Types and Distribution of Payments From Industry to Physicians in 2015

Key Points

Question  What were the types and distribution of payments from industry to physicians in 2015?

Findings  In 2015, 449 864 (approximately 48%) of US physicians were reported to have received a total of $2.4 billion in industry-related payments, with a higher likelihood and higher value of payments to physicians in surgical than to primary care specialties and to male than to female physicians.

Meaning  A substantial proportion of US physicians were reported to have received payments from biomedical and pharmaceutical industries in 2015. With increased scrutiny of the relationship between industry promotion and clinical practice, understanding physician characteristics associated with these financial relationships is important.

Importance  Given scrutiny over financial conflicts of interest in health care, it is important to understand the types and distribution of industry-related payments to physicians.

Objective  To determine the types and distribution of industry-related payments to physicians in 2015 and the association of physician specialty and sex with receipt of payments from industry.

Design, Setting, and Participants  Observational, retrospective, population-based study of licensed US physicians (per National Plan & Provider Enumeration System) linked to 2015 Open Payments reports of industry payments. A total of 933 295 allopathic and osteopathic physicians. Outcomes were compared across specialties (surgery, primary care, specialists, interventionalists) and between 620 166 male (66.4%) and 313 129 female (33.6%) physicians using regression models adjusting for geographic Medicare-spending region and sole proprietorship.

Exposures  Physician specialty and sex.

Main Outcomes and Measures  Reported physician payment from industry (including nature, number, and value), categorized as general payments (including consulting fees and food and beverage), ownership interests (including stock options, partnership shares), royalty or license payments, and research payments. Associations between physician characteristics and reported receipt of payment.

Results  In 2015, 449 864 of 933 295 physicians (133 842 [29.8%] women), representing approximately 48% of all US physicians were reported to have received $2.4 billion in industry payments, including approximately $1.8 billion for general payments, $544 million for ownership interests, and $75 million for research payments. Compared with 47.7% of primary care physicians (205 830 of 431 819), 61.0% of surgeons (110 604 of 181 372) were reported as receiving general payments (absolute difference, 13.3%; 95% CI, 13.1-13.6; odds ratio [OR], 1.72; P < .001). Surgeons had a mean per-physician reported payment value of $6879 (95% CI, $5895-$7862) vs $2227 (95% CI, $2141-$2314) among primary care physicians (absolute difference, $4651; 95% CI, $4014-$5288). After adjusting for geographic spending region and sole proprietorship, men within each specialty had a higher odds of receiving general payments than did women: surgery, 62.5% vs 56.5% (OR, 1.28; 95% CI, 1.26-1.31); primary care, 50.9% vs 43.0% (OR, 1.38; 95% CI, 1.36-1.39); specialists, 36.3% vs 33.4% (OR, 1.15; 95% CI, 1.13-1.17); and interventionalists, 58.1% vs 40.7% (OR, 2.03; 95% CI, 1.97-2.10; P < .001 for all tests). Similarly, men reportedly received more royalty or license payments than did women: surgery, 1.2% vs 0.03% (OR, 43.20; 95% CI, 25.02-74.57); primary care, 0.02% vs 0.002% (OR, 9.34; 95% CI, 4.11-21.23); specialists, 0.08% vs 0.01% (OR, 3.67; 95% CI, 1.71-7.89); and for interventionalists, 0.13% vs 0.04% (OR, 7.98; 95% CI, 2.87-22.19; P < .001 for all tests).

Conclusions and Relevance  According to data from 2015 Open Payments reports, 48% of physicians were reported to have received a total of $2.4 billion in industry-related payments, primarily general payments, with a higher likelihood and higher value of payments to physicians in surgical vs primary care specialties and to male vs female physicians.

Quiz Ref IDFinancial relationships between physicians and the biomedical industry are common.¹ Concern for financial conflicts of interest and their effect on patient care, medical research, and education^2,3 prompted the creation of the Open Payments program,^4,5 a comprehensive, nationwide public data repository reporting industry payments to physicians and teaching hospitals. Prior to Open Payments, such data were gathered from physician self-report,¹ a limited number of state databases,⁶ and third-party data sources.⁷ Open Payments, implemented under section 6002 of the Affordable Care Act and managed by the Centers for Medicare & Medicaid Services’ (CMS), requires biomedical manufacturers and group purchasing organizations (GPOs) to report all payments and ownership interests made to physicians starting in 2013. The recent release of the most current, complete, nationally representative Open Payments data⁸ presents the opportunity to analyze the distribution and variability of these financial relationships among physicians.

