The “trial within cohort design” was a pragmatic model for low-resourced settings

Martin Heine, Wayne Derman, Susan Hanekom

Research output: Contribution to journalArticleAcademicpeer-review

2 Citations (Scopus)


Objectives: To determine the feasibility of using a trial within cohort (TWIC) design as a model to study pragmatic interventions in a low-resource setting to ensure that (i) ethical concerns raised with the conventional clinical trial design could be alleviated, (ii) key parameters could be obtained that may promote implementation of interventions in low-resource settings, although retaining the methodological rigor required to assess real-world efficacy. Methods: A TWIC design was adopted to evaluate the feasibility of a community-based, patient-centered rehabilitation program, in an underprivileged South African community. Procedural aspects of the trial in relation to recruitment, retention, acceptance, and methodological rigor were evaluated. Results: A total of 74 eligible participants, 36% of those who were identified as potential participants, agreed to participate and were randomized. Acceptance of the intervention (56%) was in line with previous research, and no reports of cross-contamination were received. Key lessons were learnt in the conduct of a TWIC design in low-resource settings, among others, related to blinding of the assessor, missing data, timing of recruitment, and various resource constraints. Conclusion: The findings of this study support further exploration for the use of this design in low-resource settings, particularly in settings where the conventional randomized clinical trial is ethically challenging or where detailed information on nonacceptance is paramount.
Original languageEnglish
Pages (from-to)111-121
Number of pages11
JournalJournal of Clinical Epidemiology
Publication statusPublished - 1 Jul 2022
Externally publishedYes


  • Cohort studies
  • Developing countries
  • Noncommunicable diseases
  • Poverty areas
  • Pragmatic clinical trials
  • Rehabilitation

Cite this