Health Services Research最新文献

Objective: Develop a causal machine learning (causal ML) framework for estimating how a diagnosis (cancer in this study) affects the likelihood of receiving a specific health care service (advance care planning in this study) and associated heterogeneity.

Study setting and design: Our proposed framework leverages the causal forest method, combined with a population-weighted resampling and averaging over estimations strategy, to estimate average treatment effects (ATEs) and conditional average treatment effects (CATEs). Post hoc, we used best linear projections to identify covariates associated with variation in the CATEs. We illustrate the framework by applying it to a stratified random sample of patients, where the strata are defined by the crosstabulation of cancer diagnosis (diagnosed vs. not diagnosed) and ACP receipt (documented vs. not documented).

Data sources and analytic sample: We extracted deidentified patient data from October 2019 to October 2024 (n = 87,772) with explanatory variables in three categories: demographics, morbidity, and health care system utilization.

Principal findings: In application of the causal ML framework, we found that patients diagnosed with cancer at this health care system to be at least 17.2% more likely to have documented ACP than similar patients not diagnosed with cancer. We also found significant heterogeneity. For instance, a one standard deviation increase in in-person outpatient visits was associated with an on-average increase in the CATE estimate (by 6.1 percentage points), while a one standard deviation increase in hospital admissions, inpatient days, and surgical duration in minutes was associated with an on-average decrease in the CATE estimate (by -1.3, -5.6, and -0.5 percentage points, respectively).

Conclusions: The proposed causal ML framework enables estimation of the effect of a diagnosis on receiving a relevant health care service. In the cancer diagnosis context, it can identify patient groups less likely to receive ACP, thus informing service allocation strategies.

Objective: To analyze the conditional association between provider and organizational factors and routine cancer screening for older Medicaid enrollees before and during the COVID-19 pandemic.

Study setting and design: This study analyzed pre-pandemic (2018/2019; n = 110,882) and pandemic (2020/2021; n = 107,451) cohorts of New Jersey (NJ) Medicaid enrollees aged 50-75. Using linear probability models, we evaluated how provider and organizational characteristics, including interactions with pandemic years, influenced screening for breast, cervical, colorectal, and lung cancers. Models controlled for enrollees' demographic and clinical characteristics and geographic factors.

Data sources and analytic sample: Claims data from the 2016-2021 NJ Medicaid Management Information System were linked to Medicare Provider and Specialty files. The sample included Medicaid enrollees with an assigned primary care provider and no prior cancer diagnosis.

Principal findings: Higher patient panel sizes were consistently associated with increased screening for breast (20.4%, 95% confidence interval (CI): 13.9%-26.8%), cervical (24.1%, 95% CI: 16.6%-31.5%), and lung cancer (63.1%; 95% CI: 17.4%-108.6%) during the pandemic. Obstetrician-gynecologist providers were linked to higher screening rates for breast (50.6%, 95% CI: 41.6%-59.5%) and cervical cancers (70.5%, 95% CI: 52.3%-88.9%), even during the pandemic. Female providers improved screening rates for breast (7.6%, 95% CI: 2.8%-12.3%), cervical (3.8%, 95% CI: 0.10%-7.5%), and colorectal cancer (5.8%, 95% CI: -2.7%-14.4%) among female enrollees. Provider age was unrelated to breast, cervical, or colorectal screening; however, in 2021, lung cancer screening was 23% lower for patients of clinicians aged 62 and above.

Conclusions: Large group practices effectively maintained breast and cervical cancer screening during the pandemic while exhibiting mixed results for colorectal and lung cancers. Provider characteristics such as gender and specialty also significantly impacted screening rates. Supporting large practices and addressing barriers in smaller practices are key to improving cancer prevention, especially during crises.

Objective: To examine how mis- and disinformation about the Public Charge Ground of Inadmissibility final rule ("public charge rule") influences health care access for Latino immigrants in California as seen through the perspectives of leaders in health-serving organizations.

Study setting and design: This qualitative study included semi-structured interviews with healthcare and community-based organizational leaders serving Latino immigrants in California. Viswanath et al.'s structural influence model of communication and equity guided the analyses and interpretation of the findings.

Data sources and analytic sample: Between May 2024 and April 2025, primary data were collected from 31 organizations, resulting in 32 semi-structured interviews with 38 participants. Interviews were conducted via Zoom and transcribed verbatim. Researchers coded the data based on recurring themes using Dedoose software.

Principal findings: Participants identified the public charge rule as a significant barrier to health care access for Latino immigrants. The policy has discouraged many Latinos from accessing public benefits, particularly the state's Medicaid and Supplemental Nutrition Assistance Program. In addition, immigrants' trusted sources of information (e.g., family, friends, and attorneys) were often misinformed about the policy, which amplified confusion and fear. Organizations respond by providing accurate information and connecting individuals with reliable resources to clarify that using public benefits would not necessarily result in being classified as a public charge. However, most efforts focused on education rather than directly countering mis- and disinformation.

Conclusions: Healthcare and community-based organizations offer unique perspectives as trusted intermediaries who help Latino immigrant families navigate health care and public benefits. Their close daily interactions reveal how misinformation about the public charge rule deters families from accessing essential services and makes it more challenging for organizations to fulfill their missions. These insights underscore the need for culturally responsive outreach and policy solutions that address information gaps and the climate of fear affecting community health.

Objective: To understand the impact of the Balancing Incentive Program (BIP) and Community First Choice State Plan Option (CFC) on LTSS rebalancing as measured by the size of and balance between the community and institutional LTSS workforces.

