- Email: [email protected]
The Role of Gender, Academic Affiliation, and Subspecialty in Relation to Industry Payments to Orthopaedic Surgeons
The Role of Gender, Academic Affiliation, and Subspecialty in Relation to Industry Payments to Orthopaedic Surgeons Rafael A. Buerba, M.D., M.H.S., Armin Arshi, M.D., Danielle C. Greenberg, B.A., Nelson F. SooHoo, M.D.
Manuscript prepared for the Journal of the National Medical Association. Declaration of interests: The authors have no relevant conflicts of interest to disclose relevant to the content of this manuscript. Source of funding: None. Abstract: Background: The Physician-Payments-Sunshine-Act (PPSA) was introduced in 2010 to provide transparency regarding physician-industry payments by making these payments publicly available. Given potential ethical implications, it is important to understand how these payments are being distributed, particularly as the women orthopaedic workforce increases. The purpose of this study was thus to determine the role of gender and academic affiliation in relation to industry payments within the orthopaedic subspecialties. Methods: The PPSA website was used to abstract industry payments to Orthopaedic surgeons. The internet was then queried to identify each surgeon’s professional listing and gender. Mann-Whitney U, Chi-square tests, and multivariable regression were used to explore the relationships. Significance was set at a value of P < 0.05. Results: In total, 22,352 orthopaedic surgeons were included in the study. Payments were compared between 21,053 men and 1299 women, 2756 academic and 19,596 community surgeons, and across orthopaedic subspecialties. Women surgeons received smaller research and non-research payments than men (both, P < 0.001). There was a larger percentage of women in academics than men (15.9% vs 12.1%, P < 0.001). Subspecialties with a higher percentage of women (Foot & Ankle, Hand, and Pediatrics) were also the subspecialties with the lowest mean industry payments (all P < 0.001). Academic surgeons on average, received larger research and non-research industry payments, than community surgeons (both, P < 0.001). Multivariable linear regression demonstrated that male gender (P ¼ 0.006, P ¼ 0.029), adult reconstruction (both, P < 0.001) and spine (P ¼ 0.008, P < 0.001) subspecialties, and academic rank (both, P < 0.001) were independent predictors of larger industry research and non-research payments. Conclusions: A large proportion of the US orthopaedic surgeon workforce received industry payments in 2014. Academic surgeons received larger payments than community surgeons. Despite having a larger percentage of surgeons in academia, women surgeons received lower payments than their male counterparts. Women also had a larger representation in the subspecialties with the lowest payments. Keywords: Industry payments-Conflict of interest-Gender in orthopaedicsPhysician payment sunshine act-Academic surgery-Orthopaedic workforce
Author affiliations: Rafael A. Buerba, Department of Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, PA, USA; Armin Arshi, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Danielle C. Greenberg, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Nelson F. SooHoo, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA Correspondence: Rafael A. Buerba, M.D./M.H.S., Department of Orthopaedic Surgery, UPMC Freddie Fu Sports Medicine Center, 3200 South Water Street Pittsburgh, PA 15203, USA., email: [email protected] ª 2019 Published by Elsevier Inc. on behalf of the National Medical Association.
https://doi.org/10.1016/j.jnma.2019.09.004
INTRODUCTION
M
edical device companies have fostered strong relationships with orthopaedic surgeons due to the prominent role of implant use in the practice
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
of orthopaedic surgery.1e5 These relationships have come under scrutiny in the past several years. The PhysiciansPayment-Sunshine-Act (PPSA) was passed in 2010 as part of the Affordable Care Act in order to increase transparency regarding physician-industry relationships.6 The PPSA requires that all payments from medical device and drug companies to physicians be reported to the Center for Medicare and Medicaid Services (CMS) so that they can then become publicly available and searchable online.7 Several studies have investigated the financial ties between implant companies and orthopaedic surgeons.1e5 A study by Iyer et al. concluded that orthopaedic surgeons receive the highest median industry payments compared to all other specialties, primarily as a result of a small fraction of orthopaedists receiving substantial royalties and licensing fees.1 However, there is limited data on the surgeon characteristics that are associated with large payments. Previous studies have briefly discussed trends of industry payments to orthopaedic surgeons with regard to regional variation and subspecialty2,3,8 and academic influence.9 One recent study described the types and distributions of industry payments amongst all physicians and found that orthopaedists and neurosurgeons received the highest payments per physician and that women received lower mean payments than their male counterparts within their respective specialties.10 Another study focusing on spine surgeons found that being in an academic practice setting and being male was associated with receiving industry payments.8 However, no study has examined how gender and academic affiliation may affect industry payments within the orthopaedic subspecialties. This is of particular importance with the increasing percentage of women in recent orthopaedic graduates and trainees. The purpose of this study is thus to assess and quantify the relationship between gender and academic affiliation with industry payments within the orthopaedic subspecialties.
MATERIALS AND METHODS The study is an observational, retrospective analysis of publicly available data. This study received an Institutional Review Board exemption given that the data is publicly available.
VOL
-,
NO
-, -
2019 1
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
The CMS open payments website was queried to obtain the industry payments to orthopaedic surgeons in the 2014 fiscal year.7 Payments to each physician were aggregated by using each physician’s unique identifier (PPSA ID). The open payments website classifies physician payments as general payments, research payments or associated research payments. General payments consist of in-kind payments, royalties, travel, speaking fees, gifts, consulting fees, meals, among others. These general payments were defined as “non-research payments.” Per the open payments website, research payments refer to payments meant for research use. Associated research payments refer to payments directly paid to an institution for research purposes under a physician principal investigator (PI). The “associated” and “individual” research payments were summed by each physician’s PPSA ID and a combined industry research variable was created. These payments were defined as “research payments.” We defined academic orthopaedic surgery departments as those with an accredited (MD) residency program in the 2013e2014 academic year.11,12 When the data was collected (6/2014-12/2014), 149 orthopaedic residency programs were identified. Exclusion criteria for residency programs were lack of online faculty listings or lack of medical school affiliation. Listings were thus obtained for 132/149 (92%) programs. To identify whether surgeons listed in the CMS website were involved in academia, we manually performed internet searches and matched the surgeons to their faculty listings. We matched surgeons by name, state, city, specialty, and zip code. When the specialty was not clearly identified, the surgeon was classified as being a general orthopaedic surgeon. Gender was verified using listed demographic data and photographs during these internet searches. Statistical analysis of relationships regarding industry payments were explored with the use of chi-square, Fisher’s exact test, and Mann-Whitney U testing. Multivariable linear regression was used to determine independent correlates of industry research and non-research payments with gender, academic rank, and subspecialty as covariates. As covariates, academic ranks were compared to non-academic community surgeon status and subspecialties were compared to general orthopedics. Data analysis was performed using SPSS version 24.0 (SPSS, Chicago, IL). All tests were performed with a statistical significance set at a probability value of 0.05.
RESULTS Study population A total of 22,352 orthopaedic surgeons were included in our study (Table 1). It is important to recognize that this
2
VOL.
-,
NO
-, -
2019
number does not include the entire orthopaedic surgeon workforce in the Unites States. Rather, it reflects the number of orthopaedic surgeons who received any industry payment in 2014 that was reported to CMS. Of these surgeons, only 5.8% were identified as being women. Academic surgeons made up 12.3% of the study population. Within academia, women made up 7.5% of the academic surgeon cohort. In the community surgeon cohort, women made up 5.6% of the cohort (P < 0.001). Regarding industry payments, 99.8% of surgeons received non-research payments whereas only 4.4% received research payments.
