Recruitment and retention challenges and strategies in randomized controlled trials of psychosocial interventions for children with cancer and their parents: a collective case study

In pediatric oncology there are few examples of successful recruitment and retention strategies in psychosocial care research. This study aims to summarize experiences, challenges, and strategies for conducting randomized controlled trials (RCTs) of psychosocial intervention studies among children with cancer and their parent(s). We conducted a collective case study. To identify the cases, Pubmed and two trial registries were searched for ongoing and finished RCTs of psychosocial intervention studies for children with cancer and their parents. Online semi-structured expert interviews discussing recruitment and retention challenges and strategies were performed with principal investigators and research staff members of the identified cases. Nine studies were identified. Investigators and staff from seven studies participated, highlighting challenges and strategies within three major themes: eligibility, enrollment and retention. Regarding eligibility, collaborating constructively with healthcare professionals and involving them before the start of the study were essential. Being flexible, training the research staff, enabling alignment with the participants’ situation, and providing consistency in contact between the research staff member and the families were important strategies for optimizing enrollment and retention. All studies followed a stepped process in recruitment. Conclusion: Although recruitment and retention in some selected studies were successful, there is a paucity of evidence on experienced recruitment and retention challenges in pediatric psychosocial research and best practices on optimizing them. The strategies outlined in this study can help researchers optimize their protocol and trial-implementation, and contribute to better psychosocial care for children with cancer and their parents. Trial Registration: This study is not a clinical trial. What is Known: • Performing RCTs is challenging, particularly in pediatric psychosocial research when both the child and parent are targeted. Recruitment and retention are common concerns. In pediatric oncology, there are few examples of successful recruitment and retention strategies in psychosocial care research. What is New: • Key strategies to collaborate constructively with healthcare professionals were outlined. Being flexible, training the research staff, alignment with the participant’s situations and providing consistency in contact between the research staff member and the families were considered as essential strategies. What is Known: • Performing RCTs is challenging, particularly in pediatric psychosocial research when both the child and parent are targeted. Recruitment and retention are common concerns. In pediatric oncology, there are few examples of successful recruitment and retention strategies in psychosocial care research. What is New: • Key strategies to collaborate constructively with healthcare professionals were outlined. Being flexible, training the research staff, alignment with the participant’s situations and providing consistency in contact between the research staff member and the families were considered as essential strategies.


Introduction
Randomized controlled trials (RCTs) are the optimal way to minimize bias and optimize inference when evaluating treatments, therapies, and innovation in healthcare [1].However, performing RCTs is challenging, particularly in pediatric psychosocial research.Recruitment and retention are common concerns, in part because potential subjects may have pre-conceived notions about the interventions being studied (i.e., those that address mental health or communication of sensitive topics), and in part because these studies often involve substantial time commitments (i.e., for study procedures or questionnaires) [1,2].Approximately half of these trials require extensions of time and/or budget to successfully recruit their target sample [1].
Previous studies in children with serious illnesses have explored recruitment and retention challenges [3,4].In studies testing interventions for adolescents and young persons with chronic conditions, for example, participation may require directly identifying with their illness at a time when the subjects prefer otherwise.As a result, enrollment rates have ranged from 10 to 50% [3].Similarly, psychosocial interventions may address topics that are sensitive to healthcare professionals and parents, raising challenges in the approach and enrollment for clinical trials [4].Even when subjects do enroll, retaining them on these studies can be difficult due to the inherent burdens of study-participation; in a review of 40 clinical RCTs for parents and youth living with chronic illness, reported attrition rates were as high as 54% [5].
In pediatric oncology, there are few examples of successful recruitment and retention strategies in psychosocial care research [6].Strategies such as medical director involvement in the recruitment group and information sessions for eligible families were ineffective in one study in a large children's cancer center [7].In another study, a stepped process for recruitment was identified as an effective approach, where families were introduced to the study in a more person-centered way, spread over multiple contacts [6], and recruitment at diagnosis, face to face, and with the use of short questionnaires yielded higher participation [8].Although having a coordinator to monitor data collection across sites as a strategy to improve response rates lacked effectiveness [9], involvement of the clinical team in the contact strategy does seem to enhance retention [10].However, due to the inadequacies of current reporting practices, judging the impact of different recruitment and retention strategies is challenging [11].In general, there is a clear knowledge gap with regard to effective strategies aimed at recruitment [12].
To advance the conduct of psychosocial RCTs among children, we aimed to provide a detailed description of the experienced recruitment and retention challenges, and strategies to overcome them, from psychosocial intervention studies conducted among children with cancer and their parent(s).

