Increases in Medicare Spending and Use after Private Equity Acquisition of Retina Practices.

PURPOSE: Private equity (PE) firms increasingly are acquiring physician practices in the United States, particularly within procedural-based specialties such as ophthalmology including retina. To date, the potential impact of ophthalmology practice acquisitions remains unknown. We evaluated the association between PE acquisition and Medicare spending and use for common retina services. DESIGN: Retrospective difference-in-differences analysis using the 20% Medicare fee-for-service claims dataset from January 1, 2015, through December 31, 2019. PARTICIPANTS: Eighty-two practices acquired by PE during the study period and matched control practices. METHODS: We used novel data on PE acquisitions of retina practices linked to the 20% sample Medicare claims data. Retina practices acquired by PE between 2016 and 2019 were matched to up to 3 non-PE (control) practices based on characteristics before acquisition. Private equity-acquired practices were compared with matched control practices through 6 quarters after acquisition using a difference-in-differences event study design. Data analyses were performed between August 2022 and April 2023. MAIN OUTCOME MEASURES: Medicare spending and use of common retina services. RESULTS: Relative to control practices, PE-acquired retina practices increased the use of higher-priced anti-vascular endothelial growth factor (VEGF) agents including aflibercept, which differentially increased by 6.5 injections (95% confidence interval, 0.4-12.5; P = 0.03) per practice-quarter, or 22% from baseline. As a result, Medicare spending on aflibercept differentially increased by $13 028 per practice-quarter, or 21%. No statistically significant differences were found in use or spending for evaluation and management visits or diagnostic imaging. CONCLUSIONS: Private equity acquisition of retina practices are associated with modest increases in the use of higher-priced anti-VEGF drugs like aflibercept, leading to higher Medicare spending. This finding highlights the need to monitor the influence of PE firms' financial incentives over clinician decision-making and the appropriateness of care, which could be swayed by strong economic incentives. FINANCIAL DISCLOSURE(S): Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.

Private equity in ophthalmology: taking the long view.

PURPOSE OF REVIEW: Private equity's momentum in eye care remains controversial, even as investment continues to hasten the consolidation of ophthalmology and optometry practices. In this review, we discuss the growing implications of private equity activity in ophthalmology, drawing on updated empirical findings from the literature. We also examine recent legal and policy efforts to address private equity investment in healthcare, with implications for ophthalmologists considering sales to private equity. RECENT FINDINGS: Concerns about private equity centres around evidence that some investment entities are not just valuable sources of capital or business expertise, but that they take outright ownership and control of acquired practices in order to drive high returns on investment. Although practices may receive considerable benefits from private equity investment, empirical evidence suggests that private equity's most consistent effect on acquired practices is to increase spending and utilization without commensurate benefits on patient health. Although data on workforce effects are limited, an early study on workforce composition changes in private equity-acquired practices demonstrates that physicians were more likely to enter and exit a given practice than their counterparts in nonacquired practices, suggesting some degree of workforce flux. State and federal oversight of private equity's impact on healthcare may be ramping up in response to these demonstrated changes. SUMMARY: Private equity will continue to broaden their footprint in eye care, necessitating ophthalmologists to take the long view of private equity's net effects. For practices considering a private equity sale, recent policy developments highlight the importance of identifying and vetting a well aligned investment partner, with safeguards to preserve clinical decision-making and physician autonomy.

Mapping Rural and Urban Veterans' Spatial Access to Primary Care Following the MISSION Act.

BACKGROUND: The 2018 MISSION Act sought to improve Veterans' access to primary care by allowing Veterans living more than 30 min from VA care to utilize non-VA clinics. The impact of this legislation may vary for rural compared to urban Veterans. OBJECTIVE: Assess the extent to which the 2018 MISSION Act facilitates spatial access to primary care for Veterans living in rural versus urban Oregon. DESIGN: We identified locations of all VA and non-VA primary care clinics in Oregon then calculated 30-min drive-time catchment areas from census tract centroids to the nearest clinics. We compared measures of spatial access to primary care for Veterans in rural, micropolitan, and urban areas. PARTICIPANTS: American Community Survey data representing Oregon adults. MAIN MEASURES: Two measures of spatial access focusing on the number of clinics (supply), and an access index based on the two-step floating catchment area method (2SFCA) which accounts for number of clinics (supply) and population size (demand). KEY RESULTS: Compared to only 13.0% of rural Veterans, 83.6% of urban Veterans lived within 30 min' drive time of VA primary care. Given the MISSION Act's eligibility criteria, 81.6% of rural Veterans and ~ 97% of urban and micropolitan Veterans had spatial access to primary care. When accounting for both supply and demand, rural areas had significantly higher access scores (p < 0.05) compared to urban areas. CONCLUSIONS: Using MISSION Act guidelines for Veteran access to primary care, rural compared to urban Veterans had less spatial access based on clinic number (supply), but more access when considering clinic number and population size (supply and demand). Geographic Information System (GIS) spatial techniques may help to assess changes in access to care. However, these methods do not incorporate all dimensions of access and work is needed to understand whether utilization and quality of care is improved.

