Industry Payments to Cardiologists Are Associated With Higher Medicare Spending

ABSTRACT

Objective: To determine whether industry payments to cardiologists are associated with Medicare beneficiary spending, as financial incentives may influence clinical decision-making.

Study Design: Cross-sectional analysis comparing 2022 Open Payments data with cardiologists’ attributed traditional Medicare spending.

Methods: We used publicly available data on industry payments to providers (Open Payments) and professional fees per provider for traditional Medicare beneficiaries for 2022. We estimated a multivariable linear regression with cardiologist-attributed spending per traditional Medicare beneficiary as the outcome and nonresearch industry payments as the predictor of interest. For the main analysis, we adjusted for cardiologist characteristics and patient characteristics and included a fixed effect for state of practice.

Results: Our sample included 26,805 cardiologists, and of these, 82.3% received at least 1 payment. The mean (SD) industry payment to cardiologists was $3958 ($17,471). In the main model, a $10,000 increase in industry payments was associated with greater mean spending per Medicare beneficiary by $14.1 (95% CI, $11.9-$16.3).

Conclusions: Industry payments were positively correlated with spending per traditional Medicare beneficiary attributed to a given cardiologist. Although our analysis does not establish causation, the relationship between industry payments and attributed spending is potentially significant, even if industry payments directly influence only a fraction of the overall expenditure. Future studies should better characterize the relationship between payments from device and pharmaceutical companies and spending for those specific devices and drugs and use causal methods to measure the extent to which these payments may influence health care utilization.

Am J Manag Care. 2026;32(3):166-171. https://doi.org/10.37765/ajmc.2026.89895

Takeaway Points

Our findings reveal the financial relationship between industry payments and Medicare spending in cardiology practice, with implications for health care costs and policy decisions.

• Industry payments averaging $3958 to cardiologists correlate with higher Medicare spending, suggesting potential influence on clinical decision-making.
• A $10,000 increase in industry payments is correlated with a $14.1 increase in Medicare spending per beneficiary.
• Organizations should consider implementing policies to monitor and manage industry relationships to control health care costs.
• Greater transparency in industry-physician relationships may help inform value-based care strategies.

In 2023, industry groups paid health care providers and hospitals $3.3 billion in nonresearch payments. Payments were made to 630,384 physicians and 309,438 nonphysician practitioners.¹ Between 2013 and 2022, 57.1% of physicians received at least 1 payment.² Given concern that such payments to physicians and hospitals may impact clinical decision-making, the Physician Payment Sunshine Act was passed in 2010 as part of the Affordable Care Act. It requires medical device and pharmaceutical companies to disclose to CMS payments or other transfers of value to clinicians.³ This led to the creation of the Open Payments database, which makes most industry payments to clinicians public.

Cardiology was among the top 3 specialties for industry payments.² Prior research found that industry payments to cardiologists were highly skewed,⁴ with most cardiologists receiving less than $10,000 per year and a small subset of cardiologists receiving more than $1 million per year.⁵ Research on industry payments to cardiologists has focused on procedural subspecialties. Prior work identified a positive relationship between the level of industry payments from coronary stent manufacturers within a hospital referral region (HRR) and the rate of coronary angiograms within the HRR.⁶ A positive association was also observed between payments from antiplatelet medication pharmaceutical companies to interventional cardiologists and the rate of cardiac catheterizations.⁷ A causal relationship between industry payments and adoption of a leadless pacemaker has also been reported.⁸ The association between industry payments across cardiology subspecialties and associated health care spending more generally has not been studied. We sought to characterize the relationship between industry payments to cardiologists and their professional fees per traditional Medicare (TM) beneficiary in 2022.

METHODS

Data Sources

We utilized the Medicare Physician and Other Practitioners by Provider public use file containing provider-level data on TM patients for whom a provider received Medicare fees in 2022.⁹ This data set includes total number of TM beneficiaries, total Medicare standardized payments (adjusted for geographic differences), provider subspeciality, gender, state of practice, and rural-urban commuting area code. It also provides demographic and clinical characteristics of patients including mean age, mean Hierarchical Condition Category (HCC) score, percentage dually eligible for Medicaid, and percentage with comorbidities as defined by the CMS Chronic Conditions Warehouse (CCW).

