Feasibility of engaging child welfare-involved parents with substance use disorders in research: key challenges and lessons learned

ABSTRACT

Child welfare-involved parents provide an important perspective on service needs and program impact. Similar to other vulnerable and hard-to-reach populations, challenges exist to engaging child welfare-involved parents in community-based research and evaluation. This case study reviews recruitment and data collection strategies for effectively engaging child welfare-involved parents with substance use disorders in a survey to evaluate the Ohio Sobriety, Treatment and Reducing Trauma (START) program – a multi-county initiative implemented to support families involved in the child welfare system due to parental substance use. In this study, child welfare agencies and caseworkers played instrumental intermediary roles to inform eligible parents about research opportunities and facilitate connections with the research team. Ongoing collaboration with child welfare agencies was necessary to establish buy-in for the research, streamline recruitment, and troubleshoot recruitment challenges. Engaging parents directly required strong interpersonal skills, empathy, persistence, attention to detail, and availability during off-business hours. Recruitment strategies also accounted for the unique internet connectivity barriers of parents living in rural communities through survey completion over landline phones, and provisions for cell phone minutes. We offer several recommendations for research methods, budgeting, and staffing when conducting research with child welfare-involved parents with substance use disorders

Keywords:

• Child maltreatment
• child protection
• child welfare
• substance abuse

Although child welfare administrative data provide valuable information for researchers to track child welfare outcomes, such as maltreatment recurrence and foster care placements (Waldfogel, 2000), they often lack the nuanced program-specific measures, such as child mental health assessments, that are necessary to evaluate service needs, satisfaction, or outcomes of intervention programs. Thus, community-based child welfare research and evaluation frequently involves original data collection with caregivers in the child welfare system, with many child outcomes assessments and measures primarily relying on parent reports. Recruiting and engaging vulnerable and hard-to-reach populations in community-based child welfare research is a complex and challenging task that often involves a two-step process where researchers engage both community partners (e.g., child welfare agency, caseworkers) and study participants. Child welfare-involved parents with substance use disorders (SUDs) represent a hard-to-reach population – defined here as a population that is difficult for researchers to access due to various challenging characteristics (e.g., rare/low numbers, socially disadvantaged, hard to identify, no known sample frames, illegal or socially stigmatized behaviors) (Marpsat & Razafindratsima, 2010). Despite prior research on recruitment and engagement of other hard-to-reach populations, such as homeless individuals (Strehlau et al., 2017) and immigrants (Olukotun & Mkandawire-Valhmu, 2020), little research has examined ways to effectively recruit and engage child welfare-involved parents with SUDs in community-based child welfare research, particularly in partnership with child welfare agencies. The present case study adds to the existing literature by reviewing recruitment and data collection strategies for effectively engaging child welfare-involved parents with SUDs in a community-based child welfare evaluation study of the Ohio Sobriety, Treatment and Reducing Trauma (START) program – an intervention model to address co-occurring parental SUDs and child maltreatment.

Parental substance use and child maltreatment

A substantial portion of families involved with the child welfare system are affected by parental substance use problems (Young, Boles, & Otero, 2007). In 2018, parental alcohol abuse was a risk factor in 12.3% and parental drug abuse was a risk factor in 30.7% of the national child maltreatment cases reported to Child Protective Services in the United States (U.S. DHHS, 2020). Many other studies have also suggested that parental substance use is highly prevalent among families involved with the child welfare system (Jones, 2004; Seay, 2015; Traube, 2012; U.S. DHHS, 1999; Young et al., 2007). For example, one study found that in 2017, one in three children entered into foster care because of parental substance use (Sepulveda & Williams, 2019). Parental substance use may decrease the parent’s ability to parent effectively due to physical, cognitive, or mental impairments caused by substances (Child Welfare Information Gateway, 2014; Kepple, 2017), and limit their capacity to adequately respond to their children’s basic needs (Magura & Laudet, 1996).

