Publications

IMPORTANCE: The Centers for Medicare & Medicaid Services (CMS) Bundled Payments for Care Improvement (BPCI) program was launched in 2013 with a goal to improve care quality while lowering costs to Medicare.

OBJECTIVE: To compare changes in the quality and outcomes of care for patients hospitalized with heart failure according to hospital participation in the BPCI program.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used a difference-in-difference approach to evaluate the BPCI program in 18 BPCI hospitals vs 211 same-state non-BPCI hospitals for various process-of-care measures and outcomes using American Heart Association Get With The Guidelines-Heart Failure registry and CMS Medicare claims data from November 1, 2008, to August 31, 2018. Data were analyzed from May 2022 to May 2023.

EXPOSURES: Hospital participation in CMS BPCI Model 2 Heart Failure, which paid hospitals in a fee-for-service process and then shared savings or required reimbursement depending on how the total cost of an episode of care compared with a target price.

MAIN OUTCOMES AND MEASURES: Primary end points included 7 quality-of-care measures. Secondary end points included 9 outcome measures, including in-hospital mortality and hospital-level risk-adjusted 30-day and 90-day all-cause readmission rate and mortality rate.

RESULTS: During the study period, 8721 patients were hospitalized in the 23 BPCI hospitals and 94 530 patients were hospitalized in the 224 same-state non-BPCI hospitals. Less than a third of patients (30 723 patients, 29.8%) were 75 years or younger; 54 629 (52.9%) were female, and 48 622 (47.1%) were male. Hospital participation in BPCI Model 2 was not associated with significant differential changes in the odds of various process-of-care measures, except for a decreased odds of evidence-based β-blocker at discharge (adjusted odds ratio [aOR], 0.63; 95% CI, 0.41-0.98; P = .04). Participation in the BPCI was not associated with a significant differential change in the odds of receiving angiotensin-converting enzyme inhibitors/angiotensin receptor blockers or angiotensin receptor-neprilysin inhibitors at discharge, receiving an aldosterone antagonist at discharge, having a cardiac resynchronization therapy (CRT)-defibrillator or CRT pacemaker placed or prescribed at discharge, having implantable cardioverter-defibrillator (ICD) counseling or an ICD placed or prescribed at discharge, heart failure education being provided among eligible patients, or having a follow-up visit within 7 days or less. Participation in the BPCI was associated with a significant decrease in odds of in-hospital mortality (aOR, 0.67; 95% CI, 0.51-0.86; P = .002). Participation was not associated with a significant differential change in hospital-level risk-adjusted 30-day or 90-day all-cause readmission rate and 30-day or 90-day all-cause mortality rate.

CONCLUSION AND RELEVANCE: In this study, hospital participation in the BPCI Model 2 Heart Failure program was not associated with improvement in process-of-care quality measures or 30-day or 90-day risk-adjusted all-cause mortality and readmission rates.

The care of Black adults is highly concentrated at a limited set of US hospitals that often have limited resources. In 2011, the Medicare Hospital Value-Based Purchasing (VBP) Program began financially penalizing or rewarding hospitals based on thirty-day mortality rates for target conditions (myocardial infarction, heart failure, and pneumonia). Because the VBP Program has disproportionately penalized resource-constrained hospitals caring for high proportions of Black adults since its implementation in 2011, clinicians, health system leaders, and policy makers have worried that the program may unintentionally be widening racial disparities in health outcomes. Using Medicare claims for beneficiaries ages sixty-five and older who were hospitalized for three target conditions at 2,908 US hospitals participating in the VBP Program, we found that thirty-day mortality rates were consistently higher for two of three conditions at hospitals with high proportions of Black adults compared with other hospitals. There was no evidence of a differential change in thirty-day mortality among all Medicare beneficiaries with targeted conditions at high-proportion Black hospitals versus other hospitals seven years after the implementation of the VBP Program. However, gaps in mortality between these sites did widen in the subgroup of Black adults with pneumonia. These findings highlight that important concerns remain about the regressive nature and equity implications of national pay-for-performance programs.

IMPORTANCE: The COVID-19 pandemic led to unprecedented disruptions in health care. Little is known about whether health care access and preventive health screenings among US adults have recovered to prepandemic levels, and how patterns varied by race and ethnicity.

OBJECTIVE: To evaluate health care access and preventive health screenings among eligible US adults in 2021 and 2022 compared with prepandemic year 2019, overall and by race and ethnicity.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study used data from US adults aged 18 years or older who participated in the National Health Interview Survey in 2021 and 2022. Survey weights provided by the National Health Interview Survey were used to generate nationally representative estimates. Data were analyzed from May 23 to November 13, 2023.

