Comparison of Mortality and Therapy in Community Acquired Pneumonia

Background: Community associated pneumonia (CAP) is one the most common causes of hospital admissions, exceeding more than one million per year in the United States, contributing to 3.4% of inpatient mortality. Our objective was to compare 30-day mortality using therapies recommended for treatment of CAP. Methods: A multicenter retrospective analysis from four different hospitals was assessed from 2008 to 2013. The data was obtained from electronic medical records which included more than 70,000 patients. CAP patients were identified using discharge diagnostic codes during the years 2008-2013, as well as receiving therapy with ceftriaxone and azithromycin or a respiratory fluoroquinolone. Demographic data, antibiotic therapy, and Charlson comorbidity score was obtained to compare the study groups. Results: A total of 21,800 patients met the inclusion criteria for CAP. 1,740 patients were excluded as they received both beta-lactams and fluoroquinolones. The study included 20,600 patients. 11,201 patients (55.84%) received ceftriaxone with azithromycin, and 8,859 (44.16%) received fluoroquinolone therapy. The mortality rate for patients who received fluoroquinolone therapy was lower compared to the patients who received ceftriaxone plus azithromycin (3.56% vs 6.71%, p-value <0.001). Conclusions: Our study showed statistically significant lower 30-day mortality using fluoroquinolone therapy compared to ceftriaxone plus azithromycin for treatment of CAP. Prospective blinded randomized control trials would be needed to support this evidence. DOI: 10.18297/jri/vol2/iss2/4 Received Date: July 20, 2018 Accepted Date: August 9, 2018 Website: https://www.louisville.edu/jri Copyright: ©2018 the author(s). This is an open access article distributed under the terms of the Creative Commons AttributionNoDerivatives 4.0 International License (CC BY-ND 4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Affiliations: 1Department of Medicine, Division of Infectious Diseases, Henry Ford Hospital, Detroit MI 2Department of Epidemiology and Public Health, University of Maryland School of Medicine, Baltimore, MD 3Washington University, St. Louis, MO 4Baylor Scott & White Health, Dallas, TX 5Wayne State University School of Medicine, Detroit, MI *Correspondence To: Gina Maki, DO Work Address: Department of Internal Medicine Henry Ford Hospital 2799 West Grand Boulevard Detroit, Michigan, USA. Work Email: gmaki1@hfhs.org Work Phone: 313-916-3623 19 © ULJRI 2018 Vol 2, (2) ORIGINAL RESEARCH hospital length of stay, readmission status, and mortality within 30 days was collected from each of the four hospitals’ electronic medical records (EMR). Information was obtained from each participating centers’ EMR which was then entered into a single database. This database included approximately 70,000 patients. Data was de-identified and coded using explicit data specifications and uploaded into one large database. The study protocol was approved by the institutional review board (IRB) at each participating institution. Teleconferences, enrollments reports, and data audits were conducted between the four study sites to ensure uniform data collection. All hospitalized patients in these hospitals between January 1, 2008 and December 30, 2012 with a discharge diagnosis of pneumonia were identified. Patients were included in the present study if: 1) They were aged ≥18 years; 2) They had received either ceftriaxone plus azithromycin, levofloxacin, or moxifloxacin for the duration of the hospitalization with no other antibiotics administered. Exclusion criteria included: 1) Patients who received other antimicrobial agents; 2) Patients with diagnosis of pneumonia with no information on therapy; 3) Patients who received both therapies. The primary outcome of the study was 30-day all-cause mortality. In hospital mortality was used as a marker for 30-day mortality, as our database was unable to assess 30-day mortality. The patients’ severity of comorbidities was assessed using Charlson comorbidity index. Statistical Methods Univariate statistical analysis was used to test the association of demographic and clinical characteristics with all-cause discharge mortality. Categorical variables were analyzed using the Chi-squared test or Fisher’s exact test. Continuous variables were analyzed using an unpaired t-test. All variables with a p-value < 0.05 in the univariate analysis were included into the multivariate logistic regression. The stepwise selection method was performed to generate the final model (P≤0.05 required for variable entry, P≤0.10 required for variable removal). All p-values were two-sided. Analyses were performed by using SAS 9.4.

hospital length of stay, readmission status, and mortality within 30 days was collected from each of the four hospitals' electronic medical records (EMR). Information was obtained from each participating centers' EMR which was then entered into a single database. This database included approximately 70,000 patients. Data was de-identified and coded using explicit data specifications and uploaded into one large database. The study protocol was approved by the institutional review board (IRB) at each participating institution. Teleconferences, enrollments reports, and data audits were conducted between the four study sites to ensure uniform data collection.
All hospitalized patients in these hospitals between January 1, 2008 and December 30, 2012 with a discharge diagnosis of pneumonia were identified. Patients were included in the present study if: 1) They were aged ≥18 years; 2) They had received either ceftriaxone plus azithromycin, levofloxacin, or moxifloxacin for the duration of the hospitalization with no other antibiotics administered. Exclusion criteria included: 1) Patients who received other antimicrobial agents; 2) Patients with diagnosis of pneumonia with no information on therapy; 3) Patients who received both therapies. The primary outcome of the study was 30-day all-cause mortality. In hospital mortality was used as a marker for 30-day mortality, as our database was unable to assess 30-day mortality. The patients' severity of comorbidities was assessed using Charlson comorbidity index.

