Health Services Research最新文献

Objective: To determine whether rural hospital closures affected hospital and post-acute care (PAC) use and outcomes.

Study setting and design: Using a staggered difference-in-differences design, we evaluated associations between 32 rural hospital closures and changes in county-level: (1) travel distances to and lengths of stay at hospitals; (2) functional limitations at and time from hospital discharge to start of PAC episode; (3) 30-day readmissions and mortality and hospitalizations for a fall-related injury; and (4) population-level hospitalization and death rates.

Data sources and analytic sample: 100% Medicare claims and home health and skilled nursing facility clinical data to identify approximately 3 million discharges for older fee-for-service Medicare beneficiaries.

Principal findings: We found that hospitals that closed compared to those remaining open served more minoritized, lower-income populations, including more Medicaid and fewer commercial patients, and had lower profit margins. Following a closure, quarterly hospitalization rates (111.6 quarterly hospitalizations per 10,000 older adults; 95% CI: 53.4, 170.9) and average hospital lengths of stay increased (0.34 days; 95% CI: 0.13, 0.56 days). We observed no change in the average distance between patients' residential ZIP code and the hospital used (0.29 miles; 95% CI: -1.06, 1.64 miles); average number of standardized ADL limitations at PAC (0.08 SDs from the pre-closure average; 95% CI: -0.12, 0.28 SDs); or PAC time to start (0.02 days; 95% CI: -1.2, 1.2 days). Among more isolated hospitals, closures were associated with an increase in the likelihood of readmission (0.10 percentage-points; 95% CI: 0.00, 0.19).

Conclusions: Closures were not associated with notably worsened health care access, function, or health, potentially because closures triggered care delivery adjustments involving increased numbers of patients seeking out higher quality care.

Objective: To evaluate the impact of tort reform laws passed in 2011 capping noneconomic damages in North Carolina and Tennessee on rates and adjusted per user costs of tests, imaging, and procedures in the Medicare fee-for-service population.

Study setting and design: State-level synthetic difference-in-differences, adjusting for the percent of FFS Medicare beneficiaries in the state who were female, had ever been on Medicare Advantage, were eligible for Medicaid for at least 1 month of the year, and total state risk-adjusted, standardized per-capita costs. Analyses of North Carolina and Tennessee were performed separately. We measured the average treatment effect on the treated.

Data sources and analytic sample: Centers for Medicare and Medicaid Services Geographic Variation Public Use File, 2007-2019.

Principal findings: Our analysis showed no economically significant impact of these laws in either state, though we found a small but statistically significant increase (average treatment effect on the treated: $46, 95% confidence interval: $6-$87) in adjusted per user cost of procedures in Tennessee.

Conclusions: Our findings suggest that caps on noneconomic damages alone may be insufficient to modify physician practice habits and impact utilization. Future work should attempt to better understand the economic and noneconomic incentives that shape physician ordering decisions.

Objective: To identify barriers and facilitators to family-level comfort with health-related social needs (HRSN) data collection and documentation in the pediatric clinical setting.

Study setting and design: This qualitative study was nested within a pragmatic randomized controlled trial on social care integration in the pediatric clinical setting. We used a hybrid random-purposive strategy to sample 60 caregivers of pediatric patients ages 0-25 presenting at two primary care clinics and one emergency department affiliated with a large pediatric healthcare system between September 2022 and 2023. We developed an interview guide and codebook to explore caregiver experiences with and perceptions of HRSN data collection and documentation.

Data sources and analytic sample: We conducted semi-structured telephone interviews in English and Spanish with 60 caregivers. Interviews were conducted until thematic saturation was achieved and were transcribed verbatim. We used thematic analysis with constant comparison to code interviews and identify emerging themes.

Principal findings: Our analysis yielded several barriers to caregiver comfort with HRSN data collection and documentation: (1) stigmatization by providers and medical staff and risk of child protective services involvement, (2) providers presuming connections between documented HRSN and medical complaints, (3) permanency of documented HRSN, (4) visibility of HRSN data by pediatric patients and caregiver proxies, and (5) fear that documented HRSN could negatively impact future insurance cost and coverage. We identified four facilitators to caregiver comfort: (1) clear communication regarding the purpose of HRSN data collection and use, (2) respect for caregiver autonomy, for example, by providing the option to decline participation, (3) training of data collection personnel to ensure privacy and compassionate care, and (4) consideration of timing within the medical visit, delaying assessment until medical concerns are addressed.

