Abstract
Background
Cardiovascular magnetic resonance (CMR) has a growing role in the diagnosis and management of cardiac disease. However, there is little recent data on the availability of CMR physicians (readers) in the United States (US).
Objective
To demonstrate the geographic proximity and accessibility of patients to CMR services and CMR physicians across the US.
Methods
Using Medicare Part B data in 2022, we analyzed the number and characteristics of CMR readers, their geographical location, and the volume of CMR scans between 2013 and 2022. CMR procedure types were identified using healthcare common procedure coding system (HCPCS) codes 75557, 75559, 75561, and 75563.
Results
Among Medicare beneficiaries in 2022, there were 48,622 CMR scans, up from 17,944 in 2013 (170.9% increase). The lowest scans and reader density were in West Virginia (125.8 procedures and 2.2 readers per million beneficiaries, respectively) and the highest in the District of Columbia (4566.5 procedures and 52.9 readers per million beneficiaries, respectively). No CMR scans were billed in Puerto Rico. Among states and territories that billed for CMR, 50.8 million U.S. citizens were located more than 50 miles from CMR readers and 18.1 million were located more than 100 miles away. Out of 991 readers, 51.9% were radiologists and 48.1% were cardiologists. The median number of scans interpreted by cardiologists was higher than radiologists across all graduation year intervals, and male and female readers interpreted a similar median number of scans. The relative proportion of female readers increased markedly when assessing physicians who graduated after 2010.
Conclusion
This study highlights significant geographic disparities and barriers to accessing CMR in the US.
Keywords: Cardiac magnetic resonance (CMR), Geographic accessibility, Geographic disparities, CMR readers, Temporal trends
Graphical abstract
1. Introduction
Cardiovascular magnetic resonance (CMR) is an important technique in the diagnosis and management of cardiac disease [1]. Because of its ability to precisely assess cardiac structure and function, it has become the gold standard for evaluating multiple cardiac pathologies [2], [3]. Its use is supported by multiple class I indications in both the American and European guidelines [4].
Despite its importance, there is little contemporary data regarding access to CMR and CMR physicians (readers) in the United States (US). Disparities in the geographical distribution of cardiologists in the US have been previously documented [5]. Additionally, one prior study highlighted regional variability in access to CMR based on 2018 Medicare data [6]. However, the availability of physicians who can perform and interpret CMR studies (cardiologists or radiologists) has not been previously studied. In addition, long-term temporal trends in the utilization of CMR have not been well studied. Thus, the aim of this analysis is to assess the geographic proximity and accessibility of patients to physicians who interpreted CMR across the US.
2. Methods
2.1. Data sources
We extracted data from 2013 to 2022 involving Medicare Part B beneficiaries from the Centers for Medicare & Medicaid Services (CMS) website. Medicare Part B encompasses services such as physician visits, outpatient care, and certain preventive services. The data primarily represent beneficiaries aged 65 and older, as well as younger individuals with qualifying disabilities or end-stage renal disease (ESRD) [7]. The data provide the total number of CMR scans, the physicians who interpreted the scans, their specialties, and the state/territory where they billed. The type of CMR procedure was identified using unique Healthcare Common Procedure Coding System (HCPCS) codes (75557, 75559, 75561, and 75563 for CMR only, CMR stress without Contrast, CMR with contrast, and CMR with contrast and stress, respectively). Physicians who billed (interpreted or read) for a given HCPCS code fewer than 11 times were censored by CMS to protect the confidentiality of Medicare beneficiaries. The physicians’ year of medical school graduation and gender were obtained from the Medicare Care Compare: Doctors and Clinicians dataset [8]. The year of graduation from medical school served as a proxy for readers’ age, allowing for comparison across specialties and gender. Physician records in the two databases were linked using the National Provider Identification (NPI) number. State-level availability of CMR per million inhabitants was calculated using population sizes from total Medicare enrollment in 2022 [7]. Because the data are publicly available and doesn’t contain de-identified patient-level information, the study was exempt from institutional review board (IRB) approval.
2.2. Outcomes of interest
Main outcome variables of interest included the total number of scans performed, the number of CMR readers, the number of studies read by each physician, the geographical location where CMR studies were billed, and the distance from each US Census Tract to the nearest CMR reader.
