A systematic review of the quality of randomized controlled trials in head and neck oncology surgery

To determine the quality of randomized controlled trials (RCTs) in head and neck surgery in which surgery was a primary intervention.

Potential articles were identified in PubMed without publication date restrictions.

Articles were scored using the CONSORT checklist and the relationship between the checklist score and whether the first and/or last authors were surgeons was investigated. Differences in the checklist score based on how many surgeons were among the first and last authors of the study were analyzed using the Kruskal-Wallis test. Fisher's exact test was used to examine if there was a significant difference of the reporting of individual items from the checklist between surgeons and nonsurgeons. A nonparametric trend test was used to determine whether there was a difference in the reporting of individual items based on whether there were none, one, or two surgeons among first and last authors.

A total of 38 publications satisfied the inclusion criteria. There was a trend toward lower quality for studies in which surgeons were either first, last, or both first and last authors compared to studies that were first-authored and last-authored by nonsurgeons (P = 0.068). Nonsurgeons were more likely to report on critical elements regarding hypothesis, sample size determination, randomization, and eligibility of centers (P = 0.023-0.058).

The quality of RCTs in head and neck surgery is poor. Improved training in conducting and reporting clinical research is needed in otolaryngology residencies.