| Objective1.Differences in basic information,lifestyle,comorbidities,medication compliance,heart function,heart failure indicators,and TCM syndrome types between repeated heart failure hospitalization group and non-recurrent hospitalization group.2.Taking the risk prediction of repeated hospital admission for acute heart failure as the outcome event,identifying the risk factors through univariate analysis.3.Multi-factor logistic regression to screen independent risk factors,establish equations and calculate relevant risk coefficients.4.Use R software to draw the line chart model and verify the accuracy of the model.MethodsThrough the retrospective survey method,the relevant data of the hospitalized patients with acute heart failure in 2018 in the hospitalization system of Guangdong Provincial Hospital of Traditional Chinese Medicine were screened.Index of heart failure,classification of TCM syndromes,etc.All variables were classified and transformed,and the χ2 test was used to compare the differences in related indicators between repeated hospitalizations and non-recurrent hospitalizations for heart failure.The meaningful results were further subjected to univariate and multivariate logistic regression to calculate OR and 95%CI.Finally,the included risk factors are drawn by R software to draw a line chart,and the feasibility of the model is verified by means of C-index,correction map,and bootstrap internal verification.Results1.A total of 280 eligible cases were collected,144 in the repeated hospitalization group and 136 in the non-recurrent hospitalization group.Among them,there were 9 clinically significant differences between the two groups of data.Specific include:age group(χ2＝12.463,P=0.002),pulmonary infection(χ2=7.381,P=0.007),medication compliance(χ2=22.026,P=0.000),cardiac function classification(χ2=29.664,P=0.000),LVEF(χ2=26.155,P=0.000),Qi deficiency(χ2=8.296,P=0.004),Yang deficiency(χ2=5.411,P=0.020),Phlegm(χ2＝8.420,P=0.004),Blood stasis(χ2=6.751,P=0.009).2.The univariate analysis showed that the correlation coefficients of the nine screening indicators were all positive,and all had no protective effect on the outcome(repeated hospitalization).Among them,the OR value of medication compliance was the largest at 4.291,which means that patients with poor medication compliance had a 3.91-fold higher risk of repeated hospitalizations than patients with better medication compliance.The OR values of the remaining indicators were:age(1.776),lung infection(1.975),cardiac function classification(2.372),LVEF(2.238),qi deficiency(2.100),yang deficiency(2.182),phlegm(2.019),blood Bruise(1.917).3.Logistic stepwise regression analysis was performed for 9 indicators,and all items were found to be independent risk factors.The regression equation was:Ln[p/(1-p)]=-4.509+0.617Xage+0.919Xlung infection+1.447Xmedication compliance+0.976Xcardiac function+0 712XLVEF+1.030XQi deficiency+0.970XYang deficiency+0 952XPhlegm turbidity+0.770Xblood stasis.Among them,the age group(<50 years old,50-75 years old,>75 years old)increased the level by 1 level,the risk of repeated hospitalization increased by 0.853 times(OR=1.853,95%CI=[1.239,2.773]);lung The risk of repeat hospitalization for infection is 2.506 times that of non-pulmonary infections(OR=2.506,95%CI=[1.351,4.648]);the risk of repeat hospitalization for poor medication compliance is 4.249 times better(OR=4.249,95%CI=[1.935,9.332]);cardiac function(Ⅱ-Ⅳ)increased by 1 level,the risk of repeated hospitalization increased by 1.653 times(OR=2.653,95%CI=[1.796,3.919]);LVEF(>50%,(40-50%,<40%)For each decrease in grade,the risk of repeated hospitalizations for patients with heart failure increases by 1.038 times(OR=2.038,95%CI=[1.390,2.988]).Among the TCM syndrome types,the hazard ratios of qi deficiency,yang deficiency,phlegm turbidity,and blood stasis to repeated hospitalizations were 2.801,2.638,2.592,and 2.160,respectively.4.Draw a line chart model according to the R software function,C-index=0.852,which indicates that the model has a good prediction ability.At the same time,the actual data line and the ideal data dash line in the calibration graph basically coincide,indicating that the model has a high degree of fit.In the end,the internal verification results showed that the C-index was 0.823,which was very similar to the original index of the model,that is,the model had better stability.Conclusion1.There are five independent risk factors for repeated western hospitalization of acute heart failure,including age,pulmonary infection,medication compliance,cardiac function classification,and LVEF.There are 4 independent risk factors for TCM syndromes:Qi deficiency,Yang deficiency,phlegm turbidity,and blood stasis.2.The disease and syndrome combined with the nomogram model based on the above risk factors have a good predictability for repeated admissions of acute heart failure,which is convenient for clinicians to identify critical patients early,control the patient’s prognosis,and better doctor-patient communication. |
PDF Full Text Request