Social Determinants of Health and Informed Consent Comprehension for Pediatric Cancer Clinical Trials

Key Points Question Are social determinants of health and sociocontextual factors associated with informed consent (IC) comprehension in therapeutic childhood cancer clinical trials? Findings In this cross-sectional study that included 223 parents of children with newly diagnosed cancer, limited health literacy and use of Spanish language for medical communication were associated with lower comprehension of IC. Meaning These findings suggest that parents who have limited health literacy and who use Spanish language for medical communication may not fully comprehend IC and may not be able to make informed decisions for their children.


Introduction
Before children can participate in a clinical trial, their parents or guardians must provide informed consent (IC), which is a fundamental ethical obligation in health care.Key tenets of valid IC include information disclosure, competent decision-making, adequate comprehension, and voluntariness. 1,2[11] Health literacy is the ability to access and comprehend health information to successfully navigate health care. 12,13Limited health literacy is associated with negative health outcomes in children and adults, such as increased rates of hospitalization and lower adherence to medication regimens and preventive care. 13search investigating the role of SDOH in IC comprehension in childhood cancer clinical trials in diverse populations is lacking, 9,10,12 particularly among Hispanic individuals and those who use a language other than English for medical communication.To address these gaps, we conducted a cross-sectional study to investigate associations of SDOH and sociocontextual factors (eg, ethnicity, satisfaction with IC, and cancer type) with IC comprehension among parents enrolling their child with newly diagnosed cancer in a therapeutic clinical trial.We hypothesized that parents of Hispanic ethnicity, those who use Spanish language for medical communication, or with limited health literacy would have lower IC comprehension.

Study Participants
Our cross-sectional study included parents and legal guardians of children with newly diagnosed cancer at Rady Children's Hospital San Diego (RCHSD), a large quaternary academic pediatric hospital that serves more than 1 million children and has a large Hispanic representation (45%).This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies, which consists of a checklist of 22 items that are considered essential for good reporting of observational studies.The institutional review board for the University of California, San Diego and RCHSD approved this study.Eligible individuals were approached, and voluntary written IC was obtained from participants after a full description of study procedures, per the Common Rule.Language used for medical communication was assessed during recruitment.No one received compensation or incentive for study participation.
Children aged 0 to 17 years diagnosed with cancer between October 1, 2014, and March 31, 2021, were identified from RCHSD's cancer registry.Parents were eligible to participate if (1) their

JAMA Network Open | Ethics
SDOH and Informed Consent Comprehension for Pediatric Cancer Clinical Trials child was eligible for a therapeutic clinical trial, (2) they had consented to participate in the therapeutic clinical trial within the previous week, and (3) they understood written and spoken English or Spanish.We excluded parents who had given IC for any therapeutic clinical trial in the past (as these individuals likely had a better understanding of IC due to prior exposure), who had a child with second malignant neoplasm or relapsed disease, or who had a child previously diagnosed with cancer at an outside institution.Standard of care IC for the therapeutic clinical trial was conducted by the treating oncologist in the language used by the parent for medical communication and, if the clinician was not fluent in that language, with a medically trained interpreter either in person (preferred method) or via video conference.Informed consent documents for therapeutic clinical trials at RCHSD were available in English and Spanish.

Study Procedures
Our primary outcome was objective comprehension of the IC and related domains (purpose, procedures, and randomization; risks and benefits; alternatives; and voluntariness) and their associations with SDOH (marital status, language used for medical communication [verbal and written], educational attainment, employment, insurance type, SES, and health literacy) and sociocontextual factors (ethnicity, satisfaction with IC, and cancer type).Participants were approached in inpatient and outpatient hospital settings and completed questionnaires within 1 week of the IC discussion for the therapeutic clinical trial.Questionnaire items were related to IC comprehension, sociodemographic characteristics, health literacy, and satisfaction with IC.
5][16][17] All questionnaire questions were written with a low level of complexity and high readability. 13,15Both parents (if available) were invited to participate in the study, and they completed the questionnaires independently.We collected data from the medical record, including patient's age, sex, and cancer type.