Design
We conducted a collective case study, selecting multiple cases to generate an in-depth, multi-faceted understanding of a complex issue in its real-life context [13], and to allow us to detect patterns in recruitment and retention challenges and strategies to overcome these across the different psychosocial intervention studies.Online semi-structured expert interviews were conducted with principal investigators and research staff members of a selection of recent or ongoing trials testing psychosocial interventions for children and/or adolescents with cancer and their parents.

Selection of cases
To identify our cases, we used an elaborate search on Pubmed (Table 1) and the ClinicialTrials.govand ISRCTN.com trial registries in January 2022.Inclusion of studies was critically discussed with the core research team (ADV, KB, LD, JC, MK).Included cases (a) involved interventions that have psychosocial components (defined as any intervention that emphasizes psychological, educational, or social factors, rather than solely biological factors) [14], (b) included children/adolescents with cancer (on treatment) up until 25 years old, (c) included at least one parent in the study, (d) were either finished or ongoing studies and (e) were individually randomized controlled trials without an active comparator (Table 2).This search allowed us to identify the principal investigators (PI) of these studies and to collect relevant recruitment-and retention-related information, which was used to learn about the study before the interviews and was processed in our topic lists.The PIs were contacted by email and were allowed to invite a maximum of two other research staff members they thought would be able to give in-depth information on recruitment and retention challenges and strategies.
In March 2022, the corresponding authors were invited per email for an online interview of approximately one hour.If they did not respond after a maximum of three reminders, we contacted them via ResearchGate or contacted one of the other authors.Participants were offered a reimbursement of 60 euros for their time.

Data collection
The interviews took place in March and April 2022, via Zoom, using a self-developed topic list (see Additional File 1).Domains represented in the interviews were: 1) Actions before the study start; 2) Screening for eligibility; 3) Enrollment; 4) Attrition; 5) Encountered challenges; 6) Additional strategies to enhance recruitment; 7) Advice to other researchers.
To prepare and tailor each interview, we asked the participants to send us documents (such as information letters and protocols) they used regarding all aspects of recruitment and retention.

Analysis
All interviews were audio recorded and transcribed verbatim.The principal author analyzed all interviews, using a combination of deductive and inductive thematic analysis [15].Transcripts were read line by line and data were coded into an a priori framework of the seven domains of the topic list (deductive component).Data within each category was then coded in order to create sub-categories (inductive component).NVivo version 20 software assisted in data organization.Main findings were discussed with the research team to enhance reliability of interpretations.

Selection of cases
In total, 234 studies were identified.Of these, nine studies met the inclusion criteria (see Fig. 1).The authors of one study [16] declined participation in the interview and we could not reach the authors of one other study [17].The remaining 7 studies were included (Table 3).

Results from the interviews
The interviews lasted between 59 and 83 min.We identified challenges related to three major themes: eligibility,  enrollment and retention (discussed below)-as well as participant-suggested strategies to address these challenges (listed in Table 4) and general considerations for future investigation including illustrating quotes (Table 5).

Eligibility
We identified three challenges related to eligibility screening.First, respondents indicated that determining eligibility criteria could be challenging because they needed to balance feasibility of enrollment within a certain timeframe with the need to target experiences within specific, small populations of youth.A psychosocial intervention is often developed and tested for a specific age group with a corresponding cognitive development.This creates a tension between broadening eligibility in view of numbers and the target group specificity in view of the intervention.Some respondents indicated they changed certain criteria, such as the age of children or types of cancer, because they noticed interest in that group or because they saw this increased the feasibility of recruiting a sufficient number of participants within their timeframe.Second, gatekeeping by healthcare professionals was often mentioned as a barrier.Respondents believe this behavior was usually well-intentioned; clinicians hoped to protect families from potential stressors associated with the study-topics.Although some respondents regarded healthcare professionals watching out for the potential risks of participating in a study as a strength because they know more about the family's illness trajectory and burden, this sometimes led to disagreement between data collectors and healthcare professionals concerning the family's eligibility.Studies depending on healthcare professionals for introducing the study experienced additional challenges in maintaining awareness among the professionals about the study and their crucial task of checking families' eligibility and introducing the study to eligible participants.
Third, as childhood cancer qualifies as a rare disease [18], multiple sites must often be involved to increase the pool of eligible participants, which entailed extra challenges such as a lack of control over the approach and the procedures in the various sites.Moreover, it took a lot of time to arrange legal contracts to be able to share protected health information among the different sites, with some legal departments being extra cautious because the research involved children with cancer.In the case of the BOOST pACP study, additional legal parties had to be involved, which required extra time.

