OBJECTIVE

To determine the method of prediction of postoperative mortality in patients with hemorrhagic stroke.

MATERIAL AND METHODS

The study included 264 patients who underwent surgery for a hypertensive intracerebral hematoma at the N.V. Sklifosovsky Research Institute for Emergency Medicine from 1997 to 2020. Of these, 116 (43.9%) had putamen hematomas, 107 (40.5%) had subcortical hematomas, and 41 (15.5%) had cerebellar hematomas.

RESULTS

A retrospective analysis identified risk factors affecting the disease prognosis. Univariate and multivariate Cox regression analysis yielded a statistically significant mathematical model (p, Wald <0.0001) with good consistency (concordance coefficient=0.83). A nomogram was plotted to predict the survival of patients after surgery. The final model included the following factors that increase the probability of death: deep coma (HR=542.8; 95% CI 68.2—4321.3), decompression craniotomy (HR=7.64; 95% CI 4.06—14.39), hypertension (HR=6.48; 95% CI 1.80—23.40), recurrent hemorrhage (HR=4.78; 95% CI 2.66—8.59), postoperative infection complications (HR=3.24; 95% CI 1.56—6.71), pyramidal insufficiency (HR=3.34; 95% CI 0.85—13.04), as well as factors that reduce the risk of death: moderate coma, the presence of transverse dislocation, cerebellar symptoms, and moderate severity.

CONCLUSIONS

The analysis identified predictors of death in this category of patients, thus further narrowing the group of patients with hemorrhagic stroke for whom surgical treatment is indicated.