- Email: [email protected]
Academic Productivity of Spine Surgeons at United States Neurological Surgery and Orthopedic Surgery Training Programs
Original Article
Academic Productivity of Spine Surgeons at United States Neurological Surgery and Orthopedic Surgery Training Programs Alexander F. Post, Adam Y. Li, Jennifer B. Dai, Akbar Y. Maniya, Syed Haider, Stanislaw Sobotka, Tanvir F. Choudhri
OBJECTIVE: Spinal surgery is taught and practiced within 2 different surgical disciplines, neurological surgery and orthopedic surgery. We have provided a unified analysis of academic productivity measured using the h-index attributable to spine-focused faculty at U.S. residency programs.
-
METHODS: A total of 278 Accreditation Council for Graduate Medical Education training programs were assessed to identify 923 full-time faculty members with a spinal surgery designation, as defined by spine fellowship training or case volume >75% in spine surgery. The faculty were assessed with respect to academic rank, duration of practice in years, and academic productivity (h-index).
neurosurgical and orthopedic surgery spine faculty, with some noticeable differences. These results can be used for benchmark purposes to assess the relative productivity of its faculty and could be of interest to those pursuing academic opportunities in spine surgery.
-
RESULTS: The comparison showed a significantly greater mean h-index for neurological spine surgeons. The mean h-index for both disciplines increased significantly as faculty rank increased. Within the academic ranks of assistant and associate professor, neurological spine surgeons had significantly greater mean h-indexes. Neurological spine surgeons had a significantly lower practice duration. At all ranks except for assistant professor, the mean practice duration was not significantly different statistically between the neurological spine and orthopedic spine surgeons. A positive correlation between the h-index and practice duration was found for both spine surgical disciplines. The proportional odds models for neurological and orthopedic spine surgeons were moderately successful at predicting faculty rank according to the h-index.
-
CONCLUSIONS: We present a unified view of academic productivity as measured by the h-index among
-
INTRODUCTION
T
he academic output of an individual or a department has previously been measured by publication number, publication type (abstract, journal article, chapter), quality of research (level of evidence, prestige, impact factor of the publication journal), and associated funding (grants awarded by the National Institutes of Health). The Hirsch h-index, proposed in 2005,1 is defined as the number of papers (h) with citation number h. This metric attempts to balance the quantity of an author’s publications with an assessment of each article’s quality and relevance to the reported data. It is the most widely cited and most widely used of the citation-based bibliometrics used for longitudinal self-evaluation, comparison among members of a department or between departments, and evaluation of an individual with respect to academic promotion or funding. The h-index was first used in the neurosurgical literature by Lee et al.2 in 2009, in which an evaluation of 30 neurosurgical programs confirmed a positive correlation between the h-index and academic rank. Larger, subsequent studies by Spearman et al.3 (1120 neurosurgeons in all electronic residency application serviceelisted training programs) and Campbell et al.4 (986 faculty members in 97 academic neurosurgery departments) also supported the correlation between the h-index and academic rank. Comparisons have also been made across neurosurgical departments,5 within a particular neurosurgical subspecialty,6 and across different neurosurgical subspecialties.7
Key words Academic neurosurgery - Academic orthopedic surgery - Bibliometrics - h Index - Spine surgery
Department of Neurosurgery, Icahn School of Medicine at Mount Sinai, New York, New York, USA
Abbreviations and Acronyms NS: Neurological surgery OS: Orthopedic surgery
Journal homepage: www.WORLDNEUROSURGERY.org
-
WORLD NEUROSURGERY -: e1-e8, - 2018
To whom correspondence should be addressed: Alexander F. Post, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.09.150
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
The h-index was first used in the orthopedic literature in 2009 by Vitzthum et al.,8 who performed a scientometrics analysis of scoliosis research. Subsequent orthopedic reports have confirmed the correlation between the h-index and academic rank among academic hand surgeons and sports medicine fellowship faculty9,10 and have assessed the early scholarly output (i.e., studies reported in the first 5 years after board certification) of existing orthopedic professors.11 Khan et al.12 examined fellowship-associated adult total joint reconstruction surgeons at 66 programs and noted a positive correlation between the h-index and number of fellows, faculty academic title, years in practice, and formal fellowship training. Bibliometrics have been used to show a positive correlation between scholarly impact and National Institutes of Health funding.13 Most recently, Bastian et al.14 examined 2061 academic orthopedic surgeons from 120 departments and, again, confirmed that academic rank increased as the mean h-, m-, and e-indexes increased. The h-index can be a relevant metric when making comparisons within a given academic field. However, caution has been recommended against using the h-index to compare individuals in different academic fields.15 Nevertheless, we have not found evidence to determine whether this holds true for the same academic field within different specialties. Spinal surgery is notable in that it is taught and practiced within 2 different surgical disciplines—neurological surgery (NS) and orthopedic surgery (OS). We examined the academic productivity of spinefocused faculty at NS and OS residency programs in the United States. METHODS Selection of Programs and Surgeons Public searches of the Accreditation Council for Graduate Medical Education website were performed to identify 110 NS and 168 OS residency training programs active through 2016 and 2017. Faculty members with spine expertise were identified by examining the department websites (faculty lists, subspecialty section lists, faculty biographies) and hospital websites (faculty lists, “spine section” or “spine team” lists). The faculty members included in the present study were surgeons with neurosurgery or orthopedic surgery listed as their primary specialty. Furthermore, the included faculty had undergone fellowship training in spine surgery or had a surgeon case volume of >75% in spine surgery. Adjunct faculty with a primary specialty other than NS or OS (e.g., radiology, otolaryngology, rehabilitation medicine) and nonsurgeons (e.g., neurologists, radiologists, physiatrists, nonsurgical Doctor of Philosophy) were excluded from the present study. The faculty members were assessed for academic rank (instructor, assistant professor, associate professor, professor), duration of practice (defined as years since graduation from residency), and academic productivity (h-index as determined by Scopus). Once the spine faculty had been identified, the departments were contacted separately by electronic mail or telephone, or both. The departments were asked to verify the accuracy of lists created using publicly available data and to augment with any data available within the department. The surgeons were contacted directly in select cases. Unresponsive
e2
www.SCIENCEDIRECT.com
departments were contacted 5 times before being marked as unverified. After determining that verified and publicly available data containing unverified programs were not significantly different statistically (Figure 1), all further analyses used publicly available data. Calculation of h-Index Hirsch’s h-index, for an individual, is defined as the number of reports (h) with citation number h. For each faculty member, an automated h-index was calculated from the Scopus abstract and citation database using the “Author Search” feature with the author’s surname and first and middle initials as search input strings. Author identification included evaluation of the author’s name (with variations), location, site of academic affiliation, affiliations with other authors, journal and manuscript titles, and known publications by the author. Once the author had been appropriately identified, an automated calculation of h-index was performed using the Scopus platform. Statistical Analysis The following a priori comparisons were performed: -
h-index versus surgical discipline
-
h-index versus academic rank, surgical discipline versus academic rank
-
Publicly available data versus program-verified data
-
Surgical discipline versus practice duration
-
h-index versus practice duration
In addition to descriptive statistics, 1- and 2-way analysis of variance and Welch’s t tests were used after the Shapiro-Wilks test results showed a normal distribution with a slight skew. Tukey’s multiple comparisons post hoc test was used after analysis of variance. The proportional odds model was used to generate the correct prediction in the sample of faculty rank using the h-index. All statistical analyses were calculated using GraphPad PRISM (GraphPad, La Jolla, California, USA) and RStudio (Boston, Massachusetts, USA). An a level P < 0.05 was considered significant, and mean values are presented with the standard error of the mean. RESULTS Verification of the data was possible for 65% of the neurosurgical programs and 41% of the orthopedic programs, accounting for 63.0% of the 416 neurological spine surgeons and 48.1% of the 507 orthopedic spine surgeons. The neurosurgical spine surgeons had an overall h-index of 12.93 (0.60) using publicly available data and 13.69 (0.79) using verified data. Orthopedic spine surgeons had an overall h-index of 10.87 0.59 using publicly available data and 12.43 0.98 using verified data (Table 1). Neurological spine surgeons had a statistically greater overall h-index compared with the orthopedic spine surgeons (P ¼ 0.0241). However, the differences in the results of the comparison between the publicly available and verified data were not statistically significant (P ¼ 0.1149; Figure 1A).