Moreover, numerous studies have found that maltreated children of substance using parents have poorer child welfare outcomes. These include a longer stay in the child welfare system, higher rates of substantiated allegations and entry into foster care, longer stay in out-of-home care, lower rates of reunification, and higher rates of re-reports or re-victimization (Brook, McDonald, Gregoire, Press, & Hindman, 2010; Freisthler, Kepple, Wolf, Curry, & Gregoire, 2017; Mowbray et al., 2017; Semidei, Radel, & Nolan, 2001; U.S. Department of Health and Human Services [DHHS], 1999; Young et al., 2007). Additionally, these children when compared to maltreated children with non-substance using parents, are at higher risk for adverse developmental outcomes such as depression, aggressive behavior, and adolescent substance use (Smith & Wilson, 2016; Stein, Leslie, & Nyamathi, 2002).

Challenges to engaging child welfare-involved parents in research

To design and implement programs that result in family-centered service delivery and improved family well-being, parental participation in child welfare program evaluation and research is critical. As service recipients, parents can offer unique insights into needs and barriers to services, as well as feedback to inform program design and improvement. Furthermore, researchers and evaluators depend on parental reports of critical outcomes (e.g., their own or their children’s behaviors and symptoms) to evaluate impact. However, recruiting and engaging child welfare-involved parents with SUDs in research can be challenging. Many of the same barriers that prevent parents from receiving services can also impact their participation in research and evaluation studies (Mirick, 2016). For instance, exposure to chronic stressors (e.g., domestic violence, mental health and substance use problems, unstable housing, poverty), lack of resources, and childcare issues might prevent parents from being able to participate in new programs or study efforts.

Additionally, parents’ distrust of researchers or concerns about confidentiality (fear of disclosure) coupled with perceived pressure to participate may pose a substantial barrier to research participation among child welfare-involved parents with SUDs (Mirick, 2016). Involuntary involvement with the child welfare or substance use treatment systems, involvement in mandated child welfare services, and negative research or service experiences may also impact parents’ willingness to participate in research (Cao, Bunger, & Hoffman, 2019; Kemp, Marcenko, Hoagwood, & Vesneski, 2009; Mirick, 2016). That is, when child welfare professionals inform parents about research participation opportunities, parents might feel coerced to participate to demonstrate compliance, which may help prevent custody loss or increase the chances of reunification (Lalayants, 2013; Perez Jolles, Flick, Wells, & Chuang, 2017). This pressure can undermine the voluntary nature of human subjects research conducted with parents (Gopalan, Bunger, & Powell, 2020).

Certain living conditions and socio-environmental challenges often faced by child welfare-involved families can also hinder research participation (Collins et al., 2020; Fong, 2017; Wermeling, 2018). These challenges include variability in employment, housing, relationship dynamics, and resources (e.g., phone service or internet access). For these reasons, special efforts are needed to optimize recruitment and enrollment of this population and to provide opportunities for collaborative research engagement.

Challenges to engaging child welfare agencies/caseworkers in research

External researchers and evaluators rarely have access to child-welfare involved parents for direct recruitment and engagement in community-based studies. Rather, researchers rely on child welfare leaders and front-line professionals to help them recruit and engage parents in community-based studies. Child welfare professionals have direct access to parents, can help explain research and evaluation opportunities, and encourage their participation. However, activating child welfare professionals to assist with research and evaluation can be challenging. Child welfare professionals may be unwilling or unable to help with this additional work given high caseloads and work burdens, overwhelming competing demands, and top-down mandates that can contribute to feelings of “initiative fatigue” and disempowerment among professionals (Akin, Strolin-Goltzman, & Collins-Camargo, 2017; Cao, Bunger, Hoffman, & Robertson, 2016; Lang, Campbell, Shanley, Crusto, & Connell, 2016; Mor Barak, Nissly, & Levin, 2001).

When child welfare agencies and professionals are willing to help with research, the system context might also impact how they recruit and engage parents. For instance, front-line professionals often want to protect their clients from additional intrusions (Mirick, 2016), and might not inform all or some of the parents on their caseloads about program or data collection opportunities. In addition, high turnover rates among child welfare professionals might lead to disruptions in recruitment or parent engagement. As a consequence, evaluations and research studies might suffer from small samples, or non-representative samples.