Engaging Consumers in Medicaid Program Design: Strategies from the States.

Policy Points As Medicaid programs grow in scale and complexity, greater consumer input may guide successful program design, but little is known about the extent to which state agencies are engaging consumers in the design and implementation of programs and policies. Through 50 semistructured interviews with Medicaid leaders in 14 states, we found significant variation in consumer engagement approaches, with many common facilitators, including leadership commitment, flexible strategies for recruiting and supporting consumer participation, and robust community partnerships. We provide early evidence on how state Medicaid agencies are integrating consumers' experiences and perspectives into their program design and governance. CONTEXT: Consumer engagement early in the process of health care policymaking may improve the effectiveness of program planning and implementation, promote patient-centric care, enhance beneficiary protections, and offer opportunities to improve service delivery. As Medicaid programs grow in scale and complexity, greater consumer input may guide successful program design, but little is known about the extent to which state agencies are currently engaging consumers in the design and implementation of programs and policies, and how this is being done. METHODS: We conducted semistructured interviews with 50 Medicaid program leaders across 14 states, employing a stratified purposive sampling method to select state Medicaid programs based on US census region, rurality, Medicaid enrollment size, total population, ACA expansion status, and Medicaid managed care penetration. Interview data were audio-recorded, professionally transcribed, and underwent iterative coding with content and thematic analyses. FINDINGS: First, we found variation in consumer engagement approaches, ranging from limited and largely symbolic interactions to longer-term deliberative bodies, with some states tailoring their federally mandated standing committees to engage consumers. Second, most states were motivated by pragmatic considerations, such as identifying and overcoming implementation challenges for agency programs. Third, states reported several common facilitators of successful consumer engagement efforts, including leadership commitment, flexible strategies for recruiting and supporting consumers' participation, and robust community partnerships. All states faced barriers to authentic and sustained engagement. CONCLUSIONS: Sharing best practices across states could help strengthen programs' engagement efforts, identify opportunities for program improvement reflecting community needs, and increase participation among a population that has traditionally lacked a political voice.

Developing an Automatic System for Classifying Chatter About Health Services on Twitter: Case Study for Medicaid.

BACKGROUND: The wide adoption of social media in daily life renders it a rich and effective resource for conducting near real-time assessments of consumers' perceptions of health services. However, its use in these assessments can be challenging because of the vast amount of data and the diversity of content in social media chatter. OBJECTIVE: This study aims to develop and evaluate an automatic system involving natural language processing and machine learning to automatically characterize user-posted Twitter data about health services using Medicaid, the single largest source of health coverage in the United States, as an example. METHODS: We collected data from Twitter in two ways: via the public streaming application programming interface using Medicaid-related keywords (Corpus 1) and by using the website's search option for tweets mentioning agency-specific handles (Corpus 2). We manually labeled a sample of tweets in 5 predetermined categories or other and artificially increased the number of training posts from specific low-frequency categories. Using the manually labeled data, we trained and evaluated several supervised learning algorithms, including support vector machine, random forest (RF), naïve Bayes, shallow neural network (NN), k-nearest neighbor, bidirectional long short-term memory, and bidirectional encoder representations from transformers (BERT). We then applied the best-performing classifier to the collected tweets for postclassification analyses to assess the utility of our methods. RESULTS: We manually annotated 11,379 tweets (Corpus 1: 9179; Corpus 2: 2200) and used 7930 (69.7%) for training, 1449 (12.7%) for validation, and 2000 (17.6%) for testing. A classifier based on BERT obtained the highest accuracies (81.7%, Corpus 1; 80.7%, Corpus 2) and F1 scores on consumer feedback (0.58, Corpus 1; 0.90, Corpus 2), outperforming the second best classifiers in terms of accuracy (74.6%, RF on Corpus 1; 69.4%, RF on Corpus 2) and F1 score on consumer feedback (0.44, NN on Corpus 1; 0.82, RF on Corpus 2). Postclassification analyses revealed differing intercorpora distributions of tweet categories, with political (400778/628411, 63.78%) and consumer feedback (15073/27337, 55.14%) tweets being the most frequent for Corpus 1 and Corpus 2, respectively. CONCLUSIONS: The broad and variable content of Medicaid-related tweets necessitates automatic categorization to identify topic-relevant posts. Our proposed system presents a feasible solution for automatic categorization and can be deployed and generalized for health service programs other than Medicaid. Annotated data and methods are available for future studies.