2.3. Statistical analysis
Inferential statistics (P values from hypothesis testing and confidence intervals) were not used as the study’s aim is mainly descriptive. The total number of readers and procedures as well as temporal trends across years, the geographic distribution of CMR readers and scans, the proportion of readers according to specialty, and the median year of graduation of readers across specialty and gender of readers, were described and displayed graphically. Analyses were performed using Stata, version 17 [9]. Additionally, R version 4.4.1 (R Foundation for Statistical Computing, Vienna, Austria) was used for distance plotting [10]. The “curl” package allowed the use of database queries to obtain CMS Data [11]. The “ggplot2″ package was used to illustrate graphical displays [12]. Geocoding was performed using the “tidygeocoder” package and the ArcGIS service [13]. Distances between each census tract centroid and the nearest CMR reader were calculated in meters using the Haversine formula (distHaversine function, geosphere R package) and converted to kilometers and miles for analysis [14].
3. Results
3.1. Overall and temporal trends
In 2022, a total of 48,622 CMR scans and a median (25th–75th) of 460 (170–1480) scans were performed for Medicare beneficiaries, with the most common type being CMR with contrast (40,979; 84.3%) (Table 1). This represented a 170.9% increase in the utilization of cardiac MRI over one decade when 17,944 studies were performed in 2013. (Fig. 3).
Table 1.
Trends in CMR procedures by HCPCS code, 2013–2022.
| Year | CMR without stress or contrast (HCPCS code 75557) | CMR stress without contrast (HCPCS Code 75559) | CMR with contrast (HCPCS code 75561) | CMR with contrast and stress (HCPCS code 75563) | Total |
|---|---|---|---|---|---|
| 2013 | 2542 (14.2%) | 70 (0.4%) | 13,207 (73.6%) | 2125 (11.8%) | 17,944 |
| 2014 | 2553 (13.1%) | 67 (0.3%) | 14,667 (75.1%) | 2258 (11.6%) | 19,545 |
| 2015 | 2544 (12.0%) | 66 (0.3%) | 16,294 (76.8%) | 2296 (10.8%) | 21,200 |
| 2016 | 2882 (12.7%) | 117 (0.5%) | 19,262 (84.8%) | 2447 (10.8%) | 22,708 |
| 2017 | 3036 (10.9%) | 136 (0.5%) | 22,029 (79.2%) | 2601 (9.4%) | 27,802 |
| 2018 | 3226 (10.7%) | 118 (0.4%) | 25,970 (86.3%) | 2769 (9.2%) | 30,083 |
| 2019 | 3478 (9.2%) | 95 (0.3%) | 30,838 (81.3%) | 3531 (9.3%) | 37,942 |
| 2020 | 2880 (8.2%) | 62 (0.2%) | 28,506 (81.5%) | 3513 (10.1%) | 34,961 |
| 2021 | 2847 (6.4%) | 35 (0.1%) | 37,028 (83.4%) | 4478 (10.1%) | 44,388 |
| 2022 | 2588 (5.3%) | 22 (0.05%) | 40,979 (84.3%) | 5033 (10.3%) | 48,622 |
| Total | 28,576 (9.2%) | 788 (0.3%) | 248,780 (80.5%) | 31,051 (10.0%) | 309,195 |
This table displays the annual counts and percentages of CMR procedures from 2013 to 2022, categorized by HCPCS codes: 75557 (CMR without contrast), 75559 (CMR stress without contrast), 75561 (CMR with contrast), and 75563 (CMR with contrast and stress). Percentages indicate the proportion of each procedure type within the total number of procedures for each year. CMR cardiovascular magnetic resonance; HCPCS, healthcare common procedure coding system.
Data are annual numbers (%) of CMR procedures billed for Medicare (2013-2022).
Fig. 3.
