing the treatment summary with child’s doctors. Detailed questions: What was the name of cancer that child was diagnosed? Did the cancer treatment include chemotherapy? Did the cancer treatment include radiation therapy? Did the cancer treatment include surgery? When did the child finish cancer treatment? Has child or their parent ever shared the written summary of cancer treatment with any of child's doctors? (3) Follow-up care received (past 2 years and current) Follow-up care questions related to follow-up care received in the past 2 years for cancer- related and non-cancer medical care (7 questions), and current follow-up care (4 questions). Detailed questions: Has the child seen a doctor in the past 2 years? Was the child's last visit to a doctor within the past year? What the type (s) of doctor(s) did the child see in the past 2 years? When did the child last see a doctor for cancer related care? What the type (s) of doctor(s) has the child seen for cancer related care? When did the child last see a doctor for a routine medical care? Has the child gone to a Curandero/a, Herbalist or … for health care in the past 2 years? 7 Does the child have a doctor for check-ups or problems related to cancer / treatment? What type of doctor does the child see for check-ups related to cancer or treatment? Does the child have a doctor for regular check-ups or problems not related to cancer treatment? Does the doctor for child's regular check-ups not related to cancer treatment know about the child's cancer history? (4) Future follow-up Care There were 4 questions about plan for future follow-up care. Detailed questions: Have any of child's doctors discussed the health care that your child will need in the future? How long do you think the child should continue having cancer related follow-up care? What are the chances that the child will go to any medical doctor for a routine check-up during the next 2 years? What are the chances that the child will go for a cancer follow up visit during the next 2 years? (5) Parental/Family influence Four questions queried about parental involvement in the child’s cancer-related care and routine medical care, as well as the family influence on health care decisions. Detailed questions: Have the child and at least one of the parents discussed the health care the child will need in the future? How often does one of the parents go with the child to see a doctor (for cancer related care)? How often does one of the parents go with the child to see a doctor (routine medical care)? How much does family influence the child's health care decisions? (6) Problems in seeking care 8 Two questions dealt with the child’s problems in seeking care in transportation and referrals. Detailed questions: Has the child had any problem with getting transportation to a doctor’s appointment? Has the child been able to get a referral to a specialist if he/she need one? (7) Child health assessment Two questions queried about the child’s overall health and health problems related to cancer or cancer treatment. Responses to these items used a 5-level ordinal scale (poor, fair, good, very good, excellent). Detailed questions: How would you rate the child's general health overall? Does the child have any health problems as a result of his/her cancer and/or its treatment? (8) Health insurance Four questions asked about the type of health insurance (public, private), source of child’s medical coverage (work/school, parent, spouse, Medical/Medicaid) and difficulty in obtaining health insurance because of the child’s cancer history. Detailed questions: Has the child had difficulty obtaining health insurance because of cancer treatment history? What type of health insurance does the child currently has? What is the source of child's current medical coverage? How difficult has it been for the child or the parents to deal with a Health Insurance company or HMO during the 2 past years? 2. The Pediatric Quality of Life Inventory (PedsQL) The Pediatric Quality of Life Inventory (PedsQL) is a standardized instrument that assesses patients' and parents' perceptions of health-related quality of life (HRQoL) in pediatric patients 9 with chronic health conditions, including pediatric cancer [21]. The 23-item PedsQL TM 4.0 Generic Core Scales was used in both the child and parent surveys. The inventory consists of 4 domains: (1) Physical Functioning (8 items), (2) Emotional Functioning (5 items), (3) Social Functioning (5 items), and (4) School Functioning (5 items). Responses for each item use a 5- point ordinal scale (0= never a problem; 1 = almost never a problem; 2 = sometimes a problem; 3 = often a problem; 4 = almost always a problem). The responses are then linearly reverse-scored to a 0 (lowest health-related quality of life) to 100 (highest health-related quality of life) scale (0 = 100, 1 = 75, 2 = 50, 3 = 25, 4 = 0). The Physical Health Summary Score (8 items) equals the Physical Functioning Subscale. The Psychosocial Health Summary Score (15 items) equals the sum of the items divided by the number of items answered on the Emotional, Social, and School Functioning subscales [22]. PedsQL is reliable (total scale score: α=0.88 child self-report; α=0.93 parent proxy-report) and valid (distinguishes between healthy children and children with cancer) [21]. The minimal clinically meaningful difference (MCID) is about 4.5 points [23]. Statistical Methods Within the 8 categories of common questions, agreement between parent and child was assessed. For categorical variables, we used Cohen’s Kappa inter-rater reliability statistic as an index of the degree of agreement between child and parent response pertaining to the child’s condition. The Kappa statistic tests if agreement exceeds that which would be expected due to chance, and is suitable when the ratings of question are either nominal (e.g., “yes”, “no”) or 10 ordinal (e.g., “mild”, “moderate”, “severe”). The range of Kappa is from -1 to 1. Kappa=1 if the raters are in perfect agreement, Kappa=0 means agreement no better than that expected by chance, and Kappa<0 means agreement worse than that expected by chance (seldom seen in clinical applications) [24]. Similar to Landis and Koch’s standard [25] of grouping Kappa strength, we divided the Kappa values into 4 groups: Kappa≥0.75 indicating high agreement, 0.500≤Kappa<0.75 indicating good agreement, 0.25≤Kappa<0.5 indicating moderate agreement, and Kappa<0.25 indicating poor agreement. Univariate chi-square tests and multivariable logistic regression models were used to determine variables that were independently associated with agreement between dyads for three selected questions. Independent variables included child’s