Author + information
- Received April 2, 2015
- Revision received July 14, 2015
- Accepted August 4, 2015
- Published online October 20, 2015.
- Salim S. Virani, MD, PhD∗,†,‡,§∗ (, )
- Thomas M. Maddox, MD, MSc‖,¶,
- Paul S. Chan, MD, MSc#∗∗,
- Fengming Tang, MS#,
- Julia M. Akeroyd, MPH∗,
- Samantha A. Risch, BS††,
- William J. Oetgen, MD, MBA††,
- Anita Deswal, MD, MPH†,‡,
- Biykem Bozkurt, MD, PhD†,‡,
- Christie M. Ballantyne, MD‡,§ and
- Laura A. Petersen, MD, MPH∗
- ∗Health Policy, Quality & Informatics Program, Michael E. DeBakey Veterans Affairs Medical Center Health Services Research and Development Center for Innovations, and Section of Health Services Research, Department of Medicine, Baylor College of Medicine, Houston, Texas
- †Section of Cardiology, Michael E. DeBakey Veterans Affairs Medical Center, Houston, Texas
- ‡Section of Cardiology, Department of Medicine, Baylor College of Medicine, Houston, Texas
- §Center for Cardiovascular Disease Prevention, Houston Methodist DeBakey Heart and Vascular Center, Houston, Texas
- ‖Veterans Affairs Eastern Colorado Health Care System, Denver, Colorado
- ¶Division of Cardiology, University of Colorado Denver, Denver, Colorado
- #Saint Luke's Mid America Heart Institute, Kansas City, Missouri
- ∗∗University of Missouri, Kansas City, Missouri
- ††American College of Cardiology, Washington, DC
- ↵∗Reprint requests and correspondence:
Dr. Salim S. Virani, Michael E. DeBakey Veterans Affairs Medical Center, 2002 Holcombe Boulevard, Houston, Texas 77030.
Background The current number of physicians will not be sufficient to accommodate 30 to 40 million Americans expected to secure health coverage with Affordable Care Act implementation. One proposed solution is to use advanced practice providers (APPs) (nurse practitioners and physician assistants).
Objectives This study sought to determine whether there were clinically meaningful differences in the quality of care delivered by APPs versus physicians in a national sample of cardiology practices.
Methods Within the American College of Cardiology’s PINNACLE Registry, we compared quality of coronary artery disease (CAD), heart failure, and atrial fibrillation care delivered by physicians and APPs for outpatient visits between January 1, 2012, and December 31, 2012. We performed hierarchical regression adjusting for provider sex; panel size; duration of participation in registry; and patient’s age, sex, insurance, number of outpatient visits, history of hypertension, diabetes, myocardial infarction, and percutaneous coronary intervention or coronary artery bypass grafting in the preceding 12 months.
Results We included 883 providers (716 physicians and 167 APPs) in 41 practices who cared for 459,669 patients. Mean number of patients seen by APPs (260.7) was lower compared with that seen by physicians (581.2). Compliance with most CAD, heart failure, and atrial fibrillation measures was comparable, except for a higher rate of smoking cessation screening and intervention (adjusted rate ratio: 1.14; 95% confidence interval: 1.03 to 1.26) and cardiac rehabilitation referral (rate ratio: 1.40; 95% confidence interval: 1.16 to 1.70) among CAD patients receiving care from APPs. Compliance with all eligible CAD measures was low for both (12.1% and 12.2% for APPs and physicians, respectively) with no significant difference. Results were consistent when comparing practices with both physicians and APPs (n = 41) and physician-only practices (n = 49).
Conclusions Apart from minor differences, a collaborative care delivery model, using both physicians and APPs, may deliver an overall comparable quality of outpatient cardiovascular care compared with a physician-only model.
- cardiovascular disease
- chronic disease care
- nurse practitioners
- performance measures
- physician assistants
With implementation of the Affordable Care Act (ACA) (1), about 20 million Americans secured health care coverage between October 1, 2013 and the early part of 2014 (2). The Congressional Budget Office estimates that the number of Americans who will secure coverage as a result of the ACA implementation will likely increase to 34 million in 2016 (3). A large number of these previously uninsured Americans will likely experience many chronic diseases, including hypertension, diabetes, and cardiovascular disease (CVD).