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 1. h-Index in spine surgical disciplines for both publicly available and verified data. (A) Bar graph showing mean standard error of the mean h-index values for neurological and orthopedic spine surgeons calculated using verified or publicly available data. (BeD) Bar graphs showing mean standard error of
Once the h-index had been calculated for all faculty members, a clear trend was seen in both the neurosurgical and orthopedic faculty, with increasing academic rank correlating with an increasing mean h-index. Using publicly available data, the mean h-index for the neurosurgical, orthopedic, and combined faculty was 2.75 1.11, 2.83 1.07, and 2.82 0.92 for instructors, 6.86 0.46, 5.62 0.43, and 6.21 0.32 for assistant professors, 12.82 0.92, 8.86 0.85, and 10.91 0.65 for associate professors, and 25.80 1.42, 26.63 1.58, and 26.25 1.08 for full professors (Table 2 and Figure 1BeD). Statistically significant differences were seen when comparing professors and associate professors, professors and assistant professors, and associate professors and assistant professors in the neurosurgical, orthopedic, and combined groups (P < 0.0001). A statistically significant difference was also seen when comparing professors and instructors in the neurosurgical and combined groups
the mean h-index values calculated for spine surgeons verified by individual departments and spine surgeons listed on department websites. Surgeons were separated by faculty rank and residency training. *P value of 0.0241.
(P < 0.0001) and associate professors and instructors in only the combined group (P ¼ 0.0003). A comparison of academic rank between surgical disciplines showed a statistically greater h-index for neurosurgical spine surgeons at the associate professor rank (mean h-index for NS vs. OS, 12.82 0.92 vs. 8.86 0.85; P ¼ 0.0019) and the assistant professor rank (mean h-index for NS vs. OS, 6.86 0.46 vs. 5.62 0.43; P ¼ 0.0498). However, no statistically significance differences were found at the instructor or professor ranks. Subanalysis of the data verified through direct contact with the academic programs showed a similar pattern for both h-index versus academic rank (Table 2 and Figure 1BeD) and the comparison of academic rank pairs within a surgical subspecialty, except that a comparison of associate professors versus instructors did not achieve significance in the combined group (P ¼ 0.1938). Significance was again seen when
Table 1. Overview of h-Index for Spine Surgical Disciplines Publicly Available h-Index
Verified h-Index
n
Minimum
Maximum
Mean SEM
n
Minimum
Maximum
Mean SEM
Neurosurgical
416
0
62
12.93 0.60
257
0
62
13.69 0.79
Orthopedic
507
0
85
10.87 0.59
244
0
85
12.43 0.98
Spine Residency Type
SEM, standard error of the mean.
WORLD NEUROSURGERY -: e1-e8, - 2018
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Table 2. Overview of h-Index and Faculty Rank for Spine Surgical Disciplines Neurological Surgery h-Index
Orthopedic Surgery h-Index
Combined h-Index
n
Min
Max
Mean SEM
n
Min
Max
Mean SEM
n
Min
Max
Mean SEM
Instructor
4
1
6
2.75 1.11
23
0
17
2.83 1.07
27
0
17
2.82 0.92
Faculty Rank Publicly available data
Assistant
162
0
29
6.86 0.46
181
0
26
5.62 0.43
343
0
29
6.21 0.32
Associate
89
0
33
12.82 0.92
83
0
31
8.86 0.85
172
0
33
10.91 0.65
Professor
96
0
62
25.80 1.42
114
0
85
26.63 1.58
210
0
85
26.25 1.08
Instructor
3
1
6
3.00 1.53
8
0
17
4.00 2.63
11
0
17
3.73 1.91
Verified data
Assistant
109
0
29
7.06 0.61
90
0
26
5.16 0.65
199
0
29
6.20 0.45
Associate
56
0
32
12.50 1.16
47
0
31
8.85 1.25
103
0
32
10.84 0.87
Professor
70
0
62
26.76 1.67
70
0
85
28.21 2.19
140
0
85
27.49 1.37
Min, minimum; Max, maximum; SEM, standard error of the mean.
