Inappropriate Broad-Spectrum Antibiotic Prescriptions for Sepsis Tied to Higher Mortality

By Marilynn Larkin

April 24, 2020

NEW YORK (Reuters Health) - Both inadequate and unnecessarily broad empiric antibiotic prescribing for community-onset sepsis were associated with higher mortality in a retrospective study.

"Broad-spectrum antibiotics are frequently administered to patients with community-onset sepsis even though resistant organisms are very uncommon," Dr. Chanu Rhee of Harvard Medical School and Brigham and Women's Hospital in Boston told Reuters Health by email. "Furthermore, unnecessary administration of broad-spectrum antibiotics is associated with higher risk-adjusted mortality."

"Potential reasons for worse outcomes," he said, "include more antibiotic toxicities, such as renal failure, C. difficile infection, disruption of the gut microbiome, and higher risk of subsequently developing antibiotic-resistant infections."

Dr. Rhee and colleagues studied more than 17,000 adults with community-onset sepsis (median age, 69; 56% women) admitted to 104 US hospitals between 2009-2015.

As reported in JAMA Network Open, 16.4% of patients died in the hospital. The most common culture-positive sites were urine (52.1%), blood (40%), and the respiratory tract (16.7%) and the most common pathogens were E. coli (33.7%), S. aureus (21.3%), and Streptococcus species (13.5%).

Patients with resistant organisms were more likely than those without to have more comorbidities (median AHRQ Elixhauser Comorbidity Index score, 12 vs. 11), a pulmonary infection (41.4% vs. 29.9%) or positive respiratory cultures (29.9% vs. 12.1%), and to require vasopressors (36% vs. 30.9%), mechanical ventilation (28.3% vs. 19.2%) or ICU admission (50.1% vs. 44.4%). They were also more likely to die in the hospital (19.9% vs. 15.3% - crude OR, 1.38).

Inadequate empiric antibiotic therapy was defined as targeting one pathogen nonsusceptible to all antibiotics administered on the first or second day of treatment. Among the more than 15,000 cases in which all antibiotic-pathogen susceptibility combinations could be calculated, most patients (81.6%) received adequate empiric antibiotics.

Unnecessarily broad empiric therapy was use of antibiotics targeted against resistant organisms when none had been isolated. Empiric therapy targeted resistant organisms in 67% of cases (mostly using vancomycin and anti-pseudomonal beta-lactams). However, resistant organisms were uncommon: methicillin-resistant Staphylococcus aureus (MRSA), 11.7%; ceftriaxone-resistant gram-negative (CTX-RO),13.1%; vancomycin-resistant Enterococcus (VRE), 2.1%; and extended-spectrum beta-lactamase (ESBLs), 0.8%.

The net prevalence for at least one resistant gram-positive organism (i.e., MRSA or VRE) was 13.6%.

Similarly, for at least one resistant gram-negative organism (i.e., CTX-RO, ESBL, or carbapenem-resistant Enterobacteriaceae), the net prevalence was 13.2%.

Notably, the prevalence of resistant organisms was higher in patients with vs without septic shock (28.7% vs 24.2%).

Overall, after detailed risk adjustment, higher mortality was linked with inadequate (odds ratio, 1.19) and unnecessarily broad empiric antibiotics (OR, 1.22).

The authors conclude, "These findings underscore the need for better tests to rapidly identify patients with resistant pathogens and for more judicious use of broad-spectrum antibiotics for empiric sepsis treatment."

Dr. Gopi Patel, Associate Professor of Infectious Diseases at the Icahn School of Medicine at Mount Sinai in New York City, said the findings provide "a reason to take pause and reevaluate practice, especially in (sepsis) patients without evidence of shock."

"Major advances in molecular diagnostics have changed the way we approach antibiotic stewardship," he commented in an email to Reuters Health. "Many clinical laboratories are able to detect pathogens, especially in blood, rapidly, and clinicians can subsequently deescalate...therapy sooner, thus decreasing unnecessary broad-spectrum antibiotic exposure."

"In addition, we have seen marked changes in antibiotic resistance in our communities," said Dr. Patel, who was not involved in the study. "Regional and institutional antimicrobial susceptibility patterns should be noted, and institutional sepsis and treatment algorithms updated to accommodate changes in our community microbiomes. The authors acknowledge this limitation, as well as the fact that in many cases of sepsis we don't recover a pathogen."

"Rapid diagnostics are key," he said. "If we can identify the etiology of the sepsis, we can better treat these patients and cause the least amount of unintended harm. Clinicians may choose different empiric antibiotics for a urinary tract infection versus a pneumonia. Symptom-driven diagnostic testing and syndrome-driven antibiotic selection remain important."

"It is difficult to not throw the kitchen sink of antibiotics to a patient presenting with sepsis, but careful assessment of potential source of the infection, knowledge of local microbiology and appropriate de-escalation of antibiotics likely will improve your patient's overall outcome," Dr. Patel concluded.

SOURCE: JAMA Network Open, online April 16, 2020.