Abstract
BACKGROUND: Outpatient blood pressure variability (BPV) predicts hospitalization and death in non-surgical patients independently of hypertension. We hypothesized that preoperative BPV predicts postoperative outcomes. METHODS: We assessed 22,233 veterans undergoing CABG, colectomy, hip replacement, pancreatectomy, carotid endarterectomy or AV-fistula with ≥10 outpatient BP's over three preoperative years. Calculating BPV as SD of systolic or diastolic BP, we used logistic regression considering demographics, comorbidities, and pre-admission cardiovascular medications to estimate odds ratios for 90-day mortality or readmission, MI, CVA, renal failure, and wound infection, choosing the lowest 5%ile of systolic/diastolic BPV for reference. RESULTS: Covariate-adjusted ORs for adverse outcomes increased as BPV increased. For instance, the highest 5%ile of systolic BPV had covariate-adjusted ORs of 2.96 and 1.78 for 90-day mortality and readmission. Systolic and diastolic BPV trended together but affected outcomes independently. CONCLUSIONS: Preoperative BPV predicts postoperative outcomes. BPV should be considered in individualized risk assessment and subgroup risk stratification.