Evaluation of the productivity of hospital-based researchers: comparative study between the h-index and the h(fa)-index

The h-index is often used to evaluate researchers and to measure their respective contributions to the scientific community. However, it does not take into account the relative contribution of each author to the articles. We aimed to assess the added value of the h(fa)-index, an index introducing a weighting factor for the first author. This cross-sectional study was conducted in March 2020 in Switzerland. We selected all hospital-based senior physicians practicing internal medicine or family medicine in the six university hospitals. Using Web of Science, we retrieved the h-index and computed the h(fa)-index of all physicians with a non-zero h-index. We used Spearman coefficients to examine the correlation between the two indices and Cohen's kappa to determine the inter-rater agreement. 244 physicians were included in the study (men: 63%). They had a median h-index of 3 (IQR 11) and a median h(fa) index of 4.5 (IQR 14). The correlation between the two indices was very high (rho 0.98, p-value < 0.001). However, the inter-rater agreement was only moderate (Kappa 0.59 [95% CI 0.56-0.64]). The h(fa)-index was able to correct the categorization of the researchers in about 1/3 of the cases (agreement 63.4%). Although it is very challenging to evaluate researchers with a unique indicator, these data suggest that the h(fa)-index could be a better evaluation tool of researchers' productivity.