comparing the neurosurgical and orthopedic surgery programs at the associate professor (P ¼ 0.0349) and assistant professor (P ¼ 0.0348) ranks. Once again, no significant difference was found in the neurosurgical, orthopedic, or combined data when comparing the publicly available data with that independently verified. This also indicated that the publicly available data were acceptable for analysis; thus, all further analyses were performed using the publicly available data. A comparison of the mean duration in practice between the neurological spine surgeons (mean duration, 15.94 0.65 years) and orthopedic spine surgeons (mean duration, 17.91 0.56) showed a statistically significant difference (P ¼ 0.0221; Table 3 and Figure 2A). When comparing the duration of practice with faculty rank (Table 3 and Figure 2B), statistically significant differences were found between the rank pairs of associate professor versus assistant professor (NS, P < 0.0001; OS, P ¼ 0.0047), professor versus assistant and professor versus associate for both groups (P < 0.0001), and professor versus instructor for the orthopedic group (P < 0.0001). For the duration of practice comparison between the neurological and orthopedic spine surgeons with the same faculty rank, the only comparison that was significant was at the assistant professor level. Orthopedic assistant professors had significantly more years in practice than neurosurgical assistant professors (Figure 2B; P ¼ 0.0014). A positive correlation was seen between the duration of in practice and the h-index in the neurosurgical, orthopedic, and combined groups (P < 0.0001; Figure 2CeE). Finally, the proportional odds models showed a 62.4%, 63.4%, and 62.1% success rate in predicting the faculty rank according to the h-index for neurosurgical, orthopedic and combined spine surgery groups, respectively (Figure 3AeC). DISCUSSION When Hirsch first proposed the h-index in 2005,1 he noted that “there will be differences in typical h values in different fields,
e4
www.SCIENCEDIRECT.com
determined in part by the average number of references in a paper in the field, the average number of papers produced by each scientist in the field, and the size (number of scientists) of Table 3. Overview of Duration of Practice for Spine Surgical Disciplines Practice Duration (years) Variable
Surgeons (n) Minimum Maximum Mean SEM
Spine surgical discipline Neurosurgical
297
1
53
15.94 0.65
Orthopedic
338
1
48
17.91 0.56
Instructor
4
6
28
17.25 5.02
Assistant
122
1
35
9.41 0.66
Neurosurgical practice (years) Faculty rank
Associate
66
1
52
17.38 1.28
Professor
75
1
53
25.25 1.16
Instructor
14
1
35
12.00 2.88
Assistant
122
1
48
12.89 0.85
Associate
60
2
41
17.60 1.09
Professor
81
6
46
25.86 0.86
Orthopedic practice (years) Faculty rank
SEM, standard error of the mean.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 2. h-Index correlation with duration of practice. (A) Bar graph showing mean standard error of the mean duration of practice for neurosurgical and orthopedic spine surgeons. (B) Bar graph showing mean standard error of the mean duration of practice for neurological and
the field . . Scientists working in nonmainstream areas will not achieve the same very high h values as the top echelon of those working in highly topical areas.” In 2006, Batista et al.16 found different mean h-indexes for top Brazilian researchers in the fields of physics (h ¼ 32e37), chemistry (h ¼ 19e29), biological/ biomedical sciences (h ¼ 13e24), and mathematics (h ¼ 9e14) and proposed hI to allow for comparison between researchers in different fields. In 2013, Svider et al.15 examined 2429 academic surgeons at 20 institutions in 8 surgical specialties and noted a range of mean h-indexes across surgical specialties from 7.1 for obstetrics/ gynecology to 12.7 for urology.15 They concluded that “the use of the h-index to evaluate research productivity is most appropriate when applied to faculty members within a discipline rather than to comparison of physicians from different fields.” This statement, however, was applied to the surgical specialties as broad entities and did not inspect the various subspecialties within each parent category. Khan et al.7 evaluated 1225 academic neurosurgeons in 99 U.S. departments and found different mean h-indexes across neurosurgical subspecialties. General, spine, and pediatric neurosurgeons had the lowest mean h-indexes (h ¼ 10e14) with vascular, functional/epilepsy, radiosurgery, peripheral nerve, and neuro-oncology/skull base neurosurgeons having the greatest mean h-indexes (h ¼ 15e19). In the orthopedic literature, various subgroups have been investigated, noting mean h-indexes of 15 (95% confidence interval, 13e17) for American Orthopedic Society for Sports Medicine fellowship faculty,10 12.8 13.8 standard
WORLD NEUROSURGERY -: e1-e8, - 2018
orthopedic spine surgeons with different faculty ranks. Scatter plots with linear regression showing h-index as a function of the duration of practice for (C) neurosurgical, (D) orthopedic, and (E) combined faculty groups. *P value of 0.0221. **P value of 0.0014.
deviation (range, 0e67) for adult total joint reconstruction surgeons, and 10.2 9.9 for full-time academic hand surgeons affiliated with fellowship programs.9,12 In suggesting that a comparison of academic productivity across different scientific disciplines is impossible or ill-advised or requires increasing manipulation of bibliometric indexes to account for the relative inhomogeneity of the groups, the inherent assumption is that the groups in question are different enough to warrant separation. Our study, however, showed relative similarity between 2 subspecialty groups from different parent specialties. Our overall mean h-indexes for neurosurgery (12.93; range, 0e62) and orthopedic surgery (10.93; range, 0e85) are similar to those found by Svider et al.15 for the overall specialties of neurosurgery (12.4) and orthopedic surgery (9.1), as well as the mean h-index of spinal neurosurgeons found by Khan et al.7,12 (range, 10e12). As the academic rank increased, statistically significant increases in h-index were seen in both neurological and orthopedic surgery groups, similar to that found in multiple other studies in neurological and orthopedic surgery.2,3,7,14,17 In both specialties, the duration of practice increased as faculty rank increased from assistant professor to associate professor to professor. A positive correlation was seen between the duration of practice and the academic rank for the assistant professor through professor levels, although not at the instructor level. Although instructors had the lowest academic productivity among the spine surgeons, the orthopedic spine surgery instructors had a duration of practice similar to that for orthopedic assistant professors, and the neurosurgery spine instructors had a practice duration that
www.WORLDNEUROSURGERY.org
e5
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 3. h-Index success rates in predicting faculty rank for spine surgical disciplines. Proportional odds models for estimating probability of a specific faculty
was similar to that for neurosurgery associate professors. This might denote institutional differences in the definition of the instructor faculty position or might suggest that faculty members entering academia at the instructor position do not seek formal academic advancement of rank and publish in lower quantities. Also, only 4 neurosurgical and 23 orthopedic spine surgeons were instructors, and this limited quantity could have prevented significant conclusion from being drawn. No significant differences in practice duration were found when comparing instructors to other faculty ranks, except between orthopedic instructors and professors. The ability to predict academic rank using the h-index was nearly identical for neurological and orthopedic spine surgery. In addition to the similarities, our study also found some significant differences between these 2 subspecialty groups. Neurological spine surgery showed a greater mean h-index compared with that of orthopedic spine surgery, although with a lower number of years in practice, overall. Additional analysis showed that only neurosurgeons with the rank of assistant or associate professor had a significantly greater mean h-index compared with orthopedic assistant and associate professors. Also, neurosurgeons only had fewer mean years in practice compared with orthopedic surgeons at the assistant professor level. Because differences in academic productivity and practice
e6
www.SCIENCEDIRECT.com
rank according to the h-index values. Graphs showing model success rate of the h-index for (A) neurosurgery, (B) orthopedic, and (C) combined.
