基于泛免疫炎症值联合葡萄糖与血钾比值构建的高血压脑出血患者术后预后不良的列线图模型效能分析OA

Objective To develop a nomogram prediction model for poor postoperative prognosis in patients with hypertensive intracerebral hemorrhage(HICH)based on the pan-immune inflamma-tion value(PIV)and glucose-to-potassium ratio(GPR)and validate its efficacy.Methods A total of 370 patients with HICH who underwent surgical treatment from January 2022 to June 2025 were retro-spectively enrolled and divided into training set(259 cases)and validation set(111 cases)at a 7:3 ratio.Admission PIV and GPR were calculated.Patients were categorized into poor and good prognosis groups based on 6 month postoperative outcomes.Multivariable logistic regression analysis was used to identify factors influencing poor postoperative prognosis in HICH patients,and a nomogram predic-tion model was constructed.Model calibration was assessed using the Hosmer-Lemeshow test;pre-dictive performance was evaluated using the receiver operating characteristic(ROC)curve;internal validation was performed using the Bootstrap resampling method(1 000 repetitions),with the cali-bration curve used to assess model accuracy and the biascorrected C-index calculated;clinical utility was evaluated using decision curve analysis.Results The 6 month poor postoperative prognosis rate was 39.77％(103/259)in the training set and 36.04％(40/111)in the validation set.Multivari-able logistic regression showed that increased age,increased National Institutes of Health Stroke Scale(NIHSS)score,increased hematoma volume,elevated PIV,and elevated GPR were inde-pendent risk factors for poor postoperative prognosis in HICH patients,while increased Glasgow Co-ma Scale(GCS)score was an independent protective factor(P＜0.05).Based on these independ-ent factors,a nomogram prediction model was constructed.The Hosmer-Lemeshow test yieldedx2=12.396,P=0.134.The area under the ROC curve for predicting poor postoperative prognosis was 0.921(95％CI:0.881-0.951)in the training set and 0.901(95％CI:0.830-0.950)in the validation set.After 1 000 Bootstrap resampling iterations,the Cindex was 0.921(95％CI:0.917-0.925)for the training set and 0.889(95％CI:0.877-0.902)for the validation set.The calibration curves demonstrated good model fit,with Brier scores of 0.105 and 0.121,respec-tively.Decision curve analysis indicated that the model provided a positive net clinical benefit over a wide range of threshold probabilities.Conclusion Elevated PIV and GPR are independent risk fac-tors for poor postoperative prognosis in HICH patients.The nomogram prediction model constructed based on these factors demonstrates high predictive performance and may serve as a reference for postoperative prognostic assessment in HICH patients.