A nomogram to predict the risk of early postoperative ischemic events in patients with spontaneous intracranial hematoma

Abstract

Spontaneous intracranial hematoma (ICH) is the second leading cause of stroke and has a high risk of postoperative ischemic events (PIEs). But, the evidence on PIEs in ICH patients still lacks. Therefore, a retrospective study was carried out to screen the risk factors for PIEs and construct a visual predictive model. This was a retrospective study whose population were divided into two groups based on the occurrence of PIEs. Univariate logistic regression analysis was used to determine factors associated with PIEs. Multifactorial logistic regression analysis was used to screen risk factors and construct the early PIEs risk nomogram. In addition, impact of PIEs on patient prognosis and surgery related costs was assessed. Out of 122 ICH patients, 24 (19.7%) were diagnosed with PIEs. Coronary heart disease history, ischemic stroke history, regular shaped hematoma and platelet number were identified as risk factors for early PIEs. Early PIEs risk nomogram showed good calibration and discrimination of the data with concordance index of 0.846 (95% confidence interval, 0.747–0.945) which was confirmed to be 0.827 through bootstrapping validation. In addition, there was statistical difference in discharged Glasgow Coma Scale score (P\u2009=\u20090.046) and surgery related costs (p\u2009=\u20090.031) between PIEs group and nPIEs group. These results showed the early PIEs risk nomogram was accurate for prediction risks of PIEs and the occurrence of PIEs affects prognosis of patients, and increases surgery related costs.