duration are limited to the early career stages, this might indicate that neurosurgeons have a greater mean h-index simply owing to a longer residency (7 years vs. 5 years) and more time to perform and report research. This would allow neurosurgical spine surgeons to progress more quickly earlier in their careers from assistant to associate professor. Other possible explanations could be a greater drive for research within the neurosurgical residency or a greater drive for research early in a neurosurgeon’s academic career. One potential flaw in this reasoning is that more orthopedic surgeons than neurosurgeons in our study underwent 1 years of postresidency fellowship training. This is expected because neurosurgeons receive more spinal training during residency. Because the duration of practice was calculated using years since the completion of residency training and more orthopedic surgeons received fellowship training, some orthopedic surgeons will appear to have had a longer duration of practice, although they did not truly start practice until after their fellowship training. Finally, the similarities between neurological and orthopedic spine surgery indicate that comparing academic productivity across different specialties is possible for similar specialties. Future studies of other related fields could provide more interesting and useful information. However, the small differences in training styles we have reported and unknown factors led to
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
noticeable changes in academic output. We also understand that some differences in the composition of the neurosurgical and orthopedic spine surgery workforces were not studied and could also affect academic productivity. Therefore, future research on workforce factors should help create a more detailed and unified view of the spine surgery field. The present study was subject to several limitations. The initial identification of appropriate faculty members for inclusion was dependent on publicly available website data. We have presumed that most departments’ listed information was relatively stable and that changes in personnel and their academic rank were relatively rare. However, wide variability was found among the department and hospital websites regarding the level of specific faculty information available. Some websites did not list academic rank, year of residency completion, or fellowship training. The lack of full information made it difficult to separate those with spine “ability” or “interest” from those with spine “expertise” as defined by our selection criteria. The faculty members were identified through their online biographies, inclusion in the hospital or departmental spine group or spine section, and the pursuit of additional information from each department or the faculty member directly. Although verification was not obtained from each department, we believe that both the publicly available and the verified data are valid, and our analysis showed that remarkable concordance between the 2 (Figure 1). Every effort was made to ensure that the data used for the h-index calculations were correct. In particular, the Scopus database was used, because each individual has a unique author profile with respect to the identification of articles they have published. Scopus does not cite articles before 1996, potentially resulting in absent data unless a labor-intensive manual calculation of pre1996 citations were performed. This has been performed in other studies, which postulated that the exclusion of pre-1996 citations should become less impactful as older researchers (with pre-1996 citations) retire.7 Since 2014, Elsevier has been adding pre-1996 references to the Scopus library to enable accurate searches back to 1970. Occasionally, the Scopus database can have multiple profiles for 1 author. However, 1 of the profiles will include almost all publications and will be used for analysis, with other profiles typically limited to few and minor publications. We also understand that noncontinuous academic careers, such as time in private practice after residency training, will add some
2. Lee J, Kraus KL, Couldwell WT. Use of the h index in neurosurgery: clinical article. J Neurosurg. 2009; 111:387-392. 3. Spearman CM, Quigley MJ, Quigley MR, Wilberger JE. Survey of the h index for all of academic neurosurgery: another power-law phenomenon? J Neurosurg. 2010;113:929-933. 4. Campbell PG, Awe OO, Maltenfort MG, Moshfeghi DM, Leng T, Moshfeghi AA, et al. Medical school and residency influence on choice
CONCLUSIONS Our results suggest that the relative homogeneity of the spine surgery subspecialty might supersede the differences of the parent disciplines that made previous h-index academic comparisons of neurological and orthopedic surgery difficult. Neurological spine surgeons showed a slightly greater mean h-index than their orthopedic counterparts, with similar patterns of increasing h-index correlating to all academic ranks. The duration of practice also increased as faculty rank increased from assistant professor to professor. Publicly available and independently verified data were essentially identical for both specialties at all academic ranks. These results can be used for benchmark purposes and for each parent institution or residency program to assess the relative productivity of its faculty. Parent institutions can compare current academic output to national averages and more accurately judge the level of academic productivity of incoming faculty. This could help institutions improve their research output and impact through better selection of faculty. These data could also be of interest to those pursuing academic opportunities in spine surgery (including residency and fellowship training) by quantifying the average research output needed to obtain these positions. The results could affect their search for research opportunities and academic collaboration on bibliometric comparisons between other related fields or topics unrelated to bibliometrics.
of an academic career and academic productivity among neurosurgery faculty in the United States. J Neurosurg. 2011;115:380-386.
REFERENCES 1. Hirsch JE. An index to quantify an individual’s scientific research output. Proc Natl Assoc Sci U S A. 2005;102:16569-16572.
error to the duration of practice calculation. We also understand that forms of academic productivity other than the h-index, including patents and grants, exist and contribute to the overall productivity of institutions and individuals. These were not included owing to the high difficulty and low likelihood of obtaining complete information. Our study analyzed overall academic productivity, as measured by the h-index, which includes spine-related and nonespinerelated publications. Many physicians publish in different fields during their career, and we believe information on all publications is valuable and relevant for both physicians and parent institutions when pursuing career opportunities in specific fields. It is unknown whether different conclusions would be reached if only spine-related publications were included, which could be a topic of interest in future studies.
5. Ponce FA, Lozano AM. Academic impact and rankings of American and Canadian neurosurgical departments as assessed using the h index. J Neurosurg. 2010;113:447-457. 6. Kalra RR, Kestle JRW. An assessment of academic productivity in pediatric neurosurgery. J Neurosurg Pediatr. 2013;12:262-265. 7. Khan NR, Thompson CJ, Taylor DR, Venable GT, Wham RM, Michael LM II, et al. An analysis of publication productivity for 1225 academic neurosurgeons and 99 departments in the United States. J Neurosurg. 2014;120:746-755.
WORLD NEUROSURGERY -: e1-e8, - 2018
8. Vitzthum K, Mache S, Quarcoo D, Scutaru C, Groneberg DA, Schöffel N. Scoliosis: densityequalizing mapping and scientometric analysis. Scoliosis. 2009;4:15.
9. Lopez J, Susarla SM, Swanson EW, Calotta N, Lifchez SD. The association of the H-index and academic rank among full-time academic hand surgeons affiliated with fellowship programs. J Hand Surg Am. 2015;40:1434-1441.
10. Cvetanovich GL, Saltzman BM, Chalmers PN, Frank RM, Cole BJ, Bach BR. Research productivity of sports medicine fellowship faculty. Orthop J Sport Med. 2016;4, 2325967116679393.
www.WORLDNEUROSURGERY.org
e7
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
11. Milone MT, Bernstein J. On track to professorship? A bibliometric analysis of early scholarly output. Am J Orthop (Belle Mead NJ). 2016;45:E119-E123.
examination of 2429 academic surgeons. Laryngoscope. 2013;123:884-889.
12. Khan AZ, Kelley BV, Patel AD, McAllister DR, Leong NL. Academic productivity among fellowship associated adult total joint reconstruction surgeons. Arthroplasty Today. 2017;3:298-302. 13. Zhu E, Shemesh S, Iatridis J, Moucha C. The association between scholarly impact and National Institutes of Health funding in orthopaedic surgery. Bull Hosp Jt Dis (2013). 2017;75:257-263.
16. Batista PD, Campiteli MG, Kinouchi O, Martinez AS. Is it possible to compare researchers with different scientific interests? Scientometrics. 2006;68:179-189. 17. Ence AK, Cope SR, Holliday EB, Somerson JS. Publication productivity and experience: factors associated with academic rank among orthopaedic surgery faculty in the United States. J Bone Joint Surg Am. 2016;98:e41.
e8
www.SCIENCEDIRECT.com
Received 3 August 2018; accepted 19 September 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.09.150 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com
14. Bastian S, Ippolito JA, Lopez SA, Eloy JA, Beebe KS. The use of the h-index in academic orthopaedic surgery. J Bone Joint Surg Am. 2017;99:e14. 15. Svider PF, Pashkova AA, Choudhry Z, Agarwal N, Kovalerchik O, Baredes S, et al. Comparison of scholarly impact among surgical specialties: an
commercial or financial relationships that could be construed as a potential conflict of interest. Portions of this work were presented for oral presentation at the Spine Summit 2018 Meeting, American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Section on Disorders of the Spine and Peripheral Nerves, Orlando, Florida, March 16, 2018.
1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
Conflict of interest statement: The authors declare that the article content was composed in the absence of any
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
Academic Productivity of Spine Surgeons at United States Neurological Surgery and Orthopedic Surgery Training Programs Alexander F. Post, Adam Y. Li, Jennifer B. Dai, Akbar Y. Maniya, Syed Haider, Stanislaw Sobotka, Tanvir F. Choudhri
OBJECTIVE: Spinal surgery is taught and practiced within 2 different surgical disciplines, neurological surgery and orthopedic surgery. We have provided a unified analysis of academic productivity measured using the h-index attributable to spine-focused faculty at U.S. residency programs.
