Enhancing Clinical Study Retention Rates to Avoid Follow-up Bias: How Do We Keep Our Study Participants from "The Land of the Lost"?

Abstract

In an earlier publication, we discussed the importance of loss to follow-up, how it should be calculated, and how many patients can be lost to follow-up without mistrusting the results.1 We argued that incomplete follow-up could bias the results when the dropout rates are different between study groups or when the patients who drop out are different from those who do not drop out. Simply put, the patients lost to follow-up often have a different prognosis than those who complete the study. We discussed that properly calculating the loss to follow-up can only be done by determining the right denominator. That includes all those randomly assigned in a randomized controlled trial and all who had the procedure during a prespecified time in a cohort study. A good rule of thumb is that 20% poses serious threats to validity. However, even less than 20% loss to follow-up can be a problem. Now that we understand the significant impact that loss to follow-up can have on our study results, we want to shift our focus to strategies to limit loss to follow-up. Unfortunately for retrospective studies where the data has already been collected, the toothpaste is out of the tube; nothing can be done to improve patient participation in the study. However, with this study design it is worth documenting the flow of the patients through each stage of the study: enrollment, assignment, follow-up, and analysis. A diagram is strongly recommended (Fig. 1). Fig. 1 Example of participant flow in hypothetical retrospective case series evaluating the incidence of adjacent segment degeneration in two-level cervical artificial disk replacement (ADR). Enrollment includes the number of participants who were screened ... However, if you are involved in prospective data collection for randomized trials, longitudinal prospective cohort studies, or prospective registries, there are several safeguards and strategies that you should consider implementing from the beginning of the project. These will be discussed in this edition of Science in Spine. It is tempting to get excited about enrollment and then let your guard down during the follow-up phase. The problem is that it does not matter how many participants are enrolled if you cannot get them to the finish line. So the planning starts up front and the process continues with a diligent effort to maintain very high retention throughout the course of the study until the very last enrolled subject has finished his or her final follow-up. This result is difficult to accomplish without personnel and a commitment of resources. Every prospective study requiring follow-up beyond the perioperative period necessitates a project coordinator with experience in following participants over time, which can rarely be done effectively by clinician investigators who are busy with the day-to-day clinical responsibilities. The number of sites involved (if it is a multisite study) and the number of participants anticipated for enrollment will dictate what level of commitment you need from this project coordinator. Sometimes 25% effort is enough but often it requires 50% or more, even up to 100%, in larger multisite trials. This person needs to have strong organizational and communication skills. Ideally, the coordinator will have spent time on the “front lines” recruiting and following participants so that he or she can speak and problem solve the challenges that the study coordinators face in retaining subjects. An alternative is a central methods center that employs more than one coordinator part-time who is versed in the retention strategies. Once this person(s) is identified, and a comprehensive follow-up strategy is established, then flexibility and ingenuity are paramount as the study progresses because no study is alike. There will always be obstacles and nuances that may impact retention. Good project coordinators can respond to these challenges and enlist their clinical investigators and study coordinators along the way to help problem solve. The following sections provide a fairly self-explanatory checklist of strategies that have been employed by many successful clinical trialists. These are divided into a prestudy planning phase and a study execution phase, although there is distinct overlap in many of these. All need to be planned for and considered for Institutional Review Board (IRB) approval.