Gaps in the knowledge and study purpose

Recently, there has been a rapid increase in co-occurring child maltreatment and substance misuse fueled by the opioid epidemic that has impacted the well-being of children and families in the child welfare system (Radel, Baldwin, Crouse, Ghertner, & Waters, 2018). Therefore, it is important to engage these families in research that can inform evidence-based approaches to support them and promote family well-being. Numerous studies have documented that parents involved with the child welfare system for reasons of substance use-related child maltreatment represent a highly vulnerable population with unique challenges and co-occurring risk factors (Fong, 2017; Traube, 2012). However, it remains unclear how to build rapport, minimize participant burden, and effectively engage when conducting research with this vulnerable group. Therefore, the aims of this paper are to (1) illustrate recruitment and data collection methods employed in a survey study with the participants of the Ohio START program; and (2) review major challenges experienced in recruiting and engaging child welfare-involved parents with SUDs in research and present strategies at multiple levels – the community partner level (i.e., child welfare agencies, caseworkers) and the study participant level – to overcome these challenges.

Methods

Study Context: Ohio START

The Sobriety, Treatment and Recovery Teams (START) Model is an evidence-informed intervention that has been shown to improve safety and permanency outcomes for parents and children affected by co-occurring child maltreatment and parental substance misuse or disorders (Huebner, Willauer, & Posze, 2012). Ohio START is an adaptation of the national START model that has an approved adaptation to serve families with children of any age from 0 to 18 (the national START model serves families with at least one child 0–5 years old). Led by the Public Children Services Association of Ohio (PCSAO), Ohio START is implemented in collaboration with behavioral health providers to identify parents entering the child welfare system with substance misuse or disorders, expedite referrals and access to treatment, increase parent engagement and retention in treatment through the assistance of a family peer mentor, and enhance coordination of resources and supports for parents and children. The Ohio START program began implementation in April 2017 with 17 counties in rural and Appalachian communities (Cohort 1). It was expanded to 15 additional counties (Cohort 2) in the fall of 2018. In the fall of 2019, 14 counties (Cohort 3) were added, resulting in a total of 46 out of 88 counties across the state of Ohio.

The ohio start family survey

Study design and sample

The Ohio START family survey is a sub-component of a larger evaluation of the Ohio START program. The family survey evaluates the effectiveness of the Ohio START program in promoting child and family well-being, including positive parenting, parent–child attachment, protective factors in the family, and adaptive child functioning. The family survey employs a pre- and posttest survey design. The pretest (baseline) survey is conducted at the start of the intervention and the posttest (follow-up) survey is conducted approximately 6 months after the pretest survey. Parents who have received the Ohio START intervention (Cohorts 1 and 2) are eligible to participate in the family survey.

Measures

The information collected through the family survey (both at pretest and posttest) includes family protective factors, parenting and child-rearing attitudes, and child behavioral functioning, which are measured using the Adult Adolescent Parenting Inventory (AAPI-2; Bavolek & Keene, 2001), the Protective Factors Survey (PFS; FRIENDS, 2011), and the Strengths and Difficulties Questionnaire (SDQ; Goodman, 1997), respectively. Additionally, the participants are asked about their activity spaces (where the respondent goes for school, work, medical care, shopping, childcare, etc.), child behavioral health service utilization questionnaire (i.e., a series of questions about the type, frequency, and length of behavioral health services the child has received), and demographic information (race, education level, marital status, age, and sex). At the posttest survey, participants are also asked a series of questions about their experience in the Ohio START program.

Data collection procedures

The research team works together with child welfare agencies to identify potential participants of the family survey. As families begin working with their child welfare caseworker who is responsible for delivering the START intervention, they are asked for permission to be contacted by the family survey research team to participate in a telephone survey. Upon receiving parents’ permission, the caseworkers share contact information with the research team via password-protected excel files (i.e., participant tracking sheet) over a secured e-mail account. From there, a member of the research team aggregates all parental contacts received from the caseworkers that week by placing each parent’s name, contact number, and the START enrollment date into the participant tracking sheet. This file is shared between three research assistants (RAs) using the university’s password-protected, secured web servers. The participant tracking sheet is used to track each contact made to parents, including contact date, type (e.g., phone call and text message), scheduled surveys, completed surveys, and any notable event that occurred during the call. For instance, if a parent was no longer reachable through the number provided by the caseworker, RAs record this information in the participant tracking sheet with the corresponding date of the contact and communicate with each other by e-mail regarding this event. Next, the same RAs responsible for communicating with the caseworkers reach out to the corresponding caseworker who made the referral to attempt to obtain a new phone number for the parent. When the parents do not answer the call, RAs send text messages or leave voicemails to schedule a call/survey and record this as a contact attempt. Contact attempts are made by each RA to provide the parents with a brief explanation about the study (e.g., the purpose of the study, the nature of the questions asked, the length of the survey) and to obtain verbal consent for participation. Parents are contacted up to 10 times to attempt to schedule a call/survey. To ensure parents are not contacted more than once per day, RAs work on rotating shifts throughout the week. Any scheduling changes for that week are communicated between RAs to ensure a streamlined, continuous data collection process.