-
METHODS: A total of 278 Accreditation Council for Graduate Medical Education training programs were assessed to identify 923 full-time faculty members with a spinal surgery designation, as defined by spine fellowship training or case volume >75% in spine surgery. The faculty were assessed with respect to academic rank, duration of practice in years, and academic productivity (h-index).
neurosurgical and orthopedic surgery spine faculty, with some noticeable differences. These results can be used for benchmark purposes to assess the relative productivity of its faculty and could be of interest to those pursuing academic opportunities in spine surgery.
-
RESULTS: The comparison showed a significantly greater mean h-index for neurological spine surgeons. The mean h-index for both disciplines increased significantly as faculty rank increased. Within the academic ranks of assistant and associate professor, neurological spine surgeons had significantly greater mean h-indexes. Neurological spine surgeons had a significantly lower practice duration. At all ranks except for assistant professor, the mean practice duration was not significantly different statistically between the neurological spine and orthopedic spine surgeons. A positive correlation between the h-index and practice duration was found for both spine surgical disciplines. The proportional odds models for neurological and orthopedic spine surgeons were moderately successful at predicting faculty rank according to the h-index.
-
CONCLUSIONS: We present a unified view of academic productivity as measured by the h-index among
-
INTRODUCTION
T
he academic output of an individual or a department has previously been measured by publication number, publication type (abstract, journal article, chapter), quality of research (level of evidence, prestige, impact factor of the publication journal), and associated funding (grants awarded by the National Institutes of Health). The Hirsch h-index, proposed in 2005,1 is defined as the number of papers (h) with citation number h. This metric attempts to balance the quantity of an author’s publications with an assessment of each article’s quality and relevance to the reported data. It is the most widely cited and most widely used of the citation-based bibliometrics used for longitudinal self-evaluation, comparison among members of a department or between departments, and evaluation of an individual with respect to academic promotion or funding. The h-index was first used in the neurosurgical literature by Lee et al.2 in 2009, in which an evaluation of 30 neurosurgical programs confirmed a positive correlation between the h-index and academic rank. Larger, subsequent studies by Spearman et al.3 (1120 neurosurgeons in all electronic residency application serviceelisted training programs) and Campbell et al.4 (986 faculty members in 97 academic neurosurgery departments) also supported the correlation between the h-index and academic rank. Comparisons have also been made across neurosurgical departments,5 within a particular neurosurgical subspecialty,6 and across different neurosurgical subspecialties.7
Key words Academic neurosurgery - Academic orthopedic surgery - Bibliometrics - h Index - Spine surgery
Department of Neurosurgery, Icahn School of Medicine at Mount Sinai, New York, New York, USA
Abbreviations and Acronyms NS: Neurological surgery OS: Orthopedic surgery
Journal homepage: www.WORLDNEUROSURGERY.org
-
WORLD NEUROSURGERY -: e1-e8, - 2018
To whom correspondence should be addressed: Alexander F. Post, M.D. [E-mail: [email protected]] Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.09.150
Available online: www.sciencedirect.com 1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
www.WORLDNEUROSURGERY.org
e1
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
The h-index was first used in the orthopedic literature in 2009 by Vitzthum et al.,8 who performed a scientometrics analysis of scoliosis research. Subsequent orthopedic reports have confirmed the correlation between the h-index and academic rank among academic hand surgeons and sports medicine fellowship faculty9,10 and have assessed the early scholarly output (i.e., studies reported in the first 5 years after board certification) of existing orthopedic professors.11 Khan et al.12 examined fellowship-associated adult total joint reconstruction surgeons at 66 programs and noted a positive correlation between the h-index and number of fellows, faculty academic title, years in practice, and formal fellowship training. Bibliometrics have been used to show a positive correlation between scholarly impact and National Institutes of Health funding.13 Most recently, Bastian et al.14 examined 2061 academic orthopedic surgeons from 120 departments and, again, confirmed that academic rank increased as the mean h-, m-, and e-indexes increased. The h-index can be a relevant metric when making comparisons within a given academic field. However, caution has been recommended against using the h-index to compare individuals in different academic fields.15 Nevertheless, we have not found evidence to determine whether this holds true for the same academic field within different specialties. Spinal surgery is notable in that it is taught and practiced within 2 different surgical disciplines—neurological surgery (NS) and orthopedic surgery (OS). We examined the academic productivity of spinefocused faculty at NS and OS residency programs in the United States. METHODS Selection of Programs and Surgeons Public searches of the Accreditation Council for Graduate Medical Education website were performed to identify 110 NS and 168 OS residency training programs active through 2016 and 2017. Faculty members with spine expertise were identified by examining the department websites (faculty lists, subspecialty section lists, faculty biographies) and hospital websites (faculty lists, “spine section” or “spine team” lists). The faculty members included in the present study were surgeons with neurosurgery or orthopedic surgery listed as their primary specialty. Furthermore, the included faculty had undergone fellowship training in spine surgery or had a surgeon case volume of >75% in spine surgery. Adjunct faculty with a primary specialty other than NS or OS (e.g., radiology, otolaryngology, rehabilitation medicine) and nonsurgeons (e.g., neurologists, radiologists, physiatrists, nonsurgical Doctor of Philosophy) were excluded from the present study. The faculty members were assessed for academic rank (instructor, assistant professor, associate professor, professor), duration of practice (defined as years since graduation from residency), and academic productivity (h-index as determined by Scopus). Once the spine faculty had been identified, the departments were contacted separately by electronic mail or telephone, or both. The departments were asked to verify the accuracy of lists created using publicly available data and to augment with any data available within the department. The surgeons were contacted directly in select cases. Unresponsive
e2
www.SCIENCEDIRECT.com
departments were contacted 5 times before being marked as unverified. After determining that verified and publicly available data containing unverified programs were not significantly different statistically (Figure 1), all further analyses used publicly available data. Calculation of h-Index Hirsch’s h-index, for an individual, is defined as the number of reports (h) with citation number h. For each faculty member, an automated h-index was calculated from the Scopus abstract and citation database using the “Author Search” feature with the author’s surname and first and middle initials as search input strings. Author identification included evaluation of the author’s name (with variations), location, site of academic affiliation, affiliations with other authors, journal and manuscript titles, and known publications by the author. Once the author had been appropriately identified, an automated calculation of h-index was performed using the Scopus platform. Statistical Analysis The following a priori comparisons were performed: -
h-index versus surgical discipline
-
h-index versus academic rank, surgical discipline versus academic rank
-
Publicly available data versus program-verified data
-
Surgical discipline versus practice duration
-
h-index versus practice duration
In addition to descriptive statistics, 1- and 2-way analysis of variance and Welch’s t tests were used after the Shapiro-Wilks test results showed a normal distribution with a slight skew. Tukey’s multiple comparisons post hoc test was used after analysis of variance. The proportional odds model was used to generate the correct prediction in the sample of faculty rank using the h-index. All statistical analyses were calculated using GraphPad PRISM (GraphPad, La Jolla, California, USA) and RStudio (Boston, Massachusetts, USA). An a level P < 0.05 was considered significant, and mean values are presented with the standard error of the mean. RESULTS Verification of the data was possible for 65% of the neurosurgical programs and 41% of the orthopedic programs, accounting for 63.0% of the 416 neurological spine surgeons and 48.1% of the 507 orthopedic spine surgeons. The neurosurgical spine surgeons had an overall h-index of 12.93 (0.60) using publicly available data and 13.69 (0.79) using verified data. Orthopedic spine surgeons had an overall h-index of 10.87 0.59 using publicly available data and 12.43 0.98 using verified data (Table 1). Neurological spine surgeons had a statistically greater overall h-index compared with the orthopedic spine surgeons (P ¼ 0.0241). However, the differences in the results of the comparison between the publicly available and verified data were not statistically significant (P ¼ 0.1149; Figure 1A).
