Abstract
Objective To compare and analyze the predictive value of a variety of novel blood inflammation composite indicators in spontaneous intracerebral hemorrhage(sICH)patients'prognosis.Methods This retrospective study consecutively enrolled 329 sICH patients admitted to the Neurosurgical Intensive Care Unit of North China University of Science and Technology Affiliated Hospital between September 2022 and December 2024.Patients were categorized into favorable(with modified Rankin scale[mRS]score ≤2)or unfavorable(mRS score ≥ 3)prognosis groups based on 90-day post-onset assessments.Baseline and imaging data,included age,gender,smoking history,alcohol consumption history,admission systolic/diastolic blood pressure,medical history(diabetes,coronary heart disease,stroke),24-hour laboratory parameters after admission(albumin,hemoglobin,blood glucose,potassium,platelets,lymphocytes,monocytes,neutrophils,homocysteine,red cell distribution width[RDW]),site of hemorrhage(basil ganglia,lobes,cerebellum,brain stem),hemorrhage volume,the National Institutes of Health stroke scale(NIHSS)score.Systemic complications during hospitalization(pulmonary/urinary tract/bloodstream infections)were recorded.Head CT and CT angiography(CTA)were performed at 24 h after admission to measure hematoma volume.Calculate and analyze the comprised inflammatory indices including systemic immune-inflammation index(SII;platelets × neutrophils/lymphocytes),platelet-to-lymphocyte ratio(PLR),RDW to albumin ratio(RAR;RDW/albumin),hemoglobin to RDW ratio(HRR,hemoglobin/RDW),systemic inflammation response index(SIRI;neutrophils × monocytes/lymphocytes),lymphocyte-neutrophil-albumin ratio(LANR;lymphocytes × albumin/neutrophils),neutrophil-to-albumin ratio(NPAR;neutrophil％/albumin),glucose-to-lymphocyte ratio(GLR;glucose/lymphocytes),and glucose-to-potassium ratio(GPR,glucose/potassium).Variables with P＜0.05 in univariate analysis were included in a collinearity analysis(a tolerance＜0.1 or variance inflation factor[VIF]＞10 indicating collinearity).Significant non-collinear variables with P＜0.05 in univariate analysis were included in multivariate Logistic regression to identify factors influencing prognosis in sICH patients.Receiver operating characteristic(ROC)curves was utilized to compare the predictive power of each indictors through net reclassification improvement(NRI)and integrated discrimination improvement(IDI).With NRI/IDI＞0 indicated superiority of new models over the old ones(higher values denote greater improvement).Delong's test was performed to evaluate the area under the curve(AUC)differences.Results Among the 329 sICH patients enrolled(185 males,144 females;mean age[68±6]years,ranged 36-91),138 were categorized into the favorable prognosis group(mRS score≤ 2)and 191 into the unfavorable prognosis group(mRS score ≥ 3),all patients had received a standardized treatment.Significant differences were observed in alcohol consumption history,diabetes,prior stroke,albumin,hemoglobin,blood glucose,platelets,lymphocytes,monocytes,neutrophils,homocysteine,RDW,hematoma volume,lung infection,and NIHSS score between two groups(all P＜0.05).While no significant differences were observed in the rest baseline/imaging data between the two groups(all P＞0.05).The favorable prognosis group exhibited significantly lower GLR(4.67[3.76,5.92]vs.6.67[4.66,10.32]),SIRI(1.92[1.12,3.24]v.s.4.43[2.25,8.33]),PLR(125.11[105.11,156.20]vs.164.46[122.42,232.63]),RAR(0.28[0.27,0.29]vs.0.32[0.30,0.35]),and NPAR(0.18[0.13,0.21]vs.0.22[0.16,0.33];all P＜0.01);and significantly higher SII(1 632.90[882.18,2 429.78]vs.967.93[702.83,1 290.51]),LANR(9.63[7.75,13.10]vs.5.49[3.25,9.77]),and HRR(11.73[10.97,12.62]vs.10.89[10.00,11.82];all P＜0.01).No significant differences were found in GPR(2.15[1.80,2.65]vs.2.28[1.83,3.31])between the favorable and unfavorable prognosis groups(P=0.094).After excluding factors with multicollinearity(including,lymphocytes,platelets,neutrophils,monocytes,albumin,RDW and hemoglobin),the other factors were included in a multivariate analysis,which identified SIRI(OR,1.312,95％CI1.043-1.650,P=0.020),RAR(OR,1.578,95％CI 1.376-1.810,P＜0.01),and HRR(OR,0.641,95％CI 0.480-0.854,P=0.002)as independent predictors of poor 90-day prognosis for sICH patients.ROC analysis showed that the AUC of RAR,SIRI,and HRR were 0.862,0.739 and 0.683 respectively,with RAR demonstrating superior sensitivity(82.20％vs.SIRI 68.06％vs.HRR 51.83％)and specificity(78.26％vs.SIRI 69.57％vs.HRR 76.81％).IDI confirmed RAR's significantly greater predictive ability versus SIRI(IDI=0.210,P＜0.01)and HRR(IDI=0.263,P＜0.01),and SIRI have superior predictive ability over HRR(IDI=0.053,P=0.040).Delong test showed that RAR had significantly greater predictive power than both SIRI and HRR(both P＜0.01),while the predictive power of SIRI and HRR did not differ significantly(P=0.138).Conclusions The three blood inflammatory indicators(RAR,SIRI,HRR)are effective predictors of poor prognosis in sICH patients at 90 d from onset,and among the three composite inflammatory parameters,RAR is significantly more effective than SIRI and HRR indicators.关键词
脑出血/炎症/血液炎症复合指标/预测价值/预后Key words
Cerebral hemorrhage/Inflammation/Blood inflammation composite index/Predictive value/Prognosis