Study setting and design: Using a stacked difference-in-difference design, this paper evaluates the impact of BIP and CFC on the number of LTSS workers per 1000 persons 65+, the number of community LTSS workers per 1000 persons 65+, the number of institutional LTSS workers per 1000 persons 65+, and the proportion of all LTSS workers employed in community-based settings. We also test the impact of BIP's performance targets by separately estimating program effects for states that had yet to meet BIP rebalancing targets upon entering the program.

Data sources and analytical sample: Workforce and population data from the American Community Survey from 2005 to 2021.

Principal findings: This study finds that BIP resulted in a 13.24% (95% CI: 1.14%, 25.34%) increase in the size of the HCBS workforce in participating states, while finding no statistically significant effect for CFC (1.51%, 95% CI: -12.77%, 15.79%). The point estimate for growth in the HCBS workforce caused by BIP is twice as large in states bound by performance targets embedded within the BIP program (16.18%, 95% CI: 4.01%, 28.35%) as it is in states that are not (8.25%, 95% CI: -9.77%, 26.27%), suggesting that additional federal funding may be more effective when tied to performance targets for states. Neither program had a statistically significant effect on the size of the institutional workforce (BIP: 5.04%, 95% CI: -2.38%, 12.44%; CFC: 0.24%, 95% CI: -6.52%, 7.00%).

Conclusion: Federal policymakers seeking to increase investment in HCBS should ensure that additional funds are tied to measurable performance targets, incentivizing states to undertake expansions in HCBS that would not otherwise have taken place.

Objective: To measure discrepancies in risk adjustment scores using only Medicaid or Veterans Health Administration (VA) diagnoses for Veterans dually enrolled in VA and Medicaid.

Study setting and design: Veterans aged 18-64 enrolled in the VA and Medicaid for at least one full calendar year during 2017-2020. We compared the number and overlap of annual diagnoses derived from VA and Medicaid data. We also calculated Charlson, Elixhauser, and Centers for Medicare and Medicaid Hierarchical Condition Categories Version 21 (CMS-V21) risk scores using VA-only, Medicaid-only, and combined VA-Medicaid data for each person-year. We used intraclass correlations within risk measures to compare scores across risk measures.

Data sources and analytic sample: We used data from the VA's Assistant Deputy Undersecretary for Health's (ADUSH) enrollment files regarding age and VA Priority Group to select our cohort of VA enrollees. We used T-MSIS Analytic Files (TAF) and the Demographics and Enrollment (DE) file to determine Medicaid enrollment.

Principal findings: Our study cohort contained 183,018 dual-enrollees with service-connected disabilities representing 405,318 person years and 219,977 dual enrollees without service-connected disabilities (531,948 person years). On average, service-connected individuals had 9.1 fewer diagnoses from Medicaid-only data than from VA-only data (95% Confidence Interval (CI): [9.0, 9.1]) and 5.0 fewer for non-service-connected Veterans (95% CI: [4.9, 5.1]). Intraclass correlations between VA-only data and combined VA-Medicaid scores had higher correlations for Charlson (0.816 vs. 0.591 for service connected, 0.722 vs. 0.638 for non-service connected) and Elixhauser (0.818 vs. 0.609 for service-connected, 0.723 to 0.702 non-service-connected) scores, while Medicaid-only scores had higher correlations for CMS V21 (0.756 vs. 0.666 for service-connected, 0.795 to 0.542 for non service-connected).

Conclusions: Medicaid and VA data represent non-overlapping diagnoses data in three common risk scores. Researchers should consider combining records to calculate disease burden for dual-enrolled Veterans to ensure complete capture of risk.

Objective: To develop a novel method for enumerating the oncology specialist workforce triangulating taxonomy codes, board certification data, and clinical diagnosis codes in Medicaid claims, and to describe oncology specialists' Medicaid participation, their patient panels, and ascertain the concentration of types of cancers they treated.

Study setting and design: We identified oncology specialists using multiple data sources and conducted an exploratory analysis of their patient panels using multi-state Medicaid claims data. We used cluster analysis of diagnosis code patterns in claims to accurately determine the concentration of cancers by site in oncologists' panels.

Data sources and analytic sample: We used data from 2016 to 2020 Transformed Medicaid Statistical Information System (T-MSIS) and physician certification data. We included board-certified oncology physicians specialized in medical and radiation oncology, hematology, hematology-oncology, gynecologic oncology, and pediatric hematology-oncology. To identify surgical oncologists, we combined board certification and Medicare Provider Enrollment, Chain, and Ownership System (PECOS) data. We identified Medicaid beneficiaries with malignant neoplasms by cancer site using ICD-10-CM codes.

Principal findings: In 2016, about 89% of oncology specialists participated in Medicaid; this proportion decreased slightly to 86% in 2020. The trends in Medicaid participation and the mean number of beneficiaries differed by oncology specialty. Panels of pediatric hematologist-oncologists had a higher proportion of Hispanic Medicaid beneficiaries with cancer (26%) relative to other specialists. Cluster analysis identified 565 out of 5395 medical oncologists that had high concentration (at least 58%) of breast cancer patients in their panels. Among 6970 hematologist-oncologists, 269 had high concentrations in breast cancer (more than 60%), and 944 in hematological cancer (more than 59%).

Conclusions: Our study offers a pragmatic approach to understand the oncology specialist workforce available to Medicaid beneficiaries. The findings provide baseline estimates to track this workforce and provide policymakers with an opportunity to develop targeted strategies to improve access to cancer care.