The Association of American Medical Colleges estimates that the United States will be short 45,000 primary care physicians and 45,000 specialists by 2020 and that combined total of 90,000 will jump to >130,000 by 2025, with a shortage of nearly 65,000 specialist physicians (4,5). An aging U.S. population (especially those >65 years of age) with a high prevalence of chronic diseases could put further pressure on access to physician providers. Complicating this is the projection that nearly one-third of all physicians will retire in the next decade (6). Although it remains debatable (7), full implementation of the ACA could put pressure on the remaining physician workforce.
One of the proposed ways to deal with the projected physician shortage is to allow advanced practice providers (APPs) (i.e., nurse practitioners [NPs] and physician assistants [PAs]) to function independently. Although this nontraditional but innovative model of health care delivery has been compared in studies performed in a primary care setting in single, academic institutions (8–15), the applicability of those findings to an outpatient cardiology practice setting is not known.
Our aim was to determine whether there were any clinically meaningful differences in the quality of CVD care provided by APPs, compared with care provided by physicians in a national sample of cardiology practices. We hypothesized that there would be no clinically meaningful differences in the quality of CVD care between patients receiving care from physicians and those receiving care from APPs. As a secondary aim, we also determined whether there were any differences in the quality of care provided by practices using both physicians and APPs compared with physician-only practices.
We used the American College of Cardiology’s (ACC) National Cardiovascular Data Registry PINNACLE (Practice Innovation and Clinical Excellence) registry for the current study. PINNACLE is a prospective U.S.-based outpatient cardiac quality improvement registry with voluntary participation from academic and nonacademic cardiology practices (16–18).
Participation in this initiative is voluntary, with patient data collected at the point of care for various cardiovascular conditions, including coronary artery disease (CAD), heart failure (HF), and atrial fibrillation (AF). Performance metrics for CAD, HF, and AF are endorsed by the ACC, the American Heart Association, and/or the American Medical Association-Physician Consortium for Performance Improvement. Data in the PINNACLE registry are directly extracted using algorithms mapped to data in electronic health records (19–21). Quality control is ensured by the use of a standard data collection tool with written definitions, uniform data entry and transmission requirements, and data quality checks (18,22).
Figure 1 provides a flow diagram of study participants and exclusions. For the purpose of current analyses, we evaluated CAD, HF, or AF patients enrolled in the PINNACLE registry with an outpatient cardiology clinic visit between January 1, 2012 and December 31, 2012 among 92 U.S. practices. For each patient visit, the National Provider Identifier number of the treating practitioner was collected at the point of care. The National Provider Identifier number was searched within the National Plan and Provider Enumeration System database to identify the provider category (physician or APP) (23). Each patient was then assigned to either a physician or an APP (NP or PA). In the case of a patient having seen >1 provider during the study period, the patient was assigned to whichever provider they saw more often during the study interval.
We first identified 659,734 eligible patients enrolled in the PINNACLE registry with an outpatient visit in 2012 for CAD, HF, or AF. We then excluded 10,825 patients (10,292 patients seeing providers with missing credentials and 553 patients seen by providers with <20 patients). With these exclusions, our study cohort included 648,909 patients receiving care in 90 practices with 1,234 providers. Of these, 459,669 patients (70.8%) were receiving care in practices with both physicians and APP, whereas 189,240 patients (29.2%) were receiving care in physician-only practices. Our primary analyses compared 43,351 patients receiving care from APP with 459,669 patients receiving care from physicians in practices with both physicians and APP. We also performed secondary analyses comparing patients receiving care in practices with physicians and APPs (n = 189,240) with patients receiving care in physician-only practices (n = 459,669). There were no patients receiving care in practices with only advanced practice providers.