Total industry payments to cardiologists were identified using the CMS Open Payments data set from 2022.¹⁰ We focused on the general payments category, which excludes research payments and associated research funding. Within the general payments category, we included compensation for serving as faculty or a speaker, consulting, education, entertainment, food and beverage, gift, honoraria, and travel and lodging and excluded the acquisitions, ownership or investment interest, debt forgiveness, grant, long-term medical supply or device loan, and royalty or license categories. We summed each provider’s total relevant nonresearch industry payments as our independent variable.

Descriptive Analysis

We characterized the sample by number of cardiologists and measured average spending per TM beneficiary and general payments for all cardiologists, nonsubspecialists, and subspecialists by mean (SD) and median dollars per year. We analyzed each cardiologist’s patient population demographics (age, HCC score, dual eligibility) and medical comorbidities including alcohol/drug use disorder, tobacco use, chronic kidney disease, diabetes, ischemic heart disease, nonischemic heart failure, hypertension, atrial fibrillation, and transient ischemic attack (TIA) or stroke based on CCW definitions.

Statistical Analysis

We estimated a multivariable linear regression model with state fixed effects. The outcome was spending per TM beneficiary, and the predictor was industry payments (both 2022). The model adjusted for specialty, urban/rural status, beneficiary mean age, mean HCC score, percentage dually eligible, and total TM beneficiaries. Only cardiologists with fees for 100 or more TM beneficiaries were included. Cardiologists with no industry payments were included in the main analysis, and those with missing covariate data were excluded.

Subanalyses and Sensitivity Analyses

We performed subanalyses for generalist cardiologists and the cardiology subspecialties with the same model specifications. We conducted 2 sensitivity analyses. One model excluded cardiologists with no industry payments and winsorized for cardiologists with industry payments greater than the 95th percentile (assigned all cardiologists with industry payments above the 95th percentile as having the 95th percentile value, $16,115). The other sensitivity analysis utilized more robust adjustment for comorbidities (we did not use this as our main analysis because our main specification is more parsimonious, and the results were similar). In this model, the mean HCC score was replaced by several covariates: alcohol and/or drug use disorder, tobacco use disorder, chronic kidney disease, diabetes, ischemic heart disease, nonischemic heart failure, hypertension, atrial fibrillation, and TIA or stroke. Both models otherwise used the same specifications as the main model. Cardiologists with missing data were excluded. Analyses were conducted using Stata 16.1 (StataCorp LLC).

RESULTS

Descriptive Statistics

Our sample contained data on 26,805 cardiologists, the majority of whom (72.3%) were general cardiologists (Table 1). Of this group, 24,860 cardiologists (92.7%) had associated TM payments for at least 100 beneficiaries. A minority of cardiologists were women (14.6%) and rural providers (6.1%).

The mean (SD) payment per TM beneficiary was $244 ($318). The mean and median numbers of attributed beneficiaries per cardiologist were 936 and 784, respectively. The mean (SD) and median total industry payments to cardiologists (including those with no industry payments) in 2022 were $3958 ($17,471) and $409, respectively. Mean (SD) payments ranged from $2763 ($14,686) for general cardiologists to $8697 ($26,171) for electrophysiologists. Most cardiologists (n = 22,071; 82.4%) received at least 1 payment.

The mean (SD) age for attributed patients was 74.7 (4.1) years, and the mean (SD) HCC score was 1.8 (0.5). Alcohol or drug abuse was the least prevalent comorbidity for patients for all cardiologists (7.5%) and across cardiac subspecialties, and hypertension was the most prevalent comorbidity for patients for all cardiologists (74.3%) and across subspecialties. Prescription drug costs varied significantly by subspecialty, with heart failure cardiologists having the highest mean costs per beneficiary ($4766 vs $1320-$1340 for other subspecialties) (Table 2).