Statistical Methods
Univariate statistical analysis was used to test the association of demographic and clinical characteristics with all-cause discharge mortality. Categorical variables were analyzed using the Chi-squared test or Fisher's exact test. Continuous variables were analyzed using an unpaired t-test. All variables with a p-value < 0.05 in the univariate analysis were included into the multivariate logistic regression. The stepwise selection method was performed to generate the final model (P≤0.05 required for variable entry, P≤0.10 required for variable removal). All p-values were two-sided. Analyses were performed by using SAS 9.4.

Results
A total of 21,800 patients met our inclusion criteria for CAP, however 1,740 patients were excluded as they received both beta-lactams and fluoroquinolones. The study included 20,600 patients. 11,201 patients (55.84%) received ceftriaxone with azithromycin, and 8,859 (44.16%) received fluoroquinolone therapy. Demographic and clinical characteristics of patients are shown in Table 1. The mortality rate for patients who received fluoroquinolone therapy was lower compared to the patients who received ceftriaxone plus azithromycin (3.56% vs 6.71%, p-value <0.001).

Discussion
Our study demonstrated that the use of fluoroquinolone therapy for treatment of CAP decreased 30-day all-cause mortality compared to ceftriaxone plus azithromycin. Despite high comorbidity index in the cohort treated with fluoroquinolones, those treated with fluoroquinolones had better outcomes than those treated with ceftriaxone plus azithromycin. Our study is unique in several ways. First, the large sample size of this study strengthens the analysis as well as significantly different variables. Second, there have been limited studies conducted to compare ceftriaxone plus azithromycin to fluoroquinolone therapy with respect to mortality. The present study only included patients treated with these agents. Third, the data is collected from four different hospitals which are located in various locations throughout United States. Despite various geographic locations of the hospitals with different antibiograms, fluoroquinolones had mortality benefit over ceftriaxone plus azithromycin. Our study results are disconcordant with the meta-analysis of randomized controlled trials by Vardakas et al [7]. Their interpretation did reveal that fluoroquinolone therapy has a higher success rate in treatment of pneumonia compared to beta-lactam therapy with and without macrolides, however the data showed that there was no mortality benefit with fluoroquinolones. One of the possibilities several studies favor fluoroquinolone therapy is that there is different spectrum of coverage of fluoroquinolones including MSSA, gram-negative coverage including anti-pseudomonal activity for some agents such as levofloxacin, as well as atypical coverage [8]. Another factor may be the emergence of Streptococcus pneumonia which are resistant to macrolides. Multiple studies favor monotherapy with a fluoroquinolone over combination therapy. Querol-Ribelles et al. in a prospective observational cohort study showed improved mortality in patients treated with levofloxacin compared to those treated with combination therapy of ceftriaxone plus clarithromycin [9]. Another meta-analysis of randomized controlled trials by Raz-Pasteur et al. did not reveal a mortality benefit between monotherapy and combination therapy with a beta-lactam; however, the study was significant for decreased failure rates of therapy and treatment discontinuation in those receiving monotherapy with a fluoroquinolone [10]. Limitations of the present study include that it was a retrospective observational study with information gathered using only electronic medical records. The immune status of patients were unable to be obtained. Pneumonia severity index and CURB-65 scores, which are markers of disease severity, were unable to be obtained. Without calculating these scores, we cannot rule out the possibility of confounding by indication of the use of therapy based on disease severity. Many patients with CAP have no organism identified further limiting comparison of outcomes between the groups. Another limitation is that it is possible that activity against organisms isolated were different in one group versus another, which was not evaluated in this study. Finally, mortality during hospitalization was obtained, we did not have information on 30-day mortality, thus in hospital mortality was used as a marker for 30-day mortality.

Conclusions
Our study showed statistically significant lower 30-day mortality using fluoroquinolone therapy compared to ceftriaxone plus azithromycin for treatment of community acquired pneumonia. Further randomized controlled trials would be needed to support this evidence.
Funding Source: This collaborative research study was sponsored by Pfizer. Conflict of Interest: All authors declared no conflict of interest in relation to the main objective of this work.