Trend of percent growth in number of scans and readers. The chart shows the trend of percent growth in CMR scans in red and interpreting physicians in blue from 2013 to 2022. CMR cardiovascular magnetic resonance
Over the study period, the most commonly performed CMR procedure was CMR with Contrast, which accounted for 80.5% of all procedures. The proportion of CMR without stress or contrast declined from 14.2% in 2013 to 5.3% in 2022. Despite the slight decrease in the proportion of CMR with contrast and stress (from 11.8% to 10.3%), the number of stress CMR procedures increased markedly from 2125 in 2013 to 5033 in 2022. CMR stress without contrast (likely with dobutamine) was rarely performed across the study period (0.3%) (Table 1).
In 2022, 991 readers billed Medicare for more than 10 CMR studies. These readers read 39,896 studies. Nearly 8433 studies (17.4%) were read by readers who performed less than 11 studies annually for Medicare beneficiaries (Fig. 2). Readers who read more than 10 studies included 514 radiologists (51.9%) and 477 cardiologists (48.1%). This was a substantial increase (195.8% increase) from the 335 total readers in 2013 (Fig. 3), of whom 163 (48.7%) were radiologists and 172 (51.3%) were cardiologists (Supplementary Table 1). In 2022, Cardiologists read a slightly higher proportion of total scans than radiologists, interpreting 22,369 scans (57.4%) compared to 16,611 scans (42.6%). (Fig. 1) The median number of scans read was 32 for cardiologists and 23 for radiologists (IQR: 19 to 54 and 15 to 39, respectively).
Fig. 2.
Total number of scans read by physicians in 2022 for Medicare, categorized by annual reading volume. The graph illustrates the total number of scans read by physicians in 2022 for Medicare, categorized into four reading volume groups: those who read 50 or more scans (≥50), 26–49 scans, 11–25 scans, and fewer than 11 scans (<11). Each bar represents the total scans associated with physicians in each category, emphasizing the distribution of reading volume across these groups
Fig. 1.
Total number of CMR procedures (A) and total number of readers (B) by Specialty in 2022. These pie charts show the total number and percentage of CMR procedures billed (A), and the readers who billed for them (B), categorized by cardiology and radiology specialties. Cardiology is represented in blue, and radiology is depicted in red. Data of readers reading less than 11 studies were not reported by Medicare. CMR cardiovascular magnetic resonance
3.2. Geographic distribution
In 2022, no CMR studies were billed in Puerto Rico. The states of New York (3509), Illinois (3224), and Pennsylvania (3003) had the most scans while New Mexico (57), West Virginia (56), and Wyoming (26) had the lowest number of scans performed for Medicare beneficiaries (Supplementary Fig. 1A and Supplementary Table 2). The number of scans per million beneficiaries was the highest in the District of Columbia (4566.5), Alaska (2613.4), and Minnesota (1951.4). In contrast, New Mexico (127.6), Nevada (127.3), and West Virginia (125.8) were the lowest. (Supplementary Fig. 1B and Supplementary Table 2).
In 2022, no physicians billed Medicare for more than 10 CMR studies in Wyoming and Puerto Rico. Pennsylvania (64), New York (62), and Minnesota (52) had the greatest number of readers while Alaska (2), Delaware (2), New Mexico (2), North Dakota (2), and West Virginia (1) had the fewest number of readers (Fig. 4A and Supplementary Table 3). The number of readers per million beneficiaries was the highest in the District of Columbia (52.9), Minnesota (47.6), and Connecticut (31.1) while New Mexico (4.5), Nevada (3.5), and West Virginia (2.2) had the lowest number of readers per million beneficiaries (Fig. 3B and Supplementary Table 3).
Fig. 4.
represents the number of CMR readers by state (A), the density of CMR readers by state (B), and the distance to the nearest CMR reader in the United States (C). (A) States are shaded based on the absolute number of CMR readers, with a color gradient from white (0 readers) to dark red. Dark red indicates states with the highest number of providers, while lighter shades indicate fewer readers. (B) States are colored based on reader density, with white representing no reported readers, and dark red indicating the highest density. (C) This map illustrates the distance to the nearest CMR reader across the US, with colors representing different distance ranges in miles. Blue areas indicate regions less than 5 miles away, while orange and red areas signify increasingly remote locations. Data of readers reading less than 11 studies were not reported by Medicare. CMR cardiovascular magnetic resonance,
3.3. Distance to CMR readers
The population-weighted median distance to CMR readers reading more than 10 studies was 12.61 miles (IQR: 5.4 to 31 miles), with 50.8 million U.S. citizens living more than 50 miles away from the nearest CMR reader and 18.1 million living more than 100 miles away from CMR readers. People living in Puerto Rico (1031 miles), Wyoming (116 miles), and Montana (94 miles) had the farthest distance to the nearest CMR reader (Fig. 4C, Supplementary Fig. 2).
