Research Participation of Minor Adolescents in Foster Care

Abstract

Study Objective:

This study evaluated published studies about foster care to a) determine the types of data used; b) describe the degree to which a sexual/reproductive health topic was addressed; and c) describe the consent process.

Design:

Analysis of published literature

Setting:

PubMed was searched using “foster care” for English articles published between January 1, 2017, and September 4, 2019.

Participants:

None

Interventions:

None

Main Outcome Measures:

Articles were coded into four data source categories: primary, secondary, peripheral or perspective data. Articles with a primary data source were coded for participant ages: only ≤ 9 years-old, included 10–17-year-olds (minor adolescents) and only ≥18 years-old. Articles using a secondary data source were coded for the source of the data registry. All articles were coded for presence of a sexual/reproductive health outcome. The primary data articles that included minor adolescents were coded for the study topic and consent process.

Results:

Of the 176 articles about foster care, 72 (41%) used primary data, 53 (30%) used secondary data and 51 (29%) used peripheral/perspective data. Forty-eight of the primary data articles included minor adolescents. Secondary data sources included few national research surveys. Sexual/reproductive health outcomes were measured in 17 articles, 4 of which used primary data. The consent process for minor adolescents varied and had no consistent pattern across studies.

Conclusion:

Research best practices for consent processes and use of registries could be developed to increase research on sexual/reproductive health outcomes among adolescents in foster care.

Introduction

Adolescents in foster care are a unique population with known sexual health risks and vulnerabilities.^6–9 While there is a shared perspective among a variety of stakeholders that research is important to improve the health of youth in foster care, there continues to be limited research to develop an evidence base to diagnose, intervene and improve health conditions for adolescents in foster care.¹⁰ Youth sometimes end up in foster care due to histories of sexual maltreatment or abuse, which is a known independent risk factor for sexual risk taking in adolescence.^{11, 12} Beyond this contributing factor of sexual abuse, in general, youth in foster care have higher rates of teen pregnancy, sexually transmitted infections and sexual risk behaviors than their peers not in foster care.^6–8 Thus, it is critical to conduct research to determine and understand the specific risks, experiences, and disparities associated with being in foster care.¹³

Investigators commonly believe that enrolling “wards of the state” or even “minor adolescents” into research studies involves cumbersome consent procedures and extra protections.¹⁴ A previous study found 14 different combinations of individuals used for informed consent of minimal-risk research among minors in foster care, ranging from biological parent consent only to permission from judges, caseworkers, or caregivers with youth assent.¹⁵ There is often confusion about the appropriate consent process, when and which adult needs to provide permission for the minor, and if state or higher levels of protections need to be enacted. This complexity is even greater in the context of the sensitive research topic of sexual and reproductive health services for adolescents.

Given the questions that investigators face to recruit minor adolescents in foster care into sexual and reproductive health research studies, this study sought to examine published studies about foster care to a) determine the types of data used; b) describe the degree to which a sexual or reproductive health topic was addressed; and c) describe the consent process for studies collecting primary data from minor adolescents. The results may help investigators by providing examples of how minor adolescents in foster care have been successfully recruited into research projects about sexual and reproductive health topics.

Materials and Methods

PubMed was searched on September 4, 2019 using the term “foster care” with the following filters: past 3 recent years (Jan 1, 2017 – Sept 4, 2019), English language, human subjects. Articles not addressing foster care (n = 29) were excluded. Remaining articles were coded into four main categories to determine the types of data used to examine foster care experiences:

1. Primary data: articles with data collected directly from individuals currently or formerly in foster care. This category was further coded into three groups based on the age of participants (only ≤ 9 years, included 10–17 years – referred to as “minor adolescents” henceforward, and only ≥18 years).

2. Secondary data: articles with data reported from existing sources. This category was further coded into three groups based on the type of registry or database used (local, regional or national).

3. Peripheral data: articles with data collected from individuals who work closely with individuals in foster care (e.g., data collected from caregivers, providers)

4. Perspective data: articles that did not collect data and provided a perspective (e.g., commentaries, policy reports, letters to the editor, literature reviews, or protocols).

In order to describe the degree to which a sexual or reproductive health topic was addressed, all articles were evaluated for the inclusion of a sexual or reproductive health outcome. Articles which described a history of sexual abuse as a descriptive characteristic were not included unless the article’s outcome or focus was about a sexual or reproductive health outcome (e.g., described outcomes related to pregnancy prevention, sexually transmitted infections, sexuality).