Conclusions: Caregiver-identified barriers and facilitators should be considered in clinically based HRSN data collection efforts to ensure that these programs are equitable and family-centered.

Objective: To demonstrate the use of transportability methods to extend findings from payment model evaluations to groups of historically underserved beneficiaries.

Study setting and design: We used a simulation study to transport the effects of the Comprehensive Primary Care Plus (CPC+) model to a target population of Black fee-for-service (FFS) Medicare beneficiaries living outside the original 18 CPC+ regions. Our main outcome variable was total Medicare spending per beneficiary per year (pbpy).

Data sources and analytic sample: We simulated practice-level spending in 18 CPC+ regions and 32 non-CPC+ regions (1200 practices per region). We calibrated the simulation parameters to values from the literature and then varied four key parameters to create 16 realistic simulation scenarios. These scenarios varied the representativeness of practices in CPC+ regions that joined CPC+ (i.e., the sample) relative to the target population by changing the distribution of Black beneficiaries across practices and the distribution of practices across regions. Practices were characterized by their experience with the Medicare Shared Savings Program (SSP) and system/hospital ownership because these are known to modify the effect of CPC+ on spending.

Principal findings: Across the 16 simulation scenarios, transporting the treatment effect of CPC+ to Black FFS beneficiaries in non-CPC+ regions yielded median treatment effects that ranged from $15.5 pbpy smaller to $10 pbpy larger than in the sample. These differences are roughly the same magnitude as the estimated overall effect of $13 pbpy.

Conclusions: The Center for Medicare and Medicaid Innovation has pledged to put equity at the center of its demonstration models. However, offering models in limited geographic areas with voluntary provider participation may result in unrepresentative samples. Naively generalizing CPC+ effects from geographically limited, voluntary samples to all Black FFS beneficiaries could be misleading. Under some circumstances, transportability methods can be used to estimate effects in this target population.

Objective: To evaluate the impact of the Massachusetts Medicaid program's reimbursement policy change for perinatal depression screening on utilization rates.

Study setting and design: This study employed a difference-in-differences design to compare insurance-paid prenatal and postpartum depression screening rates as well as postpartum antidepressant receipt rates between Medicaid and privately insured individuals before and after policy implementation in May 2016.

Data sources and analytic sample: Data are from the 2014-2020 Massachusetts All-Payer Claims Database. The study included individuals with a live birth from October 10, 2014, to December 31, 2019, who were continuously insured either by Medicaid or private insurance.

Principal findings: Among 141,085 births, 42.6% were covered by Medicaid. Among those with Medicaid, 1.9% had a paid postpartum depression screening prior to the policy and 16.9% after (1.5% vs. 12.3% for prenatal screening); among privately insured, 3.8% had a paid postpartum screening prior to the policy and 10.6% after (0.9% vs. 6.7% for prenatal screening). Antidepressant receipt rose from 6.9% to 8.3% among Medicaid enrollees and from 3.3% to 4.9% among privately insured individuals after the policy. After regression adjustment, implementation of the Massachusetts Medicaid reimbursement policy was positively associated with perinatal depression screening rates with a differential increase of 10.0 percentage points (p < 0.001) for postpartum screening and 3.5 percentage points (p < 0.001) for prenatal screening among Medicaid enrollees versus privately insured. Despite increased depression screening, the policy was not associated with a statistically significant change in antidepressant receipt among Medicaid enrollees compared to privately insured individuals.

Conclusions: Separate payment for perinatal depression screening significantly improved screening rates among Medicaid beneficiaries, highlighting Medicaid's critical role in identifying mental health needs for vulnerable populations. However, the persistence of sub-optimal screening rates among perinatal individuals underscores the need for a comprehensive approach to ensure universal screening and effective treatment for perinatal depression.

Objective: To assess multi-level factors influencing the sustainability of 26 social care pilots integrating medical and social services for Medicaid enrollees across California in newly developed Medicaid benefits.

Study setting and design: This qualitative study assessed the sustainability of Whole Person Care (WPC) pilots implemented between 2016 and 2021. Pilots (n = 26) represented a majority of counties in California.

Data sources and analytic sample: Primary qualitative data were collected between June and August 2021 and included 58 hour-long, semi-structured individual and group interviews with administrators, middle managers, and frontline case management staff representing all WPC pilots. We used hybrid inductive-deductive thematic analysis to identify and analyze patterns, and outliers, in factors influencing sustainment. Deductive codes included established implementation science factors influencing the sustainability of new programs (e.g., innovation characteristics, capacity, processes and interactions, and context).