MAIN OUTCOMES AND MEASURES: Measures of health care access included the proportion of adults with a usual place for care, those with a wellness visit, and those who delayed or did not receive medical care due to cost within the past year. Preventive health screening measures included eligible adults who received blood pressure, cholesterol, or blood glucose screening within the past year (2021), as well as colorectal, cervical, breast, and prostate cancer screenings based on US Preventive Services Task Force guidelines.

RESULTS: The unweighted study population included 89 130 US adults. The weighted population included 51.6% females; 16.8% Hispanic, 5.9% non-Hispanic Asian (hereafter, Asian), 11.8% non-Hispanic Black (hereafter, Black), 62.8% non-Hispanic White (hereafter, White) individuals; and 2.9% individuals of other races and ethnicities (including American Indian, Alaska Native, Native Hawaiian or other Pacific Islander, or multiracial). After adjusting for age and sex, having a usual place for health care did not differ among adults in 2021 or 2022 vs 2019 (adjusted rate ratio [ARR] for each year, 1.00; 95% CI, 0.99-1.01). However, fewer participants had wellness visits in 2022 compared with 2019 (ARR, 0.98; 95% CI, 0.97-0.99), with the most pronounced decline among Asian adults (ARR, 0.95; 95% CI, 0.92-0.98). In addition, adults were less likely to delay medical care (ARR, 0.79; 95% CI, 0.73-0.87) or to not receive care (ARR, 0.76; 95% CI, 0.69-0.83) due to cost in 2022 vs 2019. Preventive health screenings in 2021 remained below 2019 levels (blood pressure: ARR, 0.95 [95% CI, 0.94-0.96]; blood glucose: ARR, 0.95 [95% CI, 0.93-0.96]; and cholesterol: ARR, 0.93 [95% CI, 0.92-0.94]). Eligible adults were also significantly less likely to receive colorectal cancer screening (ARR, 0.88; 95% CI, 0.81-0.94), cervical cancer screening (ARR, 0.86; 95% CI, 0.83-0.89), breast cancer screening (ARR, 0.93; 95% CI, 0.90-0.97), and prostate cancer screening (ARR, 0.86 [0.78-0.94]) in 2021 vs 2019. Asian adults experienced the largest relative decreases across most preventive screenings, while Black and Hispanic adults experienced large declines in colorectal cancer screening (ARR, 0.78; 95% CI, 0.67-0.91) and breast cancer screening (ARR, 0.83; 95% CI, 0.75-0.91), respectively. Differences in preventive screening rates across years persisted after additional adjustment for socioeconomic factors (income, employment status, and insurance coverage).

CONCLUSIONS AND RELEVANCE: Results of this cohort study suggest that, in the US, wellness visits and preventive health screenings have not returned to prepandemic levels. These findings support the need for public health efforts to increase the use of preventive health screenings among eligible US adults.

BACKGROUND: US adults often overpay for generic prescription medications, which can lead to medication nonadherence that negatively impacts cardiovascular outcomes. As a result, new direct-to-consumer online medication services are growing in popularity nationwide. Amazon recently launched a $5/month direct-to-consumer medication subscription service (Amazon RxPass), but it is unclear how many US adults could save on out-of-pocket drug costs by using this new service.

OBJECTIVES: To estimate out-of-pocket savings on generic prescription medications achievable through Amazon's new direct-to-consumer subscription medication service for adults with cardiovascular risk factors and/or conditions.

METHODS: Cross-sectional study of adults 18-64 years in the 2019 Medical Expenditure Panel Survey.

RESULTS: Of the 25,280,517 (SE ± 934,809) adults aged 18-64 years with cardiovascular risk factors or conditions who were prescribed at least 1 medication available in the Amazon RxPass formulary, only 6.4% (1,624,587 [SE ± 68,571]) would achieve savings. Among those achieving savings, the estimated average out-of-pocket savings would be $140 (SE ± $15.8) per person per year, amounting to a total savings of $228,093,570 (SE ± $26,117,241). In multivariable regression models, lack of insurance coverage (adjusted odds ratio [OR] 3.5, 95%CI 1.9-6.5) and being prescribed a greater number of RxPass-eligible medications (2-3 medications versus 1 medication: OR 5.6, 95%CI 3.0-10.3; 4+ medications: OR 21.8, 95%CI 10.7-44.3) were each associated with a higher likelihood of achieving out-of-pocket savings from RxPass.

CONCLUSIONS: Changes to the pricing structure of Amazon's direct-to-consumer medication service are needed to expand out-of-pocket savings on generic medications to a larger segment of the working-age adults with cardiovascular risk factors and/or diseases.

