Critical Assessment of Surveys in Plastic and Reconstructive Surgery: A Systematic Review

Abstract

Surveys are a fundamental research tool used in the field of plastic surgery. Although inherent biases within surveys have been recognized, the extent to which this problem exists in the plastic surgery literature remains unknown. The authors aim to (1) determine the extent to which common survey biases occur in plastic surgery surveys and (2) identify areas of improvement for survey researchers. The authors conducted a systematic review of surveys in the plastic surgery literature from 1997 to 2017. After applying their inclusion and exclusion criteria, two reviewers examined a total of 199 studies containing a survey. A modified checklist based on the published literature was used to determine the presence, type, and severity of biases in plastic surgery surveys. The mean percentage of biased questions in each survey was 24 percent. The most common biases were moderate- (12 percent) and low-impact biases (10 percent). Specifically, the most common biases were forced choice questions (6 percent), horizontal responses (6 percent), and double-barreled questions (1 percent). Only 13 percent of the articles that fit the authors' inclusion criteria contained a full-length survey. Although surveys in plastic surgery do not contain high-impact biases, the authors have identified areas of improvement for surveys used in this specialty. The authors found that the surveys contain biases in question structure and the survey design, which can be improved through minor revisions to the question stem or responses. Moreover, journals should require researchers to submit a full-text survey for research transparency.