Rapid Antimicrobial Susceptibility Testing Misses Primary Endpoint

Key Takeaways

• Rapid testing was not associated with a more favorable day 30 outcome ranking than standard testing.
• Antibiotic escalation or de-escalation occurred sooner with rapid testing, while overall time to effective therapy and other secondary outcomes were similar.
• The prespecified carbapenem-resistant subgroup had a shorter median time to effective therapy with rapid testing, but the interval remained compatible with no clear difference.

Rapid phenotypic susceptibility testing from positive blood cultures was not superior to standard testing for day 30 desirability-of-outcome ranking in hospitalized patients with gram-negative bacilli bloodstream infections, with a 48.8% probability of a more favorable outcome in the randomized comparison. Antibiotic escalation or de-escalation occurred earlier in the rapid-testing group, but the primary outcome remained below the prespecified threshold for superiority.

The FAST randomized clinical trial was an open-label randomized study. Hospitalized adults and children with gram-negative bacilli bloodstream infections were enrolled at 7 medical centers in Greece, India, Israel, and Spain from December 2023 through May 2025. Final follow-up ended on June 18, 2025, and 899 patients were randomized, with 850 included in the primary analysis. Rapid phenotypic AST directly from positive blood cultures plus standard susceptibility testing was compared with standard susceptibility testing alone, while local stewardship teams reviewed patients and provided treatment recommendations. Median age was 72 years, 43% were female, and the primary endpoint was day 30 DOOR with three categories, requiring a 95% interval above 50% for superiority.

The 95% confidence interval for the primary analysis ranged from 45.3% to 52.4%, leaving the lower bound below the superiority cutoff, so the primary outcome was not superior with rapid testing. Median time to effective antibiotic therapy was not different overall, and most other secondary outcomes were also similar. These included 30-day mortality, length of hospitalization, intensive care admission, hospital-acquired infections, and time to effective antibiotic therapy within 3 days.

One secondary timing measure differed between groups. In the as-randomized population, median time to antibiotic escalation or de-escalation was 14 hours shorter with rapid testing, with a 95% confidence interval of 6 to 22 hours. Overall time to effective therapy and broader clinical outcomes remained similar.

A prespecified subgroup analysis examined patients with carbapenem-resistant infections in settings with high antimicrobial resistance prevalence. Median time to effective therapy was 9.5 hours with rapid testing and 28 hours with standard testing, a difference of -18 hours. The 95% confidence interval ranged from -42 to 6 hours, crossing zero. The estimate favored faster effective therapy with rapid testing, but remained compatible with no clear difference.