IMPORTANCE: Medicare's Hospital Value-Based Purchasing (HVBP) program will provide a health equity adjustment (HEA) to hospitals that have greater proportions of patients dually eligible for Medicare and Medicaid and that offer high-quality care beginning in fiscal year 2026. However, which hospitals will benefit most from this policy change and to what extent are unknown.

OBJECTIVE: To estimate potential changes in hospital performance after HEA and examine hospital patient mix, structural, and geographic characteristics associated with receipt of increased payments.

DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study analyzed all 2676 hospitals participating in the HVBP program in fiscal year 2021. Publicly available data on program performance and hospital characteristics were linked to Medicare claims data on all inpatient stays for dual-eligible beneficiaries at each hospital to calculate HEA points and HVBP payment adjustments.

EXPOSURES: Hospital Value-Based Purchasing program HEA.

MAIN OUTCOMES AND MEASURES: Reclassification of HVBP bonus or penalty status and changes in payment adjustments across hospital characteristics.

RESULTS: Of 2676 hospitals participating in the HVBP program in fiscal year 2021, 1470 (54.9%) received bonuses and 1206 (45.1%) received penalties. After HEA, 102 hospitals (6.9%) were reclassified from bonus to penalty status, whereas 119 (9.9%) were reclassified from penalty to bonus status. At the hospital level, mean (SD) HVBP payment adjustments decreased by $4534 ($90 033) after HEA, ranging from a maximum reduction of $1 014 276 to a maximum increase of $1 523 765. At the aggregate level, net-positive changes in payment adjustments were largest among safety net hospitals ($28 971 708) and those caring for a higher proportion of Black patients ($15 468 445). The likelihood of experiencing increases in payment adjustments was significantly higher among safety net compared with non-safety net hospitals (574 of 683 [84.0%] vs 709 of 1993 [35.6%]; adjusted rate ratio [ARR], 2.04 [95% CI, 1.89-2.20]) and high-proportion Black hospitals compared with non-high-proportion Black hospitals (396 of 523 [75.7%] vs 887 of 2153 [41.2%]; ARR, 1.40 [95% CI, 1.29-1.51]). Rural hospitals (374 of 612 [61.1%] vs 909 of 2064 [44.0%]; ARR, 1.44 [95% CI, 1.30-1.58]), as well as those located in the South (598 of 1040 [57.5%] vs 192 of 439 [43.7%]; ARR, 1.25 [95% CI, 1.10-1.42]) and in Medicaid expansion states (801 of 1651 [48.5%] vs 482 of 1025 [47.0%]; ARR, 1.16 [95% CI, 1.06-1.28]), were also more likely to experience increased payment adjustments after HEA compared with their urban, Northeastern, and Medicaid nonexpansion state counterparts, respectively.

CONCLUSIONS AND RELEVANCE: Medicare's implementation of HEA in the HVBP program will significantly reclassify hospital performance and redistribute program payments, with safety net and high-proportion Black hospitals benefiting most from this policy change. These findings suggest that HEA is an important strategy to ensure that value-based payment programs are more equitable.

OBJECTIVE: Women and underrepresented minorities (URMs) who are at an increased risk of presenting with severe peripheral artery disease (PAD) and have different responses to treatment compared with non-Hispanic White males yet are underrepresented in PAD research.

METHODS: ELEGANCE is a global, prospective, multi-center, post-market registry of PAD patients treated with drug-eluting device that aims to enroll at least 40% women and 40% URMs. The study design incorporates strategies to increase enrollment of women and URMs. Inclusion criteria are age ≥18 years and treatment with any commercially available Boston Scientific Corporation drug-eluting device marketed for peripheral vasculature lesions; exclusion criterion is life expectancy <1 year.

RESULTS: Of 750 patients currently enrolled (951 lesions) across 39 sites, 324 (43.2%) are female and 350 (47.3%) are URMs (21.6% Black, 11.2% Asian, 8.5% Hispanic/Latino, and 5.3% other). Rutherford classification is distributed differently between sexes (P = .019). Treatment indication differs among race/ethnicity groups (P = .003). Chronic limb-threatening ischemia was higher for Black (38.3%) and Hispanic/Latino (28.1%) patients compared with non-Hispanic White (21.8%) and Asian patients (21.4%). De-novo stenosis was higher in Asian patients (92.3%) compared with Black, non-Hispanic White, and Hispanic/Latino patients (72.2%, 68.7%, and 77.8%, respectively; P < .001). Mean lesion length was longest for Black patients (162.7 mm), then non-Hispanic White (135.2 mm), Asian (134.8 mm), and Hispanic/Latino patients (128.1 mm; P = .008).

CONCLUSIONS: Analyses of data from the ELEGANCE registry show that differences exist in baseline disease characteristics by sex and race/ethnicity; these may be the result of other underlying factors, including time to diagnosis, burden of undermanaged comorbidities, and access to care.