Overall payments Industry payments to orthopaedic surgeons are summarized in Table 2. Non-research payments totaled to $402,449,706. The median non-research payment amount was $381 whereas the mean non-research payment was $18,005. The mean was highly skewed due the large standard-deviation ($197,718) as a small number of surgeons were paid significant sums above the median. Nonresearch payments were further categorized into different payment levels as shown in Table 3. The vast majority of surgeons (65%) received payments $1000 whereas only 4.5% of surgeons received payments totaling $50,000. Research payments totaled $28,640,411 (Table 2). Only 4.4% of the study population received a research payment (Table 1), thus the median and mode were $0 as shown in Table 2. The mean was highly skewed ($1281) due to large research payments to a small number of individuals.
Analyses of non-research industry payments in relation to gender There were significant gender differences regarding nonresearch payments to orthopaedic surgeons. Of the 22,307 surgeons who received a non-research payment, 5.8% were women (P < 0.001) (Tables 2e3). Men received a higher mean non-research payment amount ($18,987 vs. $2092) than women as seen in Table 2 (P < 0.001). These larger payments to men are further delineated when looking at payment levels. As shown in Table 3, it can be seen that 75.1% of women (vs. 64.4% men) received payments in the $1000 category. This is the lowest payment category and the only category that consisted of a higher percentage of women than men. The remaining higher payment level categories all have a smaller percentage of women than men (all P < 0.001). The gender difference in payments is particularly highlighted when looking at the highest payment level category (>$50,000). Only 0.4% of women surgeons received payments in this level in comparison to 4.8% of men.
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Table 1. Characteristics of orthopaedic surgeons who received industry payments in 2014.
No. (%) of Orthopaedic Surgeons (n [ 22,352) Gender Total
Male (n [ 21,053)
Practice Setting Female (n [ 1299)
p-value
Academic (n [ 2756)
Community (n [ 19,596)
p-value < 0.001
Gender Male
21,053 (94.2%)
2550 (92.5%)
18,503 (94.4%)
Female
1299 (5.8%)
206 (7.5%)
1093 (5.6%)
< 0.001
Practice Setting 2756 (12.3%)
2550 (12.1%)
206 (15.9%)
Community- based
19,596 (87.7%)
18,503 (87.9%)
1093 (84.1%)
Non-Research Payment
22,307 (99.8%)
21,009 (99.8%)
1298 (99.9%)
0.520
2738 (99.3%)
19,569 (99.9%)
<0.001
Research Payment
989 (4.4%)
976 (4.6%)
13 (1.0%)
<0.001
322 (11.7%)
667 (3.4%)
<0.001
<0.001
Orthopaedic Subspecialty
<0.001
General
15,308 (68.5%)
14,467 (68.7%)
841 (64.7%)
1537 (55.8%)
13,771 (70.3%)
Adult Reconstructive
708 (3.2%)
696 (3.3%)
12 (0.9%)
141 (5.1%)
567 (2.9%)
Foot and Ankle
575 (2.6%)
503 (2.4%)
72 (5.5%)
101 (3.7%)
474 (2.4%)
Hand
1359 (6.1%)
1214 (5.8%)
145 (11.2%)
229 (8.3%)
1130 (5.8%)
Spine
1513 (6.8%)
1488 (7.1%)
25 (1.9%)
221 (8.0%)
1292 (6.6%)
Trauma
495 (2.2%)
466 (2.2%)
29 (2.2%)
146 (5.3%)
349 (1.8%)
Pediatrics
327 (1.5%)
254 (1.2%)
73 (5.6%)
109 (4.0%)
218 (1.1%)
Sports
2067 (9.2%)
1965 (9.3%)
102 (7.9%)
272 (9.9%)
1795 (9.2%)
VOL
Bolded values represent significance of P < 0.05.
-,
NO -, -
2019 3
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Academic-based
VOL. -,
NO -, -
2019
n Non-Research Payments
Mean a
22,307
$18,005
Median
Mode
Std. Dev
Minimum
$381
$14
$197,718
$0
Maximum
Sum
$16,638,590
$402,449,076 <0.001b
Gender Male
21,009
$18,987
$394
$14
$203,641
$0
$16,638,590
$399,731,410
Female
1298
$2092
$227
$121
$17,291
$0
$568,014
$2,717,667 <0.001b
Practice Setting Academic
2738
$45,963
$1087
$0
$218,635
$0
$4,538,501
$126,673,289
Community
19,569
$14,073
$348
$14
$194,280
$0
$16,638,590
$275,775,787
$1281
$0
$0
$13,991
$0
$807,354
$28,640,411
Research Payments
p-value
989
a
<0.001b
Gender Male
976
$1353
$0
$0
$14,406
$0
$807,354
$28,484,719
Female
13
$120
$0
$0
$1830
$0
$48,002
$155,693 <0.001b
Practice Setting JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Academic
322
$3779
$0
$0
$24,581
$0
$517,007
$10,414,295
Community
667
$930
$0
$0
$11,719
$0
$807,354
$18,226,116
Bolded values represent significance of P < 0.05. a This n was not used for the statistical analysis. The number with the * represents the number of surgeons who received such payment type for clarity’s sake. The entire study population of 22,352 surgeons was used in the calculation of the mean, median, mode, SD, min, max and sum to include those who did not receive a certain payment type and provide an analysis more representative of the entire study population. b This p-value refers to the comparisons of means for the respective category (i.e. gender, practice setting).
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
4 Table 2. Summary of industry payments to orthopaedic surgeons.
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Table 3. Payment level distribution of industry non-research payments.
Gender Total n [ 22,307 (100%)
Male n [ 21,009 (94.2%)
Female n [ 1298 (5.8%)
<0.001
Practice Setting Academic n [ 2738 (12.3%)
p-value
Community n [ 19,569 (87.7%)
<0.001
<0.001
Payment Level $1000
p-value
<0.001
65.0%
64.4%
75.1%
48.7%
67.3%
$1001 - $10,000
24.2%
24.4%
21.5%
26.4%
23.9%
$10,001 - $25,000
4.0%
4.1%
2.4%
7.1%
3.6%
$25,001 - $50,000
2.2%
2.3%
0.6%
5.4%
1.8%
$50,000
4.5%
4.8%
0.4%
12.3%
3.4%
Bolded values represent significance of P < 0.05.
Analyses of research industry payments in relation to gender
payments in this category in comparison to 12.3% of academic surgeons.
Of the 989 surgeons who received a research payment, only 13 (1.3%) were women (Table 2, P < 0.001). Men received a higher mean research payment amount ($1353 vs. $120; Table 2) than womens (P < 0.001). The analysis of research payment level by gender was not performed given the low number (n ¼ 13) of womens who received a research payment.
Analyses of research industry payments in relation to practice setting
Analyses of non-research industry payments in relation to practice setting There were significant differences regarding non-research industry payments to orthopaedists based on practice setting. Of the 22,307 surgeons who received an industry non-research payment, only 2738 were academic surgeons (P < 0.001) (Tables 2 and 3). Despite consisting of only 12.3% of the study population, academic surgeons received a much higher mean non-research payment amount ($45,963 vs. $14,073) than community surgeons as seen in Table 2 (P < 0.001). When looking at these payments by payment level (Table 3), it can be seen that 67.3% of community surgeons (vs. 48.7% academic surgeons) received payments in the $1000 category. This is the lowest payment category and the only category that consisted of a higher percentage of community than academic surgeons. The remaining higher payment level categories all have a larger percentage of academic than community surgeons (all P < 0.001). This difference in payments based on practice setting is particularly highlighted when looking at the highest payment level category (>$50,000). Only 3.4% of community surgeons received
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Of the 989 surgeons who received research payments, only 322 (32.6%) were academic vs. 667 (67.4%) community surgeons (Table 2, P < 0.001). Despite a larger proportion of community surgeons receiving research payments, academic surgeons still received more research money on average ($3779 academic vs. $930 community; P < 0.001). The analysis of research payment level by practice setting was not performed given the low number of surgeons who received research payments.
