Neighborhood Distress and PCI Outcomes: A Single-Center Study.

BACKGROUND: Reliable risk adjustment underpins fair percutaneous coronary intervention (PCI) benchmarking. Whether adding neighborhood socioeconomic distress enhances in-hospital mortality prediction is uncertain. PROJECT RATIONALE: To evaluate if the Distressed Communities Index (DCI) adds discrimination beyond a validated clinical model (predicted risk of mortality [PROM]). PROJECT SUMMARY: We analyzed 1,551 PCI admissions (CathPCI Registry, Barnes Jewish Hospital) using 3 prediction models: PROM alone, DCI alone, and PROM + DCI. Discrimination was measured by area under the curve (AUC; Delong test); DCI association was assessed by nonparametric tests. In-hospital mortality occurred in 54 patients (3.5%). PROM showed excellent discrimination (AUC: 0.93); DCI alone performed poorly (AUC: 0.61); PROM + DCI showed lower discrimination than PROM alone (AUC: 0.88). Higher DCI was associated with race (P < 0.001), sex (P = 0.009), insurance type (P < 0.001), PCI indication (P < 0.001), cardiogenic shock (P = 0.0004), and in-hospital mortality (P = 0.011), but not with out-of-hospital cardiac arrest. TAKE-HOME MESSAGE: Adding DCI did not improve prediction beyond PROM; DCI may be better for equity-oriented population targeting than individual risk adjustment.