Evaluation of the Relative Citation Ratio Versus the H-Index as Measures of Citation-Based Scholarly Influence in Academic Neurosurgery

Résumé fourni par la source

BACKGROUND AND OBJECTIVES: Evaluating citation-based scholarly influence among academic faculty is foundational to academic medicine, as retention, promotion, and tenure to govern career advancement at every level. The h-index has been the standard metric across fields, including neurosurgery, but is considered ineffective due to multiple inherent deficiencies. The National Institutes of Health (NIH) developed the relative citation ratio (RCR) that addresses several of these shortcomings. We therefore sought to evaluate RCR vs the h-index for assessing citation-based scholarly influence among academic neurosurgery faculty, fellows, and residents in the United States. METHODS: A retrospective cross-sectional analysis used publicly available data on 3463 neurosurgery faculty, fellows, and residents at 115 Accreditation Council for Graduate Medical Education-accredited neurological surgery programs. RCR was calculated by NIH iCite and the h-index by Scopus. The mean RCR (m-RCR) is the average RCR across an author's publications; weighted RCR (w-RCR) is the sum of all RCR scores. RCR is benchmarked to NIH-funded publications, with an RCR of 1.0 corresponding to the median NIH-funded article in a given field and year; article-level fields are defined by cocitation networks rather than fixed journal categories. RESULTS: Men comprised 79.1% of the cohort and attendings 54.9%; most (53.8%) had 0 to 10 years of experience. Significant differences in h-index, m-RCR, and w-RCR were identified across sex, position, faculty rank, years in practice, and subspecialty (P < .001 for all). Neuro-oncology demonstrated high impact (m-RCR: 1.67; w-RCR: 76.09; h-index: 27.00) while pediatrics showed lower metrics (w-RCR: 28.55; m-RCR: 1.17). Correlation with career duration was highest for h-index (r = 0.61), followed by w-RCR (r = 0.37) and m-RCR (r = 0.11; P < .001 for all). CONCLUSION: m-RCR showed a weaker association with career duration than h-index, suggesting that it may provide a more career-stage-independent measure of citation-based scholarly influence. However, persistent subspecialty differences and the inability of both RCR-derived metrics and h-index to account for author-level contribution indicate that these bibliometrics should be interpreted as complementary tools rather than standalone measures.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.