Background: Historically, access to high quality care has been a central challenge for Medicaid programs. Prior single-year analyses demonstrated that Medicaid beneficiaries account for disproportionately high patient volumes at low-quality hospitals. Given major Medicaid shifts including expansion and increased managed care, we examined recent trends in low-quality hospital use for Medicaid beneficiaries.

Methods: Using Healthcare Cost and Utilization Project State Inpatient Databases, we compiled adult hospital discharges from 15 states in years 2016-2019 (N= 33,407,698). Hospital quality was assessed with the Agency for Healthcare Research and Quality Composite Inpatient Quality Indicator, reflecting risk-adjusted mortality for prevalent conditions. We constructed a logistic regression modeling odds of discharge from a low-quality hospital (bottom 20^th percentile by year), with payer-year interactions and covariates for patient demographics (sex, age, race/ethnicity, income), comorbidities, state, and admission type.

Results: Medicaid was associated with higher likelihood of care at low-quality hospitals (aOR=1.07, p<0.01), compared to Medicare (aOR=1.03, p<0.01) and private insurance (reference). Over time, we observed lower likelihood of discharges from low-quality hospitals across all payers (i.e., improving quality); that trend was stronger for Medicaid (aOR=0.99, p<0.01) and Medicare (aOR=0.97, p<0.01) than for private insurance.

Conclusions: This is one of the first studies examining associations between payer and inpatient care quality over time, critical information for our rapidly changing payment environment. While Medicaid patients remain more likely to be discharged from low-quality hospitals, we find promising recent trends of discharges less concentrated in low-quality facilities and less disparate by payer over time.

Background: Hospitals are increasingly consolidating. Past studies on market concentration and quality of care have shown mixed results but often fail to account for potential changes by racial/ethnic group and payer type, which may have implications for equity. We apply these lenses here.

Methods: We analyzed 2017 State Inpatient Data from 14 states, totaling 7,283,820 inpatient discharges. Competition was classified using the Herfindahl-Hirschman Index (HHI) with a 5,000 cutpoint. Quality was assessed using Prevention Quality Indicators (PQI). Multilevel logistic regression models were estimated to assess the relationship between PQI visits and HHI with interaction terms for race-competition and payer-competition, controlling for patient and hospital characteristics.

Results: Compared to less concentrated hospitals, hospitals in highly concentrated markets had higher proportions of White and Native American patients and lower proportions of all other racial/ethnic groups, and higher proportions of Medicare patients and lower proportions of commercial and Medicaid patients. In adjusted analyses, all racial minority groups had significantly higher odds of having a PQI admission compared to Whites, and Medicare and Medicaid patients had higher odds of a PQI admission compared to commercial patients. In moderation analyses, the ratio of odds ratios was positive for Asian/Pacific Islander patients and negative for all other racial minorities, and the ratio of odds ratios was negative for Medicare patients in payer analyses.

Conclusion: Competition may have positive and negative relationships with quality across different racial-ethnic groups and payers. Studies and evaluations of mergers and market changes should account for potential impacts on health equity.

Background: Since 2010, there has been an increase in state laws allowing Nurse Practitioner (NP) independence. Research is needed regarding the impact of NP independence on outpatient potentially harmful medication (PHM) prescribing to older adults. We aimed to utilize a synthetic control model to examine the effect of an October 2015 Full Scope of Practice (FSP) law on outpatient PHM prescribing to older adults in Maryland under Part D.

Methods: Medicare Public Use Files were used to quantify Part D claims. Covariates were drawn from the American Community Survey. PHM was defined using the Beer’s Criteria. A synthetic control model approach was used to develop a “Synthetic Maryland” control utilizing a weighted average of Delaware, Massachusetts, and New Jersey during the pre-policy period (2013-2015). Differences in average PHM claims per beneficiary between Maryland and Synthetic Maryland were assessed and compared to the distribution of each year after the policy change (2016-2019).

Results: The average number of PHM claims per beneficiary in the pre-policy period was 4.97 in both the Maryland and Synthetic groups. In the post-policy period (2016-2019), rates of PHM prescribing per beneficiary fell from the pre-policy period but did not differ significantly between groups (2016:β=-0.044, p =0.71; 2017:β=0.034, p=0.88; 2018:β=0.034, p=0.92; 2019: β=0.007, p=0.99).

Implications: These findings suggest that Maryland’s FSP law did not significantly impact statewide rates of PHM prescribing to Part D patients. NP independence increased access to prescribers, increasing the number of total prescriptions (2013-2015: μ=3,818,034; 2016-2019: μ=4,249,174) without worsening older adult prescribing outcomes.

Background: Effective patient-provider communication regarding treatment options and shared decision-making are crucial for building trust in healthcare. Patient-provider racial/ethnic concordance may play a vital role in facilitating this, particularly during the COVID-19 pandemic, which disrupted healthcare continuity and increased reliance on telehealth. This study aimed to evaluate the relationships between patient-provider racial/ethnic concordance, shared decision-making, and communication in 2020.

Methods: A sample of 9,634 adults from the 2020 U.S. Medical Expenditure Panel Survey was analyzed. Shared decision-making and communication were assessed based on patient inclusion in decision-making and provider explanation of treatments. Complex sampling logistic regression modeling was used to evaluate associations. Control variables included poverty, age, marital status, education, insurance, and race/ethnicity.

Results: Approximately two-thirds of the sample (66%; CI_95%: [64.27–67.43]) was seen by a racial/ethnic concordant provider. Adjusting for controls, concordance was strongly associated with provider communication of treatment options (OR_{[Concordant vs. Discordant]}=1.50; CI_95%: [1.15–2.13]). Patient inclusion in decision-making was associated only with patient race/ethnicity (OR_{[NH White vs. All others]}=1.41; CI_95%: [1.21–1.65]) and marital status (OR_{[Married vs. Not married]}=1.15; CI_95%: [1.04–1.28]).

Discussion: Providers were more likely to explain treatment options to patients of their same race/ethnicity, and shared decision-making was more common among Non-Hispanic White patients, regardless of provider race. Policymakers and public health officials should consider these findings when addressing disparities in access and outcomes among underserved populations. Further research is needed to understand the mechanisms behind these associations and the potential changes in the role of concordance in decision-making during the pandemic.