Purpose: Decision science, a relatively young academic discipline, faces challenges at institutional appointment and promotion committees. Other than the sine qua non of peer-reviewed extramural funding, scholarly publications are the coin of the realm for professional advancement. Citation analysis has emerged as an arbiter of the value of a candidate's publication record, but both numbers of articles and citations are skewed measures. Recently Hirsch (PNAS 102:16569; 2005) proposed the h-index, a statistic that attempts to reconcile publications and citations. In this study we explored the value of the h-index for the medical decision making subfield.

Methods: We used the Web of Science (WoS) to identify source documents and citations from three groups: Society for Medical Decision Making (SMDM) board members (1981-present, n=114), an academic cancer center list of principal investigators (PI's) (n=98), and a random sample of elected members of the American College of Medical Informatics (ACMI) (n=50). For each individual Google Scholar™ served as a redundant check on identity. Source articles and citations since 1964 comprised raw data; h values were measured by inspection of the ordered WoS results. Source articles and h values were compared across groups using ANOVA and conservative t-tests.

Results: SMDM board members published an average of 100.8 articles (range 12-389), PI's averaged 47.6 (1-292), and ACMI 72.8 (3-386). All pairwise comparisons were statistically significant, in all cases p<0.02. h-indexes averaged as follows: SMDM 20.3 (range 4-56), PI's 17.9 (1-84), ACMI 16.0 (2-52). The only significant pairwise comparison was between SMDM and ACMI (p<0.03). Basic scientists and bioinformatics researchers tended to have higher h and higher ratios of source documents to h. Journal articles were generally cited more frequently than scientific proceedings.

Conclusions: SMDM board members, assumed professionally successful, were highly published with strong h-indexes. Hirsch noted that promotion to associate and full professor in the physical sciences might be expected with h of 12 and 18; corresponding values in the biomedical sciences would be 17 and 25. SMDM board members h values tracked with an academic faculty group and were higher than a sample from ACMI, an elected society that ordinarily admits only senior faculty level individuals. The h index offers a statistic that can normalize publications and citations for boundary spanning academic disciplines such as medical decision making.