Is the h-Index Enough? A Comprehensive Bibliometric Analysis of 871 Academic Plastic Surgeons and Their Departments/Divisions

Abstract

Background: In plastic surgery academia, research output is heavily used as a metric of accreditation, from assessing residency applicants to evaluating faculty for promotion. The h index, defined as an author's h papers with at least h citations, is commonly used as a measure of academic success. However, the index itself disfavors junior researchers, favors publication quantity, and discounts highly cited works. Given the importance of bibliometrics within plastic surgery, there is a paramount need to adopt additional metrics to measure research productivity. The authors sought to validate the use of time-independent bibliometrics to complement the h index in measuring citation impact.

Methods: The genders and academic titles of plastic surgeons affiliated with US plastic surgery programs were recorded. Author publications were retrieved from Scopus. Bibliometrics software was used to calculate the following metrics per surgeon: h index, e index, and g index. Time-adjusted versions of these indices were used to correct for the number of years since first publication. Medians and interquartile ranges (IQRs) are reported. Departmental ranks were determined using the cumulative sum of time-corrected indices and compared with Doximity departmental research rankings. P < 0.05 was deemed significant.

Results: Indices were calculated for 871 academic plastic surgeons in 85 departments/divisions. Men had statistically greater h index (median, 13.0 [IQR, 7.0-21.0] vs 6.0 [IQR, 3.0-13]; P < 0.001), e index (18.3 [IQR, 10.0-28.7] vs 11.1 [IQR, 5.5-18.4]; P < 0.001), and g index (23.0 [IQR, 11.0-39.0] vs 11.0 [IQR, 5.0-22.0]; P < 0.001) than women. Professors had the highest median time-uncorrected indices. After adjusting for the number of years since an author's first publication, there were no significant differences in m quotient (men: 0.66 [IQR, 0.40-0.98] vs women: 0.57 [IQR, 0.33-0.90]; P = 0.05) and ec index (men: 0.93 [IQR, 0.62-1.3] vs women: 0.87 [IQR, 0.50-1.3]; P = 0.08) between genders. Departmental chairs had significantly higher indices than other faculty after correcting for time. The calculated program rankings were low to moderately correlated with that of Doximity (correlation coefficient τ = 0.49 [95% confidence interval, 0.37-0.59; P < 0.001]).

Conclusions: Men and women have statistically similar citation patterns after correcting for the time. Citation differences between academic levels are less pronounced when controlling for time, suggesting comparable research quality between academic roles.