Association of oral metoprolol exposure during heart failure hospitalization with death or readmission at 28-days and 6-months using a propensity score overlap-weighted EHR analysis.

Heart failure (HF) hospitalization is a high-risk transition period with frequent early post-discharge events; we evaluated whether oral metoprolol exposure during the index HF hospitalization is associated with lower all-cause death and/or all-cause readmission at 28-days and 6-months after discharge. An observational cohort analysis was performed using a de-identified Electronic Health Record (EHR) dataset with linked post-discharge outcomes including 2008 HF hospitalizations. The exposure was inpatient oral metoprolol (523; 26.0%) vs no oral metoprolol record (1485; 74.0%). Propensity scores for exposure were estimated via logistic regression using demographics, HF severity, comorbidities (Charlson index), vital signs, BMI, and laboratory measures. Overlap weighting improved group comparability and robust weighted Poisson models estimated adjusted risk ratios. Outcomes were all-cause death, all-cause readmission, and a composite of death or readmission at 28 and 180 days after discharge. At 28-days, metoprolol exposure was associated with a lower composite event risk (4.95% vs 9.71%; RR = 0.51; RD = - 4.76 percentage points; p = 0.001; NNT≈21). Mortality was lower (0.18% [2/523] vs 1.24% [35/1485]; RR = 0.14; RD = - 1.06 pp; p = 0.011) but deaths were rare and mortality estimates should be interpreted cautiously. Readmission was lower (4.77% vs 8.47%; RR = 0.56; RD = - 3.70 pp; p = 0.006). At 6-months, associations were attenuated: composite risk was similar (42.58% vs 42.50%; RR = 1.00; p = 0.982), readmission was similar (41.14% vs 40.18%; RR = 1.02; p = 0.744), and mortality showed a nonsignificant numerical reduction (1.43% vs 2.32%; RR = 0.62; p = 0.198). RAAS inhibitor exposure showed a similar early benefit for the 28-day composite (RR = 0.52; p = 0.014) but near-null results at 6-months (RR = 0.98; p = 0.860). In this overlap-weighted real-world cohort, inpatient oral metoprolol exposure was associated with fewer early readmissions and a directionally favorable mortality signal. However, because 28-day deaths were rare and granular time-updated in-hospital instability measures were unavailable, the mortality findings should be considered hypothesis-generating rather than definitive causal evidence.