Intra‐operative hypotension threshold models and subdistribution hazard of postoperative pneumonia in patients having non‐cardiac surgery

INTRODUCTION: Intra-operative hypotension is common during general anaesthesia, but its association with postoperative pneumonia remains uncertain, partly because hypotension has been defined using different thresholds. We evaluated whether absolute thresholds or relative reductions from baseline mean arterial pressure were associated more strongly with postoperative pneumonia after non-cardiac surgery. METHODS: We conducted a retrospective multicentre cohort study of 151,036 adults undergoing non-cardiac surgery under general anaesthesia at three Chinese hospitals. Intra-operative hypotension was assessed using 12 metrics across nine thresholds, including absolute mean arterial pressure thresholds of < 50 mmHg to < 70 mmHg and relative reductions of 20-50% from baseline. The primary outcome was postoperative pneumonia within 30 days. RESULTS: Of 151,036 patients, postoperative pneumonia occurred in 3486 (2.3%). Relative reductions in mean arterial pressure were more consistently associated with postoperative pneumonia than absolute thresholds. At a 20% reduction from baseline, the number of intra-operative hypotension episodes was associated with postoperative pneumonia (subdistribution hazard ratio (sHR) 1.079, 95%CI 1.050-1.109, p < 0.001), as was the presence of intra-operative hypotension (sHR 1.108, 95%CI 1.068-1.149, p < 0.001). Absolute thresholds were not consistently associated with postoperative pneumonia after adjustment. Duration-related metrics were generally less consistent than depth- or recurrence-related metrics. Dose-response analyses showed that risk increased rapidly at low exposure levels and then plateaued across relative thresholds of 20-40%. DISCUSSION: Intra-operative hypotension was associated with postoperative pneumonia when defined as a relative reduction from baseline mean arterial pressure, whereas conventional absolute thresholds were less informative. These findings suggest that maintaining intra-operative blood pressure close to baseline may be more relevant to pulmonary risk than relying on fixed absolute thresholds alone.