Structures and Strategies for Retaining an International Pediatric Cohort from Birth: Lessons from The Environmental Determinants of Diabetes in the Young (TEDDY) Study

Background: Retention of study participants in observational studies is essential to maintaining the representativeness of the population, minimizing selection bias, and assuring sufficient statistical power. The aim of this report is to describe the structures and strategies used to retain participants in The Environmental Determinants of Diabetes in the Young (TEDDY) Study, an observational study of children at increased genetic risk for type 1 diabetes followed in an intense protocol with frequent clinic visits from birth until age 15. Methods: A systematic review of methodologies used to retain research subjects identified four domains: barrier reduction strategies; community building strategies; follow-up/reminder strategies; and tracing strategies. Independent reviewers categorized the retention strategies implemented by the TEDDY Study into each of these domains. Strategies not fitting into any of these categories were placed into a fifth category unique to TEDDY. Results: TEDDY identified over one hundred retention strategies used during the 15 years of follow-up; most could be categorized in these domains. Those unique to TEDDY included (1) study organization and structures to support retention; (2) efforts to meet the changing developmental needs of the TEDDY population, (3) implementation of efforts to address protocol challenges in real-time; and (4) employment of a re-engagement protocol for those who had dropped out of the study. Conclusion: Pediatric cohort studies should include strategies, structures, and resources addressing retention at the study’s initiation. It is recommended that child and parent engagement in addition to the developmental needs of the child be an integrated focus of all strategies. Putting mechanisms in place to address protocol and retention challenges in real time would facilitate effectively addressing challenges as they arise. Trial registration: ClinicalTrials.gov Identifier: NCT00279318


Background
Longitudinal cohort studies have the advantage of assessing time-varying relationship between exposures collected over time and the progression of a disease in the absence of treatment (1).However, attrition can threaten the validity of any longitudinal study that seeks to understand the natural history of disease (2).The loss to follow-up can lower the ability of the study to detect true associations existing in a population and the pattern of missing data during follow up may lead to incorrect statistical inference and conclusions concerning the relation between exposure and health outcomes.Participant retention can be challenging, especially in pediatric observational cohort studies where there is no treatment, the study duration is long, and protocols require time and effort from the child and parent.Participant motivation to remain in a study may weaken over time (3).Recommended strategies to promote study retention include using experienced study coordinators, consistency of staff over time, changing strategies depending on the phase of the study, building rapport with families, tailoring strategies to the individual needs of the participant, and engaging the child with communication strategies, activities and incentives that are age-appropriate and change as the child matures (3)(4)(5)(6)(7)(8).
The Environmental Determinants of Diabetes in the Young (TEDDY) Study is an international prospective cohort study designed to identify environmental factors and gene-environment interactions that may trigger type 1 diabetes in children genetically at-risk for this disease (9).From July 2004 to February 2010, infants identi ed with increased genetic risk for type 1 diabetes were enrolled at six centers in four countries-Finland, Germany, Sweden, and the United States, with US centers in Colorado, Georgia/Florida, and Washington.All children were enrolled before 4.5 months and were followed until the development of type 1 diabetes or 15-years of age.The study protocol, designed with an intense follow up of frequent visits and various forms of data collection, presents ongoing retention challenges.
Addressing retention challenges was aided by early analyses to understand reasons why parents refused to enroll, their reasons for staying in the study (10,11) and factors that predicted early withdrawal (12,13) Findings from these analyses suggested there were both familial conditions, as well as challenges with a complex and demanding protocol, affecting enrollment and continued participation.These insights enabled the study coordinators to develop tailored interventions to reduce withdrawal (10).
As the TEDDY Study reaches the 20th year since the initial enrollment began, this milestone provides an opportunity to review the successes of retaining this unique pediatric cohort from birth to 15 years of age.This report describes structures and strategies used to retain study participants and highlights innovative strategies currently not found in the literature.