Analyses of industry payments in relation to specialty Table 1 summarizes the distribution of surgeons who received payments amongst the different subspecialties. The vast majority (68.5%) of surgeons were classified as being generalists. Pediatrics was the specialty least represented in the study population with only 327 (1.5%) surgeons having received payments in 2014. There were significant differences in the gender distribution of surgeons amongst the subspecialties. As shown in Table 1, foot and ankle, hand, and pediatrics had a significantly larger proportion of women than men (P < 0.001). On the other hand, general orthopaedics, adult reconstructive, spine, and sports medicine had a larger proportion of men than women (P < 0.001). There were also differences in the subspecialty distribution of academic vs. community surgeons. There were more generalists in the community (70.3%) than in academia (55.8%). However, there were
VOL
-,
NO
-, -
2019 5
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Figure 1. Mean research and non-research payments by subspecialty amongst academic and community surgeons.
more adult reconstructive, hand, trauma, and pediatrics specialists in academia than the community (P < 0.001). Industry payments to orthopaedic surgeons by specialty breakdown are shown in Figure 1. Adult reconstruction and spine had the highest mean industry non-research payments in both the community and academic settings. These specialties were also the specialties with the lowest proportion of women surgeons (Table 1). Foot and ankle, hand, and pediatrics had the lowest non-research payments (Figure 1). These specialties were also the specialties with the highest proportion of women surgeons as previously shown in Table 1. Similar trends were found regarding research payments. As shown in Figure 1, adult reconstruction and spine also received the highest research payments in both academic and community settings. Although academic surgeons received on average higher research payments (Table 2), there were no differences in payments between academic and community surgeons that could be delineated when breaking down research payments by subspecialties (Figure 1).
Multivariable regression Multivariable linear regression was used to ascertain the independent predictive relationship of gender, academic affiliation and rank, and orthopaedic subspecialty to
6
VOL.
-,
NO
-, -
2019
research and non-research industry payments. Male gender (P ¼ 0.006), adult reconstruction (P < 0.001) and spine (P ¼ 0.008) subspecialties, and associate professor (P ¼ 0.001), professor (P < 0.001) and chair (P < 0.001) academic rank were independently associated with larger research industry payments (Table 4). Similarly, male gender (P ¼ 0.029), adult reconstruction (P < 0.001), and spine (P < 0.001) subspecialties, and chair academic rank (P < 0.001) were independently associated with larger non-research industry payments (Table 5).
DISCUSSION Our study showed that there are differences in industry payment size to orthopaedic surgeons based on gender, practice type and subspecialty. In our study, only 1% of women surgeons received payments >$25,000 compared to 11.2% of men. This gender discrepancy can also be seen in other areas of medicine. Tringale et al. showed that male physicians were more likely to receive an industry payment and to receive larger payments than womens in nearly all specialties.10 There are several possible explanations for the difference in payments based on gender. As shown in Table 1 and Figure 1, there is a larger proportion of women than men in specialties that receive lower industry payments (foot & ankle, hand, and pediatrics).
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Table 4. Multivariable linear regression for predictors of industry research payments.
Table 5. Multivariable linear regression for predictors of industry nonresearch payments.
Covariate
Standardized Coefficient (B)
p-value
Covariate
Standardized Coefficient (B)
p-value
Female
0.019
0.006
Female
0.015
0.029
Academic rank (relative to community surgeon)
Academic rank (relative to community surgeon)
Assistant Professor
0.011
0.093
Assistant Professor
0.004
0.525
Associate Professor
0.023
0.001
Associate Professor
0.001
0.918
Professor
0.039
<0.001
Professor
0.012
0.084
Chair
0.060
<0.001
Chair
0.93
<0.001
Subspecialty (relative to general orthopedics)
Subspecialty (relative to general orthopedics)
Adult Reconstruction
0.039
<0.001
Adult Reconstruction
0.032
<0.001
Foot & Ankle
0.008
0.235
Foot & Ankle
0.002
0.817
Hand
0.013
0.057
Hand
0.014
0.035
Spine
0.018
0.008
Spine
0.046
<0.001
Trauma
0.004
0.552
Trauma
0.004
0.551
Pediatric
0.001
0.879
Pediatric
0.007
0.323
Sports Medicine
0.001
0.591
Sports Medicine
0.003
0.708
Bolded values represent significance of P < 0.05.
Despite these confounding differences, womens still receive smaller industry payments compared to men when adjusting for academic rank and subspecialty using multivariable regression. Overall, orthopaedic surgery has lagged behind other surgical subspecialties in the recruitment of women13 and it is possible that the spine and reconstruction subspecialties have been slower to adapt a larger women workforce. Regarding practice setting, our study found that academic surgeons overall receive larger industry payments that community surgeons. Historically, academic physicians have conducted the majority of pharmaceutical and biodevice research from both federal and industry funding due to their trial-design expertise, accessibility to large patient populations, university resources, and publication experience.14 Since the turn of the millennium, industryfunded pharmaceutical research has transitioned to the private sector for financial reasons.14 The trend of industry funding with regard to medical devices and implants has been less apparent. Using a limited amount of data released from a Department of Justice lawsuit in 2007 regarding biodevice funding in orthopedics, Hockenberry et al. noted an increase in the proportion of industry payments to orthopaedic surgeons with academic affiliations between 2007-2010.5 Industry support of any type has been shown to be associated with higher scholarly
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Bolded values represent significance of P < 0.05.
impact9,15,16 and thus may promote advancement in academia. Although academic surgeons received larger payments in our study, it is important to note that there was a larger proportion of women academic surgeons (7.5%) than male academic surgeons (5.6%) and that despite this, women still received lower payments than men. Previous studies regarding gender in academia have shown that women experience less research productivity earlier in their careers, receive fewer funds for research, and are slower to advance professionally in several medical and surgical subspecialties.17e23 A recent study examining industry payments within academic otolaryngology found that women receive significantly less industry funding, even when adjusted for academic rank.17 Another possible explanation for the gender discrepancy in payment could be due to women holding less patents than men overall.24 Many of the largest payments documented in the PPSA database were due to royalties. This study has several limitations that merit discussion. Although we were meticulous in our matching of surgeons by name, state, city, specialty, and zip code, it is feasible that there may be some inaccuraciesdparticularly with surgeons with the same name. There are also variables we could not control for in our study such as migration,
VOL
-,
NO
-, -
2019 7
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
retirement, name change, and death. Another source of inaccuracy would be incomplete or incorrect data from the PPSA website. Although physicians can review and dispute any incorrect data, it is unknown how often physicians are reviewing the posted payments.7 This is of relevance as previous studies have highlighted that there are many discrepancies between what surgeons disclose at academic meetings versus that the medical industry reports.25e27 Despite these limitations, the PPSA website is the most comprehensive data for industry payments at this time.7 Another limitation of our study is that it may not be representative of the entire orthopaedic workforce as it only includes orthopaedic surgeons who received some sort of industry payment in 2014.7 According to the most recent American Academy of Orthopaedic Surgeon (AAOS) biennial census, there were 29,613 practicing orthopaedic surgeons in the US in 2016.28 There are thus > 7000 practicing orthopaedic surgeons who were potentially not included in our study and indicate a selection bias in the noted findings. The subspecialty distribution in our analysis was also slightly different than the one reported on the 2016 AAOS census. Our study reported a high proportion of generalists (68.5%) vs. only 42% per the 2016 census. This discrepancy is likely due to several surgeons being classified as generalists in the PPSA database despite being specialists. Despite this, even if the correct subspecialty had been identified, it is unlikely that the overall trends in payments by subspecialty would have significantly changed as the mean non-research payment amount to generalists was still significantly smaller than the mean payment to reconstruction and spine surgeons (Figure 1). Even though there are limitations to the subspecialty analysis, it is important to note that the distribution of gender and practice setting in our study was similar to the one reported in the 2016 census. The AAOS 2016 census reports that 6.5% of practicing orthopaedic surgeons are women (vs. 5.8% in our study) and that 15% are academic surgeons (vs. 12.3% in our study), which suggests that our gender and practice setting analyses are likely a good representation of the overall US Orthopaedic workforce. Further limitations of our study are that there may be other factors that could influence industry payments such as academic rank, academic influence, years in practice, clinical volume, institution, and ethnicity, among others. A recent study by our group showed that academic influence/ impact correlated poorly with industry payment amount to orthopaedic surgeons.9 On the other hand, studies have shown a positive association between academic rank and industry payments.15,16 Given this previous research and our results, multivariable regression analyses are needed to
8
VOL.