Enrollment
Seven themes related to enrollment challenges were identified.Enrollment rates varied depending on the site and research assistant.Recruitment for psychosocial studies in this target group usually takes a long time.Within the long  The executive researcher of this project (AvD) was the interviewer Table 4 Main strategies for addressing recruitment and retention challenges

Eligibility
Broaden the inclusion criteria to the extent possible allowed by the intervention's goal and content.Determine eligibility criteria before the start of the study and make reliable estimates of the number of families adhering to these criteria.When determining the eligibility criteria, collaboration with healthcare professionals or national cancer registers is recommended Use a thorough screening process to make sure eligible families are not missed in the various steps of the study.This requires a monitoring system, informed by access to medical records and patient lists or by regular meetings with healthcare professionals Closely involve the recruitment sites before the start of the study to discuss appropriate ways of introducing the study to potential participants -Discuss inclusion criteria and implications with healthcare professionals -Take the time to "learn who your allies are" at the hospital.Who is intrinsically motivated to commit her/himself to the study tasks?-Explore whether study tasks could be a formal part of the healthcare professional's tasks.Not as an additional task, but replacing an existing task -Arrange a reimbursement for collaborating healthcare professionals (e.g. an hourly rate or per number of approaches) -Arrange research assistant/data collector access to medical records (or to the hospital system, if possible) or arrange a monthly timepoint to obtain a list of eligible people -Discuss patient eligibility regularly, or discuss beforehand the option of following an opt-out procedure: inform the primary oncologist (via email) that you will contact the eligible family about the study.Only if the oncologist disagrees, will he/she will respond (via email) asking you not to contact the family within a certain time period -Create buy-in by planning implementation and sustainability activities, such as: presentations at the ward, regular visits with small presents, or disseminating newsletters on the project -Provide the different recruitment sites with materials that facilitate introducing the study.Develop flyers and ask permission to leave them in places visited often by parents and children, such as waiting rooms, kitchens, and the reception desk If possible, apply for a special grant that provides support for the initial development of clinical trials (such as the NIH Planning Grant Program (R34)) in the USA.This money can be used for time spent on arranging contracts with the various sites you are recruiting from Enrollment Provide continuous training for the research staff.When hiring (new) research staff, remember that they need to be: 1) flexible in their availability and way of working; 2) relatable to study participants or having experience in working with them; 3) motivated and passionate about the theme.Be transparent with your (potential) staff regarding the kind of study population and how they might engage with you, and the flexibility you expect.This helps prevent staff turnover Involve the recruitment sites and the patients/families before the start of the study to determine appropriate ways of introducing the study to potential participants Communicate the experiences of families who have participated to the healthcare professionals and eligible participants (e.g. via newsletters, flyers or during presentations at the wards) Have research staff follow up closely and be present at the sites regularly Offer a relevant intervention for you target group and design the intervention in a flexible way by involving your target group beforehand Increase your study's visibility by using attractive materials (for example) and think about efficient ways to distribute them (such as patient organizations, hospital wards, local health insurance fund magazines), so that people might see information about the study beforehand Optimize the study's introduction by research staff: -Put measures in place to avoid situations (such as bad news, heavy treatment) that can make the timing of the study introduction inappropriate.
Use a personalized and stepped approach (sometimes only plant the seed and follow up quickly) -Be as flexible as possible in introducing the study.Some families do not like to be approached in the clinic because they are eager to get out of there-they prefer to be contacted at home -Make sure the families have heard about the study from healthcare professionals or via a flyer or email before your first call -Explain what the goal of the intervention is, why this is important and how this may contribute to a better quality of life and why the effectiveness of the intervention is being studied -Have a basic understanding of the diagnosis as context in approaching the family -Use a short pitch not exceeding 2 min -Use lay language -Talk to the parent and child together and engage with the child.Experience shows that, when you talk to them together, parents will participate if their child wants to participate.Respondents have often experienced that the child has the final say -Make their participation feel important.Mentioning that their participation can help other families in a similar situation can help -Acknowledge the participants' difficult situation, be sensitive and consider the family burden -Emphasize the minimal burden in participating, and explain how participation has been made as easy as possible -Respond to reasons for them not wanting to enroll Retention Be flexible when offering moments to visit the participants, and be on top of non-response by closely monitoring-and then contact the participants as soon as possible when they have missed their measurement Provide consistency in contact between the research staff member and the families recruitment period, staff turnover within the research team was inevitable, which created difficulties regarding finding new motivated and flexible members.Some respondents noticed that the enrollment rate would drop after hiring new research assistants.Finding and training suitable new staff took time and resources.New research staff members, in particular, found it often difficult to interact with families going through these difficult experiences, because they did not want to add to the family's burden.Consequently, they were extremely careful introducing the study, which sometimes negatively impacted enrollment rates.
Willingness to participate was often overestimated by the research team.Through surveys, several respondents had tried to gain insight into families' reasons for not participating in the study, but the response rate on these surveys was low.The insights obtained revealed that the families' reasons for refusing to participate were: feeling overwhelmed having to combine their family life with regular visits to the hospital; perceiving the health status of the child as not good enough; or simply not having the mental headspace to commit to something extra, regardless of the study.The children's most mentioned reason for not participating was wanting to feel normal without being preoccupied with being sick.
Respondents acknowledged that children or adolescents with cancer and their parents are an overburdened research population.Our respondents mentioned that there were often competing studies in the same target group during the time of recruitment.One study (FAMOS) could not include families of children between 6 -18 years for that reason, and were thus constricted to include only a part of their original target group (0 -18 years).Medical studies are often prioritized: for example, a child eligible for the BOOST study said she had already participated in a medical trial, preferring that over a study with psychosocial components.
According to our respondents, another factor potentially influencing enrollment rates was that families are randomized into groups.Families did not always understand the concept of randomization and did not fully know what they were signing up to.Some families also had a clear preference for one of the groups.
The dyadic nature of the enrollment and needing both the parent(s) and the child to be willing to participate, complicated this kind of research.Encountering both the child and one of the parents in the same room at the moment of introduction in the hospital often took several visits.Parents regularly assumed that their child did not want to participate, or they wanted to protect them from potential harm.Requiring the signatures of both caregivers took additional time, because they were usually not present at the hospital at the same time and they had to discuss participation at home.
Finding an appropriate time to introduce the study to the family was experienced as an important part of building up contact with the families.However, this was also seen as a prominent challenge because, as many families were in the hospital to receive news or heavy treatment, it was difficult to estimate the burden at that moment.

