Construction and verification of risk early warning model for senile chronic obstructive pulmonary disease complicated with sarcopenia

Chen Xi; Xiao Jiangqin; Huang Li

doi:10.3969/j.issn.1674-3245.2023.06.010

Abstract：Objebtive To explore the construction and verification of risk early warning model for senile chronic obstructive pulmonary disease complicated with sarcopenia.Methods A retrospective analysis of 139 elderly patients with COPD admitted to the department of respiratory medicine of our hospital from January 2020 to December 2021 was conducted.According to whether COPD patients had sarcopenia,they were divided into two groups,sarcopenia group(n=52)and non-sarcopenia group(n=87).The risk factors affecting elderly COPD complicated with sarcopenia were analyzed.The prediction model of concurrent sarcopenia risk was constructed,and 69 COPD patients from June 2020 to December 2021 were selected as the verification group to externally verify the constructed model.Results The result of univariate analysis showed that the incidence of sarcopenia in COPD patients was significantly different from age,BMI,smoking history,annual number of acute exacerbations,CAT score,FEV1/FVC%,and COPD course(P<0.05).Multivariate analysis showed that age≥80 years,BMI<25 kg/m2,number of acute exacerbations≥2,high CAT score,and long COPD course were risk factors for sarcopenia in COPD patients(P<0.05).High FEV1/FVC%was a protective factor for sarcopenia in COPD patients(P<0.05).Finally,the risk prediction model of COPD patients complicated with sarcopenia was logit P=-4.323 + 1.094×Xage + 0.914×XBMI + 0.837×Xsmoking history + 0.846×Xacute exacerbation times + 0.410×XCAT score-0.069×XFEV1/FVC%+ 0.561×XCOPD course.Hosmer-Lemeshow test was used to test the fitting degree of the prediction model of sarcopenia.The final model H-L test χ2= 4.362,P=0.823,which proved that the fitting degree of the model was good.The area under the ROC curve(AUC)was 0.756,95%CI was 0.673～0.839(P<0.05),sensitivity was 0.808,and specificity was 0.632.A total of 69 COPD patients from June 2020 to December 2021 were selected as the validation group to conduct external validation of the constructed model.H-L test χ2=8.720,P=0.366,which proved that the fitting degree of the model was good.The area under the ROC curve(AUC)was 0.808 and 95%CI was 0.691～0.926.Conclusion Age≥80 years old,BMI<25 kg/m2,annual number of acute exacerbations≥2,high CAT score,and longer COPD course are risk factors for COPD patients with sarcopenia.High FEV1/FVC%is a protective factor for COPD patients with sarcopenia.The risk prediction model of COPD patients complicated with sarcopenia has a good predictive performance and can evaluate the risk of COPD patients complicated with sarcopenia.