3.4. Characteristics of readers
The majority of CMR readers graduated from medical school between 2000 and 2009, representing 405 readers (44.5%) (Fig. 5). Among those who graduated between 1967 and 1979, cardiologists read a median of 17 scans, while radiologists read a median of 25 scans, though this group consisted only of 13 radiologists and 3 cardiologists. For graduates after 2010, the median number of scans was 31.5 for cardiologists and 19 for radiologists (Fig. 6).
Fig. 5.
represents total number of readers by year of graduation in 2022. This bar chart shows the distribution of CMR readers based on their year of graduation from medical school, categorized by different graduation periods: 1967–1979, 1980–1989, 1990–1999, 2000–2009, and after 2010. The total is 909, where 82 readers had a missing year of graduation. Data of readers reading less than 11 studies were not reported by Medicare. CMR cardiovascular magnetic resonance
Fig. 6.
represents The total number of readers by specialty and year of graduation from medical school (A) and the median number of scans read among specialty and year of graduation from medical school (B). The red bars represent the number of radiology readers per graduation period, while the blue bars represent the number of cardiology readers per graduation period (A). The red bars represent the median number of scans read by radiology readers per graduation period, while the blue bars represent the median number of scans read by cardiology readers per graduation period. (B) Data of readers reading less than 11 studies were not reported by Medicare
3.5. Readers gender disparities
In 2022, 183 (20.0%) of CMR readers were females and 726 (80.0%) were males. The highest median scan numbers for both genders were among those who graduated between 1990 and 1999, with males and females reading a median of 30 and 31 scans, respectively. Among those who graduated after 2010, median reads were 24.5 for males and 26 for females, with 151 male and 54 female readers, and a total of 205 in this group (Fig. 7).
Fig. 7.
Total number of readers by gender and year of graduation from medical school (A) and Median number of scans read among gender and year of graduation from medical school (B). The red bars represent the number of female readers per graduation period, while the blue bars represent the number of male readers per graduation period (A) The red bars represent the number of female readers per graduation period, while the blue bars represent the number of male readers per graduation period. (B) Data of readers reading less than 11 studies were not reported by Medicare
4. Discussion
4.1. Overall and temporal trends
An analysis of Medicare Part B data from 2013 to 2022 revealed important trends in CMR utilization. In 2022, 48,622 CMR scans were performed for Medicare beneficiaries, a substantial increase over the volumes seen in 2013. Although CMR volume has increased significantly, US utilization remains significantly lower than in other countries. For instance, 106,780 CMR procedures in 2018 in the United Kingdom [15]. On a population-adjusted basis, U.S. utilization appears nearly 3.7 times lower. This discrepancy likely reflects earlier and broader adoption of CMR by the European Society of Cardiology [4] as well as an increase in the number of readers who can interpret the studies and wider availability of CMR centers [16].
From 2013 to 2022, CMR imaging was performed in roughly equal proportions by radiologists and cardiologists among readers who billed for more than 10 studies. Despite that, cardiologists conducted a slightly greater share of studies, accounting for 57.4% of scans compared to 42.6% by radiologists in 2022. Medicare billing for CMR stress without contrast has declined, reaching 0.05% of the overall CMR scans in 2022. This could be related to the increased utilization of vasodilator stress as the primary mode of stress in CMR.