Study Measures
To assess objective comprehension of the basic elements of IC, we used the 20-item brief, reliable, and valid Quality of Informed Consent (QuIC) instrument. 18Its domains are based on the IC guidelines from the US Department of Health and Human Services.We analyzed overall IC comprehension and comprehension of domains related to purpose, procedures, and randomization; risks and benefits; alternatives; and voluntariness.Scores range from 0 to 100; higher scores are indicative of increased comprehension.The QuIC has been validated in adults with cancer and used among parents of children with cancer. 19,20The SDOH were assessed using a sociodemographic questionnaire, 12 which included parental age, self-reported race and ethnicity (categorized per US Census Bureau), 21 marital status, language used for medical communication (verbal and written), educational attainment, employment status and occupation, and insurance type.Socioeconomic status was calculated using the Hollingshead index (marital status, retired or employed status, educational attainment, and occupational prestige). 22The 6-item Newest Vital Sign (NVS) evaluated health literacy based on interpretation of a nutrition label. 15Scores of 0 to 3 indicate limited health literacy; 4 to 6, adequate health literacy.5][26] The 7-item satisfaction questionnaire CCG-S9901 has been used in childhood cancer 8,27 and assessed perceptions of overall satisfaction with IC discussions, clarity of explanations, quantity of information, and utility of the IC.Higher scores indicate greater satisfaction.The NVS has been validated in English and Spanish. 15][30]

Statistical Analysis
Continuous (eg, SES) and categorical variables (eg, health literacy) were summarized by mean (SD) and count (percentage), respectively.The primary outcome of interest was overall comprehension of IC, and the secondary outcomes were comprehension of related IC domains.The association of outcomes with SDOH and sociocontextual factors (ethnicity, satisfaction with IC, and cancer type) was assessed using a linear mixed-effects model.A random intercept structure was included to account for the cluster effect of parents and guardians from the same family.A parsimonious multivariable linear mixed-effects model was fit with IC comprehension and related domains as the outcome variables.Hispanic ethnicity was the primary covariate of interest and was forced into the multivariable model regardless of its significance level.The SDOH and contextual factors were included in the initial multivariable model only if they were significant in univariable models with P < .20 as the cut point.Backward elimination was used to remove insignificant variables from the model until all variables were at P < .10.Coefficient estimates (β) and 95% CIs were reported.The threshold for statistical significance for all statistical analyses was 2 sided, and significant associations were defined as P Յ .05.Sample size for enrollment was based on our a priori power calculation to achieve 80% power at the joint 2-sided P Յ .05significance level to detect a significant Cohen effect size in IC comprehension of 0.31.Missing data were minimized by intensive training of the research staff in techniques of clarifying answers and checking questionnaires while participants were on site.In case of missing data, the missing pattern was examined, and appropriate data analytic techniques were used.All analyses were conducted from January 1, 2022, to July 31, 2023, using statistical software program R, version 4.1.2(R Project for Statistical Computing).

Participant Characteristics
In total, 885 patients aged 0 to 17 years were diagnosed with cancer during the study period, and 309 parents met inclusion criteria for this study and were prospectively approached for enrollment.

Purpose, Procedures, and Randomization Domain
We assessed comprehension of the purpose, procedures, and randomization process.In multivariable analysis, older vs younger parents (mean [SD], 70.31 [11.82] vs 72.56 [13.43]; β b Scores range from 0 to 100, with higher scores indicating higher comprehension.

Risks and Benefits Domain
We assessed comprehension of risks and benefits.In multivariable analysis, limited vs adequate health literacy (mean [SD], 62.84 [20.24]

Alternatives Domain
We

Voluntariness Domain
We assessed comprehension of voluntariness.In multivariable analysis, Spanish vs English language used for medical communication (mean

