DENVER – Penicillin allergies are no reason to forego empiric beta-lactam therapy in patients whose blood is infected with gram-negative bacilli, according to results from a retrospective study.
The penicillin derivatives are "worth the risk," said lead investigator Meghan Jeffres, Pharm.D., of the department of pharmacy practice at Roseman University of Health Sciences in Henderson, Nev.
Dr. Meghan Jeffres
None of 59 patients with documented penicillin reactions had an allergic reaction to the antibiotics, which were most often cephalosporins and carbapenems. A sixtieth patient had a non-IgE-mediated rash after failing ciprofloxacin – a non-beta-lactam (NBL) – and being switched to cefcapene, she reported at the annual Interscience Conference on Antimicrobial Agents and Chemotherapy.
Meanwhile, 24% (6 of 25) of the patients who were initially treated with an NBL, but only 15.3% (9/59) of patients who received beta-lactam (BL) empiric therapy, died from bacteremia complications; the mortality difference was driven by a lack of appropriate empiric therapy in the NBL group (52% vs. 18.6%).
BLs are already considered first-line for gram-negative bloodstream infections, in part because they are safer than fluoroquinolones, aminoglycosides, and other NBL options. Still, when a patient has a penicillin allergy, "clinicians are faced with a dilemma. ‘Do I risk the allergic reaction and give a beta-lactam, or do I avoid the risk of developing an allergic reaction and give a non-beta-lactam, even though it might be inferior clinically’," Dr. Jeffres explained.
It’s an important question because there are no randomized controlled trials to guide practice, she said at the conference.
The findings were part of a larger study that compared BL empiric therapy in 494 bacteremic patients with NBL empiric therapy in 104 others.
Compared with the NBL group, hospital stays were significantly shorter in patients treated with a BL (42 days vs. 28 days), rates of clinical failure were lower (33% vs. 23%), and mortality rates were reduced (20% vs. 13%).
"Only 60% of patients started empirically on [an NBL] ended up having an antibiotic that was active against their infecting pathogen," while the offending organism was susceptible to empiric therapy in more than 70% of BL patients. "The choice of a [BL] may only be superior based on the increased likelihood of having activity against the infecting pathogen," Dr. Jeffres said.
In addition to empiric BL therapy, mechanical ventilation at the start of antibiotics and a higher APACHE II (Acute Physiology and Chronic Health Evaluation II) score were both independent risk factors for clinical failure on multivariate regression, she reported.
BL patients were more likely to be immunosuppressed, but the two groups were otherwise well matched for the presence of trauma, cirrhosis, and other comorbidities. Patients were in their mid-50s, on average, in both groups.
Dr. Jeffres said she had no financial conflicts of interest.