Early studies of Open Payments data revealed specialty variation in payments^9,10 but did not account for other physician demographics. For example, sex differences in industry relationships, especially given differential sex makeup across specialties, are largely unexplored. In this study, Open Payments data from 2015 were analyzed to evaluate variability in general payments, ownership interests, and royalty or license payments, among specialty categories and by physician sex.

The study design was an observational, retrospective, population-based analysis of physicians in the 2015 CMS National Plan & Provider Enumeration System (NPPES)¹¹ database linked to 2015 Open Payment reports of industry payments to allopathic and osteopathic physicians in the continental United States. This study was approved by the University of California San Diego Institutional Review Board.

The study population consisted of all physicians or surgeons licensed to practice in the United States per the 2015 NPPES database.⁵ The NPPES database includes all physicians with a National Provider Identifier (NPI) and is the database against which Open Payments physician records are verified; therefore, NPPES specialty counts were used to estimate the number of physicians eligible for having a report in the Open Payments database.¹⁰ The analyses were limited to physicians in allopathic and osteopathic specialties within NPPES provider taxonomy.¹¹

Data from the 2015 Open Payments program including physician general payments, ownership interests, and research payment files were linked to the August 2015 NPPES database as previously described.^9,10 Open Payments excludes medical residents (but not fellows) and physicians who are employees of a reporting entity. Inclusion and exclusion criteria of the NPPES study population and matched Open Payments physicians are shown in eFigure 1 in the Supplement. These data are generalizable to allopathic and osteopathic physicians licensed to practice in the United States.¹² Data were aggregated by specialty and grouped by nature of practice (surgery, primary care, medical subspecialists),⁹ with medical subspecialists further divided into specialists (ie, pathology, neurology) and interventionalists (ie, cardiovascular disease, anesthesiology) based on the procedural nature of practice. Specifically, the definition of interventionalists was based on data showing that these subspecialties bill more relative-value units for common procedures, which determine payment based on procedural-based codes (ie, colonoscopies, cardiac catheterization, and epidural nerve blocks).¹³ Several medical specialties, such as endocrinology and infectious disease, were categorized under internal medicine.

Physician and practice characteristics were obtained from NPPES, including sex, status as sole proprietor, and geographic spending region via zip code. Sex was self-reported in NPPES as metavariable “provider gender code” with categories M and F, so physicians were thereby categorized accordingly. To account for regional variation in physician practice setting and spending, the NPPES physician practice zip code was linked to the Dartmouth Atlas hospital referral region and corresponding most recent (2013) total price-, age-, sex-, and race-adjusted Medicare spending per enrollee.¹⁴ Practice spending region was categorized into 3 spending groups: low, average, and high.^10,15,16

Physician payments were characterized into 3 categories by payment type: general payments, research payments, and ownership interests. Payment end points included: total number and value of payments, and number of physicians reported to receive each type of payment; annual per-physician mean, median, for both number and value of payments, nature (or reason) for general payments; and total value invested and interest received for ownership interests.⁵ In addition, the proportion of physicians by specialty receiving greater than $10 000 in 2015 was evaluated because the US Department of Health and Human Services specifically identifies payments exceeding this threshold as significant conflicts of interest. Quiz Ref IDGeneral payments include all forms of payment (such as speaking fees or food and beverage) other than those classified for research purposes. Research payments are those specifically for research endeavors under a written contract or protocol, including payments for medical research writing or publication.¹⁷ Ownership interests include stocks or stock options, partnership shares, limited liability company membership, bonds, or other financial instruments secured by the reporting entity that were held by physicians. Excluded from ownership interest were payments received as compensation (until exercised), as part of a retirement plan, or interest in a publicly traded security or mutual fund. “Amount invested” is the total amount of the interest holding gained by the physician during the reporting year. “Value of interest” describes the cumulative value of ownership interest per the most recent valuation.¹⁷

Differences in industry payments among specialty categories were assessed with χ² and Kruskal-Wallis tests. Bivariable and multivariable analyses were performed to investigate the association between physician characteristics (primarily specialty, as well as sex, geographic spending region, sole proprietorship) and industry payments. Logistic regression models were also used to examine the association between physician characteristics and the odds of physicians receiving general payments, royalty or license payments, or holding ownership interests. Linear regression models were used to examine the association between physician characteristics and the value of general payments. Payment value data were log-transformed prior to analysis to account for skew.¹⁰ Negative binomial regression models examined the association between physician characteristics and the per-physician annual number of general payments. All multivariable models controlled for potentially confounding variables available for analysis,¹⁰ namely specialty, sex, geographic spending region, and sole proprietor status. To examine sex while accounting for varied sex composition across specialties, multivariable models included an interaction term between specialty and sex. A 2-tailed P value of <.001 was considered significant for all tests after Bonferroni correction for multiple comparisons (.05/86). Analyses were performed using SAS v9.4 (SAS Institute Inc).