Eligibility
Facilitate filling out surveys as much as you can.Provide electronic as well as paper versions and shorten them if possible.Prioritize having the items on your primary outcome filled out.Have a language consultant look at the items of your questionnaire to optimize understandability, especially when you have developed the items yourself In the USA studies, in particular, monetary incentives were given to the participants and, in some cases, the incentive was mentioned on the recruitment materials.However, not every respondent was convinced about the effectiveness of this.A small increase in the monetary incentive per timepoint might work best Provide for flexibility in the intervention process by offering a range of time to enable alignment with the participants' situation If your study is longitudinal and entails surveys being filled out at different timepoints, building in a run-in period can help preventing enrolling participants who will most likely drop-out.A run-in period means that, prior to randomizing a family, a family fills out several surveys to see whether they will be engaged in the study or not.Providing this opportunity was seen as less intimidating for families than, for example, committing to an 18-week intensive study for example General strategies Perform a pilot study in the target group in advance, preferably in a situation similar to that of the larger scale study (e.g.involve more recruitment sites already) This will give insight into expected enrollment rates, as well as contribute to establishing better monitoring systems to assess eligibility, ways of cooperating with healthcare professionals, and give the opportunity to slightly adapt the intervention if you find that some components are not working in practice as expected Include a plan in your protocol for changing recruitment strategies, concretizing different scenarios: e.g. if the enrollment rate is less than x%, we will use strategy number 2. Having IRB approval upfront to use these strategies, will save time Explore other recruitment 'sources' such as social media, making a video, using email lists from the university or hospital, involving patient organizations.Pilot test these other sources, if possible.When recruiting via Facebook,for example, an algorithm needs to be developed, monitored and adapted continuously, and so involving specialists might be worthwhile.Critically appraise whether your target group is appropriate: if the target group is quite narrow in terms of diagnosis and time since diagnosis, social media might not be suitable 2. Use a thorough screening process to make sure eligible families are not missed in the various steps of the study.This requires a monitoring system, informed by access to medical records and patient lists or by regular meetings with healthcare professionals "…following up with the research assistants at each site twice a week, having frequent meetings, using these reports and doing these intensive processes, we were still only able to approach 60% of people that we found to be eligible.So another 40% that became ineligible.. we were waiting to approach for various reasons and changes in their disease criteria or transferred care or died or other things.So we kind of lost big chunks as part of the process.That made it really important that we started with everybody."(3.PediQuest Response) 3. Having to involve multiple sites to increase the pool of eligible participants "In the beginning it was difficult to find people at the wards that wanted to be our main contact points during the 2 year study period.
And throughout the study, we noticed the different sites sometimes slightly adjusted the way of approaching patients to make it more feasible for them.At the beginning, we wanted to keep control but that is challenging.It takes time and effort to find ways that work for everyone" (7.Benefits of Obtaining Ownership Systematically Together in pediatric Advance Care Planning) (BOOST pACP)) 1. Create buy-in by planning implementation and sustainability activities at recruitment sites "…it depended on study site… at one site we were very reliant on oncologists going through their lists, we had one key person who had a good buy-in for this study.She made sure we had the names of people who fitted our eligibility criteria."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC)) 2. Arrange a reimbursement for collaborating healthcare professionals (f.e. an hourly rate or per number of approaches) "we kept a log of every time the contact nurses approached a patient, not necessarily if they were successful, as recruiting them, but just to approach a patient.I think for every 5, we gave them a 5 dollar Starbucks giftcard.And it was a pretty good incentive, people really liked the coffee."(2.Enhanced Psychosocial Intervention (EPSI)) "We have spent the first year of the trial just to get contracts and agreements that we can share protected health information with other sites and some of them took up over a year."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC))

