Table 3. Proportion of discontinued trials.

	REC Freiburg						Other RECs1
	Controlled trial	Single arm trial	Cohort study	Total NPSs	RCTs	Comparison total NPS vs RCTs P	RCTs
All studies
Unclear/missing = excluded	8%	16%	13%	14%	27%	<0.001	27%
Unclear/missing = completed	7%	15%	11%	13%	25%	<0.001	25%
Unclear/missing = discontinued	15%	22%	25%	22%	31%	0.014	33%
Unclear/missing = 50% completed, 50% discontinued	11%	18%	18%	17%	28%	0.004	29%
Excluding studies stopped for harm, benefit, futility
Unclear/missing = excluded	4%	13%	13%	12%	19%	0.057	22%
Unclear/missing = completed	4%	12%	11%	11%	18%	0.047	20%
Unclear/missing = discontinued	12%	19%	25%	20%	24%	0.251	29%
Unclear/missing = 50% completed, 50% discontinued	8%	16%	18%	15%	21%	0.128	25%

¹ approved by 5 research ethics committees in Switzerland and Canada [1].

Estimated proportion of discontinued studies applying different assumptions for missing data with or without excluding studies that were stopped for intervention-related reasons such as harm, benefit, and futility (see text for a rationale).

Abbreviations: NPSs, non-randomised prospective studies; RCTs, randomised controlled trials.