age at survey completion, ethnicity, cancer type, parent’s educational attainment, parent’s language, and whether or not the child was currently living with their parents. Question1 addressed when the child last saw a doctor for cancer-related care. The original responses were: (1) Within the past year (2) 1-2 years ago (3) More than 2 years ago (4) Not sure. We dichotomized the responses into “Within the past 2 years” vs. “All other” (more than 2 years, not sure). Question 2 addressed how long the child should continue having cancer-related follow-up care. The preferred answer was ‘child’s whole lifetime’ vs. any other response. Question 3 addressed the chances that the child would go for a cancer follow-up visit during the next 2 years. The preferred response was ‘very likely’ vs. any other response. For Question 1 we tested any type of agreement between parent and child; however, for Question 2 and Question 3, we only considered agreement based on the preferred 11 answer. For the multivariable models, the outcome variable of agreement was coded as 1 if both the parent and child agreed on the appropriate response for the specific question, else the agreement variable was coded as 0. For Question 1, the agreement variable was coded as 1 when both parent and child agreed that care had been received within the past 2 years ago, or if they agreed on all other combined responses. For Question 2, the agreement variable was coded as 1 when both parent and child agreed on the preferred response of ‘child’s whole lifetime’ and 0 for any other combination of response, and for Question 3, both parent and child were required to agree on the preferred response of ‘very likely’ vs. any other combination of response. The univariate association between dyad agreement and each independent variable was first examined by using the chi-square test. The independent variables that we studied included: 1) CCS age at survey completion, which was categorized into 3 groups: 15-17 years old, 18-21 years old, 21-25 years old; 2) CCS ethnicity (Hispanic, Non-Hispanic); 3) cancer type ; 4) treatment intensity (least intensive, moderately intensive, very/ most intensive [19]; 5) parent’s education (grade school/some high school, high school graduate/GED, some college/AA, college graduate/graduate school); 6) parent’s language (Spanish only, English speaking); 7) child living with parents (vs. not living with parents). After including demographics (age, and race/ethnicity) in each model, other variables that demonstrated a univariate association with agreement (p≤0.10) were selected for inclusion in the final multivariable logistic regression models. For the only continuous variable that was common to the child and parent surveys, the child’s Pediatric Quality of Life Inventory (PedsQL), we used both the intra-class correlation coefficient 12 (ICC) and Bland-Altman plots to assess agreement. We calculated the ICC to assess the magnitude of agreement of child’s health-related quality of life between child and parents [11- 12, 26]. The ICC was based on the two-way mixed effect model where (child/parent) raters were considered as fixed effects and subjects were treated as random effects. The formula: ICC (3,1) =(BMS-EMS)/(BMS+(k-1) EMS). BMS represented between-targets mean square, k represented k judges, and EMS represented error (residual) mean square [27-28]. ICCs<0.40 were considered poor to fair agreement, 0.41 to 0.60 were considered moderate agreement, 0.61 to 0.80 were considered good agreement, and 0.81 to 1.00 were considered excellent agreement [29]. The ICC analysis was conducted for the total sample (N=160 dyads) and stratified by child age groups (15-17 years, N= 50 dyads; 18-20 years, N= 54 dyads; 21-25 years, N= 56 dyads). The Bland-Altman plot is a graphical method used to evaluate the agreement between two measurements obtained on the same subject [30]. We used the Bland-Altman plot to show the differences of PedsQL score between dyads (parent score-child score) against the averages of dyads ((parent score +child score)/2). 95% confidence limits of agreement were estimated by mean difference±1.96 standard deviation of the differences. The premise of using the Bland- Altman plot is that the assumed differences between the two measurements are not different from 0, and that the magnitude of the difference is not a function of the absolute level. A linear regression line of the rater differences against the rater average is used to detect potential proportional bias (where one method gives values higher or lower than the other method by an 13 amount proportional to the size of the measurement). [31-32]. Proportional bias is evidenced if the gradient of the slope of the regression line significantly differs from zero [33]. We used SPSS Version 22.0 to calculate Intra-Class correlation coefficients (ICC) and MedCalc software to draw Bland-Altman plots of the PedsQL score. All other statistical analyses used SAS 9.4 version for Windows. 14 RESULTS Demographic and clinical characteristics of child-parent dyads The 160 CCS included in the dyads were evenly divided by gender. About 60.6% of the CCS were 10-14 years old when first diagnosed with cancer, and 30% were diagnosed with leukemia. About 51.3% of the CCS self-reported their ethnicity as Hispanic. Most of the child participants were from Children’s Hospital Los Angeles (CHLA) (84.4%). The majority of parents in this study were mothers (86.2%), and 55.6% of parents self-reported their ethnicity as Hispanic. 34.4% of parents reported grade school or some high school, and 23.8% of parents reported more than college graduate as their highest achieved educational level (Table 1). Kappa values for categories of common questions between dyads Kappa statistics indicated excellent agreement (Kappa≥0.75) for questions about cancer treatments (radiation therapy or chemotherapy) and diagnosis of child’s cancer. Kappa statistics indicated good agreement (0.50≤Kappa<0.75) for questions about treatment, including surgery, type of child’s health insurance, source of medical coverage and the time the child finished their cancer treatment. Kappa statistics indicated poor agreement (Kappa<0.25) for questions about whether doctors discussed the future need for health care, difficulty dealing with a health insurance company and family influence. The Kappa statistic was lowest for the question about 15 family influence on child's health care decisions (Kappa=0.0435). The Kappa statistics indicated moderate agreement (0.25≤Kappa<0.5) for all other questions (Table 2). Factors related to agreement for 3 selected questions 1) Agreement on Question 1 (When did the child last see a doctor for