Principal findings: Of 26 WPC pilots, 22 pilots sustained WPC by contracting with Medicaid managed care plans to provide services as part of newly developed Medicaid benefits. Three pilots chose not to sustain before the pilot period ended and one pilot decided not to sustain following completion of the full pilot. Factors influencing sustainability included: (1) program adaptability and flexibility; (2) funding structure and reimbursement requirements; (3) shared leadership with managed care plans; and (4) whether pilots chose to build out program infrastructure internally or contracted out core components to partner organizations. Many pilots, particularly those in rural areas, indicated that system and policy changes introduced as part of transitioning pilot services into Medicaid benefits reduced the sustainability of WPC for participating providers.

Conclusions: Multi-level factors including program adaptability, funding, leadership, and capacity to build out infrastructure influenced the sustainability of WPC pilots. These findings have significant implications for health equity as equitable distribution of services, resources, and benefits from these programs can be supported through sustained implementation over time.

Objective: To understand relationships between healthcare use and food and housing insecurity in Medicaid expansion members, as well as whether these relationships differ by rurality or residential segregation.

Data sources and study setting: Database of Virginia Medicaid expansion members from the Department of Medical Assistance Services. Sample included individuals who enrolled January-June 2019, were aged 19-64 years, remained continuously enrolled for 12 months, and completed a Medicaid Member Health Screening (MMHS) conducted within the first 3 months of enrollment (n = 14,735).

Study design: Retrospective cohort study. Outcomes included any primary care visits (PC) and any emergency department (ED) visits in the first 12 months of enrollment. The MMHS sample was weighted to represent all Medicaid expansion members (n = 234,296). Separate multivariable linear probability models regressed having any PC or ED visits on food and housing insecurity controlling for individual and neighborhood characteristics. Models were then stratified by rurality and racial residential segregation.

Data collection: None.

Principal findings: Food insecurity was negatively associated with having any PC visit (-2.9 percentage points (PP); p-value <0.01) and positively associated with having any ED visit (7.0 PP; p-value <0.001). No significant relationships between PC or ED visits and housing insecurity were found. Suburban and urban individuals with food insecurity were significantly less likely to have any PC visit (p < 0.05 each). Medicaid expansion members living in disproportionately low-income or mixed-income neighborhoods experiencing food insecurity were also less likely to have any PC visits (p < 0.05), and the same was not true for those living in disproportionately high-income neighborhoods.

Conclusions: Food insecurity among Medicaid expansion members is associated with less primary care and more emergency department use, but these relationships differ by the neighborhoods in which members live. Medicaid agency efforts that coordinate medical and social service benefits and also consider local context may further increase access to necessary and appropriate care.

Objective: To identify collaboration strategies used to integrate health, behavioral health, and social services for Medicaid members in California's Medi-Cal Whole Person Care Pilot program (WPC).

Data sources and study setting: WPC was a social care intervention implemented to identify and address eligible members' health, behavioral health, and social needs. Data included semi-structured key informant interviews conducted in 2018-2019 (n = 221) and 2021 (n = 167); pilot-level surveys; whole-network surveys of 507 organizations in all 25 pilots participating in WPC; and documents submitted by pilots to the state. Pilots served a total of 247,887 unique members between 2017 and 2021, the majority of whom were non-white (72%) and over half of whom experienced homelessness.

Study design/data collection: Data were collected as part of the statewide evaluation of WPC. We analyzed qualitative data to examine strategies used by pilots to integrate care, network data to identify pilots that improved cross-sector collaboration (i.e., strengthened density or multiplexity of cross-sector ties) following WPC implementation, and comparative case analysis to identify strategies that differentiated pilots that improved collaboration from those that did not.

Principal findings: Pilots used multiple strategies to facilitate the integration of care. Network analyses identified 10 pilots that significantly improved either density or multiplexity of cross-sector ties, and one pilot with high cross-sector collaboration prior to WPC. Compared to pilots that did not improve cross-sector collaboration, these pilots meaningfully engaged partners in program design and implementation, used braided funds, and leveraged WPC to support broader systems change. These pilots also reported fewer challenges in developing and managing contractual relationships and ensuring meaningful use of data-sharing infrastructure by frontline staff responsible for care coordination.

Conclusions: Data sharing is necessary but not sufficient for systems alignment. Collaboration strategies focused on addressing financial barriers to integration and strengthening normative and interpersonal integration are also needed.