What the analysis found
A systematic review and Bayesian meta-analysis published in JAMA Network Open assessed 110 randomized clinical trials comparing cefepime with other beta-lactam antibiotics. The trials included 22,608 patients. The analysis estimated a 94.4% posterior probability that all-cause mortality was higher with cefepime.
The pooled odds ratio was 1.10, with a 95% credible interval of 0.98 to 1.24. Because the interval includes 1, the data remain compatible with no difference as well as with some increase in risk. The result is a possible safety signal, not definitive evidence that cefepime caused deaths.
Why the result is uncertain
The estimated signal was stronger when published studies were analyzed alone. Adding unpublished studies reduced the estimate, while the Bayesian model continued to favor a possible increase. Differences in trial design, patient populations and available data affect how precisely the risk can be estimated.
The authors say the finding merits careful consideration in clinical guidance and future research. It is not a recommendation for patients to stop or change antibiotics. Treatment decisions should be made with a qualified clinician who can assess the individual circumstances.
Sources
Bayesian meta-analysis in JAMA Network Open (2026), DOI 10.1001/jamanetworkopen.2026.33017; 110 trials, 22,608 patients, 94.4% posterior probability, pooled OR 1.10 (95% CrI 0.98–1.24) are all stated in the study.
The 95% credible interval includes 1, so the article itself frames this as an uncertain signal, not proof of harm; the estimate weakened after unpublished trials were added, which the authors note but do not fully quantify.
The list of the 110 included trials is not published in the article; the number and identity of the unpublished studies added to the model are not specified; response from the contacted researchers not yet received.