Health Services Research最新文献

Objective: To assess how private equity ownership affects prices, service mix, and Medicaid participation in dentistry at the practice level.

Study setting and design: We utilize a proprietary dental office database linked to administrative dental commercial claims data to estimate the effects of private equity ownership on the financial and operational outcomes of dental offices, employing a staggered difference-in-differences panel design that addresses the nonrandom acquisitions of facilities by private equity firms.

Data sources: We rely on private equity transaction data from 2015 to 2021, longitudinal dental office data from the 2015 to 2017 and 2019 to 2021 American Dental Association dental office database, and aggregated commercial dental insurance price and utilization data from 2015 to 2021.

Principle findings: Following acquisition, private equity-owned dental offices increased charges for dental care services by 3.3% (95% CI: 2.3%-4.4%), although allowed prices for these services remained statistically unchanged. Dental offices acquired by private equity firms tended to shift from diagnostic and preventive procedures to generally higher reimbursement restorative, specialty, and surgical procedures. Dental offices were more likely to become multispecialty practices after being acquired by a private equity firm.

Conclusions: Allowed or negotiated prices between dentists and payers did not change in dental offices after being acquired by private equity. Nevertheless, list prices for dental services increased in private equity-owned practices, meaning higher prices can still be passed on to patients. Private equity firms can enhance dental practice revenue by shifting from preventive procedures to higher-cost restorative procedures while not reimbursing providers at a higher amount. In other words, financial enhancement of dental practices under private equity may not translate into benefits for providers or patients. Policymakers should be aware of the effects private equity acquisition has on provider and patient welfare.

Objective: To measure the impact of electronic health record (EHR) transition on burnout among Veterans Health Administration (VA) frontline clinicians using pseudorandom variation from staggered EHR implementations across VA sites.

Study setting and design: Employing a quasi-experimental design, we studied 140 VA medical center sites nationwide (including five sites that implemented the new EHR from 2019 to 2023). Explanatory measures included year, VA transition site (grouped into three cohorts by transition timing), and their interaction. Our outcome measure encapsulated two dimensions of burnout-emotional exhaustion and depersonalization (symptoms > once per week indicated burnout).

Data sources and analytic sample: Using secondary data from the 2019 to 2023 VA All Employee Survey, we aggregated survey responses on the medical-site level by year and respondent characteristics. Our analytic sample included 12,155 aggregated observations. We used a difference-in-difference approach to compare pre-post changes in burnout between VA sites implementing and not implementing the new EHR. Where available, we reported post-transition treatment effects in the short term, medium term, and long term, relative to EHR implementation.

Principal findings: Unadjusted burnout from 2019 to 2023 was 36.9% for Cohort 1, 33.0% for Cohort 2, 37.0% for Cohort 3, and 33.2% for non-transition sites. In adjusted analyses, burnout for Cohort 1 increased 4.8 percentage points (p < 0.001) in the medium term; differences in burnout dissipated in the long term. For Cohort 2, we detected a 1 percentage point increase in burnout (p = 0.004) in the short term and a 1.5 percentage point decrease (p = 0.013) in the medium term. For Cohort 3, burnout increased 3.3 percentage points (p < 0.001) in the medium term.

Conclusions: The impact of EHR transition on burnout differed across deployment sites and post-transition periods but was mild overall. Future research is needed to understand contextual and implementation process differences between sites that may explain differential effects and offer learnings to ensure a high-functioning health workforce during EHR transition.

Objective: To evaluate how adding kidney transplantation waitlisting measures-the Standardized First Kidney Transplant Waitlist Ratio for Incident Dialysis Patients (SWR) and Percentage of Prevalent Patients Waitlisted (PPPW)-affects Dialysis Facility Care Compare Star Ratings.

Study setting and design: In this observational, cross-sectional study, we calculated the difference between facilities' published (with waitlisting measures) and counterfactual (without waitlisting measures) Star Ratings. We used multinomial regression to examine associations between Star Rating changes after waitlisting measure inclusion and facility characteristics and calculated corresponding average risk differences.

