What happens if subjects in a pilot study do not complete it?

Managing Incomplete Pilot Studies: Data Handling and Implications

When subjects in a pilot study do not complete the study, researchers should document the attrition rate, provide explanations for non-completion, and analyze the impact on feasibility assessment while still including the available data in the analysis. 1

Understanding Pilot Studies and Their Purpose

Pilot studies serve a specific purpose that differs from full-scale clinical trials:

• They assess the feasibility of research approaches intended for larger studies 2
• They evaluate recruitment, randomization, retention, and assessment procedures 2
• They are not designed for hypothesis testing or efficacy evaluation 2
• They help identify needed modifications for subsequent larger studies 3

Proper Documentation of Subject Non-Completion

When participants do not complete a pilot study, the CONSORT guidelines recommend:

• Reporting the number of participants who did not complete follow-up as planned, by study group 1
• Providing explanations for why participants could not or would not complete the study 1
• Documenting reasons such as transportation issues, participant death, or dissatisfaction with the intervention 1

Impact on Feasibility Assessment

The primary purpose of a pilot study is to assess feasibility, and subject non-completion actually provides valuable information:

• High attrition rates signal potential problems with the study design that need addressing before the full-scale trial 3
• Non-completion helps identify barriers to participant retention that can be modified 2
• Retention rates from the pilot inform sample size calculations for the main study, including accounting for expected dropouts 4

Data Analysis Considerations

When analyzing pilot study data with incomplete participants:

• Include all available data from participants who did not complete the study 5
• Document the follow-up rates at all time points 1
• Use the information to perform sensitivity analyses for the main trial's sample size calculations 4
• Recognize that small sample sizes in pilots lead to imprecise estimates, requiring cautious interpretation 2

Reporting Recommendations

The American Academy of Ophthalmology Task Force recommends:

• For pilot studies, "the entire protocol should be run on these study subjects, including the data collection, data analysis, and presentation of the data" 1
• Even with incomplete data, the available information should be analyzed and reported 1

Common Pitfalls to Avoid

• Overinterpreting results: Avoid drawing conclusions about efficacy or effectiveness from incomplete pilot data 2
• Ignoring attrition: Failing to report and analyze reasons for non-completion misses valuable feasibility information 1
• Selection bias: Be aware that differential loss to follow-up between comparison groups can introduce bias 1
• Underpowered analysis: Remember that pilot studies are not designed to provide meaningful effect size estimates due to small samples 2

Implications for Main Study Design

Information from incomplete pilot studies should guide:

• Adjustments to recruitment and retention strategies 3
• Realistic sample size calculations that account for expected attrition 4
• Modifications to study protocols to address identified barriers 2
• Go/no-go decisions about proceeding with the full-scale study 3

Pilot studies with participant non-completion still provide valuable feasibility information that can significantly improve the design and conduct of subsequent full-scale trials, even when not all participants complete the protocol.