If possible, apply for a special grant that provides support for the initial development of clinical trials "They have a mechanism now, called an R34, you can now have
2-3 years to just set the trial up, so we have spent the first year of the trial just to get contracts and agreements that we can share protected health information with other sites and some of them took up over a year and some of them and I ended up travelling to some sites to review and that burned up money too."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC))

Offer a relevant intervention for your target group and design
the intervention in a flexible way by involving your target group beforehand "And I think that's the most important part, is to offer something that is relevant to people.And I think the only way to do that is to involve the families and for you study maybe the adolescents?In both how do we sort of design and set-up that's most relevant to your family and also, maybe get them on board on how to do the recruitment.…I think the fact that there was a combination of very concrete tools and then videos.And it was homebased, that was the main thing.And we were very flexible in terms of, we tried to do it on their terms.I think that were the most important things that made them feel that was comfortable for them as well."(6.Family-Oriented Support Intervention (FAMOS)) 1.Put measures in place to avoid situations that can make the timing of the study introduction inappropriate.Use a personalized and stepped approach "We do have an eligibility criteria as part of our approach checklist that asks if a family had a sort of serious conversation in the last two weeks, or if the kid is going through scans that day and might get bad news then we wouldn't approach them that day, or if we could see through the record that last week they found out about a progression, they wouldn't be eligible until 2 weeks after that.. and then sometimes information would come out of meeting a family, so we might come back to them one month later.. we might go and talk to them and say you know, can we tell you about this study.. and they will tell us 'our grandpa had just died' and we are trying to figure out the funeral next week.. and we are so overwhelmed right now.. so you could say would it be okay if I would check in with you again in a month.. in some sites we leaned more on that than in others again I would say, but that was something that we would certainly try to kind of personalize or approach each family and adopt as much as we could to make it as convenient as possible."(3.PediQuest Response) "In a survey we asked teenagers when do you want to be approached and most of them said from the time of diagnosis, so it was really based on their feedback that it was okay to talk about this stuff from the time of diagnosis.And we quickly checked with the provider, is there a particular reason we shouldn't approach today….If they were having a spinal tab, that was not the day, so sometimes we would just plant the seed, but then we came back later."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC))