An estimated 933 295 allopathic and osteopathic physicians, including 620 166 male (66.4%) and 313 129 female (33.6%) physicians, were licensed to practice in the United States in 2015 (Table 1). Among them, 449 864 physicians, including 316 022 men (51.0%) and 133 842 women (42.7%), reportedly received an industry payment or held ownership interests in 2015.

Reports of physician payments from industry totaled $2.4 billion, including approximately $1.8 billion for general payments, more than $544 million for cumulative value of ownership interests, and $75 million for research payments (Table 2).

Of all reported payments in 2015, 68% (≈$1.8 of $2.6 billion) of general payments were attributable to a physician recipient, with the remaining to teaching hospitals. The median per-physician value of general payments was $201 (interquartile range [IQR], $60-$782; Table 2). Analyses of value of general payments by nature (or reason) for payment are shown in Table 3. Although 88.7% of general payments were for food and beverage, the greatest proportion (27.3%) of value was from royalty or license payments (≈$484 million of $1.8 billion) followed by service fees (26.6%), such as faculty lectures ($472 million of $1.8 billion).

Although only 0.4% of physicians held ownership interests overall (3302 of 933 295 total physicians), 22.8% of the total value of all payments to physicians was attributable to the cumulative value of ownership interests ($544 million of $2.4 billion; Table 2). Of all reported payments in 2015, 2% ($75 million of $3.89 billion) of research payments were attributable to a physician recipient, with the remaining to teaching hospitals.

Distribution and values of general payments, research payments, and ownership interests by specialty category are shown in Table 2. Among the 4 specialty categories, interventionalists were reported to receive the highest median number (9 payments; IQR, 2-39) and value of payments per physician (≈$324, IQR; $80-$1296; P < .001). Distribution by nature of general payments significantly differed among specialties (Table 3), with surgeons receiving 54.5% of their value from royalty or license payments ($415 million of $761 million), whereas primary care physicians (41.3%), specialists (39.8%), and interventionalists (37.5%) all received the greatest proportion of payments from service fees like faculty lectures (χ² test, P < .001).

Reports of per-physician value of general payments by each specialty in 2015 are shown in Table 4. Of 29 443 cardiologist, 74.9% received payments, representing the highest proportion within physician specialties, followed by 72.2% of 15 457 gastroenterologists. Neurosurgery and orthopedics received the highest maximum reported values per-physician (≈$18 million and ≈$38 million, respectively), whereas cardiology ($862; IQR, $226-$2749) and thoracic surgery ($860; IQR, $192-$2780) received the highest median value per physician. Neurosurgery had the highest proportion of physicians to receive a personal total value of more than $10 000 (12.6%, 565 of 4468), ), followed by cardiology (12.1%; 2661 of 22 044), and neurology (11.8%; 1275 of 10 794 physicians).

Unadjusted bivariable analyses of general payments are presented in Table 5. Surgeons (61.0%) were more likely to have received general payments than primary care physicians (110 604 of 181 372; absolute difference, 13.3%; 95% CI, 13.1%-13.6%; OR, 1.72; P < .001). After log-transformation to account for the skewed distribution of per-physician payment value, the ratio of means for interventionalists was the highest (1.74; 95% CI, 1.70-1.77; P < .001) relative to primary care physicians. Interventionalists were also the only group to receive more payments per physician than primary care physicians (mean, 28.8 vs 27.7 payments; absolute difference, 1.1 payments; 95% CI, 0.7-1.6; incident rate ratio [IRR], 1.09; P < .001; Table 6).

Surgeons held 57.2% of ownership interest value invested (≈$94 million of $165 million), whereas specialists received 41.8%, the highest proportion, of cumulative interest value (≈$227 million of $544 million; Table 2). Unadjusted analyses showed that surgeons were more likely to hold ownership interests and receive royalty or license payments compared with primary care physicians (Table 7). The greatest proportion of total number (49.5%) and value (45.9%) of physician research payments went to primary care physicians (Table 2).