The TEDDY Study
To establish a birth cohort of children at high genetic risk for type 1 diabetes the TEDDY Study initiated a large-scale newborn screening protocol at each of the clinical centers.The screening protocol was a four-step process.Parental consent for genetic testing obtained at or after delivery, included the collection of a cord blood sample or a capillary sample via heel stick.All consenting parents were noti ed of the result and its meaning.Those meeting the a priori determined eligible genetic criteria were invited to enroll their infant in the follow-up phase of the study.Of the 424,788 infants whose parents consented to the newborn screening, 21,589 were eligible and invited to participate in the follow-up.Of these, 8,667 (40%) enrolled in the study, all having elevated genetic risk for type 1 diabetes.In the population with no known genetic risk or family history of T1D, 1 in 300 are likely to get type 1 diabetes.For those meeting the study's genetic criteria with no known family history, the risk increases to 3% and for those meeting the genetic criteria and having a rst-degree relative with type 1 diabetes, the risk increases to 14%.The study participants are diverse in both demographic characteristics and geographic distribution by design, permitting examination of variable levels of the multiple exposures of interest.The protocol, conducted in ve languages (English, Spanish, Finnish, Swedish, and German), included quarterly visits until the age of four years.Children who develop islet autoantibodies, indicating initiation of the autoimmune process associated with type 1 diabetes, continue with quarterly visits.Children who tested negative for islet autoantibodies had their visit frequency reduced to two visits per year after the age of 4. Study visits continue until the child is diagnosed with type 1 diabetes or reaches the age of 15 years.The demands and complexity of the study is best described by the contents of the study visits over time as described in Table 1.Study visits last 1-2 hours and include the collection of various biological samples, a blood draw, clinical measurements, interviews, and questionnaires.The study is approved by local institutional review or ethics boards and is monitored by an External Evaluation Committee formed by the National Institutes of Health (NIH).

TEDDY Structure
The operational structure of TEDDY re ects the multi-center, multi-cultural, and multi-disciplinary aspects of the study design and research objectives.There are eight scienti c committees and a study coordinator committee.All report to the Steering Committee whose voting members are the principal investigators from the clinical centers, the data coordinating center, and the NIH scienti c project o cer.The primary focus of all committees is scienti c inquiry, protocol development, implementation, and data analysis.The committees maintain this focus through participation in monthly conference calls and regular in-person or virtual meetings.
Of importance for study retention, the Study Coordinator Committee (SCC) has been involved with all aspects of protocol development from study initiation, implementation, and protocol evolution over time.
The coordinators represent the study nurses, managers, and staff conducting the study visits.They also serve as active members of the scienti c committees collaborating with study investigators to resolve any issues with protocol implementation.Having a voice at the table, the coordinators present retention and compliance challenges and solutions at the Steering Committee meetings.The SCC members focus speci cally on the participant experience, providing important observations related to participant needs and concerns that are the foundation for successful retention.
Speci c processes emerging out of this SCC structure to inform retention include: 1) using datainformed approaches to identify participants at risk for study withdrawal to develop tailored interventions to reduce attrition; 2) assessing retention and study visit compliance with recurrent allcenter review of data reports; and 3) sharing cross-center retention strategy development and experience.The aim of these structures and processes are high staff and participant retention characterized by engagement and a collaborative approach that includes participants and their families as partners in research.

TEDDY Retention
The TEDDY study de nes retention as the number of participants enrolled at any given point in time divided by the number still eligible to be followed at that time.Those diagnosed with type 1 diabetes or who died are removed from both the numerator and denominator.When participants actively withdraw from the study, the study staff collect the reason for withdrawal, contact information and permission for future contact.Passive withdrawals are those participants who have not responded to engagement attempts or have not completed any protocol elements over an extended period, but whose contact information is current.Both active and passive withdrawals are re-contacted annually and permitted to rejoin the study.Participants classi ed as lost-to-follow-up have disengaged from TEDDY and their contact information is inaccurate.
As of September 30, 2023, 65% of the 8,667 participants enrolled between 2004 and 2010, were still participating in the study (data not shown).Figure 1 describes the percentage of children withdrawn or lost-to-follow-up for the TEDDY cohort by the age of the child.The drop out is highest (11%, n = 878) at 2 years of age, with the percentage of children dropping out steadily declining thereafter to rates of less than 2% from age 9 until 15.Among those who actively withdrew from the study, the most common reasons given included distress over the blood draw, the demanding nature of the protocol, and the family being too busy or experiencing stress.Among the 2533 withdrawn, 660 (26%) participants later reenrolled in the study.

