Establishment Of Early Multi-indicator Prediction Models Of Moderately Severe Acute Pancreatitis And Severe Acute Pancreatitis

Objective To investigate the prognostic factors related to the severity of acute pancreatitis(AP)and to establish early multi-indicator prediction models of moderately severe AP(MSAP)and severe AP(SAP).Methods Clinical data of 469 patients who met the inclusion criteria between January1 st,2015 and June 30 th,2020,at the First Affiliated Hospital of Fujian Medical University and between January 1st,2012,and October 31 st,2020,at the Affiliated Union Hospital of Fujian Medical University were retrospectively collected and analyzed,including 202 cases of mild AP,240 cases of MSAP,and 27 cases of SAP.The patients data were subjected to univariate analysis and binary logistic regression analysis for independent prognostic factors of the severity of AP.The unweighted predictive score(unw Score)and weighted predictive score(w Score)for MSAP and SAP were established.The receiver operator characteristic(ROC)curves of independent prognostic factors and prediction models were produced.The cutoff values were determined while the area under the curve(AUC),the sensitivity,the specificity,the positive predictive values(PPV),the negative predictive values(NPV)and the accuracy rates were calculated to verify the predictive efficiency.Results1.Univariate analysis showed that there was statistically significant difference in white blood cell(WBC)(Z=-8.605,P=0.000),hemoglobin(Z=-3.322,P=0.001),hematocrit(Z=-2.200,P=0.028),red blood cell volume distribution width(RDW)(Z=-2.402,P=0.016),glucose(Z=-4.141,P=0.000),aspartate aminotransferase(Z=-2.249,P=0.025),lactate dehydrogenase(LDH)(Z=-8.001,P=0.000),albumin(Z=-3.271,P=0.001),total cholesterol(Z=-3.502,P=0.000),triglycerides(Z=-7.423,P=0.000),D-dimer(Z=-6.633,P=0.000),bicarbonate(Z=-5.706,P=0.000),C-reactive protein(CRP)(Z=-6.875,P=0.000),procalcitonin(Z=-6.017,P=0.000),serum calcium(Z=-5.196,P=0.000),corrected calcium(Z=-4.460,P=0.000),serum phosphorus(Z=-2.022,P=0.043),serum sodium(Z=-4.346,P=0.000)and serum potassium(Z=-2.530,P=0.011)between the two groups.There was no significant difference in platelet(PLT)(Z=-0.894,P=0.371),RDW/PLT(Z=-0.778,P=0.437),total bilirubin(Z=-1.857,P=0.063),blood urea nitrogen(BUN)(Z=-1.502,P=0.133),serum creatinine(Scr)(Z=-1.040,P=0.298)and BUN/Scr(Z=-0.129,P=0.898).2.Binary logistic regression analysis showed that WBC(OR=1.223,95% confidence interval [CI]: 1.153～1.297,P=0.000),LDH(OR=1.001,95%CI:1.000～1.001,P=0.004),CRP(OR=1.004,95%CI:1.001～1.007,P=0.005),Scr(OR=1.006,95%CI:1.000～1.012,P=0.041),TG(OR=1.078,95%CI:1.031～1.127,P=0.001),D-dimer(OR=1.186,95%CI:1.084～1.297,P=0.000),and serum potassium(OR=1.819,95%CI:1.160～2.853,P=0.009)were independent prognostic risk factors of the severity of AP,and serum calcium(OR=0.235,95%CI:0.070～0.782,P=0.018)was an independent prognostic protective factor.Respectively,AUC were 0.732,0.716,0.684,0.528,0.700,0.679,0.568 and 0.640,the sensitivity were 72.28%,78.65%,71.54%,16.10%,54.31%,65.17%,32.58% and 42.32%,the specificity were 68.32%,56.93%,64.85%,95.54%,77.72%,65.35%,85.64% and 83.17%,the PPV were 75.10%,70.71%,72.90%,82.69%,76.32%,71.31%,75.00% and 76.87%,the NPV were 34.91%,66.86%,63.29%,46.28%,56.27%,58.67%,49.01% and 52.17%,and the accuracy rates were 70.58%,69.30%,68.66%,50.32%,64.39%,65.25%,55.44% and 59.91%.3.The unw Score(AUC=0.854)and w Score(AUC=0.837)were superior to acute physiology and chronic health evaluation Ⅱ score(AUC=0.526),bedside index for severity in AP score(AUC=0.766),Ranson score(AUC=0.693)in predicting MSAP and SAP,which were equivalent to modified computed tomography severity index score(AUC=0.823).Respectively,the sensitivity were 67.42% and 70.41%,the specificity were 87.13% and 81.19%,the PPV were 87.38% and 83.19%,the NPV were66.92% and 67.48%,and the accuracy rates were 75.91%和 75.05%.Conclusions WBC,LDH,CRP,Scr,TG,D-dimer and serum potassium are independent prognostic risk factors of the severity of AP while serum calcium is an independent prognostic protective factor.The early multi-indicator prediction models of MSAP and SAP have a good predictive efficiency,which could provide a valuable clinical reference for prediction and treatment.