Temporal overlap of intraoperative hypotension and hypocapnia and postoperative outcomes after major open intracranial surgery: a time-weighted dual-hit cohort study

Abstract

Background Intraoperative hypotension reduces cerebral perfusion pressure, whereas hypocapnia may reflect either ventilation-related cerebral vasoconstriction or reduced pulmonary perfusion during a low-flow state. We evaluated whether the time-weighted overlap of hypotension and hypocapnia was associated with adverse postoperative outcomes after major open intracranial surgery. Methods We conducted a retrospective cohort study using INSPIRE version 1.4.2. Adults undergoing major open intracranial surgery were included if mean arterial pressure (MAP) and end-tidal carbon dioxide (EtCO2) had at least 80% synchronous coverage. Dual-hit cerebral hypoperfusion burden (DHCB) was defined as the proportion of synchronously valid monitoring time with simultaneous MAP <65 mmHg and EtCO2 < 30 mmHg. The primary outcome was a composite of in-hospital death, postoperative intensive care unit stay of at least 2 days, or prolonged postoperative hospitalization. Patient-clustered multivariable logistic regression simultaneously modeled hypotension-only, hypocapnia-only, and concurrent exposure burdens. Results The final cohort included 2,286 operations in 2,092 patients. Concurrent hypotension and hypocapnia occurred in 708 operations (31.0%), and the composite outcome occurred after 676 operations (29.6%). Each 5-percentage-point increase in DHCB was associated with higher odds of the composite outcome (adjusted odds ratio, 1.19; 95% CI, 1.10–1.28; P < 0.001), whereas hypotension-only and hypocapnia-only burdens were not independently associated with the outcome. Adding DHCB improved model fit but only minimally increased apparent discrimination (C-statistic, 0.764 to 0.767). Results remained consistent across progressively adjusted and sensitivity analyses. Restricted cubic spline analysis showed an overall association (P < 0.001) but no significant evidence of nonlinearity (P for nonlinearity = 0.143). The standardized risk increased from 27.3% at DHCB = 0% to 45.6% at DHCB = 30%, although estimates at higher burdens were less precise. Conclusions Greater temporal overlap between intraoperative hypotension and hypocapnia was associated with adverse short-term postoperative outcomes beyond the corresponding non-overlapping exposure burdens. DHCB should be interpreted as a hypothesis-generating marker of concurrent physiological exposure rather than evidence of two independent causal insults, and requires external validation with neurologically specific outcomes before clinical application.