RAs administer surveys immediately after obtaining the participant’s consent to participate in the family survey. To protect RAs’ privacy, Google Voice – a free telephone service application – is used by all RAs. The application allows users to select a random phone number with a local area code to send and receive phone calls, voicemails, and text messages. Prior to the survey, RAs ask the parents if they would need additional phone minutes to complete the survey and send a 15 USD e-gift card via text to those who request additional phone minutes.

Each survey takes approximately 30–45 minutes to complete. Study participants receive a 25 USD gift card to a local retail store for participation in each survey (for a total possible amount of 50 USD). Participants are given the option to receive their incentive digitally or through standard mail. To date, 616 parents from 501 families have been served through the program in Cohorts 1 and 2. We have contacted a total of 243 parents (39.4% of all parents served in Ohio START) who gave caseworkers permission to share their contact information with the research team. If both parents (from the same family) received the Ohio START intervention, they were both eligible to participate in the survey. A total of 112 telephone surveys were completed, representing 46.1% of the 243 parents contacted by the research team and 18.2% of the 616 parents received the Ohio START intervention. This study was approved by the Institutional Review Board (IRB) of The Ohio State University.

Key challenges, strategies to overcome the challenges, and successful outcomes resulting from the strategies were determined based on the events recorded in the participant tracking sheet, as well as through e-mail communication between RAs. The list of challenges recorded in the participant tracking sheet and in e-mail communication were extracted into a Word document by one member of the research team. From there, RAs provided input to challenges experienced, as well as strategies used to overcome each of these challenges. Successes of the strategies were determined primarily based on the number of additional parents who completed surveys resulting from these strategies.

Results

In the following section, we describe several key challenges experienced in engaging child welfare agencies, caseworkers, and parents in the Ohio START family survey. We also present several strategies we employed throughout the survey process to successfully identify, recruit, and engage participants to complete data collection efforts (Table 1). The challenges resulted in the development of mitigating strategies which ultimately increased pre- and posttest completion. While most of the challenges encountered related to contacting potential participants, others arose due to scheduling issues and a lack of participant resources.

Table 1. Key challenges and strategies in participant recruitment, engagement, and data collection for the Ohio START family survey.

Key Challenges	Strategies
Community Partner Level (child welfare agencies, caseworkers)
Gaining buy-in from child welfare agencies	Clearly communicated the purpose of the study and its practice implications for child welfare-involved families Emphasized the essential role of child welfare agencies in successfully conducting the study Shared preliminary study results with child welfare agencies through presentations and infographics at Ohio START consortium meetings
Identification of eligible participants through collaboration with caseworkers (a low response rate from caseworkers)	Weekly e-mails sent to remind caseworkers to provide participant contact information Escalated to START Technical Assistance (TA) consultants to remind caseworkers
Study Participant Level (Child welfare-involved parents with SUDs)
Apprehension to participate in research	Thorough verbal informed consent process Assurance in confidentiality protocols RAs used person-centered approaches to break perceived power barriers
Multiple contact attempts were needed to reach participants	Persistent recruitment efforts coordinated by a team of RAs Flexible RA schedules that allowed for contact at various times including weekdays, weeknights, and weekends Phone calls were immediately followed by personalized voicemails and text messages
Phone numbers out of service, incorrect phone numbers	Engaged caseworkers in the study process to update phone numbers or provide alternative contact information
Poor service reception, lack of phone minutes or limited data plans	Landline phone numbers were used to increase connective capacity Phone minutes were offered to each potential participant immediately prior to telephone surveys
Persistent ‘no shows’ on scheduled telephone surveys	Personalized follow-up text messaging and voicemails Empathy and understanding by research personnel facilitated rescheduling and ultimate completion of telephone surveys
Participants’ entry into rehabilitation services or prison	Meticulous documentation using participant tracking sheets

Notes. RA = research assistant; SUD = substance use disorder.