Data sources and analytic sample: We used comprehensive clinical and administrative data from the Centers for Medicare/Medicaid Services from 2021 to investigate the impact of waitlisting measure addition on Star Ratings. Facility characteristics included demographic and patient mix, area deprivation index (ADI), dialysis organization affiliation, and urbanicity.

Principal findings: 36.5% of facilities' ratings changed after waitlisting measures were added. Facility characteristics associated with a higher average risk of Star increase included location in low-ADI (0.091; 95% CI: 0.072, 0.109) or urban areas (0.061; 95% CI: 0.034, 0.087), independent/small dialysis organization affiliation (0.062; 95% CI: 0.041, 0.083), and having more PD patients (0.115; 95% CI: 0.093, 0.138). Characteristics associated with a higher average risk of Star decrease included high-ADI (0.075; 95% CI: 0.054, 0.095) or rural (0.056; 95% CI: 0.028, 0.083) location, large dialysis organization affiliation (0.058; 95% CI: 0.039, 0.078), having more patients with dual Medicare/Medicaid eligibility (0.052; 95% CI: 0.032, 0.071), and having fewer peritoneal dialysis patients (0.100; 95% CI: 0.081, 0.120).

Conclusions: Including waitlisting measures significantly impacts the Star Ratings and captures a new dimension of care quality. Worse socioeconomic status-related facility characteristics were strongly associated with worse Star Rating outcomes. These findings can inform future discussions about risk adjustment among the developers of the SWR and PPPW measures.

Objective: The objectives of this study were to evaluate associations of social deprivation with acute care utilization among patients with cancer, and to examine potential effect modification by physician network vulnerability.

Study setting and design: For this retrospective cohort study, the primary exposure variable was neighborhood-level socioeconomic disadvantage, operationalized through the social deprivation index (SDI). We assembled physician patient-sharing networks and calculated a measure of network vulnerability for each referral region to capture specialist scarcity. The two outcomes of interest were counts of emergency department (ED) visits and non-elective hospitalizations during the 12 months following cancer diagnosis. We conducted hurdle regressions, with logistic and negative binomial mixed-effects models for the zero and positive, non-zero parts of the outcome distribution, respectively, and stratified by physician network vulnerability.

Data sources and analytic sample: We analyzed 2016-2020 Surveillance, Epidemiology, and End Results (SEER)-Medicare linked data for Medicare beneficiaries diagnosed with breast, colorectal, or lung cancer.

Principal findings: The study cohort comprised 47,756 patients with breast, colorectal or lung cancer. Patients in high SDI neighborhoods (vs. low) had a higher probability of at least one ED visit across all physician network vulnerability strata (low network vulnerability-average marginal effect (AME) [95% CI]: 0.03 [0.01-0.05]; medium network vulnerability-AME [95% CI]: 0.03 [0.01-0.04]; high network vulnerability-AME [95% CI]: 0.05 [0.02-0.08]). Conditional on at least one ED visit, patients in high SDI neighborhoods (vs. low) had a greater relative risk of additional ED visits when their region was characterized by low physician network vulnerability (RR [95% CI]: 1.25 [1.09-1.43]).

Conclusions: Our findings suggest that SDI and physician network vulnerability interact to increase the probability and likelihood of ED visits, but the interaction was minimal for non-elective hospitalizations. More research is needed to better understand how social drivers of health and oncology workforce scarcity affect care utilization and outcomes in patients with cancer.

Objective: To test if state consumer protection policies reduce the share of consumers with medical debt in collections on their credit reports.

Study setting and design: This study uses a quasi-experimental research design to estimate the impact of consumer protection laws implemented between 2020 and 2022 in Illinois, Maryland, New Mexico, and Oregon on the share of consumers with medical debt in collections. These laws primarily aim to protect consumers against medical debt by expanding access to hospital financial assistance. We use a synthetic control approach to estimate changes in medical debt following the implementation of policies in treatment states relative to changes in select control states. We also assess the effects of earlier policies implemented between 2013 and 2019 in Washington, Utah, and North Carolina.

