First-day intracranial pressure correlates with ICU mortality in subarachnoid hemorrhage patients: an analysis of the MIMIC-IV database.

Background: The purpose of this study was to explore the optimal intracranial pressure (ICP) distribution of subarachnoid hemorrhage (SAH) patients on the first day in the intensive care unit (ICU) using data mining methods. Methods: Continuous ICP monitoring records of 176 SAH patients on the first ICU Day were collected from the MIMIC-IV database, comprising individuals treated at Beth Israel Deaconess Medical Center (Boston, MA) ICU between 2008 and 2019. Data underwent five-step processing: rounding (to hour point), missing value imputation, resampling, min-max normalization, and time-series clustering. Four unique clusters of the SAH cohort were identified, differing by daily average and variance of ICP. Propensity score estimation was used to determine the average treatment effect of ICP management on ICU mortality based on the X-tile recommended cut-off point. Results: The study cohort comprised 176 patients (mean age 58.8 ± 14.54 years; 58.0% [102/176] female) who met the inclusion criteria. The daily average was the only statistically significant factor in the propensity score estimation. A daily average ICP of 14 mmHg was identified as the cut-off point. The group with a daily average ICP above or below the cut-off point was an independent ICU mortality predictor in the multivariate analysis, with the largest odds ratio value among included variables. Notably, the daily average ICP was higher in ICU deaths than in survived patients under similar first-day fluid balance (ICU deaths vs. survived patients: P Conclusion: In the study cohort collected from MIMIC-IV SAH patients, using 14 mmHg as the cut-off point for the daily average ICP on the first ICU Day demonstrated favorable ICU mortality outcomes.