cancer related care?) Question 1 addressed when the child last saw a doctor for cancer-related care (Table 3). Overall 76.9% of the dyads agreed on the response (either within the past 2 years or other response) and Kappa=0.44. In the univariate chi-square test (Table 4), the agreement between parent-child dyads was significantly higher in dyads in which the CCS had lymphoma than in dyads where the CCS had other cancer types (p=0.02). The agreement was also statistically significantly higher in dyads with parents who spoke English than in dyads with parents who only spoke Spanish. The agreement was higher in dyads with parents who reported more than college graduate as their highest achieved educational level than in dyads in which the parent reported lower education levels, but the difference was not significant (p=0.07). These variables were included in the multivariable logistic regression analysis, and the only variable remaining significantly associated with dyad agreement was the cancer type of lymphoma. The adjusted odds of agreement in dyads in which the CCS had lymphoma was 4.88 (95%CI: 1.23-19.36) the odds of agreement among dyads in which the CCS had leukemia (p=0.02). There was a significant trend of increasing agreement within dyads as the parent’s education level increased (p=0.01). 16 2) Agreement on Question 2 (How long do you think the child should continue having cancer follow-up care?) Question 2 addressed how long the child should continue having cancer follow-up care (Table 5). Overall 42.5% of the dyads agreed on the preferred response (‘child’s whole lifetime’) and Kappa=0.27. In the univariate chi-square test (Table 6), agreement was statistically significantly higher in dyads with parents who spoke English than in dyads with parents who only spoke Spanish (p=0.01). In the multivariate logistic regression analysis, no factor remained significantly associated with agreement. There was a marginally significant trend of increasing agreement in dyads as parent’s education level increased (p=0.05). 3) Agreement on Question 3 (During the next 2 years, what are the chances that the child will go for a cancer follow-up visit?) Question 3 addressed the chances that the child would go for cancer follow-up visit during the next 2 years (Table 7). Overall 38.8% of the dyads agreed on the preferred response (“very likely”) and Kappa=0.45. In the univariate chi-square test (Table 8), the agreement was significantly higher in dyads with children who self-reported their ethnicity as non-Hispanic compared to those who self-reported their ethnicity as Hispanic (p=0.003). The agreement was statistically significantly higher in dyads with parents who spoke English than in dyads with parents who only spoke Spanish (p=0.004). The agreement was higher in dyads with parents who 17 reported more than college graduate as their highest achieved educational level than in dyads with parents who reported lower education levels, but the difference was not significant (p=0.07). In the multivariate logistic regression model, no factor was significantly related to agreement. There was a significant trend of increasing agreement in dyads as parent’s education level increased (p=0.01). The Pediatric Quality of Life Inventory (PedsQL) 1) Basic mean of PedsQL For the Total Scale Score, Physical Functioning Score and Psychosocial Health Summary Score, the mean child-rated scores were somewhat higher than parent-rated scores; the median child-rated scores were somewhat lower than parent-rated scores. For the Emotional Functioning and the Social Functioning Scores, both the mean and median for child-rated scores were higher than parent-rated scores. Both the mean and median for child-rated School Functioning Score were lower than parent-rated scores. With the exception of the Social Functioning Score (p=0.005), there was no statistically significant difference (p<0.05) between the child-rated and the parent rated PedsQL scores (Table 9 and Figure 1). For the Social Functioning Score, children consistently rated themselves higher than their parents did, and the mean difference (child-rated score- parent-rated score) was 5.02, which was larger than the minimal clinically meaningful difference [23] (MCID=4.5 points) (Table 9 and Figure1). 18 2) Intra-Class correlation coefficients (ICC) Most domains of the PedsQL in the total sample showed moderate agreement (0.41<ICC<0.60 defined as moderate agreement). The Physical Function Score showed good agreement (ICC=0.62), and the Social Function Score had fair agreement (ICC=0.387). After stratification by the child’s age at survey completion, the ICC was highest (ICC=0.672) for children aged 21- 25 years old at survey completion for the Physical Functioning Scale, and was lowest (ICC=0.254) for the Social Function Scale for children aged 18-20. ICCs were almost all lower for 18-20 age group than the other two age groups for each subscale of PedsQL (Table10). 3) Bland-Altman plot of PedsQL Bland-Altman plots (Figure 2-Figure7), displaying the parent-child averaged score versus the parent-child difference, were completed to determine if parent-child differences in ratings were related to the magnitude of ratings. Consistent with the comparison of means above, the Bland- Altman plots for the PedsQL™ Social Function Score (Figure 5) revealed a significant amount of variation (bias) (p=0.005), with 95% limits of agreement ranging from −48.3 to 38.3, indicating relatively poor agreement between patient self-reports versus parent proxy reports. Almost all the Bland-Altman plots (Figure 2-Figure7) had more negative outliers than positive outliers, indicating that more parents rated child’s health-related quality of life very low compared to a high rating by children (than vice versa). The absolute difference was around zero when the average score reached 100 because, in this situation, both parent and child rated 100 19 points for the score (i.e., a ceiling effect). After excluding those higher averages, we did not see any systematic association of magnitude with difference in any plot. The only significant positive linear regression line (p value for linear regression model=0.0003) for Bland-Altman plot of PedsQL™ Social Functioning Score indicated that as the parent-child averaged score increased, the parent-child differences became smaller, despite children consistently rating themselves higher than their parents did; this suggests there might be proportional bias for the Social Functioning Score. 