Data sources and analytic sample: This analysis relies on two extracts of credit bureau data from one of the country's three main credit bureau agencies. The first extract consists of random samples from June 2017 to June 2024 of approximately 125,000 consumers in each treatment state and 500,000 residents from the pool of 14 selected comparison states in each year. The second extract is based on a 2%-4% random sample of consumers in each year from 2011 to 2022.

Principal findings: We did not observe a statistically significant reduction in medical debt associated with policies implemented in these states within the study timeframe. In most states in our primary analysis, point estimates of the treatment effects are near zero, and in nearly all state-years, we can only rule out declines in medical debt larger than 1-3 percentage points following policy implementation.

Conclusions: Though we did not detect statistically significant effects of recent consumer protection policies on medical debt in collections, additional research is needed on whether these policies benefited consumers in ways that are not measured in this analysis and whether states that continue to move forward with similar laws can improve their effectiveness by extending consumer protections to a wider group of patients and providers and addressing implementation and enforcement challenges.

Objective: To estimate the effect of starting a provider-led Institutional Special Needs Plan (I-SNP) arrangement on facility-level enrollment, utilization, and quality.

Study setting and design: I-SNPs are a type of Medicare Advantage (MA) plan that allows insurers to differentiate their benefits exclusively for long-term residents in nursing homes. Since I-SNPs first became available in 2006, there has been growth in provider-led I-SNPs where nursing homes are financially integrated or partnered with an insurer to operate a plan for their own residents. We used a difference-in-differences design to estimate the effect of starting a provider-led I-SNP arrangement on several facility-level outcomes, including the share of a facility's long-stay residents who were enrolled in an I-SNP, hospitalizations, medication use, pressure ulcers, physical restraints, falls, and mortality.

Data sources and analytic sample: We used Medicare claims and nursing home resident assessments (2004-2021) to identify Medicare long-stay nursing home residents.

Principal findings: The start of a provider-led I-SNP arrangement led to a 17.0 percentage point (pp) increase (standard error [SE]: 0.006) in I-SNP enrollment among facility residents within 4 years relative to control nursing homes. We also estimate that the start of a provider-led I-SNP arrangement significantly decreased hospitalizations (-1.0 pp, SE: 0.002), increased the use of antipsychotic (0.4 pp, SE: 0.002) and hypnotic drugs (0.3 pp, SE: 0.001), and reporting of pressure ulcers (0.4 pp, SE: 0.002).

Conclusions: Provider-led I-SNPs allow nursing homes to bear financial risk for their residents. These results suggest that this form of risk bearing may successfully reduce utilization (e.g., hospitalizations), but with unclear implications for quality as increased use of sedating drugs and rates of pressure ulcers could either reflect poorer care or retention of sicker patients due to lower hospitalization rates.

Objective: To define measures of Medicare diagnosis coding intensity that capture the dynamics of changes in coding practices.

Study setting and design: Retrospective analysis of coding for risk adjustment using observational claims data from Medicare beneficiaries.

Data sources: Enrollment and claims data from 2017 and 2018 of a random 20% sample of Medicare beneficiaries were subset to those assigned to an Accountable Care Organization in 2018.

Principal findings: We decompose the prevalence of a diagnosis code into incidence (proportion of beneficiaries that newly have the code) and persistence (proportion of beneficiaries who previously had the code and continue to do so). Together these define steady-state prevalence, the hypothetical long-run prevalence implied by no changes in current rates of incidence and persistence of coding. Steady-state prevalence can help explain why observed prevalence tends to grow over time without continued behavioral change. For example, our measures suggest that the prevalence of the Specified Heart Arrhythmias diagnosis would continue to rise from 18.7% in 2018 to 28.0% without changes in coding practices.

Conclusions: Researchers and policymakers can better understand why changes in coding practices can take years to be fully reflected in data and monitor coding behavior by using our proposed measures.