-,
NO
-, -
2019
identify the factors that are associated with large industry payments.
IMPLICATIONS According to the PPSA data, there are differences in the industry payments based on gender, academic rank, and practice type among orthopaedic surgeons. In general, payments to women and to community surgeons are lower than payments to men and academic surgeons, respectively. Women surgeons also had a larger representation in the subspecialties that received the lowest industry payments.
APPENDIX A. SUPPLEMENTARY DATA Supplementary data to this article can be found online at https://doi.org/10.1016/j.jnma.2019.09.004.
REFERENCES 1. Iyer, S., Derman, P., & Sandhu, H. S. (2016). Orthopaedics and the physician payments sunshine Act. J Bone Jt Surg, 98(5), e18. 2. Cvetanovich, G. L., Chalmers, P. N., & Bach, B. R. (2015). Industry financial relationships in orthopaedic surgery: analysis of the sunshine Act open payments database and comparison with other surgical subspecialties. J Bone Joint Surg Am, 97(15), 1288e1295. 3. Lopez, J., Ahmed, R., Bae, S., et al. (2016). A new culture of transparency: industry payments to orthopedic surgeons. Orthopedics, 39(6), e1058ee1062. 4. Samuel, A. M., Webb, M. L., Lukasiewicz, A. M., et al. (2015). Orthopaedic surgeons receive the most industry payments to physicians but large disparities are seen in sunshine Act data. Clin Orthop Relat Res, 473(10), 3297e3306. 5. Hockenberry, J. M., Weigel, P., Auerbach, A., & Cram, P. (2011). Financial payments by orthopedic device makers to orthopedic surgeons. Arch Intern Med, 171(19), 1759. 6. Agrawal, S., Brennan, N., & Budetti, P. (2013). The sunshine act– effects on physicians. N Engl J Med, 368(22), 2054e2057. 7. CMS. Open Payments Centers for Medicare and Medicaid Services. 8. Weiner, J. A., Cook, R. W., Hashmi, S., et al. (2017). Factors associated with financial relationships between spine surgeons and industry: an analysis of the open payments database. Spine (Phila Pa 1976), 42(18), 1412e1418. 9. Buerba, R. A., Sheppard, W. L., Herndon, K. E., et al. (2018). Academic influence and its relationship to industry payments in orthopaedic surgery. J Bone Jt Surg, 100(9), e59. 10. Tringale, K. R., Marshall, D., Mackey, T. K., Connor, M., Murphy, J. D., & Hattangadi-Gluth, J. A. (2017). Types and distribution of
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS payments from industry to physicians in 2015. J Am Med Assoc, 317(17), 1774.
20. Eloy, J. A., Svider, P., Chandrasekhar, S. S., et al. (2013). Gender disparities in scholarly productivity within academic otolaryngology departments. Otolaryngol Neck Surg, 148(2), 215e222.
11. FREIDA Residency & Fellowship Database.
21. Zhuge, Y., Kaufman, J., Simeone, D. M., Chen, H., & Velazquez,
12. NIH Research Portfolio Online Reporting Tools (RePORT). 13. Pico, K., Gioe, T. J., Vanheest, A., & Tatman, P. J. (2010). Do men
O. C. (2011). Is there still a glass ceiling for women in academic surgery? Ann Surg, 253(4), 637e643.
outperform women during orthopaedic residency training?
22. Reed, D. A., Enders, F., Lindor, R., McClees, M., & Lindor, K. D.
Clin Orthop Relat Res, 468(7), 1804e1808.
(2011). Gender differences in academic productivity and
14. Fisher, J. A., Kalbaugh, C. A., & Ross, J. S. (2012). United States private-sector research:
a
physicians qualitative
and
pharmaceutical
study.
PLoS
Med,
contract
9(7).
leadership appointments of physicians throughout academic careers. Acad Med, 86(1), 43e47.
www.
23. Paik, A. M., Mady, L. J., Villanueva, N. L., et al. (2014). Research
15. Eloy, J. A., Kilic, S., Yoo, N. G., et al. (2017). Is industry funding
productivity and gender disparities: a look at academic plastic surgery. J Surg Educ, 71(4), 593e600.
plosmedicine.org1.
associated with greater scholarly impact among academic neurosurgeons? World Neurosurg, 103, 517e525. 16. Svider, P. F., Bobian, M., Lin, H.-S., et al. (2017). Are industry financial ties associated with greater scholarly impact among academic
otolaryngologists?
The
Laryngoscope,
127(1),
87e94. 17. Eloy, J. A., Bobian, M., Svider, P. F., et al. (2017). Association of
24. Ding, W. W., Murray, F., & Stuart, T. E. (2006). Gender differences in patenting in the academic life sciences. Science, 313(5787), 665e667. 25. Buerba, R. A., Fu, M. C., & Grauer, J. N. (2013). Discrepancies in spine surgeon conflict of interest disclosures between a national meeting and physician payment listings on device manufacturer web sites. Spine J, 13(12).
gender with financial relationships between industry and aca-
26. Hannon, C. P., Chalmers, P. N., Carpiniello, M. F., Cvetanovich,
demic otolaryngologists. JAMA Otolaryngol Neck Surg, 150(1),
G. L., Cole, B. J., & Bach, B. R. (2016). Inconsistencies between
53e60.
physician-reported disclosures at the AAOS annual meeting
18. Eloy, J. A., Svider, P. F., Cherla, D. V., et al. (2013). Gender disparities in research productivity among 9952 academic physicians. The Laryngoscope, 123(8), 1865e1875. 19. Lopez, S. A., Svider, P. F., Misra, P., Bhagat, N., Langer, P. D., & Eloy, J. A. (2014). Gender differences in promotion and scholarly impact: an analysis of 1460 academic ophthalmologists. J Surg Educ, 71(6), 851e859.
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
and industry-reported financial disclosures in the open payments database. J Bone Jt Surg, 98(20), e90. 27. Okike, K., Kocher, M. S., Wei, E. X., Mehlman, C. T., & Bhandari, M. (2009). Accuracy of conflict-of-interest disclosures reported by physicians. N Engl J Med, 361(15), 1466e1474. 28. American Academy of Orthoapaedic Surgeons. AAOS 2016 Census.