4.2. Readers' geographic distribution
Mapping the geographical distribution of CMR readers across the U.S. highlighted access disparities among Medicare beneficiaries. In 2022, Puerto Rico and Wyoming had no CMR readers that billed more than 10 studies, a modest improvement over 2017, when Hawaii, North Dakota, New Mexico, and Nevada also lacked CMR reader access [17]. Despite this modest increase in availability, our analysis shows that millions of US citizens live >50 miles away from the nearest CMR reader, highlighting that much progress needs to be made. Similarly, a previous study on positron emission tomography (PET) centers found that 44.2 million U.S. residents live more than 100 kilometers from the nearest center, highlighting similar geographic barriers in advanced imaging accessibility [18].
States like New York, Illinois, and Pennsylvania had the highest volume of scans, while New Mexico, West Virginia, and Wyoming had the lowest. Even after adjusting for population size, West Virginia recorded the fewest scans per million beneficiaries (125.8). In terms of reader availability, Alaska, Delaware, New Mexico, and North Dakota had two readers each, and West Virginia had one CMR reader billing for Medicare. These findings underscore the need for more equitable access to CMR services among Medicare beneficiaries. While remote interpretation of studies has offered a temporary solution to address limited access, the availability of CMR centers remains a critical limiting factor.
4.3. Characteristics of readers
The year of medical school graduation was used as a proxy for readers’ age, allowing for comparison across specialties. Among older readers, radiologists had a higher median read of CMR scans, whereas, among younger readers, cardiologists had a higher median number of scan interpretations. This shift may reflect an increased emphasis on CMR training in advanced imaging fellowships and CMR's growing role in cardiovascular diagnostics [19]. Another notable finding is the number of studies interpreted by readers who reviewed fewer than 11 studies in 2022 (8433, 17.4%). This falls significantly below the Society for Cardiovascular Magnetic Resonance (SCMR) [20] Level 2 maintenance of skill criteria, which recommend the primary interpretation of at least 50 scans annually. While these data are limited to the Medicare population and may underestimate physician volume, it is unlikely that Medicare beneficiaries account for less than 30% of the total annual CMR scan volume.
Concerning gender, we found that the earliest female CMR reader graduated in 1985, compared to 1969 for males, reflecting more recent entry of women readers into the field. Among younger readers who graduated between 1990 and 1999, females read a higher median number of scans than males in the same cohort. Despite this increase, male readers remain approximately three times more prevalent than female readers in cohorts graduating after 2010, highlighting persistent gender disparities.
4.4. Barriers to CMR access in the US
The observed geographic disparities in the distribution of CMR readers and scan volumes can be attributed to multiple factors. Establishing a CMR center is costly, requiring investment in advanced imaging equipment, advanced training, and recruitment of readers. Beyond the financial barriers, training board-certified imaging specialists is challenging, particularly for CMR, where there are fewer advanced training opportunities available. To our knowledge, only 64 advanced cardiovascular imaging programs are available in the United States. Although CMR scans and readers have increased, demand still outstrips supply. The limited number of training programs restricts the number of new CMR specialists graduating each year [21]. With interest in CMR growing, establishing new advanced cardiac imaging training programs focused on CMR, expanding existing programs, and encouraging physicians to pursue CMR training could help meet the demand for qualified readers.
Despite substantial advances in CMR technology, access remains uneven across the U.S., and significant barriers persist. With the rise of cardiovascular disease and the development of new treatments for cardiomyopathies such as amyloidosis and hypertrophic cardiomyopathy, CMR plays an essential role in the diagnosis and monitoring of these conditions [2], particularly during clinical trials. Addressing limitations in CMR access has therefore become increasingly critical to ensure optimal patient care and effective evaluation of new therapies. The Society for Cardiovascular Magnetic Resonance (SCMR) has developed a strategic plan called “CMR Expansion Plan” to tackle some of the challenges described. Ongoing efforts to advance CMR utilization and physician training are essential to achieving equitable access and improving cardiovascular outcomes across different patients’ populations.
5. Limitations
Our study had some important limitations. First, CMS removes records of imagers billing for less than 11 procedures which could not be accounted for by our analysis. Second, we did not correlate the local prevalence and risk of cardiovascular disease to CMR availability. Third, the dataset only provides the addresses of physicians who billed Medicare for CMR, limiting information about centers that only have access to remote readers [20]. Fourth, because the dataset is dependent on billing archives, there could be some risk of misclassification. However, it is unlikely that these random errors would systematically bias our findings. Furthermore, Medicare data primarily reflects older adults and publicly insured individuals, limiting the inclusion of younger patients and those with private insurance who undergo CMR. This also excludes data on individuals enrolled in Medicare Advantage. Nevertheless, private companies follow Medicare’s reimbursement guidelines in patient care, and most physicians who read CMR for Medicare patients will read CMR for younger and/or privately insured patients.