20 DISCUSSION Our study, based on 160 child-parent dyads is one of the largest studies to compare parent and child differences in assessment of treatment and HRQoL among CCS. Consistent with previous smaller studies, the highest agreement was found on objective questions concerning cancer diagnosis treatment and diagnosis which were factual, and unambiguous [34-35]. Poorer agreement was found for subjective questions related to family influence in heath decision or intent for future follow-up care which were emotional, unobservable, and depended more on the extent of family cohesion and communication between parent and child. [36-37]. A higher agreement rate was generally found in dyads with parents who spoke English than for dyads with parents who only spoke Spanish. Another study has shown that bilingual Hispanic parents in the United States tend to have less parent-adolescent conflict and had better communication with their English-speaking children than Spanish-speaking parents [38]. Also, some studies show that language is an indicator of acculturation, which can influence access to care and communication with physicians [39-40]. Consistent with our hypothesis and previous studies involving agreement on child’s health quality of life (HRQoL) between children with cancer and parents [7-10], children in our study reported better quality of life than did their parents with exception of the school functioning domain. This may have occurred because children with cancer tend to underreport or deny their negative symptoms (distress, depression, activity limitation, etc.), possibly through repressive 21 adaptive style (repressors consistently report lower subjective anxiety or distress despite physiological and behavior evidence that they are upset, and are of high defensiveness) in the process of adapting to their illness [36,41-42],whereas parents with mental health problems tend to be more focused on their child’s health problems [43-44]. Consistent with previous studies of agreement on HRQoL [13-14], highest agreement on the PedsQL was found in our study for the physical functioning domain but lowest for the social functioning domain. Peer social relationships play an important role in child and adolescent development [45], and support from friends or classmates could compensate for the lack of parental support [46]. The relatively poor agreement between dyads for child’s Social Functioning Score may be due to the difficulty for parents observing child’s social behavior. Parents may also observe different behaviors of the same child in different situations, and some important aspects of social functioning (e.g., social support) are best assessed from the perception of the target child [47]. The main strength of this study was the large simple size (n=160) compared to similar studies of agreement between parent and CCS [7,9,13-14,37]. In addition, 51.3% of the 160 patients in our sample were Hispanic while only 17.4% of the national population are Hispanic. The limitation of this study was that dyads including older CCS were less likely to be included because the initial contact was with the CCS and we required that the CCS invite their parent to participate. In some cases older CCS refused to have their parents contacted [20]. Among CCS <18, the initial contact was with the parent who provided permission to contact the CCS. Also 22 our study was limited in that patients from only two hospitals in Los Angeles were included, so our conclusions can only be generalized to the patients seen at those facilities and not to the entire population of Los Angeles. Our study population may not be representative of the broader population of children with cancer and their parents. Most of previous research has focused on the agreement of child’s health status or health related quality of life (HRQoL) [7,9-14,28,37,43] or health related symptoms. [28,34], but no previous study has included information about agreement on both the child’s HRQoL and follow-up care. One of the most fundamental barriers for effective long-term follow-up is lack of knowledge of long-term survivors and the non-specialist physicians caring for them [48]. The agreement between dyads on follow-up care may be an index of the need for communication between parents, children and physicians. Better communication will facilitate timely utilization of follow-up care and appropriate management of late adverse effects, thereby reducing the incidence rate of severe complications. This may also be helpful for physicians to assess child’s cancer diagnosis, treatment and physical functioning via parent proxy-reports when pediatric patients were unable or unwilling to provide self-reports, since those questions had high agreement in our study. 23 CONCLUSION In summary, our study demonstrated that better agreement between child-parent dyads for objective questions and rating of child’s quality of life in the physical functioning domain. Relatively poor agreement was found for subjective questions and assessment of child’s social functioning. Higher agreement was related to parents who spoke English (vs. parents who only spoke Spanish) and higher parent education level. Further studies are needed to establish the optimal predictors of levels of parent-child agreement. 24 REFERENCES 1. Ward, Elizabeth, et al. "Childhood and adolescent cancer statistics, 2014." CA: a cancer journal for clinicians 64.2 (2014): 83-103. 2. Howlader N, Noone AM, Krapcho M, Garshell J, Miller D, Altekruse SF, Kosary CL, Yu M, Ruhl J, Tatalovich Z,Mariotto A, Lewis DR, Chen HS, Feuer EJ, Cronin KA (eds). SEER Cancer Statistics Review, 1975-2012, National Cancer Institute. Bethesda, MD, http://seer.cancer.gov/csr/1975_2012/, based on November 2014 SEER data submission, posted to the SEER web site, April 2015. 3. Hudson, Melissa M., et al. "Health status of adult long-term survivors of childhood cancer: a report from the Childhood Cancer Survivor Study." Jama 290.12 (2003): 1583- 1592. 4. Eilertsen, Mary ‐Elizabeth B., et al. "Psychosocial health in children and adolescents surviving cancer." Scandinavian journal of caring sciences 25.4 (2011): 725-734. 5. Klassen, Anne F., et al. "Impact of caring for a child with cancer on parents’ health- related quality of life." Journal of Clinical Oncology 26.36 (2008): 5884-5889. 6. Varni, James W., Christine Limbers, and Tasha M. Burwinkle. "Literature review: health- related quality of life measurement in pediatric oncology: hearing the voices of the children." Journal of Pediatric Psychology 32.9 (2007): 1151-1163 7. Speyer, E., et al. "Agreement between children with cancer and their parents in reporting the child's health ‐related quality of life during a stay at the hospital and at home." Child: care, health and development 35.4 (2009): 489-495. 8. Russell, Kathryn MW, et al. "Assessment of health ‐related quality of life in children with cancer." Cancer 106.10 (2006): 2267-2274. 