Sample Size Parameters in Published Prosthodontic RCTs are Inadequately Reported

Objectives: In the planning of Randomized Controlled Trials (RCTs), the determination of an appropriate sample size is important. If the sample size is too small, a study may be unable to answer its research question; if too large, it may be impractical to conduct. Sample size calculation usually involves 4 parameters: type-1 error (alpha), power, minimal clinically relevant difference (MCRD), and variability. Based on these parameters, the appropriate formula is used to calculate the sample size required. The purpose of this study is to describe the reporting of sample size calculation and parameters used in RCTs published in prosthodontic journals.
Methods: A MEDLINE search via PubMed using publication type randomized controlled trial was conducted for The International Journal of Prosthodontics, The Journal of Prosthetic Dentistry, and The Journal of Prosthodontics, with publication dates 2008 - 2017. Citations retrieved were included if the study involved human subjects, included at least 2 treatment groups, and group assignment was by random allocation. Pilot and follow-up studies were excluded. RCTs identified were reviewed and the following data recorded: reporting of method used for sample size calculation and reporting of sample size parameters - alpha, power, MCRD, and variability.
Results: The search strategy retrieved 176 citations; 96 met inclusion criteria and were reviewed. Only 40% of RCTs reported the method used for sample size calculation and only 16% reported all 4 sample size parameters used. Alpha was reported in 92% of RCTs, but only 31% reported power; 21% and 11% reported MCRD and variability, respectively.
Conclusions: Sample size methodology and parameters used for sample size determination are inadequately reported in RCTs published in prosthodontic journals.