A large new review of hospital antibiotic trials has reopened a question that has followed cefepime for nearly two decades. The analysis, published in JAMA Network Open, estimated a 94.4% probability that cefepime is associated with a higher risk of death within about 30 days than other beta-lactam antibiotics, the drug family that includes penicillins and cephalosporins.
That figure does not mean cefepime raises the risk of death by 94%. Across all included trials, 6.6% of patients given cefepime died, compared with 6.2% of patients given comparison antibiotics, according to a summary by Contagion Live.
Cefepime is a commonly used, injectable antibiotic for serious infections in hospitalized patients, including pneumonia, urinary tract infections, meningitis, and febrile neutropenia, a dangerous combination of fever and very low white blood cell counts. The signal in the new review is small; it may be tied to dosing, and the researchers did not recommend that doctors stop using the drug.
Why the 94 Percent Figure Is Easy to Misread
Some coverage, including a Fox News report on the review, highlighted the 94% number. In this type of analysis, called a Bayesian meta-analysis, that number describes how likely it is that any increase in risk exists. It does not describe how large the increase is.
The pooled odds ratio was 1.10, with a 95% interval of 0.98 to 1.24. In plain terms, the best estimate is a roughly 10% relative increase in the odds of death, and the range still includes the possibility of no difference at all.
When the researchers limited the analysis to 73 published, peer-reviewed trials with 15,411 patients, the probability of higher mortality rose to 98.6%, with an odds ratio of 1.17. The signal appeared in adults but not in children. It was most pronounced in trials of febrile neutropenia and severe bacterial infections, and in trials that used higher doses.
What the 110-Trial Review Included
The review team, led by first author Dr. Sohani of the Division of Infectious Diseases and Medical Microbiology at Hôpital Maisonneuve-Rosemont in Montreal, pooled 110 trials involving 22,608 patients. Of those patients, 11,726 received cefepime and 10,882 received a comparison beta-lactam. The analysis also included 25 unpublished, industry-sponsored trials that had been provided by the FDA.
That history matters. A 2007 meta-analysis first reported higher mortality with cefepime. The FDA then examined the data, including unpublished company trials, and found no increased risk of death; FDA researchers published that analysis in 2010. The new review reaches a different conclusion using a larger set of trials and a statistical method that estimates the probability of harm rather than giving a simple yes-or-no answer on statistical significance.
"Interest in this question has recently intensified following a large observational study that suggested lower mortality with cefepime than with piperacillin-tazobactam," Sohani and colleagues wrote, referring to a 2024 study of patients with sepsis.
MedicalDaily Evidence Check: This is a systematic review and meta-analysis of existing trials, not a new clinical trial. It pooled studies that differed in design, patient populations, doses, and comparison drugs, and some date back decades. It found a high probability of a small increase in deaths from any cause, but it did not prove that cefepime itself caused those deaths, and it could not examine individual patient records. Current treatment guidance still supports using the drug.
Dosing May Explain Part of the Signal
In an accompanying commentary, Dr. Daniel Uslan and Ethan Smith, PharmD, of the David Geffen School of Medicine at UCLA, argued against abandoning cefepime. They wrote that death from any cause is a blunt tool for judging drug safety in critically ill patients, and that harm may come from giving too little or too much of the drug rather than from the drug itself.
"Cefepime is not a dangerous drug in search of a replacement; it is a useful drug in search of a dose," they wrote in their JAMA Network Open commentary.
The commentators noted that cefepime has a narrower safety margin than other beta-lactams, which could justify monitoring drug levels in the blood. The FDA has previously warned that cefepime can cause a type of seizure activity called nonconvulsive status epilepticus, particularly in patients with kidney problems whose doses were not properly adjusted.
Head-to-head trial data add context. The ACORN randomized trial, led by researchers at Vanderbilt University Medical Center in Nashville, compared cefepime with piperacillin-tazobactam in 2,511 adults with suspected infection. It found no significant difference in severe kidney injury or death, HealthDay reported, although patients given cefepime had slightly fewer days alive and free of delirium and coma. The Infectious Diseases Society of America still recommends cefepime as a first-line treatment for febrile neutropenia and for certain infections caused by bacteria that can produce a drug-inactivating enzyme called AmpC beta-lactamase.
Questions Families Can Ask at the Bedside
The patients most affected by this debate are adults with serious infections, people receiving chemotherapy, older adults, and anyone with kidney disease, because the kidneys clear cefepime from the body.
Families should not ask to stop an antibiotic without talking to the care team. An untreated serious infection is far more dangerous than the small signal described in this review. It is reasonable, however, to ask whether a relative's dose has been adjusted for kidney function, whether drug levels are being checked, and what neurological side effects staff are watching for.
Sudden confusion, unusual drowsiness, muscle twitching, or a seizure during treatment should be reported to a nurse or doctor right away. Because cefepime is given almost entirely in hospitals and clinics, dosing decisions rest with the medical team and hospital pharmacists.
What remains unknown is whether any extra deaths come from the drug itself, from dosing problems, or from differences among the patients enrolled in each trial. The researchers called for more nuanced safety guidance, and the commentators urged the release of patient-level trial data. The review does not change current FDA labeling or IDSA recommendations.
Key Questions Answered
What did the new cefepime review find? A review of 110 trials estimated a 94.4% probability that cefepime is associated with more deaths within about 30 days than other beta-lactam antibiotics. Death rates were 6.6% with cefepime and 6.2% with comparison drugs.
Does cefepime raise the risk of death by 94%? No. The 94% figure reflects how likely it is that some increase exists. The estimated increase is small, and the statistical range still includes no difference.
Who usually receives cefepime? Hospitalized patients receive it by injection for infections such as pneumonia, urinary tract infections, meningitis, and febrile neutropenia.
Should patients refuse cefepime? No. Patients should not refuse or stop a prescribed antibiotic without speaking to their care team. Infectious disease guidelines still support its use in specific situations.
What side effects should families watch for? Confusion, reduced alertness, muscle twitching, or seizures, especially in older adults and people with kidney problems, should be reported immediately.
