Association of Organizational Factors and Physician Practices' Participation in Alternative Payment Models

Abstract

Importance: Consolidation among physician practices and between hospitals and physician practices has accelerated in the past decade, resulting in higher prices in commercial markets. The resulting integration of health care across clinicians and participation in alternative payment models (APMs), which aim to improve quality while constraining spending, are cited as reasons for consolidation, but little is known about the association between integration and APM participation.

Objective: To examine the association of organizational characteristics, ownership, and integration with intensity of participation in APMs among physician practices.

Design, setting, and participants: A cross-sectional descriptive study, adjusted for sampling and nonresponse weights, was conducted in US physician practice respondents to the National Survey of Healthcare Organizations and Systems conducted between June 16, 2017, and August 17, 2018; of 2333 responses received (response rate, 46.9%) and after exclusion of ineligible and incomplete responses, the number of practices included in the analysis was 2061. Data analysis was performed from April 1, 2019, to August 31, 2019.

Exposures: Self-reported physician practice characteristics, including ownership, integration (clinical, cultural, financial, and functional), care delivery capabilities, activities, and environmental factors.

Main outcomes and measures: Participation in APMs: (1) bundled payments, (2) comprehensive primary care and medical home programs, (3) pay-for-performance programs, (4) capitated contracts with commercial health plans, and (5) accountable care organization contracts.

Results: A total of 49.2% of the 2061 practices included reported participating in 3 or more APMs; most participated in pay-for-performance and accountable care organization models. Covariate-adjusted analyses suggested that operating within a health care system (odds ratio [OR] for medical group: 2.35; 95% CI, 1.70-3.25; P < .001; simple health system: 1.46; 95% CI, 1.08-1.97; P = .02; and complex health system: 1.76; 95% CI, 1.25-2.47; P = .001 relative to independent practices), greater clinical (OR, 4.68; 95% CI, 2.28-9.59; P < .001) and functional (OR, 4.24; 95% CI, 2.00-8.97; P < .001) integration, and being located in the Northeast (OR for Midwest: 0.47; 95% CI, 0.34-0.65; P < .001; South: 0.47; 95% CI, 0.34-0.66; P < .001; and West: 0.64; 95% CI, 0.46-0.91; P = .01) were associated with greater APM participation.

Conclusions and relevance: Greater APM participation appears to be supported by integration and system ownership.

Figure 1.. Distribution of Alternative Payment Models Across 2061 Physician Practices Responding to the National Survey of Healthcare Organizations and Systems

Alternative payment models included (1) bundled or episode-based payments; (2) comprehensive primary care (CPC), CPC+, and patient-centered medical homes; (3) pay-for-performance programs; (4) capitated contracts with commercial health plans; and (5) accountable care organizations (Medicare, Medicaid, and commercial). Proportions were adjusted for sampling weights.

Figure 2.. Distribution of Alternative Payment Models (APMs) by Model Type Across 2061 Physician Practices Responding to the National Survey of Healthcare Organizations and Systems Physician Practices

Alternative payment models included (1) bundled or episode-based payments; (2) comprehensive primary care (CPC), CPC Plus (CPC+), and patient-centered medical home (PCMH); (3) pay-for-performance programs; (4) capitated contracts with commercial health plans; and (5) accountable care organizations (ACOs) (Medicare, Medicaid, and commercial). Proportions were adjusted for sampling weights. Because the outcome variable measures the number of APMs each practice reported participating in, the only bar that adds up to 100% is the one identifying practices that reported participating in a single APM. When considering the other bars identifying physician practices reporting participation in multiple APMs, the proportions are overlapping and not mutually exclusive; therefore, the sum of proportions in each of these bars exceeds 100%. For example, among physician practices reporting participation in 2 APMs (representing any combination of all 5 selected APMs, 10.6% participate in bundled or episode-based payments (the least common model), 51.5% participate in ACO models, and 61.4% participate in pay-for-performance programs (the 2 most common models). A similar description could be made for physician practices reporting participation in 3, 4, or 5 APMs.

Figure 3.. Major Barriers to Use of Evidence-Based Care Delivery Innovations by Participation in Alternative Payment Models Among 2020 Physician Practices

The survey respondents were asked to choose among potential barriers to their practice's use of evidence-based care delivery innovations (eg, care transition programs, home visits, or community health workers). For each response option, the respondents specified whether that option constituted a major barrier, a minor barrier, or did not constitute a barrier for their practice.