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 1. h-Index in spine surgical disciplines for both publicly available and verified data. (A) Bar graph showing mean standard error of the mean h-index values for neurological and orthopedic spine surgeons calculated using verified or publicly available data. (BeD) Bar graphs showing mean standard error of
Once the h-index had been calculated for all faculty members, a clear trend was seen in both the neurosurgical and orthopedic faculty, with increasing academic rank correlating with an increasing mean h-index. Using publicly available data, the mean h-index for the neurosurgical, orthopedic, and combined faculty was 2.75 1.11, 2.83 1.07, and 2.82 0.92 for instructors, 6.86 0.46, 5.62 0.43, and 6.21 0.32 for assistant professors, 12.82 0.92, 8.86 0.85, and 10.91 0.65 for associate professors, and 25.80 1.42, 26.63 1.58, and 26.25 1.08 for full professors (Table 2 and Figure 1BeD). Statistically significant differences were seen when comparing professors and associate professors, professors and assistant professors, and associate professors and assistant professors in the neurosurgical, orthopedic, and combined groups (P < 0.0001). A statistically significant difference was also seen when comparing professors and instructors in the neurosurgical and combined groups
the mean h-index values calculated for spine surgeons verified by individual departments and spine surgeons listed on department websites. Surgeons were separated by faculty rank and residency training. *P value of 0.0241.
(P < 0.0001) and associate professors and instructors in only the combined group (P ¼ 0.0003). A comparison of academic rank between surgical disciplines showed a statistically greater h-index for neurosurgical spine surgeons at the associate professor rank (mean h-index for NS vs. OS, 12.82 0.92 vs. 8.86 0.85; P ¼ 0.0019) and the assistant professor rank (mean h-index for NS vs. OS, 6.86 0.46 vs. 5.62 0.43; P ¼ 0.0498). However, no statistically significance differences were found at the instructor or professor ranks. Subanalysis of the data verified through direct contact with the academic programs showed a similar pattern for both h-index versus academic rank (Table 2 and Figure 1BeD) and the comparison of academic rank pairs within a surgical subspecialty, except that a comparison of associate professors versus instructors did not achieve significance in the combined group (P ¼ 0.1938). Significance was again seen when
Table 1. Overview of h-Index for Spine Surgical Disciplines Publicly Available h-Index
Verified h-Index
n
Minimum
Maximum
Mean SEM
n
Minimum
Maximum
Mean SEM
Neurosurgical
416
0
62
12.93 0.60
257
0
62
13.69 0.79
Orthopedic
507
0
85
10.87 0.59
244
0
85
12.43 0.98
Spine Residency Type
SEM, standard error of the mean.
WORLD NEUROSURGERY -: e1-e8, - 2018
www.WORLDNEUROSURGERY.org
e3
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Table 2. Overview of h-Index and Faculty Rank for Spine Surgical Disciplines Neurological Surgery h-Index
Orthopedic Surgery h-Index
Combined h-Index
n
Min
Max
Mean SEM
n
Min
Max
Mean SEM
n
Min
Max
Mean SEM
Instructor
4
1
6
2.75 1.11
23
0
17
2.83 1.07
27
0
17
2.82 0.92
Faculty Rank Publicly available data
Assistant
162
0
29
6.86 0.46
181
0
26
5.62 0.43
343
0
29
6.21 0.32
Associate
89
0
33
12.82 0.92
83
0
31
8.86 0.85
172
0
33
10.91 0.65
Professor
96
0
62
25.80 1.42
114
0
85
26.63 1.58
210
0
85
26.25 1.08
Instructor
3
1
6
3.00 1.53
8
0
17
4.00 2.63
11
0
17
3.73 1.91
Verified data
Assistant
109
0
29
7.06 0.61
90
0
26
5.16 0.65
199
0
29
6.20 0.45
Associate
56
0
32
12.50 1.16
47
0
31
8.85 1.25
103
0
32
10.84 0.87
Professor
70
0
62
26.76 1.67
70
0
85
28.21 2.19
140
0
85
27.49 1.37
Min, minimum; Max, maximum; SEM, standard error of the mean.
comparing the neurosurgical and orthopedic surgery programs at the associate professor (P ¼ 0.0349) and assistant professor (P ¼ 0.0348) ranks. Once again, no significant difference was found in the neurosurgical, orthopedic, or combined data when comparing the publicly available data with that independently verified. This also indicated that the publicly available data were acceptable for analysis; thus, all further analyses were performed using the publicly available data. A comparison of the mean duration in practice between the neurological spine surgeons (mean duration, 15.94 0.65 years) and orthopedic spine surgeons (mean duration, 17.91 0.56) showed a statistically significant difference (P ¼ 0.0221; Table 3 and Figure 2A). When comparing the duration of practice with faculty rank (Table 3 and Figure 2B), statistically significant differences were found between the rank pairs of associate professor versus assistant professor (NS, P < 0.0001; OS, P ¼ 0.0047), professor versus assistant and professor versus associate for both groups (P < 0.0001), and professor versus instructor for the orthopedic group (P < 0.0001). For the duration of practice comparison between the neurological and orthopedic spine surgeons with the same faculty rank, the only comparison that was significant was at the assistant professor level. Orthopedic assistant professors had significantly more years in practice than neurosurgical assistant professors (Figure 2B; P ¼ 0.0014). A positive correlation was seen between the duration of in practice and the h-index in the neurosurgical, orthopedic, and combined groups (P < 0.0001; Figure 2CeE). Finally, the proportional odds models showed a 62.4%, 63.4%, and 62.1% success rate in predicting the faculty rank according to the h-index for neurosurgical, orthopedic and combined spine surgery groups, respectively (Figure 3AeC). DISCUSSION When Hirsch first proposed the h-index in 2005,1 he noted that “there will be differences in typical h values in different fields,
e4
www.SCIENCEDIRECT.com
determined in part by the average number of references in a paper in the field, the average number of papers produced by each scientist in the field, and the size (number of scientists) of Table 3. Overview of Duration of Practice for Spine Surgical Disciplines Practice Duration (years) Variable
Surgeons (n) Minimum Maximum Mean SEM
Spine surgical discipline Neurosurgical
297
1
53
15.94 0.65
Orthopedic
338
1
48
17.91 0.56
Instructor
4
6
28
17.25 5.02
Assistant
122
1
35
9.41 0.66
Neurosurgical practice (years) Faculty rank
Associate
66
1
52
17.38 1.28
Professor
75
1
53
25.25 1.16
Instructor
14
1
35
12.00 2.88
Assistant
122
1
48
12.89 0.85
Associate
60
2
41
17.60 1.09
Professor
81
6
46
25.86 0.86
Orthopedic practice (years) Faculty rank
SEM, standard error of the mean.