Retention Strategies Classi cation
Based on a systematic review of retention strategies among 143 longitudinal cohort studies conducted over the last decade, Teague et al. identi ed 95 strategies, 44 of which were not previously mentioned in the literature (14).These individual strategies were further classi ed into four broader categories: barrier-reduction, community-building, follow-up/reminder, and tracing strategies (14).Using metaregression analyses, the impact on retention of use of these different classes of strategies was assessed.
For this report, study coordinators from the six clinical centers compiled a list of all implemented retention strategies based on reviewing study documentation, retention presentations, and meeting minutes since the initiation of the TEDDY study.Two coordinators independently classi ed each TEDDY study strategy into one of the four Teague categories.The two independent coding exercises were compared, and the rare discordant results were discussed and reconciled.Several TEDDY retention strategies did not t into one of the four Teague et al. categories.Consequently, a fth category of strategies unique to TEDDY was added to the classi cation system.

Retention Strategy Implementation
A summary of the retention strategies employed by the TEDDY study is described in Tables 2A-D.For each Teague category and individual strategy, speci c examples used in the TEDDY Study are described.Table 3 describes the strategies unique to the TEDDY Study.Teague Domain: Barrier-Reduction Strategies: Barrier-reduction strategies consist of efforts to support participants in meeting the study protocol's demands.This includes exibility in scheduling or location of visits, adapting materials to all relevant languages, and providing childcare or transportation assistance (Table 2A).Each TEDDY center developed speci c strategies aimed at making it easier for both children and parents to complete various elements of the protocol.Participant feedback is requested from the parent and child and used to inform retention efforts.Flexibility in the visit schedule, location of the visit and intermittent negotiations in data collection while maintaining delity to the protocol is a continual goal.
The approach of creating a personal connection between staff and participants is speci c to the clinical center.All TEDDY clinical centers are organized in a way suited to the speci c conditions and constraints of their environment.The professional backgrounds of the staff vary signi cantly by country and center.Some TEDDY centers have multiple staff members managing scheduling and conducting research visits.
Other centers opt for a case management approach with families assigned to one dedicated staff member who follow the family for all visits, and were responsible for scheduling, the blood draw, and the data collection.Over time, TEDDY coordinators observed the centers with a case management approach had noticeably higher retention rates.Many of the centers gravitated towards having a hybrid case-management approach where there was a consistent team in contact and an assigned clinical staff for the primary data collection.
Teague Domain: Community-building Strategies: Community-building strategies include such things as creating a study logo, gifts with the study logo, and study newsletters (Table 2B).The greatest number of TEDDY retention strategies fall in this domain.
A TEDDY logo was used in all study communications, presentations, and annual birthday gifts.As the study participants approached the age of assent, the focus of retention efforts moved away from being exclusively parent to family-child focused engagement strategies.Other strategies used by TEDDY study in this category were arranging evening meetings with the investigators to learn about study ndings, science-focused events for the families, and children had the opportunity to have a pen-pal from another TEDDY clinic.
Teague Domain: Follow Up/Reminder Strategies: This domain includes strategies to encourage participant compliance with study visits and typically involves incentives and various forms of reminders (Table 2C).TEDDY is not unique in the varied uses of incentives as a retention tool.However, there are differences between the US and EU human subjects' regulations regarding the use of cash incentives.Only the US centers use cash payments.Gift cards, vouchers, and other kinds of nancially oriented support are used more broadly, with variability based on local human subject review board considerations.
TEDDY uses multiple forms of communication to keep families engaged by text, email, letters, and phone calls.Keeping up with the family's mode of contact preferences and utilizing technology effectively requires each center to be both technologically up-to-date and creative.

Teague Domain: Tracing Strategies:
Tracing strategies involve collecting detailed information so the participant can be located even after long periods of absence from the study protocol (Table 2D).In the TEDDY study, the contact information for all primary caregivers and alternate contacts outside of the home is updated at each study visit.The European centers had the additional advantage of country-speci c uni ed registration systems to enable better tracking of participants.