9. Parsons, Susan K., et al. "Health ‐related quality of life in pediatric bone marrow transplant survivors: According to whom?." International Journal of Cancer 83.S12 (1999): 46-51. 10. Parsons, Susan K., et al. "Comparing longitudinal assessments of quality of life by patient and parent in newly diagnosed children with cancer: the value of both raters’ perspectives." Quality of Life Research 21.5 (2012): 915-923. 11. Theunissen, N. C. M., et al. "The proxy problem: child report versus parent report in health-related quality of life research." Quality of Life Research 7.5 (1998): 387-397. 12. Eiser, Christine, and Rachel Morse. "Can parents rate their child's health-related quality of life? Results of a systematic review." Quality of Life Research 10.4 (2001): 347-357. 25 13. Yeh, Chao-Hsing, Chi-Wen Chang, and Pi-Chen Chang. "Evaluating quality of life in children with cancer using children's self-reports and parent-proxy reports." Nursing Research 54.5 (2005): 354-362. 14. Fluchel, Mark, et al. "Self and proxy ‐reported health status and health ‐related quality of life in survivors of childhood cancer in Uruguay." Pediatric blood & cancer 50.4 (2008): 838-843. 15. Czyzewski, Danita I., et al. "Measurement of quality of well being in a child and adolescent cystic fibrosis population." Medical care 32.9 (1994): 965-972. 16. Edelbrock, Craig, et al. "parent-child agreement on child psychiatric symptoms assessed via structured interview." Journal of Child Psychology and Psychiatry 27.2 (1986): 181- 190. 17. Sint Nicolaas, S. M., et al. "Assessment of Quality of Life in Children with Functional Constipation: Agreement between parent and child reports." 18. González, Hector M., William A. Vega, and Wassim Tarraf. "Health care quality perceptions among foreign-born Latinos and the importance of speaking the same language." The Journal of the American Board of Family Medicine 23.6 (2010): 745-752. 19. Milam, Joel E., et al. "Cancer ‐related follow‐up care among Hispanic and non ‐Hispanic childhood cancer survivors: The Project Forward study." Cancer 121.4 (2015): 605-613. 20. Meeske, Kathleen A., et al. "Mental health disparities between Hispanic and non- Hispanic parents of childhood cancer survivors." Pediatric blood & cancer 60.9 (2013): 1470-1477. 21. Varni, Jame(Varni, Limbers et al. 2007)s W., et al. "The PedsQL™ in pediatric cancer." Cancer 94.7 (2002): 2090-2106. 22. Varni, James W., Michael Seid, and Paul S. Kurtin. "PedsQL™ 4.0: Reliability and validity of the Pediatric Quality of Life Inventory™ Version 4.0 Generic Core Scales in healthy and patient populations." Medical care 39.8 (2001): 800-812. 23. Varni, James W., et al. "The PedsQL™* 4.0 as a pediatric population health measure: feasibility, reliability, and validity." Ambulatory Pediatrics 3.6 (2003): 329-341 24. Sim, Julius, and Chris C. Wright. "The kappa statistic in reliability studies: use, interpretation, and sample size requirements." Physical therapy 85.3 (2005): 257-268. 25. Landis, J. Richard, and Gary G. Koch. "The measurement of observer agreement for categorical data." biometrics (1977): 159-174. 26. McGraw, Kenneth O., and Seok P. Wong. "Forming inferences about some intraclass correlation coefficients." Psychological methods 1.1 (1996): 30. 26 27. Shrout, Patrick E., and Joseph L. Fleiss. "Intraclass correlations: uses in assessing rater reliability." Psychological bulletin 86.2 (1979): 420. 28. Ungar, Wendy J., et al. "A parent-child dyad approach to the assessment of health status and health-related quality of life in children with asthma." Pharmacoeconomics 30.8 (2012): 697-712. 29. Bartko, John J. "The intraclass correlation coefficient as a measure of reliability." Psychological reports 19.1 (1966): 3-11. 30. Bland, J. Martin, and Douglas G. Altman. "Measuring agreement in method comparison studies." Statistical methods in medical research 8.2 (1999): 135-160. 31. Bland, Martin. An introduction to medical statistics. OUP Oxford, 2015. 32. Giavarina, Davide. "Understanding Bland Altman analysis." Biochemia medica 25.2 (2015): 141-151. 33. Zaki, Rafdzah, et al. "Statistical methods used to test for agreement of medical instruments measuring continuous variables in method comparison studies: a systematic review." PloS one 7.5 (2012): e37908. 34. Herjanic, Barbara, and Wendy Reich. "Development of a structured psychiatric interview for children: Agreement between child and parent on individual symptoms." Journal of Abnormal Child Psychology 10.3 (1982): 307-324. 35. Welner, Zila, et al. "Reliability, validity, and parent-child agreement studies of the Diagnostic Interview for Children and Adolescents (DICA)." Journal of the American Academy of Child & Adolescent Psychiatry 26.5 (1987): 649-653. 36. Rebholz, Cornelia E., et al. "Agreement between parent and child report on parental practices regarding dietary, physical activity and sedentary behaviours: the ENERGY cross-sectional survey." BMC public health 14.1 (2014): 1. 37. Levi, Rachel B., and Dennis Drotar. "Health ‐related quality of life in childhood cancer: Discrepancy in parent): 1.ental practices regarding dietary, physical activity and sedentar 38. Pasch, Lauri A., et al. "Acculturation, Parent ‐Adolescent Conflict, and Adolescent Adjustment in Mexican American Families." Family Process 45.1 (2006): 75-86. 39. Solis, Julia M., et al. "Acculturation, access to care, and use of preventive services by Hispanics: findings from HHANES 1982-84." American journal of public health 80.Suppl (1990): 11-19. 40. DuBard, C. Annette, and Ziya Gizlice. "Language spoken and differences in health status, access to care, and receipt of preventive services among US Hispanics." American Journal of Public Health 98.11 (2008): 2021-2028. 27 41. Phipps, Sean, and Deo Kumar Srivastava. "Repressive adaptation in children with cancer." Health Psychology 16.6 (1997): 521. 42. Worchel, F. F. "Denial of depression: Adaptive coping in pediatric cancer patients." Newsletter of the Society of Pediatric Psychology 13 (1989): 8-11. 43. Cremeens, Joanne, Christine Eiser, and Mark Blades. "Factors influencing agreement between child self-report and parent proxy-reports on the Pediatric Quality of Life Inventory™ 4.0 (PedsQL™) Generic Core Scales." Health and quality of life outcomes 4.1 (2006): 1. 44. Canning, Emily Harris, et al. "Maternal distress and discrepancy in reports of psychopathology in chronically ill children." Psychosomatics 34.6 (1993): 506-511. 45. Newcomb, Andrew F., and Catherine L. Bagwell. "Children's friendship relations: A meta-analytic review." Psychological bulletin 117.2 (1995): 306. 