VOL
-,
NO
-, -
2019 9
Manuscript prepared for the Journal of the National Medical Association. Declaration of interests: The authors have no relevant conflicts of interest to disclose relevant to the content of this manuscript. Source of funding: None. Abstract: Background: The Physician-Payments-Sunshine-Act (PPSA) was introduced in 2010 to provide transparency regarding physician-industry payments by making these payments publicly available. Given potential ethical implications, it is important to understand how these payments are being distributed, particularly as the women orthopaedic workforce increases. The purpose of this study was thus to determine the role of gender and academic affiliation in relation to industry payments within the orthopaedic subspecialties. Methods: The PPSA website was used to abstract industry payments to Orthopaedic surgeons. The internet was then queried to identify each surgeon’s professional listing and gender. Mann-Whitney U, Chi-square tests, and multivariable regression were used to explore the relationships. Significance was set at a value of P < 0.05. Results: In total, 22,352 orthopaedic surgeons were included in the study. Payments were compared between 21,053 men and 1299 women, 2756 academic and 19,596 community surgeons, and across orthopaedic subspecialties. Women surgeons received smaller research and non-research payments than men (both, P < 0.001). There was a larger percentage of women in academics than men (15.9% vs 12.1%, P < 0.001). Subspecialties with a higher percentage of women (Foot & Ankle, Hand, and Pediatrics) were also the subspecialties with the lowest mean industry payments (all P < 0.001). Academic surgeons on average, received larger research and non-research industry payments, than community surgeons (both, P < 0.001). Multivariable linear regression demonstrated that male gender (P ¼ 0.006, P ¼ 0.029), adult reconstruction (both, P < 0.001) and spine (P ¼ 0.008, P < 0.001) subspecialties, and academic rank (both, P < 0.001) were independent predictors of larger industry research and non-research payments. Conclusions: A large proportion of the US orthopaedic surgeon workforce received industry payments in 2014. Academic surgeons received larger payments than community surgeons. Despite having a larger percentage of surgeons in academia, women surgeons received lower payments than their male counterparts. Women also had a larger representation in the subspecialties with the lowest payments. Keywords: Industry payments-Conflict of interest-Gender in orthopaedicsPhysician payment sunshine act-Academic surgery-Orthopaedic workforce
Author affiliations: Rafael A. Buerba, Department of Orthopaedic Surgery, University of Pittsburgh, Pittsburgh, PA, USA; Armin Arshi, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Danielle C. Greenberg, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA; Nelson F. SooHoo, Department of Orthopaedic Surgery, David Geffen School of Medicine at UCLA, Los Angeles, CA, USA Correspondence: Rafael A. Buerba, M.D./M.H.S., Department of Orthopaedic Surgery, UPMC Freddie Fu Sports Medicine Center, 3200 South Water Street Pittsburgh, PA 15203, USA., email: [email protected] ª 2019 Published by Elsevier Inc. on behalf of the National Medical Association.
https://doi.org/10.1016/j.jnma.2019.09.004
INTRODUCTION
M
edical device companies have fostered strong relationships with orthopaedic surgeons due to the prominent role of implant use in the practice
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
of orthopaedic surgery.1e5 These relationships have come under scrutiny in the past several years. The PhysiciansPayment-Sunshine-Act (PPSA) was passed in 2010 as part of the Affordable Care Act in order to increase transparency regarding physician-industry relationships.6 The PPSA requires that all payments from medical device and drug companies to physicians be reported to the Center for Medicare and Medicaid Services (CMS) so that they can then become publicly available and searchable online.7 Several studies have investigated the financial ties between implant companies and orthopaedic surgeons.1e5 A study by Iyer et al. concluded that orthopaedic surgeons receive the highest median industry payments compared to all other specialties, primarily as a result of a small fraction of orthopaedists receiving substantial royalties and licensing fees.1 However, there is limited data on the surgeon characteristics that are associated with large payments. Previous studies have briefly discussed trends of industry payments to orthopaedic surgeons with regard to regional variation and subspecialty2,3,8 and academic influence.9 One recent study described the types and distributions of industry payments amongst all physicians and found that orthopaedists and neurosurgeons received the highest payments per physician and that women received lower mean payments than their male counterparts within their respective specialties.10 Another study focusing on spine surgeons found that being in an academic practice setting and being male was associated with receiving industry payments.8 However, no study has examined how gender and academic affiliation may affect industry payments within the orthopaedic subspecialties. This is of particular importance with the increasing percentage of women in recent orthopaedic graduates and trainees. The purpose of this study is thus to assess and quantify the relationship between gender and academic affiliation with industry payments within the orthopaedic subspecialties.
MATERIALS AND METHODS The study is an observational, retrospective analysis of publicly available data. This study received an Institutional Review Board exemption given that the data is publicly available.
VOL
-,
NO
-, -
2019 1
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
The CMS open payments website was queried to obtain the industry payments to orthopaedic surgeons in the 2014 fiscal year.7 Payments to each physician were aggregated by using each physician’s unique identifier (PPSA ID). The open payments website classifies physician payments as general payments, research payments or associated research payments. General payments consist of in-kind payments, royalties, travel, speaking fees, gifts, consulting fees, meals, among others. These general payments were defined as “non-research payments.” Per the open payments website, research payments refer to payments meant for research use. Associated research payments refer to payments directly paid to an institution for research purposes under a physician principal investigator (PI). The “associated” and “individual” research payments were summed by each physician’s PPSA ID and a combined industry research variable was created. These payments were defined as “research payments.” We defined academic orthopaedic surgery departments as those with an accredited (MD) residency program in the 2013e2014 academic year.11,12 When the data was collected (6/2014-12/2014), 149 orthopaedic residency programs were identified. Exclusion criteria for residency programs were lack of online faculty listings or lack of medical school affiliation. Listings were thus obtained for 132/149 (92%) programs. To identify whether surgeons listed in the CMS website were involved in academia, we manually performed internet searches and matched the surgeons to their faculty listings. We matched surgeons by name, state, city, specialty, and zip code. When the specialty was not clearly identified, the surgeon was classified as being a general orthopaedic surgeon. Gender was verified using listed demographic data and photographs during these internet searches. Statistical analysis of relationships regarding industry payments were explored with the use of chi-square, Fisher’s exact test, and Mann-Whitney U testing. Multivariable linear regression was used to determine independent correlates of industry research and non-research payments with gender, academic rank, and subspecialty as covariates. As covariates, academic ranks were compared to non-academic community surgeon status and subspecialties were compared to general orthopedics. Data analysis was performed using SPSS version 24.0 (SPSS, Chicago, IL). All tests were performed with a statistical significance set at a probability value of 0.05.
RESULTS Study population A total of 22,352 orthopaedic surgeons were included in our study (Table 1). It is important to recognize that this
2
VOL.
-,
NO
-, -
2019
number does not include the entire orthopaedic surgeon workforce in the Unites States. Rather, it reflects the number of orthopaedic surgeons who received any industry payment in 2014 that was reported to CMS. Of these surgeons, only 5.8% were identified as being women. Academic surgeons made up 12.3% of the study population. Within academia, women made up 7.5% of the academic surgeon cohort. In the community surgeon cohort, women made up 5.6% of the cohort (P < 0.001). Regarding industry payments, 99.8% of surgeons received non-research payments whereas only 4.4% received research payments.