Unique TEDDY Strategies:
In this category, retention strategies unique to the TEDDY Study are described.These strategies go beyond what has been described in the literature and include study structure to support retention, risk communication and education strategies, addressing challenging protocol elements, data informed approaches to support retention, and the re-engagement protocol (Table 3).
Study structure to support retention refers to the formation of the Engagement Committee.The collaboration between the study coordinators and the psychosocial committee resulted in the development of methods for assessing the impact of study participation and factors associated with retention.Together these efforts supported the TEDDY study in being proactive to address challenges as they emerged.The goal was to develop age-appropriate strategies for keeping child participants engaged and informed about the study and type 1 diabetes.The Child Engagement Committee was the force behind developing various child focused engagement strategies including the creation of a series of storybooks and the use of informational videos to connect children in the scienti c aspect of the TEDDY study.Initially created by the Swedish coordinators and a professional illustrator, a small picture book, "We Go To TEDDY" was translated into ve languages and adopted by all TEDDY clinical centers for the 2-3-year-old children (15).The book used colorful pictures simple sentences to explain how the character Willie participated in his TEDDY visit.A second storybook and accompanying activity book continued with the story of Will and introduced his classmate Emma, who was also a TEDDY participant.
It was developed for school-aged children and was designed to be read together with an adult (16).The book further explained type 1 diabetes, why children were in TEDDY, and how scientists studied the many samples collected in the study.The book referenced a "Junior Scientist" pin, and all study participants were given the same pin.Furthermore, artwork from the book, and other depictions of the same characters were used in study materials and holiday cards.Shortly before the age of 10, a 40 -page chapter book further explaining the type 1 diabetes disease process and a TEDDY child's risk for developing type 1 diabetes was distributed to the children (17).The same characters, aged to the new target demographic, created continuity and community, while helping children understand their role in the TEDDY study.
Risk communication and education strategies are a result of an early analysis that found inaccurate risk perception was associated with early withdrawal (12).The ndings from this analysis led the study coordinators to use varied methods for communicating and assessing the understanding of risk annually with the use of pictographs.This reinforced the importance of study participation for both the parent and child.
Original YouTube informational videos were created and shown at clinic visits, distributed through newsletters, and posted on social media.One video followed the journey of a blood sample from the TEDDY clinic to the study laboratory for processing, and the travel to the large NIH repository for storage (18).Another series of videos used the Junior Scientist illustrations and text to explain diabetes autoimmunity (19).The TEDDY Around the World video, featuring actual TEDDY children, gave a snapshot of each clinical center where the children lived and activities they might do for fun (20).The goal was to emphasize the global span of TEDDY and show children how they were part of something special.Furthermore, it sparked their imagination that TEDDY is much bigger than what they saw at their study visits.Using this video format TEDDY created a unique educational tool, that also served to build community.
Addressing challenging protocol elements includes several of the items in the TEDDY protocol such as the blood draw, stool sample collection, 3-day food and the activity meter that were timeconsuming and challenging for both parents and children.Several different strategies were developed to make it easier for the families to complete the protocol and to reduce the burden.The staff were trained, both locally and at international meetings, in different methods to reduce anxiety regarding the blood draw.Age-appropriate preparation and distraction methods were developed and used, and other methods for blood collection were presented to the families.The frequency of the stool sample collection was changed.Speci c gifts, incentives and age-appropriate information materials were designed to increase the family's motivation to collect stool samples, the 3-day food record and for the child to wear the activity meter.
The creation of the participant portal by the TEDDY Data Coordinating Center highlights an adaptation of technology to improve compliance and reduce a signi cant barrier to participation.Until this point, all questionnaires required participants to complete paper forms, requiring additional mailings prior to the study visit.Families had to remember to bring the forms to the clinic visit or lengthened the clinic visits to have the forms completed in person.The portal had the advantage of shortening clinic visit time, providing options for the parents, and eventually children, to complete the questionnaires on their personal computer or mobile device, and reducing clinician burden for data entry and follow-up of missing forms.
Data-informed approaches to support retention strategies use study to develop tailored interventions to optimize retention of the cohort.Reports and scores generated from the data are summarized at the study-wide, regional, center, and participant levels for Coordinators to use as part of their retention strategies.
High Risk for Early Withdrawal (HREW) Report is one example of a cumulative risk model for predicting early withdrawal using demographic data and questionnaire responses collected at screening and at the three-month enrollment visit (12).Nine risk factors were identi ed, and a risk > 4 classi ed a child as high risk for early withdrawal (12).All newly enrolled study participants had a risk score calculated and study coordinators were informed of those who scored > 4. The coordinators then designed a tailored intervention speci cally for each high-risk family, leading to improvement in study retention (10).
Retention-Compliance Score (RCS) Report is a second example which calculates a standardized score based on the completion of study protocol elements considered to be indicators of exposures, outcomes, and general participation.The RCS score was calculated periodically for all enrolled participants, helping to identify those in need of targeted interventions.
The Enrollment Status Report (ESR) is the primary mechanism for monitoring and measuring retention of the cohort.It describes the enrollment status (enrolled, withdrawn, lost to follow-up, died, and rejoined) and among the enrolled, the degree of activity in the previous 2 years for the total cohort, each clinical center and the site/locations within the center.This report serves to alert centers to changes in withdrawal patterns and level of engagement.
Re-engagement protocol, an additional strategy that sets the TEDDY study apart from other studies, is a speci c focus around executing a systematic approach for participants to rejoin the study after a period of withdrawal.A standard cohort surveillance protocol for accurately classifying the study endpoints is leveraged to be a mechanism for maintaining contact information, continual relationship building, and serves as a successful re-engagement retention strategy.All withdrawn participants who agree to future contact are contacted yearly to assess disease status related to study endpoints and to speci cally invite families to resume their study participation.Study staff use this approach to communicate that life circumstances, family dynamics, and the TEDDY child's stage of development could have changed, allowing study participation to become possible again.Approximately 23% of those who withdrew from the TEDDY study later rejoined.