The articles that sampled minor adolescents directly (i.e., 10–17-years-old recruited into studies) were coded for the consent processes described by the study to recruit minor adolescents. Coding was performed by two independent coders (JA, AG), with supervision. Discordance in coding was resolved through discussion for clarity between the independent coders and the first and last authors. This research does not include human subjects or human subject’s data; therefore, institutional review board approval was not sought.

Results

As seen in Figure 1, 176 articles were identified, of which 72 (41%) articles used direct primary data, 53 (30%) used secondary data, 36 (20%) used perspective data and 15 (9%) used peripheral data. Of the articles using primary data, 48/72 included minor adolescents. Of the articles using secondary data (Figure 2), most (48/53, 91%) used existing registries and databases. Four articles reported local data from electronic health records, such as a retrospective chart review or review of case files.^16–19 Most of the secondary data (32) was from regional administrative registries, which included data from state Medicaid claims,^20–23 state birth records,^24–28 regional institutional data,^{21, 25, 29–36} or regional agency data, such as state Department of Health or Child Protective Services databases or files^{22, 23, 25, 26, 28, 32, 37–49} that sometimes linked different registry data together.^{21–23, 25, 26, 28, 32} The remaining 12 articles reported secondary data from national administrative registries, which included data from national Medicaid reports^50–52 or national agency reports, such as the National Data Archive on Child Abuse and Neglect (NDACAN),⁵³ the United Nations Children’s Fund (UNICEF) reports,⁵⁴ or the National Youth in Transition Database (NYTD).⁵⁵ National research surveys included the National Survey of Child and Adolescent Well Being (NSCAW) I & II,^{52, 56–59} or the National Center for Health Statistics (NCHS),⁶⁰ and National Survey of Children’s Health (NSCH).^{52, 61} One study linked Medicaid claims from 36 states using Social Security numbers with responses to the NSCAW I.⁵²

Figure 1: Study Coding: Types of Articles about Foster Care.

*Secondary data represents articles from existing sources (Meta-analysis, Case Reports, Registries & Database Studies)

^†Perspective data represents Commentaries, Policies, Letters to the Editor, Literature Reviews, Protocols

^‡Peripheral data represents articles without data collected from current or former foster care children themselves (e.g., data collected from caregiver)

Figure 2: Types of Secondary data.

*2 studies reported data off a principle study

^†Local data included clinical electronic health record data review, retrospective chart reviews or local case files reviews

^‡Regional data included data from state Medicaid claims, state birth records, regional institutions (e.g., state department of education) or regional agency data, such as state Department of Health or Child Protective Services databases or files

^§National data included national registries (e.g., national Medicaid reports, national agency databases) and national research surveys

Overall, across the 176 articles, 17 (10%) described a sexual or reproductive health outcome or focus, exclusive of those that used a history of sexual abuse as a predictor. Four were primary data articles;^62–65 seven were secondary data articles;^{17, 19, 21, 26, 29, 44, 55} and six were peripheral / perspective articles.^66–71 Regardless of the data source, each focused or acknowledged a range of sexual or reproductive health outcome including: the difficulties youth in foster care have transitioning to adulthood, their increased risk for continued victimization, the need for greater access to education and knowledge, the vulnerabilities of sexual minorities in foster care, and the importance of predicting which adolescents or young adults were most likely to become pregnant or acquire a sexually transmitted infection. Only one of the articles focused on increasing youth access to preventive measures such as contraception.⁶²

The consent process among studies that recruited minor adolescents varied with no clearly identified pattern for who provided permission. Individuals or institutions who were used to obtain permission for minor adolescents to participate included biological or adoptive parent, guardian, foster parent, agency or Child Protective Service, and school / caseworker / or institution personnel, or caregiver not specified. Two, from the University of Southern California, waived adult permission as they studied former foster youth who were homeless at the time of the study.^{64, 72} For 14 articles, we were not able to establish what consent process was used. Seven^73–79 of these was because we were unable to identify the original, parent study to determine the consent process described in the parent study. The four primary data articles which included sexual and reproductive health outcomes^62–65 had a similar range of various consent processes, including one article that we were not able to establish which parent or caregiver provided permission,⁶⁵ one article that waived adult consent,⁶⁴ one that used the guardian⁶² and the other used case workers⁶³ to provide permission.

Conclusion

The literature on foster care includes all types of data, including primary data collection of minor adolescents, although that represents only 27% (48/176) of the entire body of recent literature. Despite the importance of sexual and reproductive health outcomes for this age group, across all types of literature focused on foster care, only 10% included a sexual and reproductive health focus as one of its outcomes.