Outcome and analyses
Our outcome of interest included compliance with CAD, HF, or AF measures among eligible patients as per the ACC/American Heart Association/Physician Consortium for Performance Improvement performance measures (19–21). For CAD, these measures included beta-blocker use in CAD patients with prior myocardial infarction (MI), antiplatelet use, smoking cessation screening and intervention, cardiac rehabilitation referral, symptoms and activity assessment, angiotensin-converting enzyme inhibitor (ACE-I) or angiotensin receptor blocker (ARB) use in CAD patients with left ventricular ejection fraction <40%, lipid-lowering therapy, and blood pressure assessment. We also evaluated the proportion of CAD patients meeting all eligible measures among the previously mentioned CAD performance measures.
For HF, the performance measures included ACE-I/ARB use, beta-blocker use among patients with left ventricular ejection fraction <40%, and the use of both ACE-I or ARB and beta-blocker in patients with left ventricular ejection fraction <40%. For AF, we assessed the use of anticoagulation (warfarin, dabigatran, or rivaroxaban) in patients with nonvalvular AF and CHADS2 score ≥2. For each measure, patients were excluded from the denominator for an individual performance measure if they had a documented contraindication to that particular therapy (e.g., beta-blocker use after MI in a patient with hypotension; patient refusal to take warfarin for AF).
We first compared demographic and clinical characteristics of patients receiving care from physician and APPs using chi-square statistic or Student t test for categorical and continuous variables, respectively. In the primary analyses, we compared compliance with each of the 8 individual CAD performance measures, and compliance with all eligible CAD measures between APPs and physician providers. The unit of analysis was the individual provider. The numerator for these rates was the total number of patients seen by a provider (physician or advanced practice) that met the performance measure (at any visit over the 12-month study period), and the denominator was the total number of patients seen by a provider group eligible for that performance measure.
To assess the association between provider type and performance rate, we constructed 3-level (with patients nested within practitioners and practitioners nested within clinic sites) hierarchical models, adjusted for provider sex; provider panel size in 2012; duration of participation in PINNACLE registry; and patient’s age, sex, insurance, number of outpatient visits, history of hypertension, diabetes, MI, and percutaneous coronary intervention (PCI) or coronary artery bypass grafting (CABG) in the preceding 12 months. These variables were chosen for the adjustment to account for provider (sex) or patient characteristics (insurance status) shown in previous studies (24–26) to be associated with quality of CVD care, characteristics that would allow more opportunities for a provider to work on risk factor modification in a CVD patient (e.g., a lower panel size, more frequent patient visits) (27), or patient characteristics that would increase a patient’s complexity or illness burden (e.g., history of concomitant hypertension, diabetes, MI, or a history of PCI or CABG in the preceding 12 months). Modified Poisson models were used to estimate rate ratio (28,29). For these analyses, the physician provider category was used as the referent category. The resultant rate ratios, therefore, can be interpreted as the rate of compliance with a performance measure among APPs divided by the rate of compliance with a performance measure among physician providers adjusting for practice characteristics; provider sex; provider panel size in 2012; duration of participation in PINNACLE registry; and patient’s age, sex, insurance, number of outpatient visits, history of hypertension, diabetes, MI, and PCI or CABG in the preceding 12 months. A rate ratio >1 would signify a higher performance on a performance measure by APPs compared with physician providers and vice versa. Similar analyses were performed for the HF and AF measures.
We also performed secondary analyses comparing compliance with CAD, HF, or AF measures among practices with both physicians and APPs (41 practices with 459,669 patients) and practices with physician-only providers (49 practices with 189,240 patients). These analyses were performed at the level of individual practice. The resultant rate ratios can be interpreted as the rate of compliance with a performance measure among practices with both physician and APPs divided by the rate of compliance with a performance measure among practices with physician-only providers adjusted for provider sex; provider panel size in 2012; duration of participation in PINNACLE registry; and patient’s age, sex, insurance, number of outpatient visits, history of hypertension, diabetes, MI, and PCI or CABG in the preceding 12 months. All analyses were performed using SAS version 9.3 (SAS Institute, Cary, North Carolina).
Our primary analyses (Figure 1) included 459,669 patients receiving care in practices with both physician and APPs (17.5% in Northeast, 32.6% in Midwest, 36% in South, and 13.9% in West). Of these patients, a total of 416,138 patients were receiving care from 716 physicians and 43,531 patients were receiving care from 167 APPs (141 NPs, 26 PAs). The number of patients with CAD, HF, or AF was 414,403, 137,747, and 144,311, respectively.