The distribution of nature of general payments significantly differed between sexes across specialties (χ² test, P < .001), eTable 1 in the Supplement. With all specialties combined, a greater proportion of men (50.8%) reportedly received a general payment than did women (42.6%; absolute difference, 8.2%; 95% CI, 8.0%-8.4%; OR, 1.39; P < .001; Table 5). Women had a lower mean value of general payments per physician than men ($1390 vs $5031; absolute difference, $3641; 95% CI, $3083-$4199; ratio of means, 1.81; P < .001; Table 5).

Multivariable models adjusted for specialty, sex, practice spending region, and sole proprietor status are presented in Table 8. The interaction between specialty and sex was significant (P < .001); thus, sex as a predictor was evaluated within each specialty category. Male physicians were significantly more likely than female physicians to receive general payments (ie, male interventionalists OR, 2.03; 95% CI, 1.97-2.10; P < .001; Table 8). Male physicians also reportedly received a greater mean per-physician payment value than female physicians within their specialty. For example, male interventionalists received 14.43% greater payment value (95% CI, 13.60-15.26; P < .001) than female interventionalists. The mean value per physician among men was $5252 (95% CI, $4895-$5609) vs $2141 among women (95% CI, $1515-$2768; absolute difference, $3111; 95% CI, $2217-$4005; eTable 2 in the Supplement). Overall, male physicians reportedly received a higher number of general payments per physician than did female physicians.

Compared with female physicians, male physicians had a higher odds of holding ownership interests and receiving royalty or license payments (Table 5). After adjusting for physician characteristics, male physicians were more likely than female physicians to receive reported royalty or license payments and to hold ownership interests in every specialty category except primary care physicians (OR, 0.84; 95% CI, 0.75-0.94; P = .002; (Table 8).

Quiz Ref IDAccording to data from the 2015 Open Payments reports, 449 864 US physicians (≈48%) were reported to have received a total of $2.4 billion in industry-related payments, including $1.8 billion for general payments, $544 million for cumulative ownership interests, and $75 million for research payments. There was a higher likelihood of receipt and higher value of reported payments to surgeons than to primary care physicians and a higher likelihood of receipt and higher value of payments to men than to women.

Quiz Ref IDThe Institute of Medicine has highlighted the tension that exists between “financial relationships with industry and the primary missions of medical research, education, and practice”³. Considerable data have shown that financial conflicts of interest, from small gifts and meals to large sums for consulting, may alter physician decision making.^2,18,19 The current population-based analysis of industry-to-physician payments in 2015 shows the far-reaching extent (more than 10 million transactions totaling $2.4 billion) of these reported financial relationships. There was considerable variability in the number, value, and nature of physician payments among specialties and by sex within specialty groups.

Similar to previous studies,^1,6 procedural specialties (surgery and interventional medicine)⁶ had a higher incidence and greater values of reported industry payments. Interventionalists, such as cardiologists and gastroenterologists, received the highest median per-physician number of reported payments in 2015 (9 payments; IQR, 2-39), suggesting frequent interactions with the vast pharmaceutical and medical device industries for cardiac²⁰ and gastrointestinal disorders.²¹ Orthopedics and neurosurgery reportedly received the highest maximum per-physician general payment values (38.4 million and 17.8 million per physician, respectively), reflecting a minority of physicians receiving large sums.²² Surgical and interventional specialties have strong industry ties, dependence on devices and equipment, and perhaps a greater influence over industry-related expenditures.^20,22 Industry presence in the operating room²³ allows representatives to form relationships with proceduralists, providing device training and income-enhancing opportunity.²⁴

Primary care physicians had frequent but lower reported value of payments, which may reflect pharmaceutical-sponsored lunches.² Recent Medicare data show that the highest proportion of total drug costs were prescribed by family practice and internal medicine.²⁵ Alternatively, pediatrics received fewer industry payments. Pediatricians often treat a relatively healthy patient population, thus have less influence over drug expenditures. Although industry-physician research payments were rare (98% of research payments went directly to an institution), primary care physicians reportedly received the majority of research payments by number and value, which may be driven by pharmaceutical-based clinical trials or their ability to access and enroll study populations.²⁶

Nature of reported payments differed among specialties, reflecting varying industry interactions. For example, the majority of general payment value to interventionalists, specialists, and primary care physicians were reported as service fees, such as faculty lectures. However, in surgical specialties, 54.5% of payment value was reported as royalty or license fees. Surgeons were significantly more likely to have been reported to receive royalty or license payments (such as a royalty payment for developing a hip implant) than other physicians. Surgical instruments comprise the largest single area of the medical device market (20.4% of $4.1 billion invested),²⁷ likely influencing the device industry’s focus on research and design.²⁸ Companies rely heavily on physicians for innovation through licensing agreements, particularly orthopedic and general surgeons who hold the highest number of medical device patents.²⁹