The secondary data articles draw mostly on administrative registries and focused on child welfare registries. Conversely, notable national research databases that many rely on for adolescent health trends, including the National Longitudinal Study of Adolescent to Adult Health (Add Health), National Health and Nutrition Examination Survey (NHANES), and the National Survey of Family Growth (NSFG), are missing entirely. National research surveys, or surveys designed to answer research questions, set the “gold standard” for best practices to collect survey data about sexual and reproductive health outcomes or behaviors. In contrast, the national administrative registries are designed to collect data to fulfill an administrative purpose to capture statistics to monitor important trends or outcomes of an entity. Thus, these datasets are collected with different initial purposes. This calls attention to a disconnect between two fields of researchers and administrators, both of which share a goal of supporting the wellbeing of vulnerable adolescents through data-driven decisions.

This disconnect may be driven by multiple factors including a lack of collaboration in the design of surveys, or researchers not knowing how to use or access administrative registries. For example, it is typical for a clinical investigator to turn to the National Longitudinal Study of Adolescent to Adult Health (Add Health), which is a national research database with over 20,000 adolescents followed over waves, to find rich data to understand behavioral or psychosocial measurements about sexual and reproductive health outcomes. In order to link those outcomes to relevant factors for those in foster care, however, it would need to have detailed information about children in the child welfare system in order to address questions unique to children in foster care. Building a bridge to link a research dataset to an administrative dataset would be ideal yet is complicated by confidentially concerns. Likewise, those creating administrative databases for the child welfare system may not have experience knowing the key sexual and reproductive measures to include to assess adolescent well-being. Thus, cross-disciplinary collaborations between researchers and administrators should be the first step to help with designing surveys collectively.

Despite acknowledging the problem of pregnant and parenting adolescents, the issues associated with access to contraception were notably missing. For example, long acting reversible contraception, the most effective form of contraception for adolescents^{80, 81} is underutilized in general⁸² and might be an ideal method for a sexually active adolescent in foster care. None of the articles addressed the practical issues with providing preventative contraceptive methods to youth in foster care. This could be particularly problematic in states with restrictive or silent regulations about adolescent access to confidential services,⁸³ such as those requiring that parents be informed about adolescent contraception initiation.

Adolescents in foster care who are part of the child welfare system live in a variety of situations for varying lengths of time and with a range of relationships to their biological relatives, which is likely a contributing factor to the lack of a uniform research consent process. Interestingly, in the primary data articles that recruited minor adolescents, regardless of type of outcome studied, the type of adult who provided permission was rarely the same. This lack of consensus, as well as the lack of detail, makes it difficult for investigators to work with their IRBs to ensure that best practices are used to foster inclusion of these adolescents. This could lead investigators to become discouraged and fail to include this vulnerable population.

Hence, establishing guidelines for best practices would aid Institutional Review Boards and investigators in their decision-making about appropriate consent processes. Minimal risk research focusing on youth in foster care should be able to be conducted with a waiver of parental permission as is suggested for all adolescents.^{10, 84} For all adolescents, certain sensitive behaviors or vulnerable groups deserve the protection to participate in research confidentially, such as research on the LGBTQ population (vulnerable groups)⁸⁵ or research on adolescent substance use (sensitive behavior).⁸⁶ These kinds of confidential standards could extend to youth in foster care, especially when the research covers similarly sensitive topics such as contraception. It is not clear from our findings, however, which “parent” or adult figure should be providing permission for the youth in foster care to participate in a minimal risk study about contraception.

The results of this study reflect a snapshot in time from PubMed. They do not include an assessment of the challenges the investigators faced to conduct their research with youth in foster care or how they resolved those challenges. It was difficult to completely understand how obtaining permission was conducted due to the concise way most methods sections were written in these publications. However, it suggests a need for greater cross disciplinary work between researchers focused on adolescent care and those focused on foster care. This seems particularly important to those who work with sexual and reproductive health services, as youth in foster care often have high risk behaviors compounded by a traumatic abuse and neglect history often resulting in negative sexual health outcomes.

Permission and rules about sexual health services vary widely across states. Advocacy on behalf of youth in foster care about their sexual health needs and services should be prioritized. Part of this advocacy should be for professional organizations (i.e., North American Society of Pediatric and Adolescent Gynecology or Society for Adolescent Health and Medicine) to develop guidelines for both the adolescent consent to participate in minimal risk research as well as the adolescent consent for confidential clinical services such as adolescent LARC (long-acting reversible contraception) procedures.

Efforts to bridge the adolescent healthcare communities with the foster care communities could help develop a consistent voice for vulnerable adolescents in foster care. Future research should continue to evaluate strategies to simplify the consent process for investigators to include minor adolescents in foster care while at the same time ensuring appropriate protections to achieve the legal and best-interest standards for informed consent for minimal-risk research¹⁵ and clinical services.⁸⁷