Baseline characteristics of patients receiving care from physician and APPs are shown in Table 1. Patients receiving care from APPs were slightly older, more often of female sex and African American race, more likely to have no insurance, and had a higher prevalence of diabetes. They were less likely to use tobacco and had a lower prevalence of MI, history of recent PCI or CABG, or CHADS2 score ≥2 (among AF patients) compared with patients receiving care from physician providers. Patients receiving care from APPs had a lower number of visits in the 12 months before their index visit. Although most of these differences were statistically significant given the large sample size, absolute differences were numerically small.
Most of the physicians (91.3%) were males as opposed to APPs (89.8% females [Table 2]). In addition, physicians had significantly larger patient panel sizes compared with APPs. The duration of participation in the PINNACLE registry was not significantly different between physicians and APPs.
Comparison of compliance with CAD performance measures among APPs and physician providers is described in Table 3, along with the unadjusted and adjusted rate ratios. Numerically, the performance on most of these measures was comparable between physician and APPs. In adjusted analyses, APPs were more likely to meet the performance measures of smoking cessation and cardiac rehabilitation referral among their CAD patients compared with physician providers. Importantly, compliance with all eligible CAD measures was low for both provider categories (12.1% and 12.2% for APPs and physician providers, respectively), which was not statistically different.
Comparison of compliance with HF and AF measures is described in Table 4. For HF, there were no differences in performance measure rates for ACE-I/ARB and beta-blocker use or both measures among HF patients. Similarly, anticoagulant agent use among high-risk AF patients with CHADS2 score ≥2 (although low for both provider categories) was not different between the 2 provider types.
We also performed stratified analyses comparing the quality of CVD care between physicians and APPs in states that allowed independent scope of practice for APPs in 2012 (30,31) and those states that did not allow independent scope of practice in 2012 (Online Tables 1 and 2). These results were qualitatively comparable with the main study results.
We also performed secondary analyses at the practice level comparing practices with both physician and APPs with physician-only practices. These analyses showed that among patients with CAD (Table 5) or HF or AF (Table 6), there were no significant differences in compliance with performance measures among practices with both physicians and APPs (41 practices with 459,669 patients) compared with physician-only practices (49 practices with 189,240 patients).
In these analyses from the ACC’s National Cardiovascular Data Registry PINNACLE registry, several observations can be made. First, the quality of outpatient CVD care measured by compliance with performance measures by APPs was equivalent to that of physician providers on most measures, and was even marginally better for APPs on some CAD measures. Second, the quality of outpatient CVD care delivered by practices with both physicians and APPs was equivalent to that delivered by physician-only practices. Last, compliance with global CAD performance measures (i.e., all eligible measures) was low for both provider categories. It is important to note that given the large sample size of our study population, some of the differences that seem to be statistically significant may not be clinically relevant.
Implications for care delivery under the ACA
Our findings indicate that a collaborative care delivery model that uses both physicians and APPs seems to provide a care quality that is equivalent to a physician-only model (Central Illustration). This is important because millions of Americans will access care under the ACA (1,2). Our results show that the use of APPs to improve health care access for these Americans is a viable option as far as meeting basic outpatient CVD performance metrics is concerned. It is important to note that our results do not apply to inpatient clinical settings where both complexity and dynamics of care are likely different from the outpatient setting. It should also be noted that our findings apply only to large integrated cardiology practices that include advanced practice clinicians and are not generalizable to other settings.
Our results also have health care policy implications. First, there has been a debate regarding the scope-of-practice laws governing APPs. Despite calls from the Institute of Medicine (32) and the Federal Trade Commission (33–35) to address the scope-of-practice laws (which fall under the purview of states), these laws remain restrictive. For example, currently only 21 states and the District of Columbia allow advanced practice NPs to independently diagnose, treat, and refer patients and prescribe medications without physician supervision. Seventeen states limit APP independence by allowing NPs to diagnose and treat but not to prescribe medications without physician involvement, and 12 states require physician oversight of NPs at all levels (diagnosis, treatment, and medication prescription) (30). Our results provide objective data to inform these discussions and should be reassuring that the quality of uncomplicated outpatient CVD care is at least equivalent between APPs and physicians even in states with independent scope-of-practice laws for APPs (albeit, in physician-owned practices), and between practices with both APPs and physicians compared with physician-only practices.