6. Conclusion
Although the use of CMR has increased markedly in the US, our analysis revealed marked geographic disparities in the availability and access of CMR readers, even in 2022. Addressing these concerns can help improve cardiovascular care and access to advanced imaging across the US This task is likely to become increasingly critical as the burden of cardiovascular disease grows and as treatment options emerge for diseases where CMR plays a critical role, such as heart failure, cardiomyopathies, ischemic heart disease, and congenital heart disease.
Funding
Dr. Al-Mallah receives research support from Siemens, unrelated to this work. He is also a consultant to General Electric and Jubilant. Others. Dr. Bucciarelli-Ducci is the Chief Executive Officer (part-time) of the Society for Cardiovascular Magnetic Resonance; she received speaker fees from Circle Cardiovascular Imaging, Siemens Healthineers, GE HealthCare, Philips, and Bayer.
Author contributions
Ahmad El: Yaman: Writing – review & editing, Writing – original draft, Methodology, Conceptualization. Ahmed Sayed: Writing – review & editing, Methodology. Maria Alwan: Writing – review & editing. Asim Shaikh: Writing – review & editing. Mahmoud Al Al Rifai: Writing – review & editing. Maan Malahfji: Writing – review & editing. Dipan J. Shah: Writing – review & editing. Ibrahim M. Saeed: Writing – review & editing. Chiara Bucciarelli-Ducci: Writing – review & editing. Mouaz H. Al-Mallah: Writing – review & editing, Supervision, Methodology, Conceptualization.
Ethics approval and consent
This study was exempted from the institutional review board’s approval because no patient-level data were analyzed, and the data were available on a public platform.
Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: Mouaz H. Al-Mallah reports a relationship with Siemens, GE Healthcare, Jubilant that includes consulting or advisory. Chiara Bucciarelli-Ducci reports a relationship with Circle Cardiovascular Imaging, Siemens Healthineers, GE HealthCare, Philips, and Bayer that includes speaking and lecture fees. Chiara Bucciarelli-Ducci reports a relationship with Society for Cardiovascular Magnetic Resonance that includes board membership. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Supplementary data associated with this article can be found in the online version at doi:10.1016/j.jocmr.2025.101921.
Appendix A. Supplementary material
Supplementary material
.
References
- 1.Shah S., Chryssos E.D., Parker H. Magnetic resonance imaging: a wealth of cardiovascular information. Ochsner J. 2009;9:266–277. [PMC free article] [PubMed] [Google Scholar]
- 2.Patel A.R., Kramer C.M. Role of cardiac magnetic resonance in the diagnosis and prognosis of nonischemic cardiomyopathy. JACC Cardiovasc Imaging. 2017;10:1180–1193. doi: 10.1016/j.jcmg.2017.08.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Al-Mallah M., Alwan M., Al Rifai M., Sayed A. Cardiac positron emission tomography and other modalities for coronary artery disease assessment: A snapshot from the medicare data. Journal of Nuclear Cardiology. doi: 10.1016/j.nuclcard.2024.102030.
- 4.von Knobelsdorff-Brenkenhoff F., Schulz-Menger J. Cardiovascular magnetic resonance in the guidelines of the European Society of Cardiology: a comprehensive summary and update. J Cardiovasc Magn Reson. 2023;25:42. doi: 10.1186/s12968-023-00950-z. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Motairek I., Chen Z., Makhlouf M.H.E., Deo S., Salerno P., Mentias A., et al. Mapping geographic proximity to cardiologists across the United States. Circ Cardiovasc Qual Outcomes. 2023;16 doi: 10.1161/CIRCOUTCOMES.123.010133. [DOI] [Google Scholar]
- 6.Li J.M., Ho D.R., Husain N., Biederman R.W., Finn J.P., Fuisz A.R., et al. Regional variability of cardiovascular magnetic resonance access and utilization in the United States. J Cardiovasc Magn Reson. 2024;26 doi: 10.1016/j.jocmr.2024.101061. [DOI] [Google Scholar]
- 7.Centers for Medicaid and Medicare Services. Medicare Physician & Other Practitioners - by Provider and Service. URL: 〈https://data.cms.gov/provider-summary-by-type-of-service/medicare-physician-other-practitioners/medicare-physician-other-practitioners-by-provider-and-service〉. Last Accessed: 08/01/2024.
