Worked example 1
A clinical trial reports 92% adherence based on patient self-report questionnaires.
(a) Outline two alternative methods of measuring adherence that would provide more objective data. [4 marks] (b) Evaluate the validity of self-report as a measure of medication adherence. [6 marks]
Show solution outline
(a) Alternative methods:
- Pill count: Pharmacist counts remaining pills vs prescribed amount — semi-objective, low cost.
- Biochemical assay: Blood test for drug metabolite levels — e.g. serum drug concentration confirms ingestion.
- MEMS cap: Electronic monitor records each bottle opening with timestamp.
- Pharmacy refill records (MPR): Medication Possession Ratio from prescription database.
(b) Evaluation of self-report validity:
- Strengths: Low cost, non-invasive, captures patient perspective and reasons for non-adherence; practical in large trials.
- Limitations:
- Social desirability bias — patients report what they think doctors want to hear; 92% likely inflated (actual adherence often ~50%).
- Recall bias — difficulty remembering missed doses over weeks.
- Lack of objectivity — cannot distinguish intentional from unintentional non-adherence without follow-up questions.
- Poor concurrent validity — self-report correlates weakly with biochemical measures (r ≈ 0.3–0.5 in meta-analyses).
- Comparison: Pill counts improve objectivity but can be gamed; biochemical tests are gold standard but expensive and impractical for routine care.
Judgement: Self-report is useful as a screening tool but should be triangulated with at least one objective measure in research. In clinical practice, non-judgemental questioning combined with pill count is most feasible.