Discussion
In longitudinal observational study designs, retention must be a central focus for maintaining the representativeness of the population.Minimizing attrition ensures high e ciency when conducting research by maintaining the statistical power to detect associations of interest.Furthermore, reducing attrition can help limit systematic bias that can lead to incorrect estimates of relationships between exposures and health outcomes (2).
Since 2003, the TEDDY Study has been a major investment of the NIH to create and sustain a multicenter collaborative effort designed to establish an international cohort of children at high genetic risk for developing type 1 diabetes.The primary goal is to identify the triggers that may promote, delay, or prevent disease progression.Among the participants enrolled between 2004 and 2010, 65% were retained.Of the enrolled cohort who had previously withdrawn from the study, 660, or 7.6% had reenrolled.Annual rates of withdrawal and loss to follow-up were highest (11%) during the rst two years of participation.As children reached the ages of 5-7 years old, the rate of withdrawal fell from 5-3%.It was at this point, 5 years into the conduct of the study, that enrollment ended, and the coordinator's focus on retention became a priority.The success of concentrating on retention during this time is evidenced by the rate of withdrawal continuing to decline to less than 2% from the ages of 9 to 15 years.
Establishing an international birth cohort to be followed intensively with a demanding data collection protocol for 15 years, across six centers, meant that the TEDDY study needed to utilize multiple retention strategies.These needed to target both child and family, in such a way as to evolve over time to meet the needs of a developing child and address changing conditions in the family.This study compiled over 100 retention strategies implemented in TEDDY and found similarities in the domains of Barrier Reduction Strategies, Community Building Strategies, Follow-up/Reminder Strategies, and Tracing Strategies published by Teague et al (14).Considerations for retention strategy development in the TEDDY study focused on age-appropriateness, gender, stage of language acquisition, cultural context, study population diversity, and identifying speci c participant needs for tailored approaches.These considerations impact everything from selection of clinic décor to the reading level of the newsletters and written materials, to the planning of the most appealing events to celebrate the participants' contribution to the study.For future pediatric cohort studies wanting to strengthen their retention efforts, strategy development and achievement is most successful when the study structure supports the coordinator role and decision-making surrounding the focus on retention.
Several previous studies presenting retention strategies conclude that nancial incentives and barrierreducing strategies are most effective (6,14,21).TEDDY employed a variety of "incentives" but only the US centers provided regular nancial incentives.The European centers could not provide nancial incentives, yet Sweden had the greatest retention over time (> 70%).This suggests the cultural norms around cash payments need to be taken into consideration.Furthermore, it takes more than cash incentives to keep families engaged and may not be the most effective strategy, especially for a longitudinal study.Other study-branded gifts served as incentives and seem to be more memorable than cash and have the added bene t of promoting community in the cohort.Conventional barrier-reducing strategies such as parking, transportation, and exible scheduling are important, though not su cient for more involved research protocols with frequent visits and the collection of multiple data points.The TEDDY study found it important to address barriers arising from the demands of the study protocol by developing a participant portal and the tailoring of the protocol visits to balance the needs of the family and the study.Strategies aimed at reducing barriers also help to guide the options to support families to re-engage to the degree that is realistic for them.Consistent with recommendations from previous studies, TEDDY has employed experienced coordinators who have established a long-term structure for collaboration permitting them to tailor and experiment with retention approaches to best meet the needs of their center-speci c cohorts and individuals (4,22).In a longitudinal study such as TEDDY, staff consistency and staff retention are critical to fostering and sustaining strong relationships with the child and parent throughout childhood.TEDDY, like the Adolescent Brain Cognitive Development (ABCD) (23) and Maternal Lifestyle Study (MLS) (5) studies, has been proactive in building rapport based on trust and respect, keeping families engaged by anticipating their needs, and being positive about their participation.
Retaining a family is much easier than trying to re-engage one.Recognizing there were many reasons for withdrawal associated with family characteristics not in the study's control resulted in the TEDDY coordinators instituting the re-enrollment protocol after the screening phase and enrollment was complete.This protocol is an important strategy for long-term engagement and data completeness.In addition, unlike standard longitudinal surveillance protocols that only assess disease status long term, our re-engagement strategies also focus on inviting families back at any time.If a family does withdraw, it is important to nd out the reason(s), since knowing these reasons can be a key for future reengagement.Gaining permission for future contact at the time of withdrawal is critical for reengagement and the surveillance of study endpoints.In a long-term study, having the ability to check back in after several years when conditions may have changed and providing different options can create a pathway to rejoining.Re-engagement protocols require extensive staff resources, sometimes with a limited return on the overall investment.Although challenging, future studies should ensure the tracking of study participation status, reasons for withdrawal and re-enrollment.This requires a welldesigned and fully utilized system that is developed before enrollment begins and implemented consistently.Future analysis is planned to review the data on who rejoins to understand how to best target re-engagement efforts.