46. Teja, Sameera, and Arnold L. Stolberg. "Peer support, divorce, and children's adjustment." Journal of divorce & remarriage 20.3-4 (1994): 45-64. 47. Alderfer, Melissa A., et al. "The social functioning of siblings of children with cancer: A multi-informant investigation." Journal of pediatric psychology 40.3 (2015): 309-319. 48. Skinner, Roderick, et al. "Long-term follow-up of people who have survived cancer during childhood." The lancet oncology 7.6 (2006): 489-498.also acceptable. 28 TABLES Table 1. Demographic and Clinical Characteristics of child-parent dyads (n=160) N(%) Child's characteristics Sex Female 81(50.6%) Male 79(49.4%) Age at first diagnosis, y <10 32(20.0%) 10-14 97(60.6%) 15+ 30(18.8%) Missing 1(0.6%) Age at survey completion, y 15-17 50(31.2%) 19-20 54(33.8%) 21-25 56(35.0%) Race/ethnicity Hispanic 82(51.3%) White, non-Hispanic 61(10.6%) Other 17(38.1%) Cancer diagnosis/site Leukemia 48(30.0%) Brain/CNS 27(16.9%) Bone 10(6.2%) Lymphoma 31(19.4%) Other 44(27.5%) Treatment Intensity 1 (Least intensive) 16(10.0%) 2 52(32.5%) 3 76(47.5%) 4 (Most intensive) 15(9.4%) Missing 1(0.6%) Hospital CHLA 135(84.4%) Long Beach 25(15.6%) Health Insurance Public 48(30.0%) Private 60(37.5%) None 24(15.0%) Unknown 22(13.7%) 29 Other 6(3.8%) Parent's characteristics Relationship to child Mother 138(86.2%) Father 18(11.3%) Other 4(2.5%) Parent ethnicity Hispanic 89(55.6%) White, non-Hispanic 49(30.6%) Other 22(13.8%) Parent Education Grade school/some High School 55(34.4%) HS graduate/GED 21(13.1%) Some college/AA 44(27.5%) College graduate/Graduate school 38(23.8%) Missing 2(1.2%) 30 Table 2. Summary of category, kappa value of dyads common questions according to Kappa range Kappa Range Category Question Kappa 95% CI P value %agreement Kappa≥0.75 Diagnosis and treatment Did the cancer treatment include radiation therapy? 0.908 0.842-0.975 <0.0001 95.63 Diagnosis and treatment Did the cancer treatment include chemotherapy? 0.878 0.790-0.966 <0.0001 95.63 Diagnosis and treatment Name of cancer child was diagnosed 0.814 0.747-0.882 <0.0001 85.00 0.500≤Kappa<0.75 Diagnosis and treatment Did the cancer treatment include surgery? 0.728 0.616-0.840 <0.0001 87.50 Health insurance Type of health insurance the child currently has 0.643 0.557-0.729 <0.0001 67.95 Health insurance Source of child's current medical coverage 0.530 0.440-0.621 <0.0001 63.75 Diagnosis and treatment When did the child finish cancer treatment? 0.512 0.374-0.650 <0.0001 80.62 0.25≤Kappa<0.5 Follow-up care received Has the child seen a doctor in the past 2 years 0.498 0.252-0.744 <0.0001 92.5 Problem in seeking care Problem of seeking care in referrals 0.4882 0.350-0.627 <0.0001 75.63 Program attendance Has the child ever gone to the LIFE Program? 0.4799 0.295-0.665 <0.0001 86.25 Diagnosis and treatment Has child or their parent ever shared the written summary of cancer treatment with any of child's doctors 0.4616 0.305-0.618 <0.0001 80.00 Problem in seeking care Problem of seeking care in transportation 0.455 0.236-0.673 <0.0001 89.38 Future follow-up care Chances that the child will go for a cancer follow up visit in next two years 0.450 0.312-0.588 <0.0001 38.75 Child health assessment Any health problems as a result of your cancer and/or its treatment 0.439 0.305-0.573 <0.0001 71.88 Follow-up care received When did the child last see a doctor for cancer related care? 0.436 0.286-0.587 <0.0001 76.88 Health insurance Difficulty obtaining health insurance because of cancer treatment history 0.416 0.248-0.584 <0.0001 80.00 Parental/Family influence How often does one of your parents go with your child to see a doctor (for cancer related care) 0.414 0.260-0.568 <0.0001 76.88 Follow-up care received Does the doctor for child's regular check-ups not related to cancer treatment know about the child's cancer history 0.408 0.2624-0.5526 <0.0001 73.13 Child Health assessment Rate the child's general health overall 0.393 0.296-0.489 <0.0001 43.13 Follow-up care Received Does the child have a doctor for regular check-ups or problems not related to cancer treatment 0.382 0.232-0.532 <0.0001 73.75 Follow-up care Received Type of doctor for check-ups related to cancer/treatment 0.382 0.382-0.493 <0.0001 58.75 Parental/Family influence How often does one of your parents go with child to see a doctor (routine med care) 0.353 0.203-0.504 <0.0001 73.13 Program attendance Has child ever gone to the Late Effects Program or Susan Shannon services? 0.343 -0.026-0.712 <0.0001 95.63 Follow-up care received Does the child have a doctor for check-ups or problems related to cancer / treatment 0.324 0.167-0.481 <0.0001 71.25 Follow-up care received Has child gone to a Curandero/a, Herbalist or … for health care in the past 2 years? 0.323 0.0288-0.6174 <0.0001 93.13 Parental/Family influence Have child and at least one of your parents discussed the health care the child will need in the future 0.308 0.1547-0.4613 <0.0001 70.00 Follow-up care received type (s) of doctor(s) child have seen for cancer related care 0.304 0.304-0.402 <0.0001 48.75 Follow-up care received Type (s) of doctor(s) child saw in the past 2 years 0.284 0.190-0.379 <0.001 43.13 Program attendance Has child ever gone to cancer clinic at other facility? 0.274 -0.007-0.554 0.0004 91.88 Future follow-up care How long do you think child should continue having cancer related follow- up care 0.272 0.122-0.423 <0.001 42.50 Future follow-up care Chances that child will go to any medical doctor for a routine check-up during the next 2 years 0.261 0.115-0.406 0.0007 63.75 Follow-up care received Was child's last visit to a doctor within the past year 0.259 0.083-0.436 <0.0001 76.25 Follow-up care received When did child last see a doctor for a routine medical care 0.256 0.144-0.369 <0.001 61.88 Kappa<0.25 Future follow-up care Have any of child's doctors discussed the health care that your child will need in the future 0.209 0.058-0.360 0.0067 63.13 Health insurance Difficult dealing with a health insurance company or HMO during the 2 past years 0.136 0.047-0.226 0.0005 32.50 Parental/Family influence How much does family influence child's health care decisions 0.044 -0.069-0.156 0.4566 45.00 31 Table 3. Question 1: When did the child last see a doctor for cancer related care? Parent Child Within past 2 years All other Total Within past 2 years 96 12 108 All other 25 27 52 Total 121 39 160 % agreement = (96+27)/160=76.88% Kappa= 0.4364 32 Table 4. Univariate analysis and Multivariable Model for characteristics associated with higher agreement on Question 1 (N=160) Univariate analysis Multivariable Model Characteristic Agree 123 (76.9%) Disagree 37(23.1%) χ 2 P value Adjusted OR [95%CI] p value CCS Age at Survey Completion 15-17 35(70.0%) 15(30.0%) 