Small sample sizes in clinical trials: a pragmatic approach to clinical research in veterinary medicine.

As evidence synthesis and guideline development efforts advance in veterinary medicine, the quality and quantity of data can be limiting factors. Aspiration and pragmatism must be balanced to ensure optimal data development and availability. Systematic reviews and meta-analyses are ideally based on multiple large randomised controlled trials from a broad range of relevant populations, but cost, time and caseload can be substantial barriers. Small randomised controlled trials can also provide useful and actionable information. Data from numerous small but robustly designed and executed, comparable randomised controlled trials can support stronger conclusions, complementing larger randomised controlled trials or providing critical important knowledge when larger randomised controlled trials are not available. The value from these small trials is in the data, not the analysis, as individual analyses within these randomised controlled trials are usually underpowered to detect reasonable and clinically relevant endpoints. A desire for perfection can inhibit progress if randomised controlled trials are not performed, or if small but potentially useful data sets remain unpublished. We encourage the veterinary scientific community, including researchers, reviewers and editors, to strive for optimal study designs and sample sizes but to be open to publishing data from small trials that may provide little insight in isolation but contribute useful data for meta-analyses.