Objective To analyze risk factors of acute respiratory failure(ARF)in patients with hypertriglyceridemia acute pancreatitis(HTG-AP)and construct a risk prediction model.Methods A total of 222 HTG-AP patients were included in this study and divided into the non-ARF group(176 cases)and the ARF group(46 cases)according to diagnostic guidelines for ARF.Clinical data of the two groups were compared and the predictive factors were screened.These selected factors were then utilized in a multivariate Logistic regression analysis to construct a Logistic regression model.Subsequent evaluation of the model′s predictive ability,accuracy and clinical utility was conducted through ROC,curve analysis,calibration plot examination and decision curve analysis(DCA),respectively.Results Compared with the non-ARF group,the levels of high density lipoprotein cholesterol(HDL-C),low density lipoprotein cholesterol(LDL)-C and albumin(ALB)were decreased in the ARF group(P＜0.05),while the levels of creatinine(Cr),urea nitrogen(BUN),aspartate aminotransferase(AST)and C-reactive protein(CRP)were increased,and the incidence of pleural fluid and ascites was also increased(P＜0.05).Multivariate Logistic regression analysis showed that higher levels of Cr and AST,lower levels of ALB,HDL-C and ascites were independent risk factors for HTG-AP complicated ARF(P＜0.05).Based on these results,a column-line prediction model for HTG-AP complicated ARF was established.After internal verification,the area under curve(AUC)of receiver operating characteristic(ROC)curve of the nomogram model was 0.952(95%CI:0.923-0.981),the Youden index was 0.808 and the sensitivity and specificity were 93.33%and 87.43%,respectively.The calibration curve showed that the probability of HTG-AP concurrent ARF predicted by the model was in good agreement with the actual probability.The DCA curve showed that the model had certain clinical value.Conclusion The nomogram prediction model combined could provide a scheme for the clinical prevention of HTG-AP complicated with ARF.