2.07 0.36 1.00(ref) - 18-20 44(81.5 %) 10(18.5%) 1.72 [0.63-4.68] 0.29 21-25 44(78.6%) 12(21.4%) 1.44[0.53-3.97] 0.48 Ptrend=0.31 CCS ethnicity Non-Hispanic 60(83.3%) 12(16.7%) 3.07 0.08 1.00(ref) - Hispanic 63 (71.6%) 25(28.4%) 0.89[0.28-2.82] 0.84 Cancer Type Leukemia 30(62.5%) 18(37.5%) 11.66 *0.02 1.00(ref) - Brain and Other Nervous System 19(70.4%) 8 (29.6%) 1.21[0.40 -3.71] 0.74 Bones and joints 8(80.0%) 2(20.0%) 2.23[0.41 -13.47] 0.34 Lymphoma 28(90.3%) 3(9.7%) 4.88[1.23-19.36] 0.02 Others 38(86.4%) 6(13.6%) 3.67[1.21-11.10] 0.02 Treatment Intensity Least intensive 14(87.5%) 2 (12.5%) 3.51 0.17 Moderately intensive 43(82.7%) 9(17.3%) Very/Most intensive 65(71.4%) 26(28.6%) Parent Education Grade school/some High School 38(69.1%) 17(30.9%) 7.17 0.07 1.00(ref) - HS graduate/GED 14(66.7%) 7(33.3%) 0.82[0.24-2.77] 0.75 Some college/AA 36(81.8%) 8(18.2%) 1.69[0.47-6.05] 0.42 College graduate/Graduate school 34(89.5%) 4 (10.5%) 3.13[0.6-16.30] 0.18 Ptrend=*0.01 Parent Language English speaking 102(81.0%) 24(19.0%) 5.55 *0.02 1.00(ref) - Spanish only 21(61.8%) 13(38.2%) 0.7[0.23-2.11] 0.53 Living with parents Not living with parents 27(81.8%) 6(18.2%) 0.57 0.45 Living with parents 96(75.6%) 31(24.4%) 33 Table 5. Question2: How long do you think the child should continue having cancer follow-up care? Parent Child Child's whole lifetime All other Total Child's whole lifetime 68 32 100 All other 24 36 60 Total 92 68 160 % agreement = 68/160=42.5% Kappa= 0.2727 34 Table 6. Univariate analysis and Multivariable Model for characteristics associated with agreement on the preferred answer of ‘child’s whole lifetime’ Characteristic Univariate analysis Multivariable Model Agree Disagree χ 2 P value Adjusted OR [95%CI] p value 68 (42.5%) 92(57.5%) CCS Age at Survey Completion 15-17 23(46.0%) 27(54.0%) 0.91 0.64 1.00(ref) - 18-20 24(44.4 %) 30(55.6%) 0.85 [0.37-1.92] 0.69 21-25 21(37.5%) 35(62.5%) 0.72[0.29-1.47] 0.30 Ptrend=0.37 CCS ethnicity Non-Hispanic 35(48.6%) 37(51.4%) 2.00 0.16 1.00(ref) - Hispanic 33 (37.5%) 55(62.5%) 0.90[0.37-2.19] 0.81 Cancer Type Leukemia 17(35.4%) 31(64.6%) 2.89 0.58 Brain and Other Nervous System 13(48.2%) 14 (51.8%) Bones and joints 5(50.0%) 5(50.0%) Lymphoma 16(51.6%) 15(48.4%) Others 17(38.6%) 27(61.4%) Treatment Intensity Least intensive 4(25.0%) 12 (75.0%) 2.1442 0.34 Moderately intensive 23(44.2%) 29(55.8%) Very/Most intensive 40(44.0%) 51(56.0%) Parent Education Grade school/some High School 16(29.1%) 39(70.9%) 6.27 0.10 1.00(ref) - HS graduate/GED 11(52.4%) 10(47.6%) 2.17[0.73-6.40] 0.16 Some college/AA 22(50.0%) 22(50.0%) 1.49[0.54-4.10] 0.44 College graduate/Graduate school 18(47.4%) 20 (52.6%) 1.29[0.39-4.27] 0.68 Ptrend=0.05 Parent Language English speaking 60(47.6%) 66 (52.4%) 6.36 *0.01 1.00(ref) - Spanish only 8(23.5%) 26(76.5%) 0.4[0.14-1.14] 0.09 Living with parents Not living with parents 12(36.4%) 21(63.6%) 0.64 0.42 Living with parents 56(44.1%) 71(55.9%) 35 Table 7. Question 3: During the next 2 years, what are the chances that child will go for a cancer follow-up visit? Parent Child Very likely All other Total Very likely 62 26 88 All other 18 54 72 Total 80 80 160 % agreement =62/160=38.75% Kappa=0.45 36 Table 8. Univariate analysis and Multivariable Model for characteristics associated with agreement on the preferred answer of ‘very likely’ Characteristic Univariate analysis Multivariable Model Agree Disagree χ 2 P value Adjusted OR [95%CI] p value 62 (38.8%) 98(61.2%) CCS Age at Survey Completion 15-17 19(38.0%) 31(62.0%) 0.57 0.75 1.00(ref) - 18-20 23(42.6 %) 31(57.4%) 1.12 [0.48-2.60] 0.79 21-25 20(35.7%) 36(64.3%) 0.81[0.35-1.89] 0.63 Ptrend=0.79 CCS ethnicity Non-Hispanic 37(51.4%) 35(48.6%) 8.81 *0.003 1.00(ref) - Hispanic 25 (28.4%) 63(71.6%) 0.46[0.19-1.14] 0.09 Cancer Type Leukemia 19(39.6%) 29(60.4%) 1.1156 0.89 Brain and Other Nervous System 10(37.0%) 17(63.0%) Bones and joints 5(50.0%) 5(50.0%) Lymphoma 13(42.0%) 18(58.1%) Others 15(34.1%) 29(65.9%) Treatment Intensity Least intensive 4(25.0%) 12(75.0%) 1.47 0.48 Moderately intensive 21(40.4%) 31(59.6%) Very/Most intensive 37(40.7%) 54(59.3%) Parent Education Grade school/some High School 14(25.4%) 41(74.6%) 6.97 0.07 1.00(ref) - HS graduate/GED 8(38.1%) 13(61.9%) 1.31[0.42-4.10] 0.65 Some college/AA 20(45.4%) 24(54.6%) 1.13[0.40-3.19] 0.82 College graduate/Graduate school 19(50.0%) 19(50.0%) 1.04[0.32-3.74] 0.95 Ptrend=*0.01 Parent Language English speaking 56(44.4%) 70(55.6%) 8.10 *0.004 1.00(ref) - Spanish only 6(17.5%) 28(82.4%) 0.43[0.14-1.34] 0.14 Living with parents Not living with parents 9(27.3%) 24(72.7%) 2.31 0.13 Living with parents 53(41.7%) 74(58.3%) 37 Table 9. Child-rated and parent-rated child quality of life on the PedsQL™ Scale N Mean(SD) Median(IQR) Mean Difference a (SD) p value for child-parent difference b Child response Total Scale(23 items) 160 74.85(17.55) 76.63(63.04-88.04) 1.44(17.5) 0.3004 Physical Functioning (8 items) 159 76.14(22.61) 81.25(62.50-93.75) 1.91(21.11) 0.2565 Emotional Functioning (5 items) 160 69.70(24.67) 75.00(52.50-90.00) 2.67(25.82) 0.1987 Social Functioning (5 items) 160 84.98(17.63) 90.00(75.00-100.00) 5.02(22.10) 0.0048 School Functioning (5 items) 158 67.74(20.99) 70.00(55.00-95.00) - 3.00(22.85) 0.1025 Psychosocial Health Summary Score (15 items) 160 74.13(17.66) 73.33(63.33-88.33) 1.30(19.29) 0.3976 Parent response Total Scale(23 items) 160 73.41(20.96) 78.33(58.33-90.22) Physical Functioning (8 items) 159 74.32(25.69) 84.38(56.25-96.88) Emotional Functioning (5 items) 156 67.73(26.56) 70.00(50.00-95.00) Social Functioning (5 items) 158 80.36(22.57) 88.75(65.00-100.00) School Functioning (5 items) 158 71.37(25.18) 80.00(50.00-93.75) Psychosocial Health Summary Score (15 items) 158 73.12(21.57) 76.67(58.93-91.07) a Difference was calculated by (child score-parent score) b Significance was ascertained by paired t-test SD: standard deviation; IQR: interquartile range. 