Impact of Subspecialty Fellowship Training on Research Productivity Among Academic Plastic Surgery Faculty in the United States

Aditya Sood, MD, MBA; Paul J. Therattil, MD; Stella Chung, BA; Edward S. Lee, MD


ePlasty. 2015;15 

In This Article


This study examined 590 academic plastic surgeons from 80 academic departments. We studied the current distribution of fellowship training among academic surgeons and whether fellowship training had an effect on their academic productivity. In the United States, approximately three-fourths (72%) of plastic surgery academic faculty members have undergone at least 1 subspecialty fellowship training program (Fig 1). This is comparable with the proportion of fellowship-trained practitioners in the fields of otolaryngology (77%), ophthalmology (83%), and neurosurgery (66%).[10–12] This phenomenon can be explained by the ability to obtain competitive professorship with postsecondary training experience. The most common training obtained by academic plastic surgeons was multiple fellowships (Fig 2). Surgeons interested in academia may pursue multiple fellowships to gain further knowledge and skills in more than 1 subspecialty. In reverse, surgeons with multiple fellowships training acquire diverse clinical and surgical proficiency, which may provide them with more opportunities to enter academia.

The h-index has been used as an important marker of scholarly impact in various fields of medicine.[2,12–29] Current metrics of academic productivity are total number of citations, total number of publications, and citations per publication, and h-index is a single statistic that incorporates all 3 factors.[9] Previous publications have identified and proved the value of h-index as a predictive role in future academic productivity, research grant funding, and academic promotion.[7,9] There are limitations of using h-index as a measurement of academic productivity. h-index is may be inflated by the following: negatively or disproved citations; favored citations of famous authors over lesser well-known individuals; and self-citations.[9,10,30–33] To minimize bias, we have excluded self-citations in the h-index calculation process.

The mean h-index of all academic surgeons was 9.1. There was no significant difference in the mean h-index between the fellowship-trained and non–fellowship-trained groups, although it approached a trend (P = .11; Fig 1). Similar findings have been reported from the fields of urology and neurosurgery.[11,34] In these surgical specialties, h-index may not be affected by fellowship training itself but by gender, type of residency program, posttraining academic position, or institution the surgeon is affiliated with. Our results differed from a previously published study of academic plastic surgeons that reported a significantly higher mean h-index in the fellowship-trained group.[1] The discrepancy is most likely due to excluding faculty from independent plastic surgery institutions in our analysis. A significantly higher h-index in fellowship-trained physicians has also been reported from the fields of ophthalmology and otolaryngology.[1,10] Factors that are likely to be associated with a higher h-index were male gender, faculty from frequently represented training programs, no private practice affiliation, and academic rank.[1,9] The variation noted among different surgical specialties may be impacted by several factors such as the number of surgeons, research emphasis within a specific field, and inherent bias within the data collection process. Therefore, comparison within a given field rather than between specialties is currently recommended.[10]

Despite the lack of significant correlation between fellowship training and h-index, a trend was observed upon stratification by fellowship type. Research fellowship-trained surgeons comprised only 8% of the faculty, yet they had the highest mean h-index (Fig 3). Both personal interest and emphasis on research and publishing in the research group are evident factors leading to a high h-index. Research fellowship provides time and resources for a surgeon to focus on innovative research. This is possible in clinical fellowships as well; however, practice is often busier with increased clinical duties required to generate revenue and sustain a practice.[1] Furthermore, research fellowships are not restricted to a particular subspecialty or time frame. While clinical fellowships are typically 1- or 2-year programs, postresidency research fellowships can last for any number of years, thus allowing the surgeons to complete experiments and/or trials to publish their findings. The second highest h-index was seen in the multiple fellowships-trained cohort. Many surgeons in this category also have research background that increases their scholarly impact.

We further distributed fellowship-trained surgeons by academic position. We found that 70% of professors, 72% of associate professors, and 73% of assistant professors have completed fellowship training. Similar proportions across all ranks indicate that postsecondary training experience has been traditionally important and is not a recent phenomenon. Professors had the highest mean h-index of 15.3 and continued to maintain a significantly higher mean h-index than their juniors across all fellowships, with exception to burn and wound healing (Fig 5). Our findings are consistent with previously published studies showing higher h-indices associated with higher academic ranks.[9,12,34]

Burn and research fellowship-trained surgeons had the greatest proportion of both professor and associate professor faculty members, whereas cosmetic and microsurgery fellowship-trained surgeons had the greatest proportion of assistant professors (Fig 4). This may indicate a current shift of plastic surgery practice to cosmetic surgery and microsurgery as well as increased academic opportunities for surgeons specializing in these fields. Alternatively, one may speculate that cosmetic surgery and microsurgery are newer fellowships that were less available to senior plastic surgeons at the time of their residency completion.

One of the criticisms of using h-index as a static measurement of academic productivity is its lack of control for the time a publication may take into account.[6,7] We did indeed observe that the duration of medical practice as determined by the year of medical school graduation had an overall positive effect on h-index growth (Fig 6). The graduation year of plastic surgeons ranged from 1950s to 2000s. The highest h-index was observed in the 1960s group rather than the 1950s group. We believe that the small sample size (n = 2) in the 1950s group is the cause of deviation from an otherwise linear trend. Inflation and deflation of h-index must be noted. Some surgeons have obtained degrees such as PhD and DMD after medical school. The additional years in school have been unaccounted for, due to the limited information available online, leading to inflation of actual years of medical practice. In addition, h-index of senior faculty may be deflated since Scopus includes only cited publications before 1996. Overall, the time of practice versus h-index trend is relevant and meaningful in illustrating that time is needed for h-index growth. With time, practitioners gain in-depth knowledge and become more clinically proficient, increasing their scholarly impact.

There are other limitations to this study. Eloy et al[10] have extensively discussed the limitations in using Scopus as a single database for h-index calculation. Several other databases exist with inherent variations, such as Google Scholar, but studies show that one particular database is unlikely to result in inaccurate conclusions.[10,15] In addition, a surgeon's academic productivity is a multifaceted dimension. It not only comprises research work but also includes clinical expertise and teaching.[9]h-index does not encompass all of the mentioned factors but is frequently used as an index to quantify a physician's scholarly impact.[9] Plastic surgery is unique in that surgeons from various residencies such as otolaryngology, oral and maxillofacial surgery, and orthopedic surgery participate in plastic surgery fellowship training. Therefore, fellowship population and data may be more variable than other surgical fields.

In this study, we showed the current status of fellowship-trained academic plastic surgeons. We examined their academic research output distributed by type of fellowship, academic rank, gender, and time. A surgeon's scholarly impact is an important tool in academic job decisions such as hiring and promotion, and it also has a role in predicting grant funding.[35] Differences in h-indices among fellowship-trained faculty should encourage more mentoring and educational resources in underrepresented fellowships. As time proved to be an important factor in one's scholarly impact, we expect that junior faculty will see an eventual increase in their academic productivity. The same principle can be applied to fellowship categories associated with lower h-indices.