By Jane Osindo, Elizabeth Kemigisha, Stephen Gakuo Maina, Joshua Eliud Omondi, Ann Wanjiku Kigondu, Richard Sanya, Abdhalah Ziraba
Longitudinal studies are crucial for generating scientific evidence and offer several advantages. They allow researchers to identify and link events to specific exposures (unintended exposure or deliberate intervention), defining these exposures by their presence, timing, and duration while also tracking changes and outcomes over time in the same individuals. However, longitudinal studies also face several challenges, particularly participant attrition. This occurs when participants drop out of the study for reasons such as fatigue, loss of interest, change in place of residence, illness, death, or other unforeseen circumstances. A significant loss of participants can reduce the study’s power, introduce bias, and compromise the overall quality of the results. One common issue associated with attrition is selection bias, where the remaining participants may differ significantly from the original group in ways that compromise the validity of estimates derived from the studies. These challenges can lead to wasted financial resources and time, especially if the study is discontinued or produces unreliable results. Anticipating, recognizing and addressing these challenges is critical for maintaining the integrity of longitudinal studies. In this blog, as an example, we use one of our recent studies to describe some of the challenges and how they were addressed to mitigate attrition.
Since 2001, the African Population and Health Research Center has been running the Nairobi Urban Health and Demographic Surveillance (NUHDSS), a research platform that records and follows up all residents of a defined geographical area over time for various research purposes. The Determined Resilient Empowered Aids-free Mentored Safe (DREAMS) Impact Evaluation project was a longitudinal HIV program evaluation cohort study nested within the NUHDSS, allowing for follow-up of individuals and assessment of their outcomes. The study’s goal was to assess the impact of the DREAMS interventions, a comprehensive package aimed at preventing HIV among adolescent girls and young women (AGYW) aged 10-24 in resource-poor settings. The study followed a cohort of 1,687 AGYW from 2017 to 2022.
Challenges experienced during participant follow-up in the DREAMS study
- Participant Migration
Generally, there is high population mobility in informal settlements, also known as slums. This makes it challenging to track research participants for follow-up interviews and measurement of outcomes. As individuals and families transition through life they can migrate for various reasons including schooling, marriage, work, economic pressures and security among others. During the DREAMS study, frequent relocations posed significant difficulties in tracing participants, especially when they moved to distant areas or lost contact with their families. The COVID-19 pandemic further exacerbated these challenges. Lockdown measures led to widespread economic hardship, causing many families to lose their livelihoods. As a result, some participants returned to rural areas, making it even more difficult to maintain contact and continue follow-up efforts.
- Maintaining Participants’ Contact Information
The initial telephone contact information provided by participants changed during follow-up, making them difficult to reach. Some phone numbers were disconnected, while others went unanswered. Additionally, service providers reassigned some numbers to new individuals who did not recognize the participants or were unfamiliar with the research team. In some instances, participants gave numbers for people they were not related to, and those individuals moved, making it impossible to reach them.
- Respondent Fatigue
After more than 20 years of data collection for the NUHDSS and other nested projects, some participants became hesitant to continue their involvement in the study. Many felt fatigued from being interviewed multiple times across various studies. Some participants declined to participate, others missed appointments, while others were discouraged by family members who provided misleading information about their whereabouts. This situation was particularly common among families expecting tangible benefits from the research, including DREAMS interventions.
- Community Social Disorder
The DREAMS study also coincided with two national election periods which came with election-related disruptions. During this time, heightened political tensions, suspicion, and occasional civil disturbances led some participants to avoid the research team due to fears of political affiliation. Others became involved in supporting election campaigns, which interfered with their availability. Some participants left the slums to vote or escape unrest, further disrupting follow-up and contributing to increased attrition.
Innovations implemented to ensure high retention in the DREAMS study
The DREAMS study leveraged the NUHDSS database for participant recruitment, utilizing existing biodata and contact information to track participants. Key reference points, such as names, household information, and residence locations, facilitated participant follow-up.
Before the study, community engagement meetings were held with various local stakeholders, including representatives from local administration, religious leaders, and Community Health Promoters (CHPs). These efforts raised community awareness about the study’s significance, fostering a sense of buy-in among community members. The stakeholders voluntarily provided information regarding participants’ movements within their villages and assisted in identifying members who could help with participant tracing.
Tracking methods included collecting participant phone numbers and alternative contacts, conducting follow-ups via phone, consulting household heads and trusted contacts, and employing active tracing through family members and neighbors. CHPs also assisted field interviewers in locating participants within the community.
To improve the participant experience and retention, interviews were conducted in a central private location rather than participants’ homes. This approach offered a comfortable and conducive environment, which helped participants feel more relaxed and confident. Extra support was provided for those traveling with children. Incentives included refreshments, modest reimbursements, and transportation support for relocated participants. Remote interviews via phone or WhatsApp were offered to participants who had temporarily moved away from the study sites but were still willing to participate.
Deploying experienced field interviewers was another key factor in ensuring high retention. Many of the interviewers were familiar with the community and had been involved in previous rounds of data collection for the DREAMS study. Their experience enabled them to effectively manage sensitive situations while engaging with participants and to easily follow up with them. During politically sensitive periods, fieldwork was paused to ensure the safety of both the participants and field teams. This approach also helped maintain trust with the community, ensuring that the study remained independent and unbiased.
These innovations, combined with practical tracking methods, community involvement, flexible interview formats, and thoughtful incentives, significantly contributed to high retention rates throughout the DREAMS study.
Lessons for longitudinal studies
Several key lessons from the DREAMS project can inform similar longitudinal studies, particularly regarding participant retention. These include:
- Anticipating and proactively planning to minimize and manage attrition
- Continuously monitoring and analyzing reasons for attrition
- Implementing community and staff-level mechanisms to reduce attrition
- Collecting adequate contact information from participants
- Building and maintaining strong community relationships
- Providing thorough training for field staff involved in data collection
- Engaging communities regularly by sharing project updates and findings
- Leveraging existing resources, such as CHPs, to help track participants
- Offering non-monetary, community-level incentives to support follow-up when feasible
By applying these innovative strategies, the DREAMS project successfully addressed attrition, achieving participant retention rates of 80% in 2018, 79.7% in 2019, and 84% in 2022 over a six-year follow-up period.