38 Table 10.Intra-Class correlation coefficients (ICC) a between child-rated and parent rated child PedsQL CCS Age at Survey Completion PedsQL TM Total (n=160) 15-17 years (n=50) 18-20 years(n=54) 21-25 years(n=56) ICC(95%CI) ICC(95%CI) ICC(95%CI) ICC(95%CI) Total Scale(23 items) *0.590 (0.479,0.682) *0.604 (0.393,0.754) *0.560 (0.3460.719) *0.603 (0.406,0.746) Physical Functioning (8 items) *0.620 (0.515,0.707) *0.608 (0.400,0.757) *0.462 (0.348,0.720) *0.672 (0.497,0.795) Emotional Functioning (5 items) *0.484 (0.354,0.595) *0.539 (0.309,0.709) 0.400 (0.148,0.604) *0.509 (0.279,0.684) Social Functioning (5 items) *0.397 (0.258,0.521) *0.473 (0.227,0.662) 0.254 (-0.015,0.489) *0.451 (0.213,0.638) School Functioning (5 items) *0.512 (0.387,0.619) *0.502 (0.262,0.683) *0.553 (0.334,0.715) *0.480 (0.243,0.663) Psychosocial Health Summary Score (15 items) *0.518 (0.394,0.624) *0.573 (0.353,0.733) *0.510 (0.280,0.684) *0.478 (0.246,0.659) a Two-way mixed effects ICC coefficient model. Significance was ascertained by an F-test for difference between the observed ICC coefficient and ICC coefficient = 0. * p<0.001 39 FIGURES Figure 1. Bar chart of mean PedsQL score from parent and child response 74.85 76.14 69.7 84.98 67.74 74.13 73.41 74.32 67.73 80.36 71.37 73.12 0 10 20 30 40 50 60 70 80 90 Total Scale Physical Functioning Emotional Functioning Social Functioning School Functioning Psychosocial Health Summary Score Child Parent 40 Figure 2 Bland-Altman plot of PedsQL Total Score including regression line -60 -40 -20 0 20 40 60 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean -1.4 -1.96 SD -35.7 +1.96 SD 32.9 ----regression line 41 Figure 3 Bland-Altman plot of PedsQL Physical Functioning Score including regression line -80 -60 -40 -20 0 20 40 60 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean -1.9 -1.96 SD -43.3 +1.96 SD 39.5 ----regression line 42 Figure 4 Bland-Altman plot of PedsQL Emotional Functioning Score including regression line -100 -80 -60 -40 -20 0 20 40 60 80 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean -2.7 -1.96 SD -53.3 +1.96 SD 47.9 ----regression line 43 Figure 5 Bland-Altman plot of PedsQL Social Functioning Score including regression line -80 -60 -40 -20 0 20 40 60 80 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean -5.0 -1.96 SD -48.3 +1.96 SD 38.3 ----regression line 44 Figure 6 Bland-Altman plot of PedsQL School Functioning Score including regression line -60 -40 -20 0 20 40 60 80 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean 3.0 -1.96 SD -41.8 +1.96 SD 47.8 ----regression line 45 Figure 7 Bland-Altman plot of PedsQL Psychosocial Health Summary Score including regression line -80 -60 -40 -20 0 20 40 60 0 20 40 60 80 100 120 Average((parent score +child score)/2) Difference(parent score-child score) Mean -1.3 -1.96 SD -39.1 +1.96 SD 36.5 ----regression line
Abstract (if available)
Abstract
Background: The number of childhood cancer survivors (CCS) has been increasing greatly over recent decades due to longer term survival. Both CCS and their parents experience health problems caused by the child’s cancer and cancer treatment. Parents often make health decisions as a proxy for their children when children are too ill or young. Knowing how well the parent’s view of the child’s health compares to the child’s perspective is important, however previous studies have not provided consistent conclusions about the agreement between child-parent dyads on ratings of child’s health-related quality of life (HRQoL) and have been based on small numbers of dyads. The objectives of this study were: (1) to assess the consistency of responses between child-parent dyads to questions regarding use of health care and health assessment of the child
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Childhood cancer survivorship: parental factors associated with survivor's follow-up care behavior and mental health
PDF
Association between informed decision-making and mental health-related quality of life in long term prostate cancer survivors
PDF
Associations between exercise and quality of life in adult survivors of non-Hodgkin's lymphoma and colorectal cancer
PDF
Early childhood health experience & adult phenotype in twins
PDF
Associations between isoflavone soy protein (ISP) supplementation and breast cancer in postmenopausal women in the Women’s Isoflavone Soy Health (WISH) clinical trial
PDF
Melanoma in children, adolescents, and young adults
PDF
Diet quality and pancreatic cancer incidence in the multiethnic cohort
PDF
Red and processed meat consumption and colorectal cancer risk: meta-analysis of case-control studies
PDF
Multilevel sociodemographic correlates of the health and healthcare utilization of childhood cancer survivors
PDF
Racial/ethnic differences in colorectal cancer patient experiences, health care utilization and their association with mortality: findings from the SEER-CAHPS data
PDF
An analysis of disease-free survival and overall survival in inflammatory breast cancer
PDF
Predictive factors of breast cancer survival: a population-based study
PDF
The study of germ cell tumors and related conditions: an analysis of self-reported data with characterization and comparison of Family History Questionnaires respondents and non-respondents
PDF
Native American ancestry among Hispanic Whites is associated with higher risk of childhood obesity: a longitudinal analysis of Children’s Health Study data
PDF
Evaluation of factors influencing Los Angeles Tiered-Dispatch System’s improvement on bystander CPR rate and inter reliability between electronic patient care report (ePCR) and 911 call review on...
PDF
Street connectivity and childhood obesity: a longitudinal, multilevel analysis
PDF
Association of subclinical atherosclerosis with plasma B-vitamin, cysteine, homocysteine, and cysteinyl glycine in a cardiovascular disease-free population
PDF
Multilevel influences of care engagement and long-term survival among childhood, adolescent, and young adult cancer survivors
PDF
Validation of the Children’s International Mucositis Evaluation Scale (ChIMES) in pediatric cancer and SCT
PDF
Health-related quality of life in preschool children with strabismus or amblyopia
Asset Metadata
Creator
Zhuang, Xueyan
(author)
Core Title
Factors associated with agreement between childhood cancer survivors (CCS) and their parents on knowledge and attitudes about health-related quality of life, treatment and follow-up care
School
Keck School of Medicine
Degree
Master of Science
Degree Program
Applied Biostatistics and Epidemiology
Publication Date
07/28/2016
Defense Date
07/25/2016
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
cancer,Childhood,follow-up care,health-related quality of life,OAI-PMH Harvest,parent agreement,survivorship
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Hamilton, Ann (
committee chair
), Mack, Wendy (
committee chair
), Milam, Joel (
committee member
)
Creator Email
xueyanzh@usc.edu,zxytessa@gmail.com
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-283772
Unique identifier
UC11280313
Identifier
etd-ZhuangXuey-4650.pdf (filename),usctheses-c40-283772 (legacy record id)
Legacy Identifier
etd-ZhuangXuey-4650.pdf
Dmrecord
283772
Document Type
Thesis
Format
application/pdf (imt)
Rights
Zhuang, Xueyan
Type
texts
Source
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 a...
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
Tags
follow-up care
health-related quality of life
parent agreement
survivorship