Health Policy Perspective: Medicaid and State Politics Beyond COVID.

The COVID-19 pandemic is poised to drastically alter the Medicaid program. While state Medicaid programs are currently expanding coverage policies and enrollment to address acute public health needs, states will soon face significant budget shortfalls. These impending changes may renew partisan debates about restrictive policies like work requirements, which generally require beneficiaries to verify their participation in certain activities-such as employment, job search, or training programs-in order to receive or retain coverage. We argue that restrictive Medicaid policies are driven, to a great extent, by political party affiliation, highlighting the outsized role of partisanship in Medicaid policy adoption. To combat these dynamics, additional efforts are needed to improve community-informed decision-making, strengthen evaluation approaches to tie evidence to policymaking, and boost participation in and understanding of the political processes that affect policy change.

Access to Primary, Mental Health, and Specialty Care: a Comparison of Medicaid and Commercially Insured Populations in Oregon.

OBJECTIVE: To describe how access to primary and specialty care differs for Medicaid patients relative to commercially insured patients, and how these differences vary across rural and urban counties, using comprehensive claims data from Oregon. DESIGN: Cross-sectional study of risk-adjusted access rates for two types of primary care providers (physicians; nurse practitioners (NPs) and physician assistants (PAs)); four types of mental health providers (psychiatrists, psychologists, advanced practice NPs or PAs specializing in mental health care, behavioral specialists); and four physician specialties (obstetrics and gynecology, general surgery, gastroenterology, dermatology). PARTICIPANTS: 420,947 Medicaid and 638,980 commercially insured adults in Oregon, October 2014-September 2015. OUTCOME: Presence of any visit with each provider type, risk-adjusted for sex, age, and health conditions. RESULTS: Relative to commercially insured individuals, Medicaid enrollees had lower rates of access to primary care physicians (- 11.82%; CI - 12.01 to - 11.63%) and to some specialists (e.g., obstetrics and gynecology, dermatology), but had equivalent or higher rates of access to NPs and PAs providing primary care (4.33%; CI 4.15 to 4.52%) and a variety of mental health providers (including psychiatrists, NPs and PAs, and other behavioral specialists). Across all providers, the largest gaps in Medicaid-commercial access rates were observed in rural counties. The Medicaid-commercial patient mix was evenly distributed across primary care physicians, suggesting that access for Medicaid patients was not limited to a small subset of primary care providers. CONCLUSIONS: This cross-sectional study found lower rates of access to primary care physicians for Medicaid enrollees, but Medicaid-commercial differences in access rates were not present across all provider types and displayed substantial variability across counties. Policies that address rural-urban differences as well as Medicaid-commercial differences-such as expansions of telemedicine or changes in the workforce mix-may have the largest impact on improving access to care across a wide range of populations.

Characteristics of Twitter Use by State Medicaid Programs in the United States: Machine Learning Approach.

BACKGROUND: Twitter is a potentially valuable tool for public health officials and state Medicaid programs in the United States, which provide public health insurance to 72 million Americans. OBJECTIVE: We aim to characterize how Medicaid agencies and managed care organization (MCO) health plans are using Twitter to communicate with the public. METHODS: Using Twitter's public application programming interface, we collected 158,714 public posts ("tweets") from active Twitter profiles of state Medicaid agencies and MCOs, spanning March 2014 through June 2019. Manual content analyses identified 5 broad categories of content, and these coded tweets were used to train supervised machine learning algorithms to classify all collected posts. RESULTS: We identified 15 state Medicaid agencies and 81 Medicaid MCOs on Twitter. The mean number of followers was 1784, the mean number of those followed was 542, and the mean number of posts was 2476. Approximately 39% of tweets came from just 10 accounts. Of all posts, 39.8% (63,168/158,714) were classified as general public health education and outreach; 23.5% (n=37,298) were about specific Medicaid policies, programs, services, or events; 18.4% (n=29,203) were organizational promotion of staff and activities; and 11.6% (n=18,411) contained general news and news links. Only 4.5% (n=7142) of posts were responses to specific questions, concerns, or complaints from the public. CONCLUSIONS: Twitter has the potential to enhance community building, beneficiary engagement, and public health outreach, but appears to be underutilized by the Medicaid program.

State-Based Marketplaces Outperform Federally-Facilitated Marketplaces.