Our results may have implications in terms of how ACA is implemented. Title V of the ACA (pages 1244 to 1500) (1) calls for increasing the supply of the health care workforce via support from federal grants in education, training, and continuous education activities. Although Title V broadly calls for an increase in government funding for physicians, PAs, nurses, and NPs, specific allocation of funds among these various constituents is not described. Instead, the ACA calls for the creation of a National Health Care Workforce Commission to guide Congress, the President, states, and localities by making recommendations on the appropriate use of these resources on the basis of research evaluating effectiveness of various care models. Our results indicate that potential return on investment of these funds for training and education of APPs in addition to physicians will likely expand the network of clinicians who can meet the expected growing shortages in the delivery of chronic disease care.
Our findings that only 1 in 8 patients with CAD met all eligible CAD measures (whether they received care from APPs or physicians) is concerning and has been described in previous published data (36). Because these performance measures were created to include treatments shown to improve both cardiovascular morbidity and mortality, our results highlight a need to improve global performance on all eligible measures in patients with CAD.
These analyses only assessed one aspect of the quality of care (i.e., effectiveness of cardiac care delivery). It is quite possible that there could be differences between physician and APPs when other domains of quality of care (e.g., safety, efficiency, or patient-centeredness) are assessed. Our study assessed the outpatient CVD care for patients receiving care in cardiology practices enrolled in the PINNACLE registry. These results, therefore, may not apply to other cardiology or primary care practices. It is also important to note our results may not be generalizable to APP-only practices because there were no practices with only APPs. For some measures (e.g., aspirin therapy), over-the-counter use of medications not captured in the patient’s medical record could have led to an under-ascertainment. Although we adjusted for several patient, provider, and practice characteristics to account for selection bias, there could still be unmeasured confounding for which we could not account. Last, these analyses do not account for collaboration between physician and APPs, which can happen in a practice while taking care of a complex patient. Indeed, this collaboration is something that would be desirable as more patients seek access to health care.
A collaborative care delivery model using both physicians and APPs may deliver quality of outpatient CVD care comparable with a physician-only model. These results have implications for how resources are allocated for cardiovascular care delivery as more Americans access care under the ACA. Only 1 in 10 patients with CAD regardless of the provider type met all eligible performance measures. These results highlight a need to improve global performance on all eligible measures in CAD patients.
COMPETENCY IN SYSTEMS-BASED PRACTICE: The quality and effectiveness of outpatient care provided through collaboration of physicians with advanced practice providers can be comparable with that delivered solely by physicians.
TRANSLATIONAL OUTLOOK: Future studies should address whether other quality domains, such as the safety, efficiency, or patient-centeredness of care, can be enhanced through collaboration between physicians and APPs managing chronic comorbidities associated with cardiovascular disease.
Dr. Virani is supported by a Department of Veterans Affairs Health Services Research and Development (HSR&D) Service Career Development Award (09-028); American Heart Association Beginning Grant-in-Aid; and the American Diabetes Association Clinical Science and Epidemiology award. Dr. Maddox is supported by a Department of Veterans Affairs HSR&D Service Career Development Award (08-021). Dr. Chan receives funding (1R01HL123980) from the National Heart, Lung, and Blood Institute. Dr. Deswal has received research support from Novartis as site PI of a multicenter clinical trial. Dr. Ballantyne has received grant/research support (all paid to the institution, not individual) from Abbott Diagnostic, Amarin, Amgen, Bristol-Myers Squibb, Eli Lilly, Esperion, Merck, Novartis, Pfizer, Regeneron, Roche Diagnostic, Sanofi-Synthelabo, and Takeda Development Centers of America, National Institutes of Health, American Diabetes Association, and American Heart Association; is a consultant to Abbott Diagnostic, Amarin, Amgen, AstraZeneca, Eli Lilly, Esperion, Genzyme, Matinas BioPharma Inc., Merck, Novartis, Pfizer, Regeneron, and Sanofi-Synthelabo; and has a provisional patent for “Biomarkers to Improve Prediction of Heart Failure Risk” (patent no. 61721475), filed by Baylor College of Medicine and Roche. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- Abbreviations and Acronyms
- Affordable Care Act
- American College of Cardiology
- angiotensin-converting enzyme inhibitor
- atrial fibrillation
- advanced practice practitioner
- angiotensin receptor blocker
- coronary artery bypass grafting
- coronary artery disease
- cardiovascular disease
- heart failure
- myocardial infarction
- nurse practitioners
- physician assistants
- percutaneous coronary intervention
- Received April 2, 2015.
