Multicenter Clinical Study Of Community-acquired Pneumonia In Adults With Initial Treatment Failure

Objective: To investigate the clinical features and prognosis of CAP patients with initial treatment failure,analysis the correlation factors of treatment failure,to assist clinicians identify patients at risk of treatment failure.Methods: Observational retrospective multicenter study of 6056 adult hospitalized patients with CAP admitted to admitted to 13 centers in northern,central and southern China,from January to December 2014.Data on clinical measurements,underlying disease,severity of pneumonia,laboratory findings,imaging studies,microbiological investigations and outcome were collected.Evaluated effects of treatment after 72 h therapy,dividing into treatment succeeded group with 4464 cases and treatment failure group with 1592 cases.Retrospective study analyzed the correlation factor of initial treatment failure of CAP.Results: From a total 6056 consecutive patients with CAP who were enrolled during the study period,4464(73.7%)were classified as treatment successes,treatment failure occurred in 1592 of 6056 patients(26.3%),and 1482(93.1%)were nonresponding pneumonia,progressive pneumonia were fulfilled in 110 patients(6.9%).1)Comparison of demographic characteristics between the two groups:males,advanced age(>65 years),aspiration,pre-hospital drug-treatment history(long-term use of glucocorticoid,application of antibiotics before admission or immunosuppressive therapy),a positive history for chronic obstructive pulmonary disease(COPD),bronchial asthma,coronary atherosclerotic heart disease,congestive heart failure,hypertension,cerebrovascular diseases,liver cirrhosis,chronic kidney diseases were significantly more frequent inpatients who failed to improve,the difference was statistically significant(P<0.05).2)For comparison of pneumonia severity within groups: CURB-65 score and pneumonia severity index(PSI)score were independent risk factor for treatment failure in CAP patients(P<0.001).3)Comparison of main clinical syndromes and signs:there was significant difference in fever,dyspnea,disturbance of consciousness,cyanosis,wet rales,and lower extremity edema(P<0.05).4)Compared with the two groups of laboratory tests:the difference of white blood cell count,neutrophil count,hemoglobin and albumin and so on between the two groups were statistically significant(P<0.05).5)Comparison of imaging characteristics between the two groups:multilobar CAP,pleural effusion and consolidation associated with treatment failure(P<0.05).6)Distribution of pathogens in the two groups:6056 patients with CAP,of 4407 patients had some types of microbiological testing,and 1051 patients(23.8%)were positive for pathogens.A total of 1602 strains of pathogens were detected in 1051 CAP patients with positive pathogens.Among them,665 strains were detected in the treatment failure group and 937 strains were detected in the treatment effective group.The highest number of Candida albicans(36.1%)was detected in the treatment failure group,followed by 60 strains of Pseudomonas aeruginosa(9.0%),47 strains of Klebsiella pneumoniae(7.1%),and 32 strains of Aspergillus(4.8%),27 strains of Escherichia coli(4.1%),26 strains of Staphylococcus aureus,only 9 strains(1.4%)of Streptococcus pneumoniae 8 strains(1.2%)Mycoplasma pneumoniae were detected,only 2 strains(0.3%)of Haemophilus were detected.7)In the pathogen test program,Only 5.5% of patients had a nucleic acid detection and 2.7% of patients had antigen detection.And only5.4% of patients had viral nucleic acid detection and 2.3% of patients had a urinary antigen testing for S pneumoniae.7)Compared with the two groups of compliance with Guidelines:28.2% of treatment failure of CAP patients had not followed the guidelines and 18.1% of treatment successes of CAP patients had not followed the guidelines(P<0.001).8)The treatment failure patients have significantly longer hospital stays,higher in overall healthcare costs,higher rates of mortality,and with increasing risk of complications such as respiratory failure,cardiac insufficiency and septic shock.9)The multivariable logistic regression model showed that aspiration(OR 1.879,95%CI 1.666-2.120,P=0.001),a positive history for cerebrovascular disease(OR 1.353,95%CI 1.138-1.608,P=0.001),cirrhosis(OR 2.453,95%CI 1.179-5.107,P=0.016),dyspnea(OR 1.298,95%CI 1.130-1.490,P=0.002),disturbance of consciousness(OR 2.024,95%CI 1.563-2.621,P=0.001),application of antibiotics before admission(OR 1.239,95%CI 1.125-1.364,P=0.002),multilobar CAP(OR 1.150,95%CI 1.016-1.301,P=0.027),pleural effusion(OR 1.440,95%CI 1.246-1.664,P=0.012),CURB-65 score(OR 1.723,95%CI 1.420-2.092,P=0.002),PSI score(OR 1.127,95%CI 1.059-1.199,P=0.001),complicating with respiratory failure(OR 3.498,95%CI 2.026-6.041,P=0.002)and ARDS(OR 3.498,95%CI 2.026-6.041,P=0.002)were independent risk factors of clinical failure related to CAP on admission to the hospital.while bronchial asthma(OR 0.726,95%CI 0.544-0.986,P=0.029)was a protective factor for CAP treatment failure.conclusions: Aspiration,a positive history for cerebrovascular disease,cirrhosis,dyspnea,disturbance of consciousness,application of antibiotics before admission,multilobar CAP,pleural effusion,CURB-65 score,PSI score,complicating with respiratory failure and ARDS were independent risk factors of clinical failure related to CAP on admission to the hospital.while bronchial asthma is a protective factor for CAP treatment failure.Treatment failure in CAP was associated with high mortality and morbidity,remains a significant burden to both patients and healthcare systems.