Overall payments Industry payments to orthopaedic surgeons are summarized in Table 2. Non-research payments totaled to $402,449,706. The median non-research payment amount was $381 whereas the mean non-research payment was $18,005. The mean was highly skewed due the large standard-deviation ($197,718) as a small number of surgeons were paid significant sums above the median. Nonresearch payments were further categorized into different payment levels as shown in Table 3. The vast majority of surgeons (65%) received payments $1000 whereas only 4.5% of surgeons received payments totaling $50,000. Research payments totaled $28,640,411 (Table 2). Only 4.4% of the study population received a research payment (Table 1), thus the median and mode were $0 as shown in Table 2. The mean was highly skewed ($1281) due to large research payments to a small number of individuals.
Analyses of non-research industry payments in relation to gender There were significant gender differences regarding nonresearch payments to orthopaedic surgeons. Of the 22,307 surgeons who received a non-research payment, 5.8% were women (P < 0.001) (Tables 2e3). Men received a higher mean non-research payment amount ($18,987 vs. $2092) than women as seen in Table 2 (P < 0.001). These larger payments to men are further delineated when looking at payment levels. As shown in Table 3, it can be seen that 75.1% of women (vs. 64.4% men) received payments in the $1000 category. This is the lowest payment category and the only category that consisted of a higher percentage of women than men. The remaining higher payment level categories all have a smaller percentage of women than men (all P < 0.001). The gender difference in payments is particularly highlighted when looking at the highest payment level category (>$50,000). Only 0.4% of women surgeons received payments in this level in comparison to 4.8% of men.
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Table 1. Characteristics of orthopaedic surgeons who received industry payments in 2014.
No. (%) of Orthopaedic Surgeons (n [ 22,352) Gender Total
Male (n [ 21,053)
Practice Setting Female (n [ 1299)
p-value
Academic (n [ 2756)
Community (n [ 19,596)
p-value < 0.001
Gender Male
21,053 (94.2%)
2550 (92.5%)
18,503 (94.4%)
Female
1299 (5.8%)
206 (7.5%)
1093 (5.6%)
< 0.001
Practice Setting 2756 (12.3%)
2550 (12.1%)
206 (15.9%)
Community- based
19,596 (87.7%)
18,503 (87.9%)
1093 (84.1%)
Non-Research Payment
22,307 (99.8%)
21,009 (99.8%)
1298 (99.9%)
0.520
2738 (99.3%)
19,569 (99.9%)
<0.001
Research Payment
989 (4.4%)
976 (4.6%)
13 (1.0%)
<0.001
322 (11.7%)
667 (3.4%)
<0.001
<0.001
Orthopaedic Subspecialty
<0.001
General
15,308 (68.5%)
14,467 (68.7%)
841 (64.7%)
1537 (55.8%)
13,771 (70.3%)
Adult Reconstructive
708 (3.2%)
696 (3.3%)
12 (0.9%)
141 (5.1%)
567 (2.9%)
Foot and Ankle
575 (2.6%)
503 (2.4%)
72 (5.5%)
101 (3.7%)
474 (2.4%)
Hand
1359 (6.1%)
1214 (5.8%)
145 (11.2%)
229 (8.3%)
1130 (5.8%)
Spine
1513 (6.8%)
1488 (7.1%)
25 (1.9%)
221 (8.0%)
1292 (6.6%)
Trauma
495 (2.2%)
466 (2.2%)
29 (2.2%)
146 (5.3%)
349 (1.8%)
Pediatrics
327 (1.5%)
254 (1.2%)
73 (5.6%)
109 (4.0%)
218 (1.1%)
Sports
2067 (9.2%)
1965 (9.3%)
102 (7.9%)
272 (9.9%)
1795 (9.2%)
VOL
Bolded values represent significance of P < 0.05.
-,
NO -, -
2019 3
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Academic-based
VOL. -,
NO -, -
2019
n Non-Research Payments
Mean a
22,307
$18,005
Median
Mode
Std. Dev
Minimum
$381
$14
$197,718
$0
Maximum
Sum
$16,638,590
$402,449,076 <0.001b
Gender Male
21,009
$18,987
$394
$14
$203,641
$0
$16,638,590
$399,731,410
Female
1298
$2092
$227
$121
$17,291
$0
$568,014
$2,717,667 <0.001b
Practice Setting Academic
2738
$45,963
$1087
$0
$218,635
$0
$4,538,501
$126,673,289
Community
19,569
$14,073
$348
$14
$194,280
$0
$16,638,590
$275,775,787
$1281
$0
$0
$13,991
$0
$807,354
$28,640,411
Research Payments
p-value
989
a
<0.001b
Gender Male
976
$1353
$0
$0
$14,406
$0
$807,354
$28,484,719
Female
13
$120
$0
$0
$1830
$0
$48,002
$155,693 <0.001b
Practice Setting JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Academic
322
$3779
$0
$0
$24,581
$0
$517,007
$10,414,295
Community
667
$930
$0
$0
$11,719
$0
$807,354
$18,226,116
Bolded values represent significance of P < 0.05. a This n was not used for the statistical analysis. The number with the * represents the number of surgeons who received such payment type for clarity’s sake. The entire study population of 22,352 surgeons was used in the calculation of the mean, median, mode, SD, min, max and sum to include those who did not receive a certain payment type and provide an analysis more representative of the entire study population. b This p-value refers to the comparisons of means for the respective category (i.e. gender, practice setting).
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
4 Table 2. Summary of industry payments to orthopaedic surgeons.
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Table 3. Payment level distribution of industry non-research payments.
Gender Total n [ 22,307 (100%)
Male n [ 21,009 (94.2%)
Female n [ 1298 (5.8%)
<0.001
Practice Setting Academic n [ 2738 (12.3%)
p-value
Community n [ 19,569 (87.7%)
<0.001
<0.001
Payment Level $1000
p-value
<0.001
65.0%
64.4%
75.1%
48.7%
67.3%
$1001 - $10,000
24.2%
24.4%
21.5%
26.4%
23.9%
$10,001 - $25,000
4.0%
4.1%
2.4%
7.1%
3.6%
$25,001 - $50,000
2.2%
2.3%
0.6%
5.4%
1.8%
$50,000
4.5%
4.8%
0.4%
12.3%
3.4%
Bolded values represent significance of P < 0.05.
Analyses of research industry payments in relation to gender
payments in this category in comparison to 12.3% of academic surgeons.
Of the 989 surgeons who received a research payment, only 13 (1.3%) were women (Table 2, P < 0.001). Men received a higher mean research payment amount ($1353 vs. $120; Table 2) than womens (P < 0.001). The analysis of research payment level by gender was not performed given the low number (n ¼ 13) of womens who received a research payment.
Analyses of research industry payments in relation to practice setting
Analyses of non-research industry payments in relation to practice setting There were significant differences regarding non-research industry payments to orthopaedists based on practice setting. Of the 22,307 surgeons who received an industry non-research payment, only 2738 were academic surgeons (P < 0.001) (Tables 2 and 3). Despite consisting of only 12.3% of the study population, academic surgeons received a much higher mean non-research payment amount ($45,963 vs. $14,073) than community surgeons as seen in Table 2 (P < 0.001). When looking at these payments by payment level (Table 3), it can be seen that 67.3% of community surgeons (vs. 48.7% academic surgeons) received payments in the $1000 category. This is the lowest payment category and the only category that consisted of a higher percentage of community than academic surgeons. The remaining higher payment level categories all have a larger percentage of academic than community surgeons (all P < 0.001). This difference in payments based on practice setting is particularly highlighted when looking at the highest payment level category (>$50,000). Only 3.4% of community surgeons received
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Of the 989 surgeons who received research payments, only 322 (32.6%) were academic vs. 667 (67.4%) community surgeons (Table 2, P < 0.001). Despite a larger proportion of community surgeons receiving research payments, academic surgeons still received more research money on average ($3779 academic vs. $930 community; P < 0.001). The analysis of research payment level by practice setting was not performed given the low number of surgeons who received research payments.