Quiz Ref IDAlthough few physicians held ownership interests in industry, surgeons gained the highest reported value in 2015 ($94 million of $165 million, 57.2%). This could reflect ownership of ambulatory surgical centers, which have been shown to create conflicts of interest around physician self-referral.³⁰ Among urologists and orthopedic surgeons, ownership of ambulatory surgical centers is associated with greater use.^30,31 Medical specialists held the highest cumulative value of ownership interest ($227 million of $544 million, 41.8%). The vast majority was held by radiologists ($209 million, 92.2% of interest value received by specialists), perhaps related to diagnostic imaging equipment.³²

Overall, male physicians had a higher likelihood of engaging with biomedical industry and at higher values than females. Female physicians may have different preferences for industry engagement³³ and have been shown to exhibit different clinical practice patterns with more patient-centered care.³⁴ Industry-based marketing may target male physicians more than female physicians, specifically by engaging with key opinion or thought leaders to establish product credibility and promotion.³⁵ Men were more likely than women to have been reported to receive royalty or license payments across all specialties, with the greatest difference among surgeons. Sex discrepancies in patenting show that women hold fewer patents than men.³⁶ Furthermore, women who obtain patents are less likely to have them commercialized or licensed than men.³⁷ Male physicians were also more likely than female physicians to hold ownership interests in most specialties. Previously, male sex has been shown to be associated with physician ownership among general surgeons and emergency physicians.^38,39

Although physicians may consider themselves committed to ethical practice and professionalism, many do not recognize the subconscious bias that industry relationships have on their decision making.⁴⁰ Companies may preferentially market to profitable specialties such as cardiologists because these physicians influence the prescribing practices of nonspecialists.⁴¹ However, industry also appears to target and influence primary care physicians who accepted frequent low-value payments. A recent study found that physicians who received industry-sponsored meals were more likely to prescribe brand-name medications, with mixed findings across sexes.² Some medical centers have restricted access of industry representatives to physicians.⁴⁰ Further studies are needed to understand the effects of industry payments on specialty- and sex-specific prescribing or operating practices.

This study has several limitations. The Open Payments and NPPES databases may have inaccuracies and physicians may be unaware of their reported payments. The CMS suggests that all physicians review and dispute any incorrect data prior to publication in Open Payments; such vigilance will help improve the accuracy of the data. Yet, Open Payments remains the most comprehensive data for industry payments to date, and there have been improvements in accuracy and reporting in the most recent 2015 release compared with earlier years.⁸ Although a minority of physicians in Open Payments did not have NPIs, 94.5% of Open Payments physicians were matched to NPPES, the same database CMS uses to verify all Open Payments records. It is possible that the extent of industry payments may actually be underreported in the current study. Certain potential confounding variables were unavailable in NPPES including age, career duration, or race/ethnicity. Specialty was grouped to facilitate multivariable analyses, classifying physicians by nature of practice because certain practice types have different relationships with industry. The direct association of payment receipt and drug or device use was not analyzed in this study and should be an area of active investigation. Although specialty categories were not completely homogeneous, they enabled the analysis of sex-specific differences, which to our knowledge has not been performed in previous studies.^1,7

According to data from 2015 Open Payments reports, 48% of physicians were reported to have received industry-related payments, primarily involving general payments, with a higher likelihood and higher value of payments to physicians in surgical than primary care specialties and to male than female physicians.

Corresponding Author: Jona A. Hattangadi-Gluth, MD, University of California, San Diego, Moores Cancer Center, Department of Radiation Medicine and Applied Sciences, 3960 Health Sciences Dr, #0865, La Jolla, CA 92093-0843 (jhattangadi@ucsd.edu).

Accepted for Publication: March 31, 2017.

Author Contributions: Ms Tringale had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Tringale, Marshall, Mackey, Murphy, Hattangadi-Gluth.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Tringale, Connor, Murphy, Hattangadi-Gluth.

Statistical analysis: Tringale, Marshall, Connor, Murphy, Hattangadi-Gluth.

Obtained funding: Hattangadi-Gluth.

Administrative, technical, or material support: Tringale, Mackey, Hattangadi-Gluth.

Supervision: Hattangadi-Gluth.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Hattangadi-Gluth reported receiving research grants from Varian Medical Systems. No other financial disclosures were reported.