Discussion
This cross-sectional study investigated SDOH and sociocontextual factors associated with IC comprehension among parents who had consented to their child's participation in a therapeutic clinical trial for cancer.One of our key findings was that, on multivariable analysis, limited health literacy was consistently associated with lower overall comprehension of IC and all IC comprehension domains analyzed (purpose, procedures, and randomization; risks and benefits; alternatives; and voluntariness).Health literacy is closely linked to literacy and entails the individual's knowledge and skills to access, comprehend, and apply health information and thereby make informed health decisions. 13A review of over 100 IC documents for clinical trials found an average grade 12 reading level, suggesting that limited literacy poses a barrier to IC comprehension. 31Health literacyconcordant medical communication is critical for patients to successfully navigate the complexities of clinical trials. 12,31We previously reported that limited health literacy is associated with decreased perception of voluntariness during IC for clinical trials in leukemia. 12In the present study, we found that limited health literacy was associated with lower comprehension of IC voluntariness, suggesting that parents may not understand the key concept of providing voluntary IC.
Consistent with our findings, past research has revealed challenges with parental comprehension of essential elements of IC.Knifed et al 32 reported that 33% of parents did not recall any risks associated with the clinical trial or recalled only 1 to 4 risks, which tended to be generic.
Many parents had difficulty understanding study procedures, randomization, and alternatives to the clinical trial. 3,4,8,32One study 33 showed that only 32% of parents had an understanding of the

JAMA Network Open | Ethics
SDOH and Informed Consent Comprehension for Pediatric Cancer Clinical Trials purpose behind the clinical trial, and most of these parents were non-Hispanic White individuals with higher SES.Therefore, implementation of strategies to enhance IC comprehension is critical.
There is a dramatic underrepresentation of children of Hispanic and Spanish-speaking parents participating in clinical trials for childhood cancer. 34,35Our study included a significant proportion of Hispanic parents (49.8%) and individuals who used Spanish for medical communication (37.8% of Hispanic parents), allowing assessment of outcomes in these understudied populations.Use of Spanish language for medical communication was associated with lower overall IC comprehension and lower comprehension of the purpose, procedures, and randomization and voluntariness domains, highlighting the importance of language-concordant IC delivery to ensure IC comprehension.Among Hispanic cultural norms, respect may affect IC comprehension, 36 as Hispanic parents may be hesitant to ask clarifying questions to the clinician.In a qualitative study, parents who communicated in a language other than English were less likely to ask questions compared with those who used English when presented with IC forms for cancer clinical trials. 9Similarly to culture, language discordance plays an important role in the delivery of IC.Discussions of IC have been reported to be about equal in duration for both participants who use an interpreter and those who use English, which implies that clinicians may shorten the IC discussion for individuals who do not use English for medical communication. 9Also, studies have shown that parents who use a language other than English had greater difficulty understanding "their right to withdraw from research studies" or randomization, 9 even 6 months after their child was enrolled in a clinical trial. 37Our finding of lower comprehension of the purpose, procedures, and randomization domain being associated with use of Spanish language for medical communication is congruent with the literature. 9,10 also found that unmarried marital status, older parental age, and lower SES were associated with lower IC comprehension.Associations between marital status and IC are not well delineated.
Discussion among partners or "teach back" (where individuals explain health information in their own words 38 ) to each other may improve IC comprehension.Moreover, unmarried parents and those with lower SES may be burdened by other responsibilities that detract from focus and time spent understanding IC, or they may be more likely to come from underserved backgrounds and sociocultural contexts may influence their comprehension. 10,11Our results support identifying unmarried, older parents and those with lower SES who may need additional support.These findings contribute to the scant literature on adverse SDOH and associations with outcomes in parents of children with cancer. 10,12,35 our study, less satisfaction with IC was associated with lower IC comprehension.This association is not well understood.In contrast to prior research that reported that parents who did not comprehend basic elements of IC were still satisfied with the IC process, 4,8 our findings suggest that there may be a direct association between overall perceived satisfaction with IC and IC comprehension that warrants further investigation.Based on our results, we postulate that parental satisfaction with IC is a proxy for effective IC delivery by the clinician, and effective delivery is essential for comprehension.
Obtaining IC in the oncology setting poses unique challenges.With the devastating news of a childhood cancer diagnosis, the shock experienced by parents can affect IC comprehension. 11,27ditionally, the time-sensitive nature of starting cancer treatment creates a high-stakes environment that is not conducive to comprehension. 39The length of IC documents has also increased by 10-fold within a span of 3 decades. 31,39Additionally, there is a reported lack of formal training for clinicians to successfully deliver IC. 11 To improve IC comprehension for clinical trials in childhood cancer, it is imperative that IC discussions are tailored to the language, health literacy, and cultural needs of families.Furthermore, using in-person interpreters facilitates a better understanding of IC content in different languages. 40,41For parents who may hesitate to ask questions, engaging them in decision-making and assessing their understanding using techniques such as the teach back method can be helpful. 36,39Moreover, a tiered, staged IC approach 42 and interventions using decision aids and

