Develop A Tool For Evaluating Severity Of Community-acquired Pneumonia Adapted For Patients With Type 2 Diabetes

Community-acquired pneumonia(CAP)is a common respiratory disease.CAP is the 6th leading cause of death in the world and threatens patients'lives.In the past three decades,the prevalence of diabetes increased dramatically.In China,the number of adults with diabetes was about 116 millions in 2019.Type 2 diabetes mellitus(T2DM)is the most common type.Therefore,CAP with T2DM is more common in clinical practice.Respiratory system is common infection site in patients with diabetes.Lung is another target organ damaged by diabetes.Pneumonia can be considered as a pulmonary complication of diabetes.In recent years,more and more attention has been paid to the relationship between diabetes and pneumonia.In August 2020,?diagnosis and treatment pathway for pneumonia in patients with diabetes mellitus:a Chinese experts'consensus?was developed by Infections Group of Respiratory Diseases Branch of Chinese Medical Association.The consensus was used to guide doctors to treat pneumonia in patients with diabetes.More and more research suggest that T2DM was associated with CAP.However,some questions remain unclear.Question one,epidemiological surveies show that the incidence of CAP was higher in patients with T2DM.But the effects of T2DM on the mortality of CAP were inconsistently in different studies.Question two,the outcomes varied enormously between different CAP patients.Therefore,it is very important to evaluate the severity of CAP.CURB-65 and Pneumonia Severity Index(PSI)are two classical tools for evaluating the severity of CAP.CURB-65 and PSI were recommended by CAP guidelines.Can CURB-65 and PSI be used to evaluate the severity of CAP in patients with T2DM?Question three,how to develop a tool for evaluating severity of CAP adapted for patients with T2DM.We explored these questions in these study.Part One The effect of type 2 diabetes on the prognosis of infectious inpatients and the cause of death in type 2diabetesObjective:The aim of the study was to explore the relationship between T2DM and mortality in infectious inpatients and analyze the major cause of death in infectious inpatients with T2DM.Methods:All subjects were inpatients from the First Hospital of Qinhuangdao between January 2015 and December 2018.The inclusion criteria included the following:1)all patients were hospitalized due to infections,2)subjects were men and women over 18 years of age.The exclusion criteria included the following:1)subjects with type 1 diabetes,2)subjects with other specific types diabetes,3)subjects with no clear type classification,4)subjects with pre-diabetes,5)pregnancy and 6)obstetric infection.All subjects were devided into non-diabetic group and T2DM group.Mortality records were collected.Results:This study enrolled 13916 patients.In these patients,2236patients(16.1%)were T2DM.The mortality was higher in patients with T2DM than patients without T2DM(T2DM 4.3%vs non-diabetes 1.7%,?~2=59.560,P<0.001).In multiple logistic regression analysis,death was considered as the dependent variables with sex,age,ethnicity,T2DM,coronary heart disease,heart failure,chronic kidney disease,hypoproteinemia and cerebrovascular disease as independent variables.T2DM was an independent risk factor of death in infectious inpatients(OR=1.539,95%CI:1.181?2.006,P=0.001).Lower respiratory tract(51.3%),digestive system(21.1%),urinary and genital tract(8.8%),skin and soft tissue(6.8%),peritoneal cavity(4.1%)ranked in the top five infection site in T2DM group.Lower respiratory tract(79 patients,81.4%,community-acquired pneumonia:47 patients,48.5%),digestive system(7 patients,7.2%),peritoneal cavity(2 patients,2.1%)ranked in the top three cause of death in T2DM group(97 patients death).Summary:1.The in-hospital mortality was higher in infectious patients with T2DM compared with non-diabetic patients.2.Lower respiratory tract was the common infection site in inpatients with T2DM and ranked in the top one in death,especially community-acquired pneumonia.Part Two The effect of type 2 diabetes on the prognosis of community-acquired pneumoniaObjective:The aim of the study was to explore the effects of T2DM on the severity and prognosis in CAP patients.Methods:All subjects were CAP inpatients from the First Hospital of Qinhuangdao between January 2015 and December 2018.The inclusion criteria included the following:1)all patients were hospitalized due to CAP,2)subjects were men and women over 18 years of age.The exclusion criteria included the following:1)subjects with type 1 diabetes,2)subjects with other specific types diabetes,3)subjects with no clear type classification,4)subjects with pre-diabetes,5)pregnancy and 6)obstetric infection.All subjects were devided into non-diabetic group and T2DM group.The information of pneumonia and mortality records were collected.Results:This study enrolled 2615 CAP patients.In these patients,582patients(22.3%)were T2DM.The frequencies of septic shock(6.4%vs 3.5%,?~2=9.380,P=0.002)and respiratory failure(22.7%vs 16.5%,?~2=11.866,P=0.001)was higher in patients with T2DM than patients without T2DM.The mortality was higher in patients with T2DM than patients without T2DM(T2DM 8.1%vs non-diabetes 4.2%,?