Publications

INTRODUCTION: The coronavirus 2019 pandemic disrupted medical care delivery for cerebrovascular disease. Little is known about the long-term changes in healthcare utilization and mortality among people who suffered cerebrovascular events.

METHODS: We used 2018-2022 data to identify all Medicare fee-for-service and Medicare Advantage beneficiaries with a diagnosis code for an acute cerebrovascular event (stroke, transient ischemic attack, hemorrhage). We used Poisson-lognormal regression models with an offset term (beneficiary years) to model incident rate ratios (IRR) for late COVID (2021-2022) versus pre-COVID (2018-2019) for outpatient visits, emergency department (ED) visits, hospitalizations, and mortality in the year following the index event, adjusting for age, sex, and race. We examined patterns overall and by rurality (urban, micropolitan, rural) and quartile (q) of social vulnerability index.

RESULTS: We identified 7,071,072 and 7,065,171 unique beneficiaries in the pre-COVID and late-COVID periods, respectively, with an acute cerebrovascular event. Our cohort was predominantly non-Hispanic white (73%), female (54%), and urban (83%). Compared to pre-COVID, outpatient care utilization in the year following the acute event was higher in 2021-2022 (IRR = 1.14, 95% confidence interval [CI]: 1.13-1.14), while ED visits (IRR = 0.92, 95% CI: 0.91-0.92) and inpatient hospitalizations (IRR = 0.94, 95% CI: 0.94-0.95) were lower. Mortality was higher during the late-COVID versus pre-COVID period (IRR = 1.24, 95% CI: 1.23-1.25). The mortality differences were similar across rural strata but higher in more socially disadvantaged areas (IRR 1.20, 1.22, 1.24, 1.27 for q1 [reference], p interaction = q2 = 0.12, q3 = 0.01, and q4 < 0.001).

CONCLUSIONS: Medicare beneficiaries who sustained acute cerebrovascular events had marked differences over the subsequent year in healthcare utilization patterns and increased mortality in the late-COVID period compared with their utilization and outcomes pre-COVID. The differences in mortality were the most pronounced for beneficiaries in socially disadvantaged areas.

BACKGROUND: The real-world uptake of the 2021 Kidney Disease: Improving Global Outcomes (KDIGO) blood pressure (BP) guideline, which lowered the systolic BP target to <120 mmHg for patients with chronic kidney disease (CKD), is poorly understood. We examined the adoption of the 2021 KDIGO systolic BP target in clinical practice and its association with clinical outcomes.

METHODS: The study was based on data from the Mass General Brigham healthcare network, an integrated healthcare system spanning primary to tertiary care in New England. In serial cross-sectional analysis, we identified ∼50 000 patients with CKD stage 3-4 in each year from 2020 to 2024 and assessed the annual proportion of patients within the 2021 KDIGO systolic BP target. In longitudinal analysis, we identified 18 996 patients with incident CKD stage 3-4 in 2014-2019 and evaluated the association between systolic BP above vs. within the target and clinical outcomes.

RESULTS: In serial cross-sectional analysis, 18.3% of patients with CKD had systolic BP within the 2021 KDIGO target in 2020 (pre-guideline). The proportion changed only marginally after the guideline's publication-18.0% in 2021 (absolute difference, -0.3% [95% CI, -1.2 to 1.3]), 19.3% in 2022 (absolute difference, 1.0% [95% CI, -0.1 to 2.0]), 20.0% in 2023 (absolute difference, 1.7% [95% CI, 0.2 to 3.1]), and 21.9% in 2024 (absolute difference, 3.6% [95% CI, 1.9 to 5.3]). In longitudinal analysis, patients with systolic BP above the 2021 KDIGO target exhibited higher risks of cardio-kidney end points, lower risk of hypotension, and no differences in other safety end points compared to those within the target.

CONCLUSIONS: Adoption of the 2021 KDIGO BP guideline remained limited in real-world practice. As of 2024, nearly 4 in 5 patients with CKD had systolic BP above the new guideline target.

Background: Although guidelines recommend invasive management for non-ST-elevation myocardial infarction (NSTEMI), there is considerable variability in the application of these recommendations across different hospitals, reflecting a lack of standardized clinical pathways and highlighting ongoing uncertainty in real-world practice. We sought to describe site-level variability in the use and timing of invasive angiography for NSTEMI and their association with in-hospital outcomes. Methods: Using NCDR Chest Pain-MI registry data (2019-2024), the rates and timing of invasive coronary angiography, if any, were characterized among patients with NSTEMI. Hierarchical logistic regression models were created to describe hospital-level variability in management using median odds ratios (MORs), adjusted for patient and site characteristics. Inverse probability weighting was used to estimate the association between treatment strategy and in-hospital outcomes. Results: We included 287,275 patients with Type-1 NSTEMI from 541 hospitals (age 67.6±13.3 years, 36.4% women). Invasive coronary angiography was performed in 87.1%, of whom 56.9% within 24 hours. Among those treated invasively, 66.1% received percutaneous coronary intervention. Older patients with more comorbidities were paradoxically more likely to receive conservative management or delayed intervention (>24 hours). Site-level variability for invasive strategy (vs. conservative) was large [MOR 2.85 (2.64-3.10)], as was early invasive treatment [MOR 1.67 (1.62-1.74)], particularly on weekends/holidays [MOR 1.89 (1.81-1.98)]. The use of any invasive strategy was associated with lower in-hospital mortality versus conservative management [weighted OR 0.36 (0.31-0.42)]. This finding was consistent across all baseline risk categories (P-interaction <0.001). Conclusions: Patients with Type-1 NSTEMI and higher-risk clinical profiles were not consistently prioritized to undergo early invasive management with substantial variability across hospitals. Invasive management was associated with lower in-hospital mortality compared with conservative treatment. Future randomized studies in the modern PCI era are needed to confirm our findings, and identify which patients benefit most and when intervention should occur.