In response to regulatory changes at the federal level, states that run their own marketplaces have taken steps to stabilize their individual markets. In this comparison of state-based and federally-facilitated marketplaces from 2016-2018, we find that SBMs had slower premium increases (43% vs. 75%), and fewer carrier exits, than FFMs. The total population participating in FFMs declined by 10%, while the enrolled population in SBMs remained largely stable, increasing by 2%. We find that the performance of the ACA marketplaces varies by state and appears to cluster around marketplace types.

Inaccuracies in provider directories persist for long periods of time.

A growing literature has identified substantial inaccuracies in consumer-facing provider directories, but it is unclear how long these inaccuracies persist. We re-surveyed inaccurately listed Pennsylvania providers (n = 5170) between 117 to 280 days after a previous secret-shopper survey. Overall, 19.0% (n = 983) of provider directory listings that had been identified as inaccurate were subsequently removed, 44.8% (n = 2316) of provider listings continued to show at least 1 inaccuracy, and 11.6% (n = 600) were accurate at follow-up. We were unable to reach 24.6% (n = 1271) of providers. Longer passage of time was associated with reductions in directory inaccuracies, particularly related to contact information, and to a lesser degree, with removal of inaccurate listings. We found substantial differences in corrective action by carrier. Together, these findings suggest persistent barriers to maintaining and updating provider directories, with implications for how well these tools can help consumers select health plans and access care.

Life cycle of private equity investments in physician practices: an overview of private equity exits.

Private equity (PE) firms acquire and grow physician practices through add-on consolidation, generating outsized returns on the sale of the acquisition in 3-8 years ("exit"). Private equity's abbreviated investment timeline and exit incentives may deter long-term investments in care delivery and workforce needed for high-quality care. To our knowledge, there has been no published analyses of the nature or duration of PE exits from physician practices. We address this knowledge gap by using novel data to characterize PE exits from dermatology, ophthalmology, and gastroenterology, physician specialties with the largest number of acquisitions between 2016 and 2020. Of 807 acquisitions, over half (51.6%) of PE-acquired practices underwent an exit within 3 years of initial investment. In nearly all instances (97.8%), PE firms exited investments through secondary buyouts, where physician practices were resold to other PE firms with larger investment funds. Between investment and exit, PE firms increased the number of physician practices affiliated with the PE firm by an average of 595% in 3 years. Findings highlight the rapid scale of ownership change and consolidation under PE ownership and motivate evaluations by policymakers on the effects of PE ownership over the life cycle of PE investments.

Rural and frontier access to mental health prescribers and nonprescribers: A geospatial analysis in Oregon Medicaid.

OBJECTIVE: Medicaid enrollees in rural and frontier areas face inadequate access to mental health services, but the extent to which access varies for different provider types is unknown. We assessed access to Medicaid-participating prescribing and nonprescribing mental health clinicians, focusing on Oregon, which has a substantial rural population. METHODS: Using 2018 Medicaid claims data, we identified enrollees aged 18-64 with psychiatric diagnoses and specialty mental health providers who billed Medicaid at least once during the study period. We measured both 30- and 60-minute drive time to a mental health provider, and a spatial access score derived from the enhanced 2-step floating catchment area (E2SFCA) approach at the level of Zip Code Tabulation Areas (ZCTAs). Results were stratified for prescribers and nonprescribers, across urban, rural, and frontier areas. RESULTS: Overall, a majority of ZCTAs (68.6%) had at least 1 mental health prescriber and nonprescriber within a 30-minute drive. E2SFCA measures demonstrated that while frontier ZCTAs had the lowest access to prescribers (84.3% in the lowest quintile of access) compared to other regions, some frontier ZCTAs had relatively high access to nonprescribers (34.3% in the third and fourth quartiles of access). CONCLUSIONS: Some frontier areas with relatively poor access to Medicaid-participating mental health prescribers demonstrated relatively high access to nonprescribers, suggesting reliance on nonprescribing clinicians for mental health care delivery amid rural workforce constraints. Efforts to monitor network adequacy should consider differential access to different provider types, and incorporate methods, such as E2SFCA, to better account for service demand and supply.

Factors Influencing Turnover and Attrition in the Public Behavioral Health System Workforce: Qualitative Study.