In the multivariable linear regression model, a $10,000 increase in industry payments to a cardiologist was associated with greater mean spending per TM beneficiary of $14.1 (95% CI, $11.9-$16.3) (Table 3). For the subanalyses by specialty type, the associated greater mean spending per TM beneficiary ranged from $12.3 (95% CI, $9.5-$15.2) for general cardiologists to $26.7 (95% CI, $20.5-$32.9) for interventional cardiologists and was not statistically significant for heart failure subspecialists (Table 3). Geographic analysis showed that counties with higher levels of per-user cardiologist spending or industry payments to cardiologists had similar general utilization patterns, with elevated inpatient and emergency department payments compared with national averages (Table 4).

The sensitivity analyses showed a similar relationship. However, the magnitude of the relationship between industry payments and TM spending per beneficiary was larger in the model that excluded cardiologists who did not receive any industry payments and winsorized at the 95th percentile: A $10,000 increase in industry payments was associated with $77.4 (95% CI, $67.0-$87.8) greater spending per beneficiary (Table 3). The magnitude of the relationship in the sensitivity analysis that adjusted for a broad range of comorbidities was similar to that in the main analysis ($12.8; 95% CI, $10.6-$14.9) (Table 3).

The only variables with missing data were urban/rural status (0.2% missing) and proportion of a physician’s patients with ischemic heart disease (1.6% missing).

DISCUSSION

We sought to characterize the relationship between industry payments to cardiologists and spending per TM beneficiary attributed to them. Our analysis showed a modest positive relationship that was robust to sensitivity analyses. Our sensitivity analysis excluding cardiologists without payments and winsorizing at the 95th percentile showed an association of a greater magnitude than the main model, suggesting that the relationship we observed was not driven by outlier physicians (with very high industry payments and spending per beneficiary). Subanalyses for general cardiologists, electrophysiologists, and interventional cardiologists yielded a similar relationship to the main analysis. Taken together, these findings suggest an association between industry payments and cardiology practice patterns with important implications for understanding health care utilization and expenditures.

The Open Payments data set has been used to characterize industry payments to physicians. Evidence suggests that the creation of the Open Payments data set in 2013 was associated with a decrease in meal-related payments to physicians.¹¹ Another study found an associated decrease in the proportion of physicians receiving payments across specialties from 2014 through 2018.¹² For cardiologists specifically, there was a decrease in the value of total payments, number of payments, and mean payment per physician from 2014 through 2019.⁵

The association between industry payments and health care utilization or spending is not well understood. One study using provider-level data from 2016 found that a 10% increase in industry payments was associated with 1.3% higher medical and 1.8% higher drug costs.¹³ Another study found that prescription rates of marketed relative to nonmarketed oral anticoagulants and noninsulin diabetes drugs were positively correlated with payments from the pharmaceutical company at the HRR level.¹⁴ In terms of cardiologists at the HRR level, rates of percutaneous coronary interventions were positively associated with payments from stent manufacturers in 2018.⁶ At the individual cardiologist level, higher industry payments for antiplatelet drugs in 2016 were associated with higher costs for cardiac procedures, diagnostic catheterization volumes, and rates of stenting in 2017.⁷ Finally, an analysis in a working paper suggests that industry payments from pacemaker companies from 2014 to 2019 led to a large increase in leadless pacemaker utilization when indicated and a small increase in situations for which the indication was unclear.⁸ Our study contributes by examining cardiologist-associated spending across all cardiology subspecialties rather than focusing on specific medications or procedures, allowing for broader inferences about the relationship between industry payments and health care utilization.

The economic implications of our findings are noteworthy. Currently, there are 33.4 million TM enrollees, so a $14.1 increase in spending per beneficiary is associated with $470.9 million in greater health care spending. Medicare Advantage (MA) enrollees currently number 34.1 million,¹⁵ and reimbursement for MA patients is similar to that for TM patients.¹⁶ Thus, if our results were extrapolated to this population, one might anticipate a similar magnitude in terms of health care expenditures. The TM and MA population makes up only 18.5% of the US population, and 68.7% of the US population has commercial insurance.¹⁷ Thus, even if a small proportion of the US population received cardiology services in 2022 and a small proportion of the observed relationship is causal, these findings are noteworthy.

Limitations

Our study has several limitations. First, our data on spending attributed to cardiologists include payments for TM but not MA, commercially insured, or Medicaid patients. Therefore, our findings may not be generalizable to cardiologists with a smaller proportion of TM patients in their panel.