Unanticipated circumstances (e.g. COVID-19 pandemic)
"also a key part here is the face to face contact.With COVID and so on made it much more challenging."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC)) "our visibility and presence in the hospitals would have been better if it wasn't for the COVID pandemic.We would have been more present at the hospital wards and personal contact always helps in making things happen.The COVID pandemic gave us a false start."(7.Benefits of Obtaining Ownership Systematically Together in pediatric Advance Care Planning) (BOOST pACP)) 1. Increase your study's visibility by using attractive materials (for example) and think about efficient ways to distribute them (such as patient organizations, hospital wards, local health insurance fund magazines magazines), so that people might see information about the study beforehand "Hanging flyers in the hospital ward and writing a testimonial from a family that participated in a local magazine of a patient organization and promoting this via the social media (such as LinkedIn or Facebook) of the participating hospitals helps increasing the visibility of your study.This won't necessarily lead to a higher number of participants, but families may recognize the study when they are formally introduced for the first time and this might contribute to them saying "yes" to participation a bit quicker."(7.Benefits of Obtaining Ownership Systematically Together in pediatric Advance Care Planning) (BOOST pACP)) Table 5 (continued) Challenges and illustrating quotes (study from Table 3) Strategies and illustrating quotes (study from Table 3) Retention 1. Non-response and drop-out "We may have had a slightly inflated rate of drop out from our intervention arm, so maybe for some people that was almost like they hoped for the easiest study experience that took the least time and energy for them."(3.PediQuest Response) "We followed our intervention group monthly.This was not the case for the control group, we got them a year later.We didn't retain as many in the control group because the people we were following on a monthly basis were engaged.In the control group, we may have lost as many as 25%." (2.Enhanced Psychosocial Intervention (EPSI)) "And I think that was part of the reason why we had attrition, especially in the control group.It was a lot of questionnaires for the parents to fill out, without sort of receiving any intervention."(6. Family-Oriented Support Intervention (FAMOS)) 1. Be flexible when offering moments to visit the participants, and be on top of non-response by close monitoring "The people that followed-up the families were very persistent.If the family said today is not a good day, they said: "can I call you tomorrow or can I check up on you in 2 weeks, and put that in my calendar?".In reports it is known that at some number of phone calls there is a diminishing return, they are not going to call you back after 4 or 5 phone calls, but up until that point, people were busy.So be very flexible and persistent, but not just one voicemail and then they are lost to follow-up."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC)) "On every Monday and Thursday, we receive from the system whoever did not answer a survey.As the project manager, I would contact the research assistant at that site and say please follow up with so and so.The next survey is on a Tuesday so let's call them when the next one is active, and that it might be a good time to pop by the clinic and give them the gift card.So it was a lot of monitoring to be on top of it when somebody missed a survey to reel them back in."(3.PediQuest Response) 2. Provide consistency in contact between the research staff member and families "I think the main piece of retention that was important was engaging the research assistants to form relationships with the families."(3.PediQuest Response) "I think that was a good strategy.That it was the same person that day that they had contact with along the way."(6.Family-Oriented Support Intervention (FAMOS)) 3. Facilitate filling out surveys as much as you can "But I think of course, that shortening the questionnaires as much as possible is the best strategy.And at that time, we only had paperbased questionnaires and I think it helps today with the electronic ones as well."(6.Family-Oriented Support Intervention (FAMOS)) 4. If your study is longitudinal and entails surveys being filled out at different timepoints, building in a run-in period can help preventing enrolling participants who will most likely drop-out "We have the run-in period built in that we have described as like a three survey prior to randomizing a family to kind of prove that they are going to be engaged in the study or not, and so we really invited people to test it out, which was a little less intimidating than committing to an 18-week, sometimes intensive study experience… it kinds of helps the quality of the data we are getting… one thing we have learned is that people who did not respond the first 2 surveys, were not going to respond after.. and instead, we decided to do this run-in period, and get rid of the non-responders, because they would not give you any valuable data."(3.PediQuest Response) 5.In the USA studies, in particular, monetary incentives were given to the participants and, in some cases, the incentive was mentioned on the recruitment materials.However, not every respondent was convinced about the effectiveness of this.A small increase in the monetary incentive per timepoint might work best "I think they do help decrease drop-out at certain timepoints, so that's how we used it.My strategy now is different than what we did in this study.If I would do it over, I would have had an increasing incentive, at each data assessment point that it would increase as the study goes, I would have had a small incentive for T1, just to get them going, I would not have had an incentive with completing the intervention, because of that early attrition we would had in the beginning in the intervention group, we offered them an incentive when they completed the intervention."(5.Legacy Intervention) Lastly, the COVID-19 pandemic was detrimental to many ongoing studies, because it further negatively influenced families' willingness to participate.Moreover, some respondents felt that in-person recruitment worked better (addressing families and being present in the wards), which was not possible during the lockdown periods.

Retention
Two themes in retention challenges were identified.All respondents faced non-response and drop-out in their studies.In some studies, the drop-out rate was higher in the intervention group compared to the control group.Respondents hypothesized that this could be due to the intervention taking a lot of time or not being sufficiently user-friendly.The respondents from studies in which drop-out was higher in the control group speculated that this could be caused by them being contacted and engaged less often throughout the study period or by them being less motivated because they did not receive an intervention in return.
Another challenge impacting retention was keeping perprotocol timelines versus required flexibility.Once families had been enrolled, it was challenging to follow the protocol and keep to the predefined time window, especially when psychosocial interventions consisted of several intervention sessions and questionnaires.Some children were admitted to the intensive care unit (ICU) during the study period, or sessions needed to be rescheduled because parents were exhausted or had other obligations, which impacted the planning of the follow-up sessions.Therefore, the risk of families having to drop-out is increased and a certain amount of flexibility is required.

General challenges experienced by the respondents
Three overarching challenges were elaborated on during the interviews.For an indication of how many recruitment sites are required and how much time is needed to enroll the target number of participants, most respondents performed a pilot test.However, in some cases -due to different circumstances in the larger study-the numbers in the pilot study were vastly different from the numbers eventually enrolled in the larger scale trial.
In addition, some studies had to pause enrollment because another recruitment strategy they wanted to implement was not included in their initial, IRB-approved, protocol.
Lastly, ending up with enough recruitment sites willing and able to participate in the study was mentioned to be challenging.Therefore, one respondent recruited via social media (Facebook) instead.Although this eventually worked well after gaining expertise in creating and running an advertising campaign, the respondent argued that this is not applicable to every population and success is difficult to predict.