Gaining buy-in from child welfare agencies

The first challenge encountered during the recruitment phase was gaining buy-in from child welfare agencies to allow their Ohio START families to participate in the family survey. Engaging community agencies, including child welfare agencies, in research can be challenging if there is a lack of perceived relevance or usefulness of the research (Mirick, 2016). PCSAO included a requirement to cooperate with program evaluation activities when contracting with each of the county child welfare agencies for the project. However, a contract requirement alone was often insufficient for gathering evaluation data and therefore, more intensive local efforts were necessary to explain the purpose and procedures of the family survey. To emphasize the importance of the family survey study in producing valuable information for Ohio START participants, we delivered multiple in-person presentations to child welfare agency directors and workers. These presentations allowed us to clearly describe the purpose of the study, its practical relevance, and anticipated benefits to Ohio START families. The critical role of the child welfare agencies to successfully conduct this study was also underscored. Furthermore, preliminary findings from the family survey were shared with the child welfare agencies via annual reports, infographics, conference presentations, and in-person presentations at Ohio START consortium meetings. These clear communications and transparency about the purpose and implications of the study proved to be effective strategies in successfully gathering buy-in to engage child welfare agencies.

Identification of eligible participants through collaboration with caseworkers

Collaboration with child welfare agencies and the Ohio START caseworkers who were responsible for delivering the intervention was a critical first step in the recruitment process. The START caseworkers served as key points-of-contact to identify potential participants and connect them to the research staff. A major challenge we experienced during this process was a low response rate from START caseworkers. That is, in 2017, out of three cases enrolled in the START program, only one case was referred to the family survey (33.3% referral rate). It is possible that at the beginning of the study, many caseworkers were unaware of the family survey or failed to remember to share parent contact information with the research team. To address this challenge, research staff sent weekly e-mail reminders to caseworkers who, in turn, provided the contact information of individuals who had consented to be contacted by the research team. In addition, child welfare agencies whose caseworkers were not providing weekly referrals or responding to the weekly e-mail reminders were contacted by START Technical Assistance (TA) consultants to either update the contact list of caseworkers or to personally remind these caseworkers to discuss the START family survey study with each parent enrolled in the START program. The weekly reminders helped increase participation numbers and facilitated regular communication between key research personnel and caseworkers. As a result, we received 55 new potential participant referrals from 177 families served in 11 counties from Cohort 1 in 2018 (31.1% referral rate). In 2019, referrals increased again for a total of 125 new potential participants from three additional counties in Cohort 1 and 10 counties in Cohort 2 (55.6% referral rate). By the time Cohort 2 began enrolling clients in the intervention, the procedures had been modified and streamlined such that asking intervention participants if they were interested in taking the family survey was a routine part of the process. These engagement strategies with caseworker staff resulted in an average of eight new referrals each week by the end of 2020.

Apprehension to participate in research

Another challenge encountered was that some parents were initially apprehensive to be involved in this research. Based on the participant tracking sheet which documented the RA’s comments, at least eight participants voiced uncertainty about the purpose of the study or expressed apprehension or confusion about survey items. It is possible that this apprehension was the byproduct of poor experiences or multiple contacts with the child welfare or criminal justice systems. Participant mistrust may have also had historical roots, as researchers in the United States have historically engaged in unethical conduct when working with vulnerable populations (Jones, 1981).

RAs played a large role in helping to overcome challenges relates to mistrust and misunderstanding. Each RA was trained to address respondent questions or to direct questions to the principal investigator (PI). Additionally, RAs were trained by the PI or other RAs to follow the scripted consent protocol addressing the purpose of the study, how the results will inform service availability in the state, confidentiality, and voluntary participation. RAs carefully addressed individuals’ questions about the survey and about the Ohio START project and completed a thorough verbal informed consent protocol by phone prior to beginning the survey. Time taken to address potential participants’ questions and concerns about the study was a crucial component that established trust not only in the recruitment stage but also in during the data collection process.

Multiple contact attempts needed to reach participants

One of the primary difficulties the research staff faced was in reaching potential participants by telephone. Although we have no systematically collected data on this, it is probable that variable employment schedules, SUD treatment or rehabilitation obligations, and childcare obligations contributed to difficulties reaching participants during the first several contact attempts, particularly during standard business hours. To resolve this issue, additional RAs were hired (for a total of three) to better coordinate data collection efforts. Additional staff allowed the research team to make contact attempts at various times including weekdays, weeknights, and weekends. It is noteworthy that contacts made on nights and weekends generated approximately 20 pre- or posttest surveys of the 112 completed surveys (17.9%).