- 8.Centers for Medicaid and Medicare Services. Care Compare: Doctors and Clinicians National Downloadable File. URL: 〈https://data.cms.gov/provider-data/dataset/mj5m-pzi6〉. Last Accessed: 08/10/2024.
- 9.StataCorp . StataCorp LLC,; College Station, TX: 2021. Stata statistical software. [Google Scholar]
- 10.R Core Team . R Foundation for Statistical Computing,; Vienna, Austria: 2024. _R: a language and environment for statistical computing_.〈https://www.R-project.org/>〉 [Google Scholar]
- 11.Ooms J. (2024). _curl: A Modern and Flexible Web Client for R_. R package version 5.2.1, 〈https://CRAN.R-project.org/package=curl〉.
- 12.Wickham H. Springer-Verlag,; New York: 2016. ggplot2: elegant graphics for data analysis. [Google Scholar]
- 13.Cambon J. 〈https://cran.r-project.org/web/packages/tidygeocoder/tidygeocoder.pdf〉. 2025.
- 14.Hijmans R.J. Spherical Trigonometry. 〈https://cran.r-project.org/web/packages/geosphere/geosphere.pdf〉. 2024.
- 15.Keenan N., Captur G., McCann G., Berry C., Myerson S., Fairbairn T., et al. 103 UK national and regional trends in cardiovascular magnetic resonance usage - the British Society of CMR 2019 survey results. Heart. 2020;106:A80–A82. doi: 10.1136/heartjnl-2020-BCS.103. [DOI] [Google Scholar]
- 16.Betemariam T., Aleka A., Ahmed E., Worku T., Mebrahtu Y., Androulakis E., et al. Barriers to cardiovascular magnetic resonance imaging scan performance and reporting by cardiologists: a systematic literature review. Eur Heart J Imaging Methods Pract. 2025;3 doi: 10.1093/ehjimp/qyaf010. [DOI] [Google Scholar]
- 17.Goldfarb J.W., Weber J. Trends in cardiovascular MRI and CT in the U.S. medicare population from 2012 to 2017. Radiol Cardiothorac Imaging. 2021;3 doi: 10.1148/ryct.2021200112. [DOI] [Google Scholar]
- 18.Sayed A., Al-Mallah M. Geographic proximity to cardiac positron emission tomography facilities across the United States. JACC Cardiovasc Imaging. 2024;17:1513–1515. doi: 10.1016/j.jcmg.2024.07.019. [DOI] [PubMed] [Google Scholar]
- 19.Narula J., Chandrashekhar Y.S., Dilsizian V., Garcia M.J., Kramer C.M., Malik S., et al. COCATS 4 task force 4: training in multimodality imaging. J Am Coll Cardiol. 2015;65:1778–1785. doi: 10.1016/j.jacc.2015.03.024. [DOI] [PubMed] [Google Scholar]
- 20.Kim R.J., Simonetti O.P., Westwood M., Kramer C.M., Narang A., Friedrich M.G., et al. Guidelines for training in cardiovascular magnetic resonance (CMR) J Cardiovasc Magn Reson. 2018;20:57. doi: 10.1186/s12968-018-0481-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Parwani P., Chen T., Allen B., Kallianos K., Ng M.Y., Kozor R., et al. Challenges and opportunities for early career medical professionals in cardiovascular magnetic resonance (CMR) imaging: a white paper from the Society for Cardiovascular Magnetic Resonance. J Cardiovasc Magn Reson. 2023;25:65. doi: 10.1186/s12968-023-00968-3. [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
Supplementary material