Second, referral patterns limit the extent to which Medicare payments per beneficiary serve as a proxy for cardiologist-directed utilization. This precludes us from adjusting for hospital characteristics. Professional fees are attributed to the billing provider rather than the ordering provider, potentially underestimating the impact of ordering patterns on spending.

Third, our methodology does not allow for causal conclusions. The COVID-19 pandemic’s impact on 2020-2021 practice patterns limited our ability to use longitudinal data, restricting our analysis to 2022 data. Although we found an association between higher spending and industry payments, this could reflect pharmaceutical companies preferentially engaging with cardiologists who already prescribe newer therapies. However, it is unlikely that industry groups would persist in making such payments without financial benefit. Additionally, industry payments may lead to appropriate use of newer therapies through physician education, benefiting patients. Our analysis cannot distinguish between appropriate and inappropriate increases in spending.

CONCLUSIONS

We found a statistically significant, economically meaningful, and robust relationship between industry payments to cardiologists and cardiologist-associated spending for TM patients. A significant body of literature supports the notion that physicians’ clinical decisions may be influenced, in part, by economic factors.^18-23 Future studies should better characterize the relationship between payments from device and pharmaceutical companies and spending for those specific devices and drugs and evaluate the extent to which there is a causal relationship between industry payments and clinical decision-making in cardiology and what influence such relationships may have on patient care and outcomes.

Author Affiliations: Department of Medicine, Johns Hopkins University School of Medicine (MIE, JEM), Baltimore, MD; Mimilabs, Inc (YP), Atlanta, GA; Department of Computer Science, Emory University (YP), Atlanta, GA; Department of Health Sciences, Bouvé College of Health Sciences, Northeastern University (BP), Boston, MA.

Source of Funding: Dr Ellenbogen was supported by the Agency for Healthcare Research and Quality (AHRQ) K08HS028673. Dr Post is supported by AHRQ K01HS029278. The funding agent had no role in the study design, conduct, drafting, or decision to publish the manuscript.

Author Disclosures: Dr Ellenbogen was a consultant for Garner Health in the care navigation/physician quality assessment space at the time this research was done and subsequent to finishing the research has taken a full-time job at Garner Health and owns stock in the company. Drs Ellenbogen and Post received the above-mentioned AHRQ Career Development Awards. The remaining authors report no relationship or financial interest with any entity that would pose a conflict of interest with the subject matter of this article.

Authorship Information: Concept and design (MIE, JEM, BP); acquisition of data (MIE, YP); analysis and interpretation of data (MIE, JEM, YP, BP); drafting of the manuscript (MIE, YP); critical revision of the manuscript for important intellectual content (MIE, JEM, BP); statistical analysis (MIE, YP); and obtaining funding (MIE).

Address Correspondence to: Michael I. Ellenbogen, MD, Johns Hopkins University School of Medicine, 600 N Wolfe St, Meyer 8-134P, Baltimore, MD 21287. Email: Mellenb6@jhmi.edu.

REFERENCES

1. Facts about Open Payments data. CMS. Accessed September 3, 2024. https://openpaymentsdata.cms.gov/summary

2. Sayed A, Ross JS, Mandrola J, Lehmann LS, Foy AJ. Industry payments to US physicians by specialty and product type. JAMA. 2024;331(15):1325-1327. doi:10.1001/jama.2024.1989

3. Richardson E. Health policy brief: the Physician Payments Sunshine Act. Health Affairs. October 2, 2014. doi:10.1377/hpb20141002.272302

4. Annapureddy A, Murugiah K, Minges KE, Chui PW, Desai N, Curtis JP. Industry payments to cardiologists. Circ Cardiovasc Qual Outcomes. 2018;11(12):e005016. doi:10.1161/CIRCOUTCOMES.118.005016

5. Zhang R, Purkayastha S, Ying X, et al. Trends in industry payments to cardiologists from 2014 to 2019. Circ Cardiovasc Qual Outcomes. 2023;16(5):e009820. doi:10.1161/CIRCOUTCOMES.122.009820

6. Muluk SL, Lin GA, Anderson TS. Association of device industry payments, physician supply, and regional utilization of orthopedic and cardiac procedures. J Gen Intern Med. 2023;38(11):2501-2510. doi:10.1007/s11606-023-08101-x

