Evaluating the Landscape of Clinical Research in Neurosurgery.

Abstract

Many clinical trials and observational research never reach publication in peer-reviewed journals. Unpublished research results, including neutral study findings, hinder generation of new research questions, utilize healthcare resources without benefit, and may place patients at risk without benefit. To examine the publication of neurosurgery trials listed in ClinicalTrials.gov. Clinical neurosurgery research was identified by searching the registry and categorized by study type. Associated publications were identified on Pubmed.gov. Among the 709 studies identified, spine (292, 41.2%) studies were most common, followed by tumor and cranial (each 114, 16.1%). Funding was predominantly private (482, 68.0%), followed by industry (135, 19.0%) and National Institutes of Health (9, 1.3%). A lower proportion of published studies (vs unpublished) received private funding in functional (33.3 vs 65.3%) and tumor (80.0 vs 68.7%). Only 104/464 (22.4%) studies had an associated publication. The mean time from listed study completion to first publication was 31.0 ± 27.5 mo. Most published studies had significant study differences between treatment arms (n = 72, 69.2%); studies with neutral findings were less likely to be published (n = 13, 12.5%). Surgical discipline (P = .1), funding source (P = .8), patient age (P = .4), planned enrollment (P = .1), phase of trial (P = .3), and study type (P = .2) did not affect publication rates. However, the interaction between study category and funding source significantly affected publication rate (P = .04, generalized linear model, R2 = 0.05). Publication timing (1-way analysis of variance, P = .5) and frequency (chi-square, P = .2) did not differ among disciplines. Clinical trials and observational research in neurosurgery are often not published promptly, especially if results were nonsignificant or the trial had private funding.