Pilot randomized controlled trials in pediatric critical care: a systematic review.

Duffett M, Choong K, Cupido C, Hartling L, Menon K, Thabane L, Cook DJ.

Aims & objectives

There are substantial challenges to conducting random- ized controlled trials (RCTs) in pediatric critical care. Pilot RCTs can be used to evaluate the feasibility of, and inform the design and con- duct of, larger RCTs. Our objective was to systematically identify and describe pilot RCTs in pediatric critical care.

Methods

We searched the Evidence in Pediatric Intensive Care database (epicc.mcmaster.ca) for published trials (1986 to July 1, 2012) that were described by the authors as pilot, feasibility, proof-of- concept, phase 2 or preliminary studies. We then used Web of Sci- enceÒ and the World Health Organization’s Clinical Trials Search Portal to search for trials informed by these pilot trials.

Results

We identified 20 pilot RCTs out of a total of 225 RCTs. The number of children randomized ranged from 6 to 56; the median number was smaller than non-pilot RCTs (26 vs. 50, p\0.001). Three trials (15 %) explicitly reported feasibility as an outcome and none reported criteria for success. Blinding was reported in 9 RCTs (45 %). We found 3 registered trials (2 listed as recruiting and 1 as terminated) in trial registries that were similar to 4 pilot trials. 17 pilot trials were indexed in Web of Science; these were cited a median number of 15 (IQR 4–30) times in total and 2.7 (IQR 0.9–6.9) times per year since publication.

Conclusions

Pilot trials are an underutilized research strategy in Pediatric Critical Care research. Published pilot trials uncommonly report explicit feasibility outcomes.

Download the poster | Visit the journal website

This poster was presented at the ESPNIC Annual Meeting (June 2013; Rotterdam, Netherlands).