Analyses of industry payments in relation to specialty Table 1 summarizes the distribution of surgeons who received payments amongst the different subspecialties. The vast majority (68.5%) of surgeons were classified as being generalists. Pediatrics was the specialty least represented in the study population with only 327 (1.5%) surgeons having received payments in 2014. There were significant differences in the gender distribution of surgeons amongst the subspecialties. As shown in Table 1, foot and ankle, hand, and pediatrics had a significantly larger proportion of women than men (P < 0.001). On the other hand, general orthopaedics, adult reconstructive, spine, and sports medicine had a larger proportion of men than women (P < 0.001). There were also differences in the subspecialty distribution of academic vs. community surgeons. There were more generalists in the community (70.3%) than in academia (55.8%). However, there were
VOL
-,
NO
-, -
2019 5
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Figure 1. Mean research and non-research payments by subspecialty amongst academic and community surgeons.
more adult reconstructive, hand, trauma, and pediatrics specialists in academia than the community (P < 0.001). Industry payments to orthopaedic surgeons by specialty breakdown are shown in Figure 1. Adult reconstruction and spine had the highest mean industry non-research payments in both the community and academic settings. These specialties were also the specialties with the lowest proportion of women surgeons (Table 1). Foot and ankle, hand, and pediatrics had the lowest non-research payments (Figure 1). These specialties were also the specialties with the highest proportion of women surgeons as previously shown in Table 1. Similar trends were found regarding research payments. As shown in Figure 1, adult reconstruction and spine also received the highest research payments in both academic and community settings. Although academic surgeons received on average higher research payments (Table 2), there were no differences in payments between academic and community surgeons that could be delineated when breaking down research payments by subspecialties (Figure 1).
Multivariable regression Multivariable linear regression was used to ascertain the independent predictive relationship of gender, academic affiliation and rank, and orthopaedic subspecialty to
6
VOL.
-,
NO
-, -
2019
research and non-research industry payments. Male gender (P ¼ 0.006), adult reconstruction (P < 0.001) and spine (P ¼ 0.008) subspecialties, and associate professor (P ¼ 0.001), professor (P < 0.001) and chair (P < 0.001) academic rank were independently associated with larger research industry payments (Table 4). Similarly, male gender (P ¼ 0.029), adult reconstruction (P < 0.001), and spine (P < 0.001) subspecialties, and chair academic rank (P < 0.001) were independently associated with larger non-research industry payments (Table 5).
DISCUSSION Our study showed that there are differences in industry payment size to orthopaedic surgeons based on gender, practice type and subspecialty. In our study, only 1% of women surgeons received payments >$25,000 compared to 11.2% of men. This gender discrepancy can also be seen in other areas of medicine. Tringale et al. showed that male physicians were more likely to receive an industry payment and to receive larger payments than womens in nearly all specialties.10 There are several possible explanations for the difference in payments based on gender. As shown in Table 1 and Figure 1, there is a larger proportion of women than men in specialties that receive lower industry payments (foot & ankle, hand, and pediatrics).
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
Table 4. Multivariable linear regression for predictors of industry research payments.
Table 5. Multivariable linear regression for predictors of industry nonresearch payments.
Covariate
Standardized Coefficient (B)
p-value
Covariate
Standardized Coefficient (B)
p-value
Female
0.019
0.006
Female
0.015
0.029
Academic rank (relative to community surgeon)
Academic rank (relative to community surgeon)
Assistant Professor
0.011
0.093
Assistant Professor
0.004
0.525
Associate Professor
0.023
0.001
Associate Professor
0.001
0.918
Professor
0.039
<0.001
Professor
0.012
0.084
Chair
0.060
<0.001
Chair
0.93
<0.001
Subspecialty (relative to general orthopedics)
Subspecialty (relative to general orthopedics)
Adult Reconstruction
0.039
<0.001
Adult Reconstruction
0.032
<0.001
Foot & Ankle
0.008
0.235
Foot & Ankle
0.002
0.817
Hand
0.013
0.057
Hand
0.014
0.035
Spine
0.018
0.008
Spine
0.046
<0.001
Trauma
0.004
0.552
Trauma
0.004
0.551
Pediatric
0.001
0.879
Pediatric
0.007
0.323
Sports Medicine
0.001
0.591
Sports Medicine
0.003
0.708
Bolded values represent significance of P < 0.05.
Despite these confounding differences, womens still receive smaller industry payments compared to men when adjusting for academic rank and subspecialty using multivariable regression. Overall, orthopaedic surgery has lagged behind other surgical subspecialties in the recruitment of women13 and it is possible that the spine and reconstruction subspecialties have been slower to adapt a larger women workforce. Regarding practice setting, our study found that academic surgeons overall receive larger industry payments that community surgeons. Historically, academic physicians have conducted the majority of pharmaceutical and biodevice research from both federal and industry funding due to their trial-design expertise, accessibility to large patient populations, university resources, and publication experience.14 Since the turn of the millennium, industryfunded pharmaceutical research has transitioned to the private sector for financial reasons.14 The trend of industry funding with regard to medical devices and implants has been less apparent. Using a limited amount of data released from a Department of Justice lawsuit in 2007 regarding biodevice funding in orthopedics, Hockenberry et al. noted an increase in the proportion of industry payments to orthopaedic surgeons with academic affiliations between 2007-2010.5 Industry support of any type has been shown to be associated with higher scholarly
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
Bolded values represent significance of P < 0.05.
impact9,15,16 and thus may promote advancement in academia. Although academic surgeons received larger payments in our study, it is important to note that there was a larger proportion of women academic surgeons (7.5%) than male academic surgeons (5.6%) and that despite this, women still received lower payments than men. Previous studies regarding gender in academia have shown that women experience less research productivity earlier in their careers, receive fewer funds for research, and are slower to advance professionally in several medical and surgical subspecialties.17e23 A recent study examining industry payments within academic otolaryngology found that women receive significantly less industry funding, even when adjusted for academic rank.17 Another possible explanation for the gender discrepancy in payment could be due to women holding less patents than men overall.24 Many of the largest payments documented in the PPSA database were due to royalties. This study has several limitations that merit discussion. Although we were meticulous in our matching of surgeons by name, state, city, specialty, and zip code, it is feasible that there may be some inaccuraciesdparticularly with surgeons with the same name. There are also variables we could not control for in our study such as migration,
VOL
-,
NO
-, -
2019 7
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS
retirement, name change, and death. Another source of inaccuracy would be incomplete or incorrect data from the PPSA website. Although physicians can review and dispute any incorrect data, it is unknown how often physicians are reviewing the posted payments.7 This is of relevance as previous studies have highlighted that there are many discrepancies between what surgeons disclose at academic meetings versus that the medical industry reports.25e27 Despite these limitations, the PPSA website is the most comprehensive data for industry payments at this time.7 Another limitation of our study is that it may not be representative of the entire orthopaedic workforce as it only includes orthopaedic surgeons who received some sort of industry payment in 2014.7 According to the most recent American Academy of Orthopaedic Surgeon (AAOS) biennial census, there were 29,613 practicing orthopaedic surgeons in the US in 2016.28 There are thus > 7000 practicing orthopaedic surgeons who were potentially not included in our study and indicate a selection bias in the noted findings. The subspecialty distribution in our analysis was also slightly different than the one reported on the 2016 AAOS census. Our study reported a high proportion of generalists (68.5%) vs. only 42% per the 2016 census. This discrepancy is likely due to several surgeons being classified as generalists in the PPSA database despite being specialists. Despite this, even if the correct subspecialty had been identified, it is unlikely that the overall trends in payments by subspecialty would have significantly changed as the mean non-research payment amount to generalists was still significantly smaller than the mean payment to reconstruction and spine surgeons (Figure 1). Even though there are limitations to the subspecialty analysis, it is important to note that the distribution of gender and practice setting in our study was similar to the one reported in the 2016 census. The AAOS 2016 census reports that 6.5% of practicing orthopaedic surgeons are women (vs. 5.8% in our study) and that 15% are academic surgeons (vs. 12.3% in our study), which suggests that our gender and practice setting analyses are likely a good representation of the overall US Orthopaedic workforce. Further limitations of our study are that there may be other factors that could influence industry payments such as academic rank, academic influence, years in practice, clinical volume, institution, and ethnicity, among others. A recent study by our group showed that academic influence/ impact correlated poorly with industry payment amount to orthopaedic surgeons.9 On the other hand, studies have shown a positive association between academic rank and industry payments.15,16 Given this previous research and our results, multivariable regression analyses are needed to
8
VOL.