Discussion
This collective case study gives an overview of the most commonly experienced recruitment and retention challenges, plus strategies to overcome them, while conducting RCTs of psychosocial intervention studies for children with cancer and their parents, sub-divided into three themes: eligibility, enrollment and retention.Although many studies were found on this target group, they mostly used different designs, such as pretest-posttest designs without a control group, cluster RCTs, or used an attention control group.Other studies we found stated that they did not randomize due to ethical and practical considerations.The limited number of studies meeting our inclusion criteria, may indicate that conducting individually randomized controlled trials without an active comparator in this field and target group is rare, possibly because it is very challenging.All of the respondents followed a stepped process for recruitment, using a personalized approach, with several visits per family.Bradford et al. (2021) [6] estimated it takes an average of 98 min for an experienced research member to screen, approach and recruit one participant in a pediatric oncology setting, emphasizing the importance of training all research staff (involved in referral, data collection, and intervention) [19].This training could entail ways to perform systematic screening of patient lists, thoughtful messaging to make the study relevant, and how to deal with flexible protocols to accommodate patients' needs [20].Furthermore, communication with families in the pediatric oncology setting-and striking the balance between respectfully recognizing their burden without being too careful introducing the study-should be part of the training.As participants, including children and adolescents, reported benefits including the opportunity to talk, be altruistic, reflect, have a safe space, gain understanding or perspective, and express emotions [21], potential advantages of these kinds of studies may be presented to eligible families also.More transparency on the content of these trainings from trials in this setting would be of added value for researchers to use as a reference.
One of the overarching recommendations from our respondents was to perform a pilot study prior to the larger scale trial.Testing feasibility and acceptability by conducting a pilot study is widely accepted as a great value, if not a necessity, for RCTs [22] as they can "inform decisions about whether further testing of an intervention is warranted".Being aware of which features of the intervention and sample in the pilot study are not scalable or generalizable to the next stage of testing in a larger effectiveness trial-might optimize trial implementation [23].Context and implementation are critical to the extent of an intervention's effect, especially in complex interventions [24].Therefore, different approaches regarding effectiveness research have been suggested to improve our understanding of generalizability, stressing the need for process evaluations to understand an intervention's mechanisms of action [25].Once identified, then larger trials, with integral process evaluations can be conducted [25].Especially missing data (due to drop out, attrition or item or questionnaire non-response) can introduce problems of selection bias, particularly if they are influenced by what arm study participants are allocated to, affecting generalizability [26].As recruitment and retention rates of a largescale trial can differ from the pilot study, an additional strategy in the development phase of an intervention is pre-emptively generating strategies to address potential recruitment and retention difficulties via focus groups and interviews with the target group [27].
In general, there is little evidence that helps researchers to make well-informed decisions about how to do their trials [28].Performing a Study Within A Trial (SWAT)-which is a self-contained research study that has been embedded within a host trial with the aim of evaluating or exploring alternative ways of delivering or organizing a particular trial process-might contribute to more evidence-based decisions within trials.However, especially in trials testing psychosocial interventions for children with cancer and their parents this might be challenging because undertaking SWATs includes challenges such as increased complexity and management burden, which are more profound in pediatric populations due to a perceived increased patient vulnerability or risk [29].
A strength of this study is the high participation rate from the researchers (75%) we approached, showing that recruitment and retention are experienced as challenging and recognized as important.Moreover, our search strategy to select cases limited the risk of missing relevant trials.Including the PI as well as other research staff in the interviews allowed us to gain an in-depth understanding of the recruitment and retention challenges and strategies.
A limitation of this study is that the overall number of included studies was low, so drawing conclusions must be done with caution.Additionally, we included studies that were registered from 2011 onwards, leading to an increased risk of recall bias for some studies.However, we noticed that, in most interviews, the respondents complemented each other when not remembering something, and some respondents could compare experiences from the selected case to follow-up trial studies of the intervention they were undertaking now, enriching the information.Another limitation is that we based our selection of cases on a search string in English, possibly missing relevant studies reported on in other languages.A final limitation is that the analysis was done by one researcher, although main findings were discussed with the core research team to enhance reliability of interpretations.

Conclusion
We identified a paucity of evidence on experienced recruitment and retention challenges and best practices on optimizing them.Respondents indicated that even after years of experience of conducting studies in this field, it is difficult to know in advance what strategies will have desired effects.Effectiveness of 'new' strategies is challenging to map because they are often used in combination with other strategies and in pediatric studies, the number of eligible participants is generally low and effectiveness cannot be directly attributed to the use of only that particular strategy or to other factors such as more eligible participants in that particular period for example.Additionally, what may have worked in a former variant of the study might not necessarily work in a different variant of the study due to contextual factors.However, some studies had high enrollment and retention rates and managed to overcome challenges successfully, which indicates that suitable recruitment and retention is possible.The close follow-up of families and maintaining regular contact with the key persons at the recruitment sites while continuing a flexible attitude to the extent possible was emphasized most often.The strategies outlined in this study can help researchers to optimize their protocol and implementation of the trial and contribute to better psychosocial care for children with cancer and their parents.
accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments and in compliance with our institutional guidelines.
Consent to participate Informed consent was obtained from all individual participants included in the study.
Consent to publish All respondents provided informed consent for publication of the study characteristics and use of quotes.