~2=14.319,P<0.001).In multiple logistic regression analysis,death was considered as the dependent variables with sex,age,ethnicity,lung cancer,chronic obstructive pulmonary disease,asthma,T2DM,coronary heart disease,heart failure,chronic kidney disease,hypoproteinemia and cerebrovascular disease as independent variables.T2DM was an independent risk factor of death in CAP inpatients(OR=1.701,95%CI:1.158?2.499,P=0.007).The mortalities between those without T2DM and those with T2DM were stratified by age:?39 years 1.1%vs0.0%,40?49 years 0.6%vs 0.0%,50?59 years 0.8%vs 4.7%,60?69 years3.2%vs 6.6%,70?79 years 5.6%vs 8.2%and 80?years 8.3%vs 11.9%.In univariate logistic regression analysis,the odds ratios for mortality between CAP patients with and without T2DM declined after 50?59 years.Summary:1.CAP Patients with T2DM were more severe than CAP patients without T2DM.2.The mortality was higher in CAP patients with T2DM compared with non-diabetic patients.Whether in those with T2DM or in those without T2DM,the mortalities increases with age.However,the increased mortality appeared earlier in those with T2DM.Part Three Compare the accuracy of traditional tools for evaluating severity of community-acquired pneumonia in patients with and without type 2diabetesObjective:The aim of the study was to explore the accuracy of traditional tools(CRB-65,CURB-65 and PSI)for evaluating severity of community-acquired pneumonia in patients with and without type 2 diabetes.Methods:CAP patients in section two were included.The informations of CURB-65 and PSI were extracted.Patients without clinical data about CURB-65 and PSI were further excluded.CURB-65 and PSI were calculated.The area under curves(AUCs)were used to evaluate the abilities of CRB-65,CURB-65 and PSI class for predicting in-hospital mortality in patients with CAP.Results:This study enrolled 2365 CAP patients.In these patients,531patients(22.5%)were T2DM.The AUCs for predicting in-hospital mortality were 0.705?0.782(P<0.001).The AUC of PSI class was higher than the AUCs of CRB-65 and CURB-65(P<0.05).The AUCs for predicting in-hospital mortality were 0.728?0.798 in patients without T2DM and0.641?0.716 in patients with T2DM(P<0.001).The AUCs of CRB-65,CURB-65 and PSI class were lower in patients with T2DM than in patients without T2DM(PSI score,P=0.047).Summary:1.The performance of CRB-65,CURB-65 and PSI class for predicting in-hospital mortality in patients with CAP were moderate.CRB-65 and CURB-65 were simple but the accuracy is slightly poor.The performance of PSI was superior to CRB-65 and CURB-65 but too complicated.2.Compared with non-diabetes patients,the predictive performance of CRB-65,CURB-65 and PSI decreased in patients with T2DM.Part Four Develop a tool for evaluating severity of community-acquired pneumonia based on albumin adapted for patients with type 2diabetes(APUA model)Objective:The aim of our study was to developed a tool for predicting in-hospital mortality of CAP adapted for patients with type 2 diabetes.Methods:CAP patients with T2DM in section three were included.The primary outcome was in-hospital mortality.Variables to develop the nomogram were selected by using multiple logistic regression analysis.Predictive model was constructed by rms package of R software.Discrimination was evaluated using receiver operating characteristic curve(ROC curve).Calibration was evaluated using the calibration plot.Clinical effectiveness was evaluated using the decision curve analysis(DCA).Results:Multiple logistic regression analysis showed that Age,Pulse,Urea and Albumin were independent risk predictors.Based on these results,we developed nomogram(APUA model)for predicting in-hospital mortality of CAP in T2DM patients.A web calculator of APUA model was developed(https://doctorma.shinyapps.io/Dyn Nomapp/).In training set,the area under curve(AUC)of APUA model was 0.814(95%CI:0.770?0.853,P<0.001).The AUC of APUA model was higher than AUCs of albumin alone(AUC=0.695)and PSI(AUC=0.710)(P<0.05).The calibration plot showed excellent agreement between the predicted possibility in APUA model(P=0.802).In decision curve analysis,APUA model showed better net benefit than albumin alone and PSI class.In the validation sets,the results were similarly to the results of training sets.Summary:1.Based on albumin and combined with age,pulse and urea,APUA model is a simple and accurate tool for predicting in-hospital mortality of CAP adapted for patients with T2DM.2.According to APUA model,CAP patients can be devided into three group:low risk was defined as death risk<5%,intermediate risk was defined as death risk 5?19.9%,high risk was defined as death risk?20%.Part Five APUA model predicting long-term prognosis of community-acquired pneumonia in patients with type 2 diabetesObjective:The aim of our study was to explored the relationship between APUA model and long-term prognosis of CAP in patients with T2DM.Methods:CAP patients with T2DM in section three were included.Forty-seven patients died in hospital and were excluded.Four hundred and eighty-four patients were visited.According to APUA model,CAP patients were devided into three group:low risk group(in-hospital death risk<5%),intermediate risk group(in-hospital death risk 5?19.9%),high risk group(in-hospital death risk?20%).The primary outcome was all-cause death.Cox proportional hazard model was used to investigate the relationship between APUA model and long-term prognosis of CAP in patients with T2DM.Results:The median follow-up duration was 42.5 months.One hundred and nineteen patients died during the follow-up duration.The long-term mortality increased with the risk stratification of APUA model(low risk group15.7%,intermediate risk group 30.9%,high risk group 46.2%,P<0.05).Compared with patients in low risk group,patients in intermediate risk group and high risk group had significantly increased risk of long-term death,HRs(95%CIs)were 1.476(0.967?2.253,P=0.071)and 2.362(1.360?4.102,P=0.002),respectively,after adjusting for sex,age,lung cancer,chronic obstructive pulmonary disease,asthma,coronary heart disease,heart failure,chronic kidney disease and cerebrovascular disease.Summary:Risk stratification of APUA model was associated with long-term death,and can be used for evaluating the long-term prognosis in T2DM patients with CAP.Conclusions:CAP was the common cause of death in infection patients with T2DM.CAP patients with T2DM have higher mortality and worse outcome.Compared with non-diabetes patients,the predictive performance of traditional tools(CURB-65 and PSI)decreased in patients with T2DM.Based on albumin,APUA model is a simple and accurate tool for predicting in-hospital mortality of CAP adapted for patients with T2DM.APUA model can also be used for evaluating the long-term prognosis in T2DM patients with CAP.(WebcalculatorofAPUAmodel:https://doctorma.shinyapps.io/Dyn Nomapp/).