Summary
A medRxiv preprint pooling two randomized trials found no statistically significant difference in 90-day mortality between intravenous cefazolin and antistaphylococcal penicillins for adults with MSSA bacteremia. Estimates for kidney injury and treatment discontinuation favoured cefazolin in direction but were imprecise.
A systematic review and meta-analysis posted as a medRxiv preprint found no statistically significant difference in 90-day mortality between intravenous cefazolin and antistaphylococcal penicillins for adults with methicillin-susceptible Staphylococcus aureus (MSSA) bacteremia. The analysis combined two randomized controlled trials involving 1,656 participants. Its findings support cefazolin as a potential definitive treatment option under appropriate clinical conditions, while the small number of trials and imprecise safety estimates call for cautious interpretation.
MSSA bacteremia is a bloodstream infection caused by S. aureus that is susceptible to methicillin. The review compared definitive intravenous cefazolin with antistaphylococcal penicillins: oxacillin, cloxacillin, flucloxacillin or nafcillin. The authors searched five databases and included randomized-trial evidence only.
What the two trials found
The trials were SNAP, with 1,341 participants, and CloCeBa, with 315. Of the 1,656 participants randomized, 1,579 contributed to the analysis of death from any cause within 90 days. The pooled risk ratio was 0.90, with a 95% confidence interval of 0.57 to 1.41 and no measured statistical heterogeneity (I² = 0%). The result did not show a statistically significant difference. The confidence interval remains compatible with benefit or harm, so it should not be read as evidence that the treatments are equivalent.
Both trials’ results for acute kidney injury (AKI) favoured cefazolin in direction. But the random-effects pooled estimate was exceptionally imprecise: risk ratio 0.26, with a 95% confidence interval of 0.00 to 939,970.59; heterogeneity was high (I² = 81.7%). That estimate does not provide a reliable measure of how much the risk might differ. Treatment discontinuation because of adverse events also favoured cefazolin directionally, but its estimate was imprecise (risk ratio 0.24, 95% CI 0.00–24.41; I² = 36.2%).
The authors did not pool microbiological outcomes. The trials used different definitions and assessment time points, so those results were summarized narratively instead.
What the findings mean for treatment evidence
The analysis addresses a practical question in antibiotic treatment: whether cefazolin is a potential option alongside antistaphylococcal penicillins for definitive therapy of MSSA bacteremia. The authors conclude that randomized evidence supports considering cefazolin when adequate source control, appropriate antimicrobial dosing and clinical monitoring can be achieved. Source control means addressing the infection’s source, for example by managing an infected focus; the preprint makes this a condition of its conclusion rather than a result that the pooled comparison tested separately.
The main finding is narrower than a demonstration that one antibiotic is superior. The mortality analysis detected no significant difference, while the available kidney-injury and treatment-discontinuation estimates are too imprecise for confident comparisons. The report is a preprint, and its conclusions rest on only two randomized trials. Further evidence could refine the comparison, particularly for safety outcomes.
The study was posted on medRxiv on October 13, 2026.