Competing interests
The authors declare no competing interests.

Table 2
Inclusion-and exclusion criteria for the selection of studies The child / adolescent and parent(s) are randomized in an intervention group and control group with care as usual 4. Studies registered before 1 January 2011 5. Studies are registered after 1 January 2011

Table 3
Key characteristics of selected studies for which we held interviews

Table 3 (continued) Selected studies Function of the participants in the interview Target population Country, sites and timing Timepoints of surveys Intervention goal and content Original target goal and time window of recruitment as defined in protocol/trial register Actual number of participants enrolled and time window of recruitment. Extra time/sites required? Enrollment rate a and retention rate b and extra time or sites required
a Enrollment rate: defined as the proportion of people who enrolled out of all people determined to be eligible b Retention rate: defined as the proportion of people who enroll and who complete the study c

Table 5
Challenges in recruitment and retention and strategies to overcome them per major theme

Broaden the inclusion criteria to the extent possible allowed by your intervention's goal and content
"For some other trials currently, at some of the sites they have systems where the patient primary oncology provider required them to get it approved and we have run into issues where oncologists will arbitrarily say no, and of course it's not completely arbitrarily, but we have had a scenario in some sites where there were 50% of eligible patients and the oncologists were just like 'I think they are too distressed' or 'this kid is going to college soon'.You have to kind of figure out ways to like.. respectfully listen to those things, and try to give every family the opportunity to say no to us rather than someone else opt them out."(1.Promoting Resilience in Stress Management (PRISM))

Be transparent with your (potential) staff regarding
the kind of population, how the population might engage with you and what kind of flexibility you expect.This contributes to preventing staff turnover "transparency with your staff is key, from the beginning be like this is the population, this is how they like to engage with you, here is what we expect in flexibility.The flexibility piece was major in helping us with recruitment and retention.As we mentioned earlier, there are weekends and evenings, and home visits where people felt safe.
Because you might have staff members that say I cannot meet those expectations, but it might help if you tell them early on, their brain gets programmed in a way, that this is how I'm going to engage in it and interact with the participants from the study."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC)) 2.

Making realistic estimations of willingness to participate
"We didn't really appreciate how stressful the actual study might be to people, even if we know what might be helpful… It's really a factor of us overestimating people's willingness to participate" (1.Promoting Resilience in Stress Management (PRISM)) "Our study was on symptoms and some people don't want to think about the symptoms.Because if you felt nauseous, would you sit and talk about how you are nauseous?Sometimes people felt like acknowledging the symptoms and thinking about them made them feel worse."(3.PediQuest Response) 1.

Table 5
(continued) kid, who you think might be super excited but you have a parent that doesn't want them to do it for some reason, so it's hard to get everybody on board.It's also hard: the nature of children with advanced cancer.The illness is such a rollercoaster, it's like how the stars are aligned, if they are feeling good and schedules align or whatever to have them complete the intervention session.(5.Legacy Intervention) "Ideally, we wanted the nurses to talk to parents in the hospital, but some of the parents would be at home or just coming for a quick check-up.It was difficult to catch them physically."(6.Family-Oriented Support Intervention (FAMOS))1.

Close follow-up by research staff and being present at the sites regularly
"Compared to a former FACE study, the recruitment was more difficult in this study.The staff this time had a lot of other things to do, although of course they were also very dedicated.But the first staff was only focused on the FACE study.They had more hands-on follow-up, or maybe they were on the ground a little bit more, in the clinic areas."(4.Family-Centered pediatric advance care planning for teens with cancer (FACE-TC)) 2.

Talk to the parent and child together and engage with the child
"if we were able to convince the child, then it was a go.So the parent would only do it of the child wanted to do it, never did you see the opposite.So that was interesting….the other thing is that sometimes with parents, when they are overwhelmed, assume sometimes that their child does not want to participate.So I found it was really good to be talking to them together about the study.And sometimes they would just project their own.. to the child and the child was sometimes like, yeah I have got nothing else to do, I'm here in bed anyway."(2.Enhanced Psychosocial Intervention (EPSI))

Table 5 (continued) Challenges and illustrating quotes (study fromTable 3 )
Strategies and illustrating quotes (study from

Table 3 ) 6. Finding an appropriate time to introduce the study to the family
"It's SO dependent on the family and it's really hard to tell from a chart note or even like a conversation with the provider, because I think that different providers have different interpretations of when a family is ready or if they think they are receptive."(1.Promoting Resilience in Stress Management (PRISM))

Finding a sufficient number of recruitment sites that are able to or want to participate
for that are that some of the sites initiated some sort of automatic to palliative care teams for patients that met their criteria and made the patients ineligible."(3.PediQuest Response) 1.