Research Design Service: East Midlands
« back

Top ten tips for developing and using progression criteria for internal pilot studies

  Created on Tuesday, 11 Feb 2020. Posted in Design Tips

From the BMJ Journals: Informing efficient randomised controlled trials: exploration of challenges in developing progression criteria for internal pilot studies

1. A traffic light system of green (go), amber (amend) and red (stop) might be preferable to a simple stop/go approach when specifying progression criteria for internal pilot studies;

2. Pre-specified progression criteria agreed with funders need to strike a careful balance between being firm enough to promote ambition in the trial team yet being flexible enough to allow opportunities to remedy early problems;

3. Recruitment progression criteria should be based on rates per centre per unit time (eg, per month) that can be easily extrapolated, rather than specifying that an absolute number should be reached by a specific date, due to the unpredictability of opening sites;

4. When recruitment falls behind, it is essential to explore screening logs to determine whether insufficient participants were approached, insufficient participants passed eligibility criteria or insufficient eligible participants agreed to randomisation;

5. The assessment of intervention adherence, cross-over and outcome event rates should take into account the duration from randomisation to timing of primary outcomes if sufficient data are to be gleaned in time to inform progression;

6. When assessing missing data, it is important to explore the degree of missing data within key outcomes as well as the percentage of participants with missing data;

7. Trial teams should involve both their funders and their Trial Steering Committee in assessing their progression criteria;

8. Pilot study recruitment sites should be representative of sites that recruit into the main study;

9. Triallists may be able to take the opportunity to assess whether changes to existing technologies have occurred since the original study was planned, so that new technologies can be considered with funders, such as using an adaptive design;

10. Pilot studies need to be reported fully. An extension to CONSORT guidelines for pilot and feasibility studies is now available.