7. Yanagisawa M, Blumenthal DM, Kato H, Inoue K, Tsugawa Y. Associations between industry payments to physicians for antiplatelet drugs and utilization of cardiac procedures and stents. J Gen Intern Med. 2022;37(7):1626-1633. doi:10.1007/s11606-021-06980-6

8. Everhart AO, Landon BE, Normand SLT. The persuasive and informative roles of manufacturer payments to physicians: evidence from leadless pacemaker adoption. May 2023. Accessed July 6, 2024. https://static1.squarespace.com/static/5f6e2c5c4200c7308344f24d/t/647f633875e8e43dccbc848a/1686070074064/ELN+Micra+Payment+2023_05_05.pdf

9. Medicare physician & other practitioners – by provider. CMS. Accessed July 1, 2024. https://data.cms.gov/provider-summary-by-type-of-service/medicare-physician-other-practitioners/medicare-physician-other-practitioners-by-provider

10. Dataset downloads. CMS. Accessed July 3, 2024. https://openpaymentsdata.cms.gov/datasets/download

11. Cheng S, Duan W, Zhou W. Assessing the impact of the Physician Payments Sunshine Act on pharmaceutical companies’ payments to physicians. PloS One. 2024;19(8):e0306886. doi:10.1371/journal.pone.0306886

12. Marshall DC, Tarras ES, Rosenzweig K, Korenstein D, Chimonas S. Trends in industry payments to physicians in the United States from 2014 to 2018. JAMA. 2020;324(17):1785-1788. doi:10.1001/jama.2020.11413

13. Mejia J, Mejia A, Pestilli F. Open data on industry payments to healthcare providers reveal potential hidden costs to the public. Nat Commun. 2019;10(1):4314. doi:10.1038/s41467-019-12317-z

14. Fleischman W, Agrawal S, King M, et al. Association between payments from manufacturers of pharmaceuticals to physicians and regional prescribing: cross sectional ecological study. BMJ. 2016;354:i4189. doi:10.1136/bmj.i4189

15. Medicare enrollment dashboard. CMS. Accessed October 15, 2024. https://data.cms.gov/tools/medicare-enrollment-dashboard

16. Trish E, Ginsburg P, Gascue L, Joyce G. Physician reimbursement in Medicare Advantage compared with traditional Medicare and commercial health insurance. JAMA Intern Med. 2017;177(9):1287-1295. doi:10.1001/jamainternmed.2017.2679

17. U.S. health care coverage and spending. Congressional Research Service. March 20, 2024. Accessed October 2, 2024. https://web.archive.org/web/20240919113719/https://sgp.fas.org/crs/misc/IF10830.pdf

18. Clemens J, Gottlieb JD. Do physicians’ financial incentives affect medical treatment and patient health? Am Econ Rev. 2014;104(4):1320-1349. doi:10.1257/aer.104.4.1320

19. Post B, Alinezhad F, Mukherjee S, Young GJ. Hospital-physician integration is associated with greater use of cardiac catheterization and angioplasty. Health Aff (Millwood). 2023;42(5):606-614. doi:10.1377/hlthaff.2022.01294

20. Bouck Z, Ferguson J, Ivers NM, et al. Physician characteristics associated with ordering 4 low-value screening tests in primary care. JAMA Netw Open. 2018;1(6):e183506. doi:10.1001/jamanetworkopen.2018.3506

21. Hong AS, Ross-Degnan D, Zhang F, Wharam JF. Clinician-level predictors for ordering low-value imaging. JAMA Intern Med. 2017;177(11):1577-1585. doi:10.1001/jamainternmed.2017.4888

22. Shah BR, Cowper PA, O’Brien SM, et al. Association between physician billing and cardiac stress testing patterns following coronary revascularization. JAMA. 2011;306(18):1993-2000. doi:10.1001/jama.2011.1604

23. Ellenbogen MI, Marine JE, Arbab-Zadeh A, Pathiravasan CH, Swann J, Brotman DJ. Relationship between insurance status and receipt of cardiac tests and procedures during hospitalization: a cross-sectional study. J Am Heart Assoc. 2024;13(19):e035797. doi:10.1161/JAHA.124.03579