Prediction model construction of disease severity in patients with active ulcerative colitis based on inflammatory indicators and intestinal flora indicators
Objective To construct the prediction model of disease severity in patients with active ulcerative colitis based on inflammatory indicators and intestinal flora indicators to provide more reference for follow-up clinical diagnosis and individualized treatment plan.Methods Totally 361 patients with active ulcerative colitis diagnosed and treated in our hospital from January 2018 to January 2024 were retrospectively included and were divided into mild group(158 cases),moderate group(104 cases)and severe group(99 cases)according to the disease severity.Univariate and multivariate analyses of predictors of severity of active ulcerative colitis.Construction of prediction model for severity of active ulcera-tive colitis and analysis of prediction efficacy.Results Tumor necrosis factor-α[(107.53±20.30)ng/L vs(125.82±31.37)ng/L vs(159.23±44.60)ng/L]and interleukin-1β[(34.77±12.04)ng/L vs(48.83±14.10)ng/L vs(75.99±18.57)ng/L],interleukin-6[(26.39±5.05)ng/L vs(32.42±7.33)ng/L vs(53.03±12.68)ng/L],neutrophil/platelet ratio[(13.95±1.39)vs(18.29±1.90)vs(23.07±2.26)],Escherichia coli[(3.53±0.60)CFU/g vs(4.17±0.85)CFU/g vs.(4.85±1.07)CFU/g],Bifidobacterium[(1.79±0.65)CFU/g vs(1.22±0.43)CFU/g vs(0.49±0.10)CFU/g],Lactobacillus[(6.53±1.40)CFU/g vs(5.67±1.15)CFU/g vs(3.79±0.87)CFU/g]levels were significantly different(P＜0.05).The results of the ordered logistic regression model show:Tumor necrosis factor-α,interleukin-1β,interleukin-6,neutrophil/platelet ratio,Escherichia coli,bifidobacterium and Lactobacillus were all independent factors influencing the severity of active ulcerative colitis(OR=1.012,1.057,1.064,1.072,2.950,0.471,0.651,P＜0.05).Tumor necrosis factor α,interleukin1β,interleu-kin6,neutrophil/platelet ratio,Escherichia coli,bifidobacterium,lactobacillus and P-value prediction probability were used to predict the severity of active ulcerative colitis.The area under the curve was 0.632,0.632,0.593,0.589,0.764,0.607,0.595,0.845.Conclusion Both inflammatory laboratory indicators and intestinal flora charac-teristics can be used to predict the severity of active ulcerative colitis.
NETL
NSTL
万方数据