Additionally, RAs were consistent and persistent in data collection efforts. In most cases, calls were made weekly to potential participants. Among the completed surveys, approximately 4.4 contact attempts were made per participant to complete pretest surveys (n = 206 contact attempts/91 completed pretests), and about 1.8 contact attempts were made to complete posttest surveys (n = 37 contact attempts/21 completed posttest surveys). All contact attempts, including voicemails and text messages, were tracked using a password-protected, shared participant tracking sheet which was updated immediately after each contact attempt. This allowed for seamless participant tracking and prevented research staff from duplicating data collection efforts.

Phone numbers out of service, incorrect phone numbers

Another key challenge to reaching participants was that many phone numbers changed or were out of service during the time of the study (n = 64). To address this challenge, we engaged Ohio START caseworkers to locate and update participant phone numbers whenever possible. RAs also attempted to locate the correct phone numbers if family members or acquaintances answered the phone. These efforts resulted in the completion of 6 additional surveys of the 112 surveys completed. Engaging collaborators is a central approach to overcoming barriers in recruitment and data collection (Forsythe et al., 2016; Patel, Doku, & Tennakoon, 2003). Our persistence and creativity in obtaining participant phone numbers, as well as engaging Ohio START caseworkers, proved successful.

Poor service reception, lack of phone minutes, or limited data plans

Another barrier to reaching participants was poor service reception among participants. This issue was common given the sparse Internet and cellular service access in the rural communities where Ohio START has been implemented. In response, RAs requested that participants who had consented to participate provide a landline phone number that could be used in the case of service connectivity issues.

Similarly, lack of phone minutes or limited data plans hindered research participation for some participants. To address this, the cost for participant phone minutes ($15 Visa gift card per each participant that the participant can use to put more minutes on their phone; approximately 45–60 phone minutes) was preemptively budgeted. This consideration facilitated the participation of individuals with limited minute or data plans, who otherwise would not have been able to participate in the research. At the beginning of each call, RAs provided the opportunity to each participant to receive phone minutes as needed. This resulted in an additional six participants of the 112 surveyed (5.4%) who opted to receive phone minutes and who subsequently completed pretest surveys.

Persistent ‘no shows’ on scheduled telephone surveys

Another challenge was the issue of ‘no shows’; participants who had a telephone survey scheduled, but who could not be reached at the time of their scheduled appointment. Three strategies were used to address this issue. First, RAs sent personalized text messages and voicemails to participants to remind them of their scheduled survey appointment. Second, RAs demonstrated patience, empathy, and flexibility in rescheduling telephone surveys. Third, RAs communicated with each other to assist in survey completion by coordinating rescheduled surveys around times that worked with the other RA’s schedule. This resulted in the completion of five additional pre- and posttest surveys (4.46%).

Participants’ entry into rehabilitation services or prison

Finally, another barrier encountered was the entry of potential participants into residential SUD treatment or incarceration. In the case of one potential participant, an RA was informed about incarceration through a family member who answered the phone. The RA then recorded the projected release date in the tracking sheet so that survey contact attempts could be made at a later date. Meticulous tracking of each participant, including detailed notes about each attempted contact, and the nature of each call enabled RAs to locate participants and determine the next steps in data collection. From here, at least one additional participant was able to be contacted while attending in-patient treatment.

Discussion

Lessons learned

Several lessons were learned while recruiting and engaging child-welfare involved parents with SUDs through the Ohio START family survey that can inform future research efforts with this unique population. First, it is imperative that key personnel – including research staff and community partners – have strong interpersonal communication and organizational skills since these are key to participant identification, engagement, and retention in research (Patel et al., 2003). Utilizing interpersonal communication skills – such as active listening, avoidance of jargon, use of appropriate and non-judgmental language, and showing respect for study participants – was one of the most crucial components to building rapport, easing participant apprehension, and helping the participants feel safe during their involvement in the study (Gaglio, Nelson, & King, 2006; Rosa, Babino, Rosario, Martinez, & Aijaz, 2012). In addition, organizational skills among research staff were crucial. Coordinating and tracking e-mail communication with caseworkers, as well as in tracking all participant contact attempts, pre- and posttest eligibility and completion dates, and updating participant telephone numbers helped to reduce duplication of efforts, locate potential participants, and complete telephone surveys (Patel et al., 2003).

