Perioperative Risk Factors Associated with Unplanned Neurological Intensive Care Unit Events Following Elective...

Our aim of this study was to identify risk factors and develop a prediction model for unplanned neurological intensive care unit (NICU) events after elective infratentorial brain tumor resection in order to propose an individualized admission to the NICU tailored to patient needs.

Patients admitted to our NICU between September 2018 and May 2021 after elective infratentorial brain tumor resection were reviewed. Prolonged NICU stays and unplanned NICU admissions were defined as unplanned NICU events. The prognostic model of unplanned NICU events was developed using a forward stepwise logistic regression analysis, and external validation was evaluated. The C-statistic was used to assess discrimination, and a smooth, nonparametric calibration line was used to assess calibration graphically in the model.

Of the 1,710 patients in the development cohort, unplanned NICU events occurred in 162 (9.5%). Based on the lesion type, a Karnofsky Performance Status score <70 at admission, longer duration of surgery, bleeding in the operative area evident on postoperative computed tomography, higher fibrinogen and blood glucose levels at admission, and more intraoperative blood loss were independently associated with unplanned NICU events. The external validation test showed good discrimination (C-statistic = 0.811) and calibration (Hosmer-Lemeshow P = 0.141) for unplanned NICU events.

Several patient and operative characteristics are associated with a greater likelihood of the occurrence of unplanned NICU events. In the future, we may be able to provide better help for the resource allocation of NICUs according to these risk factors and prediction models.