SeriesSample size slippages in randomised trials: exclusions and the lost and wayward
Section snippets
Exclusions before randomisation
Investigators can exclude participants before randomisation. The eventual randomised treatment comparison will remain unbiased (good internal validity), irrespective of whether researchers have well-founded or whimsical reasons for exclusion of particular individuals. However, exclusions at this stage can hurt extrapolation, the generalisability, of the results (external validity). For most investigations, we therefore recommend that eligibility criteria be kept to a minimum, in the spirit of
Exclusions after randomisation
Exclusions made after randomisation threaten to bias treatment comparisons. Randomisation itself configures unbiased comparison groups at baseline. Any erosion, however, over the course of the trial from those initially unbiased groups produces bias, unless, of course, that erosion is random, which is unlikely. Consequently, for the primary analysis, methodologists suggest that results for all patients who are randomly assigned should be analysed, and, furthermore, should be analysed as part of
Discovery of participant ineligibility
In some trials, participants are enrolled and later discovered not to have met the eligibility criteria. Exclusions at this point could seriously bias the results, since discovery is probably not random. For example, participants least responsive to treatment or who have side-effects might draw more attention and, therefore, might be more likely to be judged ineligible than other study participants. Alternatively, a physician who had treatment preferences for certain participants might withdraw
Postrandomisation, pretreatment outcome
Researchers sometimes report exclusion of participants on the basis of outcomes that happen before treatment has begun or before the treatment could have had an effect. For example, in a clinical trial of a specific drug's effect on death rates, investigators withdrew as non-analysable data on all patients who died after randomisation but before treatment began or before they had received at least 7 days of treatment. This winnowing seems intuitively attractive, because none of the deaths can
Protocol deviations
Deviations from assigned treatment happen in many trials. Some investigators suggest that participants who deviate substantially from the allotted treatment should be excluded in the final analysis, or should be included only up to the point of deviation. Although this approach seems attractive, it has a serious flaw: “the group which deviates from one protocol and the group which deviates from the other protocol may be so different […] that the treatment comparison in the remaining patients
Loss to follow-up
Losses to follow-up are perhaps the most vexing of the proffered reasons for exclusions after randomisation. Participants might move or might refuse to continue participating in the trial. Participants lost to follow-up could still be included in the analysis if outcome information could be obtained from another source, such as gathering data from a national death registry. Such opportunities, however, rarely arise. Without outcomes from those lost to follow-up, investigators have little choice
Conclusion
Trialists should endeavour to minimise exclusions after randomisation and to do intent-to-treat analyses. They should also follow the CONSORT statement for reporting.18, 19 The flow diagram (trial profile) helps particularly to track the progress of participants through a trial.
For readers, non-reporting of exclusions results in interpretation difficulties, such as the exclusion paradox, which misleads readers about trial quality. Moreover, mishandling of exclusions causes serious
References (23)
- et al.
Generation of allocation sequences in randomised trials: chance, not choice
Lancet
(2002) - et al.
Allocation concealment in randomised trials: defending against deciphering
Lancet
(2002) Statistical considerations in the intent-to-treat principle
Control Clin Trials
(2000)- et al.
An overview of clinical research: the lay of the land
Lancet
(2002) - et al.
Content of reports on clinical trials: a critical review
Control Clin Trials
(1984) - et al.
The CONSORT statement: revised recommendations for improving the quality of reports or parallel-group trials
Lancet
(2001) - et al.
Design and analysis of randomized clinical trials requiring prolonged observation of each patient, I: introduction and design
Br J Cancer
(1976) - et al.
Design and analysis of randomized clinical trials requiring prolonged observation of each patient, II: analysis and examples
Br J Cancer
(1977) On some prerequisites for a successful clinical trial
- et al.
Evidence-based medicine: how to practice and teach EBM
(2000)
Blinding and exclusions after allocation in randomised controlled trials: survey of published parallel group trials in obstetrics and gynaecology
BMJ
Cited by (455)
Determining the follow-up rates for shoulder arthroplasty: a retrospective review of a multicenter study of 4700 shoulders
2024, Seminars in Arthroplasty JSESHow robust are clinical trials in primary and secondary ankle sprain prevention?
2023, Physical Therapy in SportThe effect of ear acupressure on sexual functioning among lactating women: A randomized sham controlled trial
2023, European Journal of Integrative MedicineThe prospective association between the use of E-cigarettes and other psychoactive substances in young people: A systematic review and meta-analysis
2023, Neuroscience and Biobehavioral Reviews