Abstract
This study examines whether payments from a left ventricular assist device manufacturer to cardiologists performing percutaneous coronary intervention were associated with any use of the devices.
Mechanical circulatory support devices are increasingly used by physicians for patients receiving percutaneous coronary intervention (PCI).1 Use of intra-aortic balloon pumps (IABPs) has decreased, whereas use of intravascular microaxial left ventricular assist devices (LVADs) grew from 4.1% to 9.8% of all PCIs for acute myocardial infarction complicated by cardiogenic shock from 2015 to 2017,1 despite limited evidence demonstrating safety or clinical effectiveness2,3 and higher cost. Payments—even relatively small payments—from industry to physicians have been associated with greater likelihood of brand-name drug prescribing and medical device use.4,5 We examined whether payments from the LVAD manufacturer to cardiologists performing PCI were associated with any use of LVADs.
Methods
This cross-sectional study used data from the Open Payments program database linked to Part B claims for a 20% sample of Medicare fee-for-service beneficiaries to identify cardiologists performing at least 1 PCI in any year from 2016 to 2018 at hospitals capable of placing both LVADs and IABPs. Institutional review board approval was obtained at the University of California, San Francisco. Informed consent was not required given the use of publicly available data.
We used descriptive statistics to characterize industry payments from the single LVAD manufacturer (LVAD payments) to cardiologists, excluding research and royalty payments. We then used multivariable Poisson regression analyses with robust SEs to assess association between receiving LVAD payments vs not receiving payments and use of any LVAD (vs no LVAD use) in the same year and, in separate models, use of any LVAD in the year after payment. Payments in the same year as use may reflect physicians’ being rewarded by industry, whereas payments the following year may reflect payment influence on physician use. We also conducted these analyses among physicians who received their first payment during a year in the study (after not previously receiving payments) to assess whether new payments were associated with LVAD use and stratified by tertile of payment value among physicians who received LVAD payments to assess whether higher payments were associated with LVAD use. Covariates (physician demographics, PCI volume, placement of any IABPs, and payments from IABP manufacturers) and methodological details are provided in the eMethods in Supplement 1.
Statistical significance was defined as 2-sided 95% CIs not including 1. We performed analyses with SAS 9.4 and Stata 18.
Results
The cohort included 6398 cardiologists (median age, 52 years [IQR, 43-60 years]; 4.1% female), comprising 17 558 cardiologist-year observations. Of these cardiologists, 3796 (59.3%) placed at least 1 mechanical circulatory support device and 1871 (29.2%) at least 1 LVAD. Overall, 3586 (56.0%) received LVAD payments (median value, $164; IQR, $64-$451).
Among the 17 558 cardiologist-year observations, 6336 cardiologists received LVAD payments in individual years. Of these cardiologists, 1470 (23.2%) performed at least 1 LVAD, whereas among the 11 222 cardiologists not receiving LVAD payments, 1068 (9.5%) performed at least 1 LVAD (Table 1).
Table 1. Association of Cardiologist Characteristics With Use of Intravascular Microaxial Left Ventricular Assist Device.