Many states are experiencing a behavioral health workforce crisis, particularly in the public behavioral health system. An understanding of the factors influencing the workforce shortage is critical for informing public policies to improve workforce retention and access to care. The aim of this study was to assess factors contributing to behavioral health workforce turnover and attrition in Oregon. Semistructured qualitative interviews were conducted with 24 behavioral health providers, administrators, and policy experts with knowledge of Oregon's public behavioral health system. Interviews were transcribed and iteratively coded to reach consensus on emerging themes. Five key themes emerged that negatively affected the interviewees' workplace experience and longevity: low wages, documentation burden, poor physical and administrative infrastructure, lack of career development opportunities, and a chronically traumatic work environment. Large caseloads and patients' high symptom acuity contributed to worker stress. At the organizational and system levels, chronic underfunding and poor administrative infrastructure made frontline providers feel undervalued and unfulfilled, pushing them to leave the public behavioral health setting or behavioral health altogether. Behavioral health providers are negatively affected by systemic underinvestment. Policies to improve workforce shortages should target the effects of inadequate financial and workplace support on the daily work environment.

Racial and Ethnic Disparities in Geographic Availability of Buprenorphine.

OBJECTIVES: Overdose mortality has risen most rapidly among racial and ethnic minority groups while buprenorphine prescribing has increased disproportionately in predominantly non-Hispanic White urban areas. To identify whether buprenorphine availability equitably meets the needs of diverse populations, we examined the differential geographic availability of buprenorphine in areas with greater concentrations of racial and ethnic minority groups. METHODS: Using IQVIA longitudinal prescription data, IQVIA OneKey data, and Microsoft Bing Maps, we calculated 2 outcome measures across the continental United States: the number of buprenorphine prescribers per 1000 residents within a 30-minute drive of a ZIP code, and the number of buprenorphine prescriptions dispensed per capita at retail pharmacies among nearby buprenorphine prescribers. We then estimated differences in these outcomes by ZIP codes' racial and ethnic minority composition and rurality with t tests. RESULTS: Buprenorphine prescribers per 1000 residents within a 30-minute drive decreased by 3.8 prescribers per 1000 residents in urban ZIP codes (95% confidence interval = -4.9 to -2.7) and 2.6 in rural ZIP codes (95% confidence interval = -3.0 to -2.2) whose populations consisted of ≥5% racial and ethnic minority groups. There were 45% to 55% fewer prescribers in urban areas and 62% to 79% fewer prescribers in rural areas as minority composition increased. Differences in dispensed buprenorphine per capita were similar but larger in magnitude. CONCLUSIONS: Achieving more equitable buprenorphine access requires not only increasing the number of buprenorphine-prescribing clinicians; in urban areas with higher racial and ethnic minority group populations, it also requires efforts to promote greater buprenorphine prescribing among already prescribing clinicians.

Composition of buprenorphine prescribing networks in Medicaid and association with quality of care.

INTRODUCTION: Despite Medicaid's outsized role in delivering and financing medications for opioid use disorder (MOUD), little is known about the extent to which buprenorphine prescriber networks vary across Medicaid health plans, and whether network characteristics affect quality of treatment received. In this observational cross-sectional study, we used 2018-2019 Medicaid claims in Oregon to assess network variation in the numbers and types of buprenorphine prescribers, as well as the association of prescriber and network characteristics with quality of care. METHODS: We describe prescribers (MD/DOs and advanced practice providers) of OUD-approved buprenorphine formulations to patients with an OUD diagnosis, across networks. For each patient who initiated buprenorphine treatment during 2018, we assigned a "usual prescriber" and assessed four measures of quality in the 180d following initiation: 1) continuous receipt of buprenorphine; 2) receipt of any behavioral health counseling services; 3) receipt of any urine drug screen; and 4) receipt of any prescription for a benzodiazepine. We used multivariable linear regressions to examine the association of prescriber and network characteristics with quality of buprenorphine care following initiation. RESULTS: We identified 645 providers who prescribed buprenorphine to 20,739 eligible Medicaid enrollees with an OUD diagnosis. The composition of buprenorphine prescriber networks varied in terms of licensing type, specialty, and panel size, with the majority of prescribers providing buprenorphine to small panels of patients. In the 180 days following initiation, a third of patients were maintained on buprenorphine; 69.9 % received behavioral health counseling; 88.4 % had a urine drug screen; and 11.3 % received a benzodiazepine prescription. In regression analyses, while no single network characteristic was associated with higher quality across all examined measures, each one unit increase in prescriber-to-enrollee ratio was associated with a 1.18 p.p. increase in the probability of continuous buprenorphine maintenance during the 180 days following initiation (95 % confidence interval = [0.21, 2.15], p = 0.017). CONCLUSIONS: Medicaid plans may be able to leverage their networks to provide higher quality care. Our findings, which should be interpreted as descriptive only, suggest that higher prescriber-to-enrollee ratio is associated with increased buprenorphine maintenance. Future research should focus on isolating the causal relationships between MOUD prescribing network design and patient outcomes.