In many longitudinal studies, rates of dropout are not uniform, and obvious missing data patterns are important to understand as early as possible so they may be addressed in real-time.It is unfortunate that many studies report drop-out rates or rates of missing data from protocol violations at the time of data analysis instead of using these data to improve retention or protocol compliance on an ongoing basis.TEDDY was fortunate in this regard, as early analyses provided data to inform strategy development (11,13).Concerns about the blood draw and being too busy to participate were given as reasons for refusing enrollment and later were reasons for withdrawal.(12,13) Early in the study, the di culty children and parents had with the blood draws became apparent, and many interventions were put in place to reduce child and parent distress.Similarly, once the factors associated with study drop-out were identi ed (10), a strategy was put in place to identify those at high risk so these individuals could receive an individually tailored intervention to reduce their risk (13).This avoided employing retention strategies targeted at speci c exposures or groups that could have resulted in a systematic bias.
Learning what motivates participants to stay is also important.Based on parent surveys, among those enrolled, 76% had never thought of leaving the TEDDY study, where having someone watching the child for development of type 1 diabetes was the most important reason for staying in the study.Other reasons for staying included helping scientists discover the causes of diabetes, getting the child's test results from the blood draw, seeing the same staff at TEDDY visits, and the possibility of a prevention trial in the future.A minority (24%) of parents acknowledged some thoughts of leaving TEDDY and cited the blood draws, being too busy/not having enough time, the demanding protocol, and the collection of food records as their reasons for considering leaving.Flexibility around scheduling, visit reminders, free parking, and help with transportation were suggestions to make participation easier (11).
Finally, TEDDY actively changed its communication and engagement strategies to meet the needs of the child's stage of development and the changing family circumstances.These efforts to use TEDDY data, collected in real-time, to meet the needs and challenges of this unique cohort are critical components of the TEDDY study success in retaining its study cohort over such a long study duration.However, there were certainly limitations and missed opportunities in the TEDDY study retention strategies.During the rst few years of TEDDY, the focus was on screening and enrollment.It was not until after the screening phase was complete that the implementation of the dedicated and innovative retention efforts began.Future studies should emphasize the importance of retention from the beginning with dedicated team members concentrating on speci c strategic planning.Another limitation is that the TEDDY study implemented multiple retention strategies simultaneously and with different timing across the clinical centers.As a multi-center study with distinct cultures and constraints, we had to remain exible in the timing and approach in the implementation of different strategies.While Robinson suggests studies should employ multiple strategies to best improve retention (24), it is di cult to accurately measure the effectiveness of any one strategy at a given point.Fortunately, it was apparent that the dropout of participants continued to decline once the retention strategies were consistently implemented.
Dnro168/2004, Germany's Bayerischen Landesärztekammer (Bavarian Medical Association) Ethics Committee 04089, Sweden's Regional Ethics Board in Lund, Section 2 (2004Section 2 ( -2012) )  Authors contributions PG, JM, JB, CC, RK designed the study, proposed the analysis, interpreted the ndings, and wrote the manuscript.JM, CC and RK performed the analysis and contributed to the manuscript.PG, JM, JB, CC, RK, MK, DH, AA, JS, and SBJ reviewed and edited the manuscript.PG, JM, and JB are the guarantors of this work and, as such, had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.All authors read and approved the nal manuscript.Legend not included with this version.
and Lund University Committee for Continuing Ethical Review (2013-2021), Swedish Ethical Review Authority (2022-present) 217/2004.Consent for publication Not applicable The TEDDY Study is funded by U01 DK63829, U01 DK63861, U01 DK63821, U01 DK63865, U01 DK63863, U01 DK63836, U01 DK63790, UC4 DK63829, UC4 DK63861, UC4 DK63821, UC4 DK63865, UC4 DK63863, UC4 DK63836, UC4 DK95300, UC4 DK100238, UC4 DK106955, UC4 DK112243, UC4 DK117483, U01 DK124166, U01 DK128847, and Contract No. HHSN267200700014C from the National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK), National Institute of Allergy and Infectious Diseases (NIAID), Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD), National Institute of Environmental Health Sciences (NIEHS), Centers for Disease Control and Prevention (CDC), and JDRF.This work is supported in part by the NIH/NCATS Clinical and Translational Science Awards to the University of Florida (UL1 TR000064) and the University of Colorado (UL1 TR002535).The content is solely the responsibility of the authors and does not necessarily represent the o cial views of the National Institutes of Health.