JAMA Network Open | Ethics
SDOH and Informed Consent Comprehension for Pediatric Cancer Clinical Trials patient navigation may improve IC comprehension and decision-making. 43,44Finally, strategies at the clinician and health care system levels are also urgently needed. 11,35,45

Limitations
This study has some limitations.Although our overall sample was representative of the population we serve in our catchment area, it was limited by the inclusion of more mothers than fathers, a small proportion of Asian or Pacific Islander and Black individuals, and the inclusion only of English-and Spanish-speaking parents.Hispanic participants were primarily from Mexico, and although individuals of Mexican descent are the largest Hispanic group (62%) in the US, 46 this may constrain the generalizability of our findings to other Latin American subgroups.Our analysis accounted for the cluster effect of parents from the same family; however, we did not have sufficient parental couples to perform paired cross-parent comparisons, thereby potentially causing bias or precluding comparisons among subgroups.Future research should include a larger and more diverse sample to allow for subgroup comparisons and sensitivity analyses.Our sample included mostly parents of patients with hematological malignant neoplasms, which are the most common pediatric cancer; nevertheless, this may limit the applicability of our findings to parents of children with solid tumors.
Of note, during the study period, there were more therapeutic clinical trials open for enrollment for hematological than for solid tumors.Additionally, if the participant completed the IC comprehension questionnaire prior to completing the satisfaction with IC questionnaire, this could have affected satisfaction scores.Furthermore, we did not survey those who declined to participate in the therapeutic clinical trial to determine reasons for declining.Thus, our findings may not represent those who have baseline hesitancy about research participation.Another limitation is that participants with lower health literacy might have had greater difficulty understanding the questionnaires as well as the IC, which might have contributed to the observed differences by health literacy.8][49] We acknowledge that we did not conduct a comprehensive assessment of all domains of SDOH and analyzed mostly those at the individual level, limiting the understanding of the role of community and neighborhood factors.

Conclusions
In this cross-sectional study, although IC comprehension is conceptually complex, the significant associations we found for limited health literacy and Spanish used for medical communication with

6
Figure.Study Flow Diagram

Future
research should include inclusive evaluations of SDOH, characterization of cultural and linguistic patient-clinician concordance, theory-guided qualitative studies to further understanding of barriers and facilitators to adequate IC, assessments of longitudinal retention of the information in the IC form, and evaluation of the role of clinician training in cultural and health literacy competencies to enhance IC delivery.Last, inclusion of sociocontextual factors in the same multivariable model including SDOH may obscure or exacerbate the associations of variables with the outcomes.
low IC comprehension highlight the critical need to timely identify and support individuals with adverse SDOH, including limited health literacy and/or use of a language other than English during clinical trial recruitment.Future research should be conducted to ensure that all clinical trial participation is voluntary and that all elements of IC are fully comprehended by the participant or proxy decision-maker.Our findings support the investigation across pediatric disciplines of the potential role of culturally, linguistically, and health literacy-concordant interventions to improve IC comprehension and decision-making and ensure more equitable research participation in racial and ethnic minority populations.
a Of the potential 410 eligible parents, only 309 were approached.JAMA Network Open | Ethicsa Unless otherwise indicated, data are expressed as b Includes single, divorced, widowed, and separated.c Six parents selected "prefer not to say." e Scores range from 7 to 21, with higher scores indicating greater satisfaction.f Includes 178 children (86 non-Hispanic and 92 Hispanic).

Table 2 .
Outcome Measures of Study Participants a a Data are presented as mean (SD).Demographics did not vary significantly by parental study participation vs decline.