News Release

Antibiotic therapy for nearly 1 in 4 adults with pneumonia does not work

Peer-Reviewed Publication

American Thoracic Society


image: Antibiotic treatment ineffective for one in four adults with pneumonia. view more 

Credit: ATS

ATS 2017, WASHINGTON, DC -- Approximately one in four (22.1 percent) adults prescribed an antibiotic in an outpatient setting (such as a doctor's office) for community-acquired pneumonia does not respond to treatment, according to a new study presented at the 2017 American Thoracic Society International Conference.

"Pneumonia is the leading cause of death from infectious disease in the United States, so it is concerning that we found nearly one in four patients with community-acquired pneumonia required additional antibiotic therapy, subsequent hospitalization or emergency room evaluation," said lead author James A. McKinnell, MD, an LA BioMed (Los Angeles, CA) lead researcher and infectious disease specialist. "The additional antibiotic therapy noted in the study increases the risk of antibiotic resistance and complications like C. difficile ("C diff") infection, which is difficult to treat and may be life-threatening, especially for older adults."

Dr. McKinnell and colleagues conducted this study because current community-acquired pneumonia guidelines from the American Thoracic Society and the Infectious Disease Society of America, published in 2007, provide some direction, but large-scale, real-world data are needed to better understand and optimize antibiotic choices and to better define clinical risk factors that may be associated with treatment failure.

The researchers examined databases containing records for 251,947 adult patients who were treated between 2011 and 2015 with a single class of antibiotics (beta-lactam, macrolide, tetracycline, or fluoroquinolone) following a visit to their physician for treatment for community-acquired pneumonia. The scientists defined treatment failure as either the need to refill antibiotic prescriptions, antibiotic switch, ER visit or hospitalization within 30 days of receipt of the initial antibiotic prescription. The total antibiotic failure rate was 22.1 percent, while patients with certain characteristics -- such as older age, or having certain other diseases in addition to pneumonia -- had higher rates of drug failure. After adjusting for patient characteristics, the failure rates by class of antibiotic were: beta-lactams (25.7 percent), macrolides (22.9 percent), tetracyclines (22.5 percent), and fluoroquinolones (20.8 percent).

"Our findings suggest that the community-acquired pneumonia treatment guidelines should be updated with more robust data on risk factors for clinical failure," said Dr. McKinnell. "Our data provide numerous insights into characteristics of patients who are at higher risk of complications and clinical failure. Perhaps the most striking example is the association between age and hospitalization: Patients over the age of 65 were nearly twice as likely to be hospitalized compared to younger patients when our analysis was risk adjusted and nearly three times more likely in unadjusted analysis. Elderly patients are more vulnerable and should be treated more carefully, potentially with more aggressive antibiotic therapy."

Dr. McKinnell also stated that his study found substantial regional variations in treatment outcomes, which are not addressed in a specific way in the community-acquired pneumonia guidelines. In addition, the study showed that thousands of patients who suffer from other conditions - such as chronic obstructive pulmonary disease, cancer or diabetes - were not treated with combination antibiotic therapy or respiratory fluoroquinolone, as the guidelines recommend.

"While certain aspects of the guidelines need to be updated, some prescribers also have room for improvement and should implement the current guidelines," he concluded.


Abstract 8450

Clinical Predictors of Antibiotic Failure in Adult Outpatients with Community-Acquired Pneumonia

Authors: J. McKinnell1, P. Classi2, P. Blumberg3, S. Murty3, G. Tillotson2; 1UCLA - Los Angeles, CA/US, 2Cempra Pharmaceuticals - Chapel Hill, NC/US, 3Xcenda LLC - Palm Harbor, FL/US

Rationale: Antibiotic failure for community-acquired pneumonia (CAP) is associated with substantial morbidity and mortality and results in significant medical expenditures. Current CAP guidelines provide some direction on antibiotic selection in the outpatient setting, but large-scale, "real-world" data are needed to better understand and optimize antibiotic choice and to better define clinical risk factors which may be associated with treatment failure.

Methods: We conducted a retrospective cohort analysis of outpatient CAP from MarketScan® Commercial & Medicare Supplemental Databases between 2011 and 2015. Patients were ?18 years old and received antibiotic treatment following an outpatient visit for CAP (based on ICD-9-CM codes). Patients were required to have a monotherapy antibiotic prescription claim for one of the following drug classes: macrolides, fluoroquinolones, beta-lactams or tetracyclines. Treatment failure was defined as any of the following CAP related events occurring within 30 days of initial antibiotic fill: antibiotic refill, antibiotic switch, ER visit or hospitalization. A multivariable logistic regression model was employed to determine predictors of treatment failure.

Results: 251,947 adult outpatients met inclusion criteria. Mean age was 52.2 years with 47.7% male and 21.5% Medicare enrollees. The majority of patients were prescribed azithromycin (n=101,492, 40.3%) followed by levofloxacin (n=95,019, 37.7%). Total antibiotic failure rate was 22.1% (n=55,741/251,947) and comprised of: antibiotic refill (n=11,493/55,741, 20.6%), antibiotic switch (n=39,397/55,741, 70.7%), ER visit (n=1,835/55,741, 3.3%) and hospitalization (n=3,015/55,741, 5.4%). Multivariate predictors of antibiotic failure included: diagnosis of pneumococcal pneumonia (p<0.02), older age (p<0.0001), and female gender (p<0.0001). Various comorbidities were associated with higher rates of antibiotic failure including: hemiplegia/paraplegia (OR=1.33 [1.17-1.51]), rheumatologic disease (OR=1.28 [1.21-1.35]), chronic pulmonary disease (OR=1.25 [1.21-1.29]), cancer (OR=1.14 [1.09-1.18]), diabetes (OR=1.07 [1.04-1.10]) and asthma (OR=1.05 [1.01-1.10]). With each increasing Charlson Comorbidity Index (CCI) score, the probability of antibiotic failure increased (OR=1.16 [1.13-1.20] for CCI=1, OR=1.22 [1.18-1.26] for CCI=2, OR=1.44 [1.39-1.49] for CCI ?=3) compared to CCI=0. After adjusting for baseline patient characteristics, beta-lactams were associated with the highest antibiotic failure rate (25.7%), followed by macrolides (22.9%), tetracyclines (22.5%), and fluoroquinolones (20.8%).

Conclusion: Approximately one in four adult outpatients prescribed antibiotic monotherapy for CAP fail treatment. We identified multiple predictors of failure and clinically meaningful differences between antibiotic classes. Prescribers should be aware of those CAP patients at risk for poor outcomes and consider these factors to guide a comprehensive treatment plan, including more appropriate antibiotic treatment.


Dacia Morris
ATS Office 212-315-8620 (until May 17)
Cell Phone 917-561-6545

Session: A95 Acute Pneumonia: Clinical Studies
Abstract Presentation Time: Sunday, May 21, 3:15 p.m. ET
Location: West Salon G-1 (South Building, Street Level), Walter E. Washington Convention Center

Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.