Objective To construct a Cox proportional hazards prediction model for secondary intracranial hypertension in patients with acute subdural hematoma (ASDH) during the perioperative period and validate its effectiveness.
Methods Clinical data of 78 patients with ASDH were retrospectively collected and divided into secondary group (25 cases with secondary intracranial hypertension during perioperation) and control group (53 cases without secondary intracranial hypertension during perioperation). Differences in demographic indicators, comorbidities, clinical biochemical indicators, and imaging data between the two groups were compared. The Cox proportional hazards model was used to perform a multivariate analysis of independent risk factors that may affect secondary intracranial hypertension in ASDH patients during the perioperative period. A prediction model for secondary intracranial hypertension in ASDH patients during the perioperative period was established, and Harrell′s C index was calculated to assess the predictive accuracy of the model. The degree of agreement between the model prediction and actual risks was evaluated through a nomogram and calibration curve.
Results The six-month follow-up rate was 89.74% (70/78). Age, smoking history, hypertension, diabetes, preoperative Glasgow Coma Scale (GCS) score, Glasgow Outcome Scale (GOS) score, complex hematoma, intracranial hematoma volume, mean arterial pressure, glycated hemoglobin (HbA1c), international normalized ratio (INR), interleukin-6 (IL-6), and procalcitonin (PCT) in the secondary group showed statistically significant differences compared with the control group (P < 0.05). Age(OR=2.895; 95%CI, 1.264 to 6.633; P=0.022), smoking history (OR=2.146; 95%CI, 1.029 to 4.475; P=0.036), GOS score (OR=0.288; 95%CI, 0.112 to 0.741; P=0.015), HbA1c (OR=3.325; 95%CI, 1.243 to 8.894; P=0.028), INR (OR=2.746; 95%CI, 1.203 to 6.267; P=0.027), and PCT (OR=3.426; 95%CI, 1.335 to 8.795; P=0.019) were independent influencing factors for secondary intracranial hypertension in ASDH patients during the perioperative period. Harrell′s C index was 0.812 (95%CI, 0.789 to 0.872). The nomogram and calibration curve showed good consistency between the actual risk and the model prediction.
Conclusion Cox proportional hazards model for patients with acute subdural hematoma has high accuracy in predicting the risk of secondary intracranial hypertension during the perioperative period and is suitable for clinical promotion.