Second, engaging the child welfare caseworkers who were responsible for delivering the Ohio START intervention was critical to study success. This is consistent with previous research demonstrating that harnessing community partners’ knowledge, skills, and resources can facilitate participate engagement, retention, and data collection in hard-to-reach populations, populations with mental health problems, and populations involved in health research programs (Forsythe et al., 2016; Kaiser, Thomas, & Bowers, 2017; Stirman et al., 2010).

Furthermore, flexibility and adaptability demonstrated by all project personnel helped us effectively engage child welfare-involved parents with SUDs in the family survey. For instance, child welfare-involved parents with SUDs are often required to receive mandated treatment or services, such as substance use treatment. Therefore, it is critical that research staff be flexible in working with the participants and be sensitive to their unique needs. Flexibility in scheduling appointments, rapid follow-up methods, and swift incentive provision can enhance participant engagement and should be factored into research protocols from the outset of the study (Zweben, Fucito, & O’Malley, 2009).

Limitations

This study has some limitations that need to be considered. First, we used a small sample of adults drawn from child welfare agencies in a single Midwestern state. Thus, the findings of the study may not be generalizable to the broader population of child welfare-involved parents with SUDs. Second, we could not examine how recruitment and engagement challenges and strategies might work differently for different groups (e.g., males vs. females; racial/ethnic minorities). This was largely due to the homogeneous nature of the study sample that consisted of overwhelmingly White, female participants. Future research may benefit from examining potential group differences in engagement challenges and strategies via subgroup analysis. Third, we did not use formal quantitative or qualitative research techniques in this study for data analysis. Yet, this was deemed appropriate because of the nature of the study (i.e., a methodological paper vs. hypothesis-testing research) that desired to share the authors’ experiences with recruitment/data collection methodologies to offer helpful methodological recommendations for future studies with the same or similar populations of interest. Despite these limitations, the findings of the study have offered valuable insight into methods and strategies for effectively recruiting and engaging child welfare-involved parents with SUDs and other vulnerable populations in community-based research.

Implications for future research

We offer several recommendations for researchers and evaluators in conducting future research with child-welfare involved parents with SUDs. First, making budgetary allowances to hire additional RAs to assist with data collection efforts, may be necessary to bolster participant enrollment numbers and ensure successful study implementation. It is also important to consider time and effort costs for the research team, beyond what was initially thought to be necessary. In our study, we had multiple RAs working simultaneously to recruit participants. Additional staffing proved essential to adding time and scheduling flexibility to increase contact attempts, engage parents in the study, and ultimately increasing the number of completed pre- and posttest surveys.

Second, hiring research staff (e.g., RAs) with the appropriate skill sets is essential to study activities. Research staff members who have strong interpersonal, organizational, and communication skills, and experience conducting research with individuals with SUDs or who are engaged in child welfare can help engage parents and address participation barriers. Notably, it is also essential that research staff carry out their recruitment and data collection efforts in a non-coercive manner that respects the principle of voluntary participation in human subjects research. Specifically, we employed students with experience conducting research with hard-to-reach and hard-to-engage populations. Because of their previous experiences, RAs were comfortable contacting and speaking with child-welfare involved parents with SUDs, which undoubtedly strengthened data collection efforts. RA’s strong interpersonal skills, including empathy, flexibility, active listening, understanding of vulnerable and socially disadvantaged populations (e.g., child welfare-involved parents with SUDs), and the ability to communicate clearly and credibly, not only helped to engage parents in the study, but also the child welfare caseworkers to identify potential participants.

Third, fostering a culture of teamwork and collaboration on study activities is essential for successful data collection. All three RAs worked together to contact parents, ensure that incentives were sent out in a timely manner, and update the tracking sheet with all contact attempts and call notes. When new RAs were hired, the existing RA(s) provided training on study procedures. Following training, all RAs communicated by e-mail and text message to answer each other’s questions and to ensure study procedures were carried out judiciously.