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 2. h-Index correlation with duration of practice. (A) Bar graph showing mean standard error of the mean duration of practice for neurosurgical and orthopedic spine surgeons. (B) Bar graph showing mean standard error of the mean duration of practice for neurological and
the field . . Scientists working in nonmainstream areas will not achieve the same very high h values as the top echelon of those working in highly topical areas.” In 2006, Batista et al.16 found different mean h-indexes for top Brazilian researchers in the fields of physics (h ¼ 32e37), chemistry (h ¼ 19e29), biological/ biomedical sciences (h ¼ 13e24), and mathematics (h ¼ 9e14) and proposed hI to allow for comparison between researchers in different fields. In 2013, Svider et al.15 examined 2429 academic surgeons at 20 institutions in 8 surgical specialties and noted a range of mean h-indexes across surgical specialties from 7.1 for obstetrics/ gynecology to 12.7 for urology.15 They concluded that “the use of the h-index to evaluate research productivity is most appropriate when applied to faculty members within a discipline rather than to comparison of physicians from different fields.” This statement, however, was applied to the surgical specialties as broad entities and did not inspect the various subspecialties within each parent category. Khan et al.7 evaluated 1225 academic neurosurgeons in 99 U.S. departments and found different mean h-indexes across neurosurgical subspecialties. General, spine, and pediatric neurosurgeons had the lowest mean h-indexes (h ¼ 10e14) with vascular, functional/epilepsy, radiosurgery, peripheral nerve, and neuro-oncology/skull base neurosurgeons having the greatest mean h-indexes (h ¼ 15e19). In the orthopedic literature, various subgroups have been investigated, noting mean h-indexes of 15 (95% confidence interval, 13e17) for American Orthopedic Society for Sports Medicine fellowship faculty,10 12.8 13.8 standard
WORLD NEUROSURGERY -: e1-e8, - 2018
orthopedic spine surgeons with different faculty ranks. Scatter plots with linear regression showing h-index as a function of the duration of practice for (C) neurosurgical, (D) orthopedic, and (E) combined faculty groups. *P value of 0.0221. **P value of 0.0014.
deviation (range, 0e67) for adult total joint reconstruction surgeons, and 10.2 9.9 for full-time academic hand surgeons affiliated with fellowship programs.9,12 In suggesting that a comparison of academic productivity across different scientific disciplines is impossible or ill-advised or requires increasing manipulation of bibliometric indexes to account for the relative inhomogeneity of the groups, the inherent assumption is that the groups in question are different enough to warrant separation. Our study, however, showed relative similarity between 2 subspecialty groups from different parent specialties. Our overall mean h-indexes for neurosurgery (12.93; range, 0e62) and orthopedic surgery (10.93; range, 0e85) are similar to those found by Svider et al.15 for the overall specialties of neurosurgery (12.4) and orthopedic surgery (9.1), as well as the mean h-index of spinal neurosurgeons found by Khan et al.7,12 (range, 10e12). As the academic rank increased, statistically significant increases in h-index were seen in both neurological and orthopedic surgery groups, similar to that found in multiple other studies in neurological and orthopedic surgery.2,3,7,14,17 In both specialties, the duration of practice increased as faculty rank increased from assistant professor to associate professor to professor. A positive correlation was seen between the duration of practice and the academic rank for the assistant professor through professor levels, although not at the instructor level. Although instructors had the lowest academic productivity among the spine surgeons, the orthopedic spine surgery instructors had a duration of practice similar to that for orthopedic assistant professors, and the neurosurgery spine instructors had a practice duration that
www.WORLDNEUROSURGERY.org
e5
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
Figure 3. h-Index success rates in predicting faculty rank for spine surgical disciplines. Proportional odds models for estimating probability of a specific faculty
was similar to that for neurosurgery associate professors. This might denote institutional differences in the definition of the instructor faculty position or might suggest that faculty members entering academia at the instructor position do not seek formal academic advancement of rank and publish in lower quantities. Also, only 4 neurosurgical and 23 orthopedic spine surgeons were instructors, and this limited quantity could have prevented significant conclusion from being drawn. No significant differences in practice duration were found when comparing instructors to other faculty ranks, except between orthopedic instructors and professors. The ability to predict academic rank using the h-index was nearly identical for neurological and orthopedic spine surgery. In addition to the similarities, our study also found some significant differences between these 2 subspecialty groups. Neurological spine surgery showed a greater mean h-index compared with that of orthopedic spine surgery, although with a lower number of years in practice, overall. Additional analysis showed that only neurosurgeons with the rank of assistant or associate professor had a significantly greater mean h-index compared with orthopedic assistant and associate professors. Also, neurosurgeons only had fewer mean years in practice compared with orthopedic surgeons at the assistant professor level. Because differences in academic productivity and practice
e6
www.SCIENCEDIRECT.com
rank according to the h-index values. Graphs showing model success rate of the h-index for (A) neurosurgery, (B) orthopedic, and (C) combined.