Figures
Figures

Table 2 A
-D: Mapping of TEDDY Retention Strategies used 2004-2020 to Meta-Analysis Typology of 4 Strategy Domains and 144 Individual Strategies Developed by Teague, et al 14 Standard o ce refreshments (e.g.Coffee, snacks); lite meal after Oral Glucose Tolerance Test Consistency in research staff Center-speci c approaches included: 1) Assigned clinician as single point of contact for data collection, scheduling, results communication; 2) Participant able to request speci c clinician; 3) Consistency in clinical team working with participants Extended data collection window Changes in data collection frequency -e.g.stool samples from monthly to quarterly, Incentive (cash/vouchers) US centers only paid participants for completion of different elements of the study protocol Incentives ra es/competitions Select centers implemented ra es to increase stool sample compliance Increased incentive for hard-toreach Some US Centers used additional pay for completing study visit after extended time of not attending Limiting number of calls etc. based on participants' response Contact attempts were limited to those who were did not respond Medical assistance (e.g., diagnostic testing) Results from blood draw and were shared with participant after each visit (e.g., type 1 diabetes, celiac and thyroid disease autoantibodies) Phone Follow-up Phone calls to discuss positive results Provide referrals, e.g., medical or legal Referral to specialist for needle phobia /referral to psychologist based on questionnaire response to psychosocial measures Email reminder Email used for scheduling, study follow up, reminder of visit Reminders (unspeci ed) Sent reminders for speci c data items multiple times Follow-up/Reminder Strategies not used in TEDDY: Incentive increasing value over time; Resend survey once, Resend survey multiple times; SMS follow-up; Website follow-up; Face-to-face reminder (e.g., home visit)

Table 3
Retention Strategies Unique to TEDDY with Speci c Implementation Examples Implementation of High Risk for Early Withdrawal (HREW) with tailored intervention, Retention-Compliance Score (RCS) Report, Enrollment Status Report (ESR)Re-engagement protocolYearly mailed retention materials to re-engage inactive and withdrawn participants, offer of a one-time blood draw.