A mixed methods and triangulation model for increasing the accuracy of adherence and sexual behaviour data: the...

Area of Science:

• Clinical Research
• HIV Prevention Research
• Social Science in Medicine

Background:

• Accurate data on adherence and sexual behavior are critical for HIV-related research, particularly microbicide trials.
• Participant self-reporting is the primary method for collecting such sensitive data, but often lacks a "gold standard" for validation.
• Existing methods for data collection in clinical trials have limitations that can impact accuracy.

Purpose of the Study:

• To describe and evaluate a mixed-method, triangulation model for improving the accuracy of adherence and sexual behavior data in a multi-center vaginal microbicide clinical trial.
• To address the challenges of collecting reliable data on sensitive topics within the context of clinical research.
• To identify and resolve discrepancies in self-reported data through a comprehensive data collection strategy.

Main Methods:

• A mixed-method approach was employed, integrating quantitative and qualitative data collection techniques.
• Data were gathered from a subsample of 725 women using structured interviews, coital diaries, in-depth interviews, returned gel applicator counts, focus group discussions, and ethnography.
• Two levels of triangulation were implemented: resolving discrepancies with participants and integrating individual data with broader qualitative findings.

Main Results:

• The model successfully identified and, to a large extent, resolved discrepancies in data collected through various methods.
• Triangulation of data revealed insights into "clinical trial culture" and practices that can affect data accuracy.
• Collaboration between clinical trialists and qualitative social scientists, alongside preparatory studies, was crucial for the model's success.

Conclusions:

• Integrating qualitative social science with mixed methods and triangulation is feasible and effective in clinical trials.
• This approach can uncover and resolve inaccuracies in data concerning adherence and sensitive behaviors.
• The study highlights the impact of "trial culture" on data accuracy and suggests potential solutions.