Antibiotic De-escalation and 30-day mortality in patients with suspected bacterial culture-negative Sepsis

The implementation of early antibiotic de-escalation and its outcomes among patients with suspected bacterial culture-negative sepsis remain poorly characterized. The objective of this study was to evaluate whether antibiotic de-escalation was associated with 30-day mortality while accounting for early changes in severity of illness.

We conducted a retrospective cohort study of adult patients with community-onset sepsis and negative bacterial cultures using Duke Health System data between 2016 and 2023. Sepsis was defined using the CDC Adult Sepsis Event criteria. Antibiotic change was assessed within 5 days of onset. De-escalation was defined as a reduction in antibiotic rank across 4 spectrum-based categories, discontinuation of anti-MRSA antibiotics, or reduction in the number of antibiotics. The cohort was stratified by SOFA score change from day 1 to day 3 as SOFA-improved (≤ - 2), SOFA-unchanged (-1 to 1), or SOFA-deteriorated (≥ + 2).

A total of 7741 patients met the inclusion criteria; 33% showed improved SOFA scores, 59% had unchanged scores, and 8% had worsened scores. Overall, antibiotic de-escalation occurred in 4806 patients (62%). The crude 30-day mortality rate was 17%. In multivariable regression analyses, antibiotic de-escalation was not associated with 30-day mortality (adjusted OR 0.92; 95% CI 0.79-1.07). Findings were similar across SOFA change strata.

Approximately two thirds of patients with suspected bacterial culture-negative sepsis underwent antibiotic de-escalation, suggesting potential opportunities to optimize antimicrobial stewardship. Antibiotic de-escalation was not associated with increased 30-day mortality across SOFA change groups.