-,
NO
-, -
2019
identify the factors that are associated with large industry payments.
IMPLICATIONS According to the PPSA data, there are differences in the industry payments based on gender, academic rank, and practice type among orthopaedic surgeons. In general, payments to women and to community surgeons are lower than payments to men and academic surgeons, respectively. Women surgeons also had a larger representation in the subspecialties that received the lowest industry payments.
APPENDIX A. SUPPLEMENTARY DATA Supplementary data to this article can be found online at https://doi.org/10.1016/j.jnma.2019.09.004.
REFERENCES 1. Iyer, S., Derman, P., & Sandhu, H. S. (2016). Orthopaedics and the physician payments sunshine Act. J Bone Jt Surg, 98(5), e18. 2. Cvetanovich, G. L., Chalmers, P. N., & Bach, B. R. (2015). Industry financial relationships in orthopaedic surgery: analysis of the sunshine Act open payments database and comparison with other surgical subspecialties. J Bone Joint Surg Am, 97(15), 1288e1295. 3. Lopez, J., Ahmed, R., Bae, S., et al. (2016). A new culture of transparency: industry payments to orthopedic surgeons. Orthopedics, 39(6), e1058ee1062. 4. Samuel, A. M., Webb, M. L., Lukasiewicz, A. M., et al. (2015). Orthopaedic surgeons receive the most industry payments to physicians but large disparities are seen in sunshine Act data. Clin Orthop Relat Res, 473(10), 3297e3306. 5. Hockenberry, J. M., Weigel, P., Auerbach, A., & Cram, P. (2011). Financial payments by orthopedic device makers to orthopedic surgeons. Arch Intern Med, 171(19), 1759. 6. Agrawal, S., Brennan, N., & Budetti, P. (2013). The sunshine act– effects on physicians. N Engl J Med, 368(22), 2054e2057. 7. CMS. Open Payments Centers for Medicare and Medicaid Services. 8. Weiner, J. A., Cook, R. W., Hashmi, S., et al. (2017). Factors associated with financial relationships between spine surgeons and industry: an analysis of the open payments database. Spine (Phila Pa 1976), 42(18), 1412e1418. 9. Buerba, R. A., Sheppard, W. L., Herndon, K. E., et al. (2018). Academic influence and its relationship to industry payments in orthopaedic surgery. J Bone Jt Surg, 100(9), e59. 10. Tringale, K. R., Marshall, D., Mackey, T. K., Connor, M., Murphy, J. D., & Hattangadi-Gluth, J. A. (2017). Types and distribution of
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
DISPARITIES IN INDUSTRY PAYMENTS TO ORTHOPAEDIC SURGEONS payments from industry to physicians in 2015. J Am Med Assoc, 317(17), 1774.
20. Eloy, J. A., Svider, P., Chandrasekhar, S. S., et al. (2013). Gender disparities in scholarly productivity within academic otolaryngology departments. Otolaryngol Neck Surg, 148(2), 215e222.
11. FREIDA Residency & Fellowship Database.
21. Zhuge, Y., Kaufman, J., Simeone, D. M., Chen, H., & Velazquez,
12. NIH Research Portfolio Online Reporting Tools (RePORT). 13. Pico, K., Gioe, T. J., Vanheest, A., & Tatman, P. J. (2010). Do men
O. C. (2011). Is there still a glass ceiling for women in academic surgery? Ann Surg, 253(4), 637e643.
outperform women during orthopaedic residency training?
22. Reed, D. A., Enders, F., Lindor, R., McClees, M., & Lindor, K. D.
Clin Orthop Relat Res, 468(7), 1804e1808.
(2011). Gender differences in academic productivity and
14. Fisher, J. A., Kalbaugh, C. A., & Ross, J. S. (2012). United States private-sector research:
a
physicians qualitative
and
pharmaceutical
study.
PLoS
Med,
contract
9(7).
leadership appointments of physicians throughout academic careers. Acad Med, 86(1), 43e47.
www.
23. Paik, A. M., Mady, L. J., Villanueva, N. L., et al. (2014). Research
15. Eloy, J. A., Kilic, S., Yoo, N. G., et al. (2017). Is industry funding
productivity and gender disparities: a look at academic plastic surgery. J Surg Educ, 71(4), 593e600.
plosmedicine.org1.
associated with greater scholarly impact among academic neurosurgeons? World Neurosurg, 103, 517e525. 16. Svider, P. F., Bobian, M., Lin, H.-S., et al. (2017). Are industry financial ties associated with greater scholarly impact among academic
otolaryngologists?
The
Laryngoscope,
127(1),
87e94. 17. Eloy, J. A., Bobian, M., Svider, P. F., et al. (2017). Association of
24. Ding, W. W., Murray, F., & Stuart, T. E. (2006). Gender differences in patenting in the academic life sciences. Science, 313(5787), 665e667. 25. Buerba, R. A., Fu, M. C., & Grauer, J. N. (2013). Discrepancies in spine surgeon conflict of interest disclosures between a national meeting and physician payment listings on device manufacturer web sites. Spine J, 13(12).
gender with financial relationships between industry and aca-
26. Hannon, C. P., Chalmers, P. N., Carpiniello, M. F., Cvetanovich,
demic otolaryngologists. JAMA Otolaryngol Neck Surg, 150(1),
G. L., Cole, B. J., & Bach, B. R. (2016). Inconsistencies between
53e60.
physician-reported disclosures at the AAOS annual meeting
18. Eloy, J. A., Svider, P. F., Cherla, D. V., et al. (2013). Gender disparities in research productivity among 9952 academic physicians. The Laryngoscope, 123(8), 1865e1875. 19. Lopez, S. A., Svider, P. F., Misra, P., Bhagat, N., Langer, P. D., & Eloy, J. A. (2014). Gender differences in promotion and scholarly impact: an analysis of 1460 academic ophthalmologists. J Surg Educ, 71(6), 851e859.
JOURNAL OF THE NATIONAL MEDICAL ASSOCIATION
and industry-reported financial disclosures in the open payments database. J Bone Jt Surg, 98(20), e90. 27. Okike, K., Kocher, M. S., Wei, E. X., Mehlman, C. T., & Bhandari, M. (2009). Accuracy of conflict-of-interest disclosures reported by physicians. N Engl J Med, 361(15), 1466e1474. 28. American Academy of Orthoapaedic Surgeons. AAOS 2016 Census.
VOL
-,
NO
-, -
2019 9