Conclusion

Child welfare-involved parents with SUDs are a highly vulnerable and hard-to-reach population. Through the illustration of key challenges experienced and the strategies employed to overcome these challenges in the Ohio START family survey study, we show that though challenging, engaging this population in research is feasible. Our study highlights strong interpersonal communication and organizational skills of study personnel, collaboration and teamwork between the study team and community partners (i.e., child welfare agencies/caseworkers), and flexibility and adaptability among study personnel as guidelines for successful recruitment and engagement of substance using, child welfare-involved parents in community-based research.

Acknowledgments

This research was funded by an internal grant from The Ohio State University, College of Social Work. The content is solely the responsibility of the authors and does not necessarily represent the official views of The Ohio State University, College of Social Work. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions

Additional information

Funding

This work was supported by the Public Children Services Association of Ohio (PCSA) and The Ohio State University College of Social Work Seed Grant. This study was made possible in collaboration with PCSAO, whose mission is to advocate for safe children, stable families, and supportive communities. This publication was supported by grant number G-2021-06-0415 awarded by the Ohio Department of Job and Family Services.

Notes on contributors

Susan Yoon

Susan Yoon, PhD. is Assistant Professor at the Ohio State University, College of Social Work. Dr. Yoon’s research seeks to promote resilience and well-being in children who have experienced childhood trauma, including child maltreatment. She is particularly interested in identifying factors, mechanisms and pathways that promote or hinder resilient development following child maltreatment. Dr. Yoon’s research has been funded by the NIH National Institute of Child Health and Human Development (NICHD).

Kathryn Coxe

Kathryn Coxe is a PhD Candidate in the College of Social Work at The Ohio State University. Her research focuses on the implementation of traumatic brain injury (TBI) screening and interventions in behavioral healthcare settings. She also studies determinants to mental health treatment access, integration, and service utilization among adults with co-occurring TBI, substance use, and mental health conditions.

Alicia Bunger

Alicia Bunger, PhD. is Associate Professorthe Ohio State University, College of Social Work.Dr. Bunger’s research addresses how human service organizations and professionals work together to improve service access, quality, and outcomes for the communities they serve. Her work is set within the context of behavioral health and child welfare systems and focuses on collaboration and implementation of evidence-based practices. Currently, she is focusing on strategies for effective implementation of interventions that require collaboration across systems, and is interested in developing practical tools to support executive leaders.

Bridget Freisthler

Bridget Freisthler, PhD. is Professor and Associate Dean for Research and Faculty Development at the Ohio State University, College of Social Work. Dr. Freisthler’s expertise is in studying how substance use affects parenting, particularly child abuse and neglect.  She is currently leading interventions designed to reduce child welfare involvement among substance-affected families and opioid overdose deaths in vulnerable communities. She is also interested in how the substance use environment (e.g., alcohol outlet and medical marijuana dispensaries) is related to a variety of harms, such as child maltreatment and how the places where parents spend time may increase or decrease use of aggressive discipline, punitive and neglectful parenting.

Elinam Dellor

Elinam Dellor, PhD. is a Senior Researcher and the Ohio State University, College of Social Work. Dr. Dellor conducts research on the effect of early life trauma on adult behavior and health, with expertise in biological markers of childhood trauma.  She is also interested in substance use behavior among youth who age out of foster care and in the use of technology in service linkage for child welfare involved families. She currently directs an intervention aimed at improving safety and permanency outcomes for families affected by substance use and child welfare involvement.

Anika Langaigne

Anika Langaigne.is a PhD student at The Ohio State University. She is also a Licensed Clinical Social Worker (LCSW) and Licensed Independent Social Worker (LISW). Anika received her BSW from the University of West Florida in 2014 and a joint MSW/MPA from Florida State University in 2016. Prior to entering the PhD program, Anika spent two years working as a mental health therapist in a residential facility for male juvenile offenders. Anika’s research interests include adverse childhood experiences, juvenile delinquency, and sexual violence.

Jennifer Millisor

Jennifer Millisor, MPA. is the Ohio START Technical Assistance Program Manager. While in this role she has develop the Ohio START infrastructure to support provision of technical assistance to Ohio START Public Children’s Services Agencies (PCSA). She leads the Ohio START regional technical assistance team that guides Ohio START PCSAs towards model fidelity and continuous quality improvement to attain certification for Family First Prevention Services. Prior to PCSAO, Ms. Millisor served for five years at Hamilton County Children Services and twelve years in Behavioral Health.