duration are limited to the early career stages, this might indicate that neurosurgeons have a greater mean h-index simply owing to a longer residency (7 years vs. 5 years) and more time to perform and report research. This would allow neurosurgical spine surgeons to progress more quickly earlier in their careers from assistant to associate professor. Other possible explanations could be a greater drive for research within the neurosurgical residency or a greater drive for research early in a neurosurgeon’s academic career. One potential flaw in this reasoning is that more orthopedic surgeons than neurosurgeons in our study underwent 1 years of postresidency fellowship training. This is expected because neurosurgeons receive more spinal training during residency. Because the duration of practice was calculated using years since the completion of residency training and more orthopedic surgeons received fellowship training, some orthopedic surgeons will appear to have had a longer duration of practice, although they did not truly start practice until after their fellowship training. Finally, the similarities between neurological and orthopedic spine surgery indicate that comparing academic productivity across different specialties is possible for similar specialties. Future studies of other related fields could provide more interesting and useful information. However, the small differences in training styles we have reported and unknown factors led to
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
noticeable changes in academic output. We also understand that some differences in the composition of the neurosurgical and orthopedic spine surgery workforces were not studied and could also affect academic productivity. Therefore, future research on workforce factors should help create a more detailed and unified view of the spine surgery field. The present study was subject to several limitations. The initial identification of appropriate faculty members for inclusion was dependent on publicly available website data. We have presumed that most departments’ listed information was relatively stable and that changes in personnel and their academic rank were relatively rare. However, wide variability was found among the department and hospital websites regarding the level of specific faculty information available. Some websites did not list academic rank, year of residency completion, or fellowship training. The lack of full information made it difficult to separate those with spine “ability” or “interest” from those with spine “expertise” as defined by our selection criteria. The faculty members were identified through their online biographies, inclusion in the hospital or departmental spine group or spine section, and the pursuit of additional information from each department or the faculty member directly. Although verification was not obtained from each department, we believe that both the publicly available and the verified data are valid, and our analysis showed that remarkable concordance between the 2 (Figure 1). Every effort was made to ensure that the data used for the h-index calculations were correct. In particular, the Scopus database was used, because each individual has a unique author profile with respect to the identification of articles they have published. Scopus does not cite articles before 1996, potentially resulting in absent data unless a labor-intensive manual calculation of pre1996 citations were performed. This has been performed in other studies, which postulated that the exclusion of pre-1996 citations should become less impactful as older researchers (with pre-1996 citations) retire.7 Since 2014, Elsevier has been adding pre-1996 references to the Scopus library to enable accurate searches back to 1970. Occasionally, the Scopus database can have multiple profiles for 1 author. However, 1 of the profiles will include almost all publications and will be used for analysis, with other profiles typically limited to few and minor publications. We also understand that noncontinuous academic careers, such as time in private practice after residency training, will add some
2. Lee J, Kraus KL, Couldwell WT. Use of the h index in neurosurgery: clinical article. J Neurosurg. 2009; 111:387-392. 3. Spearman CM, Quigley MJ, Quigley MR, Wilberger JE. Survey of the h index for all of academic neurosurgery: another power-law phenomenon? J Neurosurg. 2010;113:929-933. 4. Campbell PG, Awe OO, Maltenfort MG, Moshfeghi DM, Leng T, Moshfeghi AA, et al. Medical school and residency influence on choice
CONCLUSIONS Our results suggest that the relative homogeneity of the spine surgery subspecialty might supersede the differences of the parent disciplines that made previous h-index academic comparisons of neurological and orthopedic surgery difficult. Neurological spine surgeons showed a slightly greater mean h-index than their orthopedic counterparts, with similar patterns of increasing h-index correlating to all academic ranks. The duration of practice also increased as faculty rank increased from assistant professor to professor. Publicly available and independently verified data were essentially identical for both specialties at all academic ranks. These results can be used for benchmark purposes and for each parent institution or residency program to assess the relative productivity of its faculty. Parent institutions can compare current academic output to national averages and more accurately judge the level of academic productivity of incoming faculty. This could help institutions improve their research output and impact through better selection of faculty. These data could also be of interest to those pursuing academic opportunities in spine surgery (including residency and fellowship training) by quantifying the average research output needed to obtain these positions. The results could affect their search for research opportunities and academic collaboration on bibliometric comparisons between other related fields or topics unrelated to bibliometrics.
of an academic career and academic productivity among neurosurgery faculty in the United States. J Neurosurg. 2011;115:380-386.
REFERENCES 1. Hirsch JE. An index to quantify an individual’s scientific research output. Proc Natl Assoc Sci U S A. 2005;102:16569-16572.
error to the duration of practice calculation. We also understand that forms of academic productivity other than the h-index, including patents and grants, exist and contribute to the overall productivity of institutions and individuals. These were not included owing to the high difficulty and low likelihood of obtaining complete information. Our study analyzed overall academic productivity, as measured by the h-index, which includes spine-related and nonespinerelated publications. Many physicians publish in different fields during their career, and we believe information on all publications is valuable and relevant for both physicians and parent institutions when pursuing career opportunities in specific fields. It is unknown whether different conclusions would be reached if only spine-related publications were included, which could be a topic of interest in future studies.
5. Ponce FA, Lozano AM. Academic impact and rankings of American and Canadian neurosurgical departments as assessed using the h index. J Neurosurg. 2010;113:447-457. 6. Kalra RR, Kestle JRW. An assessment of academic productivity in pediatric neurosurgery. J Neurosurg Pediatr. 2013;12:262-265. 7. Khan NR, Thompson CJ, Taylor DR, Venable GT, Wham RM, Michael LM II, et al. An analysis of publication productivity for 1225 academic neurosurgeons and 99 departments in the United States. J Neurosurg. 2014;120:746-755.
WORLD NEUROSURGERY -: e1-e8, - 2018
8. Vitzthum K, Mache S, Quarcoo D, Scutaru C, Groneberg DA, Schöffel N. Scoliosis: densityequalizing mapping and scientometric analysis. Scoliosis. 2009;4:15.
9. Lopez J, Susarla SM, Swanson EW, Calotta N, Lifchez SD. The association of the H-index and academic rank among full-time academic hand surgeons affiliated with fellowship programs. J Hand Surg Am. 2015;40:1434-1441.
10. Cvetanovich GL, Saltzman BM, Chalmers PN, Frank RM, Cole BJ, Bach BR. Research productivity of sports medicine fellowship faculty. Orthop J Sport Med. 2016;4, 2325967116679393.
www.WORLDNEUROSURGERY.org
e7
ORIGINAL ARTICLE ALEXANDER F. POST ET AL.
ACADEMIC PRODUCTIVITY OF U.S. NEUROLOGICAL/ORTHOPEDIC SPINE SURGEONS
11. Milone MT, Bernstein J. On track to professorship? A bibliometric analysis of early scholarly output. Am J Orthop (Belle Mead NJ). 2016;45:E119-E123.
examination of 2429 academic surgeons. Laryngoscope. 2013;123:884-889.
12. Khan AZ, Kelley BV, Patel AD, McAllister DR, Leong NL. Academic productivity among fellowship associated adult total joint reconstruction surgeons. Arthroplasty Today. 2017;3:298-302. 13. Zhu E, Shemesh S, Iatridis J, Moucha C. The association between scholarly impact and National Institutes of Health funding in orthopaedic surgery. Bull Hosp Jt Dis (2013). 2017;75:257-263.
16. Batista PD, Campiteli MG, Kinouchi O, Martinez AS. Is it possible to compare researchers with different scientific interests? Scientometrics. 2006;68:179-189. 17. Ence AK, Cope SR, Holliday EB, Somerson JS. Publication productivity and experience: factors associated with academic rank among orthopaedic surgery faculty in the United States. J Bone Joint Surg Am. 2016;98:e41.
e8
www.SCIENCEDIRECT.com
Received 3 August 2018; accepted 19 September 2018 Citation: World Neurosurg. (2018). https://doi.org/10.1016/j.wneu.2018.09.150 Journal homepage: www.WORLDNEUROSURGERY.org Available online: www.sciencedirect.com
14. Bastian S, Ippolito JA, Lopez SA, Eloy JA, Beebe KS. The use of the h-index in academic orthopaedic surgery. J Bone Joint Surg Am. 2017;99:e14. 15. Svider PF, Pashkova AA, Choudhry Z, Agarwal N, Kovalerchik O, Baredes S, et al. Comparison of scholarly impact among surgical specialties: an
commercial or financial relationships that could be construed as a potential conflict of interest. Portions of this work were presented for oral presentation at the Spine Summit 2018 Meeting, American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Section on Disorders of the Spine and Peripheral Nerves, Orlando, Florida, March 16, 2018.
1878-8750/$ - see front matter ª 2018 Elsevier Inc. All rights reserved.
Conflict of interest statement: The authors declare that the article content was composed in the absence of any
WORLD NEUROSURGERY, https://doi.org/10.1016/j.wneu.2018.09.150