Written by Peter Liu
Spoon Feed
A procalcitonin-based protocol for discontinuation of antibiotics in sepsis care resulted in fewer antibiotic days, within a noninferior mortality margin of 5.4%.
Procalcitonin protocols encourage earlier antibiotic discontinuation
Sepsis care requires timely, precise interventions in a high-stakes environment with high mortality. In the ADAPT-Sepsis RCT, researchers compared biomarker-guided protocols based on procalcitonin (PCT) or C-reactive protein (CRP) to standard care to study whether these protocols reduced antibiotic duration safely in critically ill adults with suspected sepsis across all organ systems. Involving 2,760 patients across 41 UK intensive care units, the study compared daily PCT-guided, CRP-guided, and standard care approaches. The PCT protocol reduced antibiotic duration significantly (mean difference 0.88 days, p=0.01) without increasing 28-day all-cause mortality within the noninferiority safety margin of 5.4%. CRP-guided care did not clearly reduce antibiotic use and exceeded the noninferiority safety margin of 5.4%. These findings support PCT use in helping to guide antibiotic discontinuation. A baseline PCT less than 0.25ug/L would strongly support antibiotics discontinuation, and a PCT fall >80% from baseline or PCT less than 0.5ug/L would support discontinuation.
There are many concerns with this trial design. First, estimates for mortality were higher for both biomarker-based protocols. The PCT 28-day mortality difference vs standard care was 20.9% vs 19.4% (+1.57%, 95%CI -2.18% to 5.32%). While the noninferiority analysis confidently concluded that PCT-protocol mortality was not more than 5.4% worse than standard care (e.g. noninferior based on the prespecified noninferiority margin), we cannot confidently say that PCT-protocols do not cause worse mortality at a lower level. Second, the design of the RCT applied a test (i.e. PCT) that has not been validated for all causes of sepsis and performed poorly in diagnosing UTI in a hospitalized older population. Lastly, the system interface involved clinically-blinded antibiotic advice from a local research team, which does not reliably reproduce a normal clinical setting, at least not in the United States.
How does this change my practice?
Overall, these findings prompt me to consider utilizing procalcitonin to guide antibiotic discontinuations for septic immunocompetent patients with particularly high risks associated with antibiotic exposure. I remain concerned about its effects on infection-related mortality.
Source
Biomarker-Guided Antibiotic Duration for Hospitalized Patients With Suspected Sepsis: The ADAPT-Sepsis Randomized Clinical Trial. JAMA. 2024 Dec 9. doi: 10.1001/jama.2024.26458. Epub ahead of print. PMID: 39652885