| No. (%) of cardiologist-years (n = 17 558)a | No. (%) of cardiologist-years in category performing ≥1 LVADb | Any LVAD performed, ARR (95% CI)c | |
|---|---|---|---|
| Receipt of any payments from LVAD manufacturer | |||
| No | 11 222 (63.9) | 1068 (9.5) | 1 [Reference] |
| Yes | 6336 (36.1) | 1470 (23.2) | 2.04 (1.89-2.20) |
| Cardiologist characteristics | |||
| Sex | |||
| Male | 16 890 (96.2) | 2464 (14.6) | 1 [Reference] |
| Female | 668 (3.8) | 74 (11.1) | 0.86 (0.69-1.09) |
| Years since medical school graduationd | |||
| 0-9 | 791 (4.5) | 122 (15.4) | 1 [Reference] |
| 10-20 | 5629 (32.1) | 1064 (18.9) | 1.09 (0.92-1.29) |
| >20 | 10 097 (57.5) | 1272 (12.6) | 0.77 (0.65-0.92) |
| Physician census regiond | |||
| Northeast | 2946 (16.8) | 421 (14.3) | 1 [Reference] |
| Midwest | 3875 (22.1) | 567 (14.6) | 1.06 (0.93-1.20) |
| South | 7584 (43.2) | 1097 (14.5) | 1.01 (0.91-1.13) |
| West | 3144 (17.9) | 452 (14.4) | 1.03 (0.91-1.17) |
| Volume of PCIs per year, quartilee | |||
| 1 | 3372 (19.2) | 122 (3.6) | 1 [Reference] |
| 2 | 4230 (24.1) | 344 (8.1) | 2.05 (1.67-2.51) |
| 3 | 4585 (26.1) | 648 (14.1) | 3.36 (2.77-4.07) |
| 4 | 5371 (30.6) | 1424 (26.5) | 5.91 (4.91-7.13) |
| Performed any IABPs | |||
| No | 13 143 (74.9) | 1604 (12.2) | 1 [Reference] |
| Yes | 4415 (25.1) | 934 (21.2) | 1.35 (1.26-1.46) |
| Payments from IABP manufacturers | |||
| No | 17 186 (97.9) | 2459 (14.3) | 1 [Reference] |
| Yes | 372 (2.1) | 79 (21.2) | 1.17 (0.96-1.43) |
Abbreviations: ARR, adjusted risk ratio; IABP, intra-aortic balloon pump; LVAD, intravascular microaxial left ventricular assist device; PCI, percutaneous coronary intervention.
Column percentages, or percentage of observations in each category.
Row percentages, or percentage of observations within each category performing LVAD.
Denominator is 6398 cardiologists, of whom 5186 (81.1%) were included in the data set for all 3 years, 788 (12.3%) for 2 years, and 424 (6.6%) for 1 year.
Years since medical school graduation was missing for 1041 cardiologists. Years since medical school graduation data are based on the final year of available data per physician. Census region was missing for 9 cardiologists.
Quartile volume of data for PCIs per year is as follows: quartile 1, less than or equal to 3; quartile 2, 4 to 7; quartile 3, 8 to 13; and quartile 4, greater than or equal to 14.
Adjusted risk ratios are adjusted for physician sex, postgraduate year (categorized as <10, 10-20, and >20 years in training, as well as missing), census region, volume of PCIs per year (categorized in quartiles), study year (2016, 2017, or 2018), placement of any IABPs, and payments from IABP manufacturers.
In multivariable analyses, receipt of LVAD payments (vs no receipt) was associated with any LVAD use in the year of payment (adjusted risk ratio [ARR], 2.04; 95% CI, 1.89-2.20) and in the subsequent year (ARR, 1.90; 95% CI, 1.74-2.07) (Table 2). Results were consistent among cardiologists who received their first LVAD payment after not receiving payment, and the highest ARR was observed in the highest tertile of LVAD payment values.
Table 2. Sensitivity Analyses of Payments to Cardiologists With Intravascular Microaxial LVAD Use in the Year of Payment and Year After Paymenta.