- Revision received July 14, 2015.
- Accepted August 4, 2015.
- American College of Cardiology Foundation
- ↵U.S. Department of Health and Human Services. Available at: http://www.hhs.gov/healthcare/rights/law. Accessed February 12, 2015.
- ↵Insurance Coverage Provisions of the Affordable Care Act - CBO’s January 2015 Baseline Washington, DC: Congressional Budget Office, January 2015. Available at: http://www.cbo.gov/sites/default/files/cbofiles/attachments/43900-2015-01-ACAtables.pdf. Accessed February 17, 2015.
- ↵Association of American Medical Colleges. Physician supply and demand through 2025: key findings. Available at: https://www.aamc.org/download/153160/data/physician_shortages_to_worsen_without_increases_in_residency_tr.pdf. Accessed February 12, 2015.
- U.S. Department of Health and Human Services, Health Resources and Services Administration; Exhibit 51, Baseline FTE Supply Projects of Active Physicians, and Exhibit 52, Baseline Physician Requirements Projections, December 2008. Available at: http://bhpr.hrsa.gov/healthworkforce/reports/physwfissues.pdf/. Accessed February 12, 2015.
- ↵Help wanted: more U.S. doctors projections indicate America will face shortage of M.D.s by 2020. AAMC Center for Workforce Studies, September 27, 2010. Available at: https://www.aamc.org/download/82874/data/helpwanted.pdf. Accessed February 17, 2015.
- ↵The Commonwealth Fund. Available at: http://www.commonwealthfund.org/publications/issue-briefs/2015/feb/how-will-aca-affect-use-health-services. Accessed April 1, 2015.
- Lenz E.R.,
- Mundinger M.O.,
- Hopkins S.C.,
- et al.
- Aparasu R.R.,
- Hegge M.
- Hooker R.S.,
- McCaig L.F.
- Lenz E.R.,
- Mundinger M.O.,
- Kane R.L.,
- et al.
- Hira R.S.,
- Kennedy K.,
- Nambi V.,
- et al.
- Chan P.S.,
- Oetgen W.J.,
- Buchanan D.,
- et al.
- American College of Cardiology Foundation/American Heart Association/American Medical Association–Physician Consortium for Performance Improvement
- Bonow R.O.,
- Bennett S.,
- Casey D.E. Jr..,
- et al.
- Estes N.A. 3rd.,
- Halperin J.L.,
- Calkins H.,
- et al.
- Messenger J.C.,
- Ho K.K.,
- Young C.H.,
- et al.,
- NCDR Science and Quality Oversight Committee Data Quality Workgroup
- ↵National Plan and Provider Enumeration System. Available at: https://nppes.cms.hhs.gov/NPPES/Welcome.do Accessed February 17, 2015.
- Smolderen K.G.,
- Spertus J.A.,
- Tang F.,
- et al.
- Zou G.
- Greenland S.
- ↵The American Association of Nurse Practitioners. Available at: http://www.aanp.org. Accessed June 2, 2015.
- Institute of Medicine
- ↵Federal Trade Commission. Letter from DeSanti SS, Farrell J, Feinstein RA to Jeanne Kirkton, Missouri House of Representatives, March 27, 2012.
- Federal Trade Commission. Letter from DeSanti SS, Farrell J, Feinstein RA to Gary Odom, Tennessee House of Representatives, September 28, 2011.
- Federal Trade Commission. Letter from Gavil AI, Feinstein RA, Shelanski H to Theresa Conroy, Connecticut House of Representatives, March 19, 2013.