| % (No./total No.) performing any LVAD during index year | Any LVAD performed during index year, ARR (95% CI)b | % (No./total No.) performing ≥1 LVAD in year after index year | Any LVAD performed in year after index year, ARR (95% CI)c | |
|---|---|---|---|---|
| Receipt of any LVAD payments | ||||
| No | 9.5 (1068/11 222) | 1 [Reference] | 11.5 (829/7179) | 1 [Reference] |
| Yes | 23.2 (1470/6336) | 2.04 (1.89-2.20) | 24.9 (991/3981) | 1.90 (1.74-2.07) |
| Receipt of any LVAD payments among physicians not receiving payments in prior year | ||||
| No | 8.1 (448/5530) | 1 [Reference] | 9.6 (533/5530) | 1 [Reference] |
| Yes | 11.8 (195/1649) | 1.26 (1.08-1.48) | 18.0 (296/1649) | 1.65 (1.45-1.89) |
| Receipt of LVAD payments by tertile of payment value, $ | ||||
| None | 9.5 (1068/11 222) | 1 [Reference] | 11.5 (829/7179) | 1 [Reference] |
| 10.30-61.60 | 18.7 (391/2091) | 1.73 (1.55-1.92) | 19.2 (273/1420) | 1.51 (1.33-1.71) |
| 61.70-182.00 | 18.4 (397/2155) | 1.69 (1.52-1.87) | 23.8 (303/1273) | 1.87 (1.66-2.10) |
| 182.10-99 045.10 | 32.6 (682/2090) | 2.68 (2.45-2.93) | 32.2 (415/1288) | 2.33 (2.09-2.59) |
Abbreviations: ARR, adjusted risk ratio; IABP, intra-aortic balloon pump; LVAD, intravascular microaxial left ventricular assist device.
Risk ratios are adjusted for physician sex, postgraduate year (categorized as <10, 10-20, and >20 years in training, as well as missing), census region, volume of percutaneous coronary interventions per year (categorized in quartiles), study year (2016, 2017, or 2018), placement of any IABPs, and payments from IABP manufacturers. Index year refers to the year of payment.
Denominator is 17 558 cardiologist-years, constructed from 6398 cardiologists meeting inclusion criteria (most cardiologists were included in multiple years).
Denominator is 11 160 cardiologist-years, constructed from 5974 cardiologists with 2 contiguous years of claims data.
Discussion
Intravascular microaxial LVAD manufacturer payments (often modest) to cardiologists were associated with increased use of LVADs by cardiologists who perform PCIs. It is concerning that payments from the manufacturer to cardiologists may be associated with increased use of more expensive medical devices not demonstrated to be more safe or effective.2,3
Study limitations include observational design and inability to determine causality; cardiologists preferentially placing LVADs may seek opportunities to receive payments. Use of a 20% Medicare fee-for-service data sample means other Medicare beneficiaries and those with other insurance were not included. Although attention has been paid to marketing of pharmaceuticals,4,6 marketing payments to physicians for medical devices, including LVADs, warrant further scrutiny, given potential risk and high cost.
Section Editors: Kristin Walter, MD, and Jody W. Zylke, MD, Deputy Editors; Karen Lasser, MD, Senior Editor.
eMethods. Methodological Details
Data Sharing Statement
References
- 1.Dhruva SS, Ross JS, Mortazavi BJ, et al. Use of mechanical circulatory support devices among patients with acute myocardial infarction complicated by cardiogenic shock. JAMA Netw Open. 2021;4(2):e2037748. doi: 10.1001/jamanetworkopen.2020.37748 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Dhruva SS, Ross JS, Mortazavi BJ, et al. Association of use of an intravascular microaxial left ventricular assist device vs intra-aortic balloon pump with in-hospital mortality and major bleeding among patients with acute myocardial infarction complicated by cardiogenic shock. JAMA. 2020;323(8):734-745. doi: 10.1001/jama.2020.0254 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Miller PE, Bromfield SG, Ma Q, et al. Clinical outcomes and cost associated with an intravascular microaxial left ventricular assist device vs intra-aortic balloon pump in patients presenting with acute myocardial infarction complicated by cardiogenic shock. JAMA Intern Med. 2022;182(9):926-933. doi: 10.1001/jamainternmed.2022.2735 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Mitchell AP, Trivedi NU, Gennarelli RL, et al. Are financial payments from the pharmaceutical industry associated with physician prescribing? a systematic review. Ann Intern Med. 2021;174(3):353-361. doi: 10.7326/M20-5665 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.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 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.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 [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods. Methodological Details
Data Sharing Statement