Sepsis is a critical medical emergency that requires prompt treatment with broad-spectrum antibiotics (BSAs) to cover potential multidrug-resistant organisms (MDROs). While early empiric BSA therapy has been associated with reduced mortality in high-risk patients and is recommended by the 2021 Surviving Sepsis Campaign Guidelines, prolonged exposure to these agents carries significant risks. These include antibiotic-associated adverse events, Clostridioides difficile infection, and population-level antimicrobial resistance.
A recent study aimed to evaluate outcomes associated with BSA de-escalation compared with continuation in patients hospitalised with community-onset sepsis. The study included adult patients aged 18 years and older hospitalised with community-onset sepsis who initiated empiric BSA therapy without evidence of MDRO infection.
Two separate target trial emulations were conducted: one assessing de-escalation of antimethicillin-resistant Staphylococcus aureus (MRSA) antibiotics, and another evaluating de-escalation of anti-Pseudomonas aeruginosa (PSA) or other gram-negative bacteria coverage. Time zero, the point of enrolment, was defined as the end of the third calendar day of hospitalisation. Patients were compared based on whether they underwent de-escalation (no BSA on day 4) or continued BSA therapy (BSA on day 4).
You might also like: Efficacy, Appropriateness of Antibiotic Use in the ICU
The study analysed 36,924 patients with community-onset sepsis, with a median age of 71 years. Only 43.2% of eligible patients underwent anti-MRSA de-escalation and 22.4% underwent anti-PSA de-escalation. This finding highlights substantial variation in clinical practice, with more than two-fold differences in de-escalation rates across the 67 participating hospitals. The anti-MRSA de-escalation rate ranged from 27.3% to 61.7%, whilst anti-PSA de-escalation varied from 6.9% to 37.7% across institutions.
The primary outcome (90-day all-cause mortality) revealed reassuring results. After weighting, both anti-MRSA and anti-PSA de-escalation were associated with mortality rates similar to those with continued BSA therapy. This finding demonstrates that de-escalation does not compromise patient survival.
Secondary outcomes revealed additional benefits of de-escalation. Both anti-MRSA and anti-PSA de-escalation were associated with fewer days of antibiotic therapy through day 14 and shorter length of hospitalisation. In-hospital mortality, 30-day mortality, and rates of C. difficile infection were similar between groups. Anti-PSA de-escalation, but not anti-MRSA de-escalation, was associated with lower 90-day readmission rates. The authors hypothesise this may relate to the preservation of gut microbiome integrity, as anti-PSA coverage often includes antianaerobic agents that can be particularly deleterious to intestinal flora.
Another finding emerged from the subgroup analysis of clinically stable patients on day 3. Clinical stability was defined as absence of vasopressor therapy, no invasive mechanical ventilation, and no more than one abnormal vital sign. Among these 2,161 patients in the anti-MRSA cohort and 3,344 in the anti-PSA cohort, de-escalation of anti-MRSA coverage was associated with lower 90-day mortality. While the point estimate for anti-PSA de-escalation also suggested lower mortality, this did not reach statistical significance in this smaller subset.
This finding addresses a key concern raised by sceptics of de-escalation that apparent benefits might result from reverse causation, where improving clinical status increases the likelihood of de-escalation rather than de-escalation causing improvement.
These findings support guideline recommendations for BSA de-escalation in clinical practice. By adjusting for predicted mortality and clinical stability, the study addresses limitations of previous observational research that cast doubt on de-escalation safety. The results suggest that, among patients with community-onset sepsis and no positive MDRO test results, de-escalation is not only safe but also potentially beneficial, with reduced antibiotic exposure and shorter hospitalisation.
The study provides robust evidence supporting the safety and potential benefits of BSA de-escalation on day 4 amongst patients with MDRO-negative community-onset sepsis. The findings demonstrate similar mortality outcomes alongside reduced antibiotic exposure and shorter hospital stays. The substantial variation in de-escalation practices across hospitals suggests significant opportunity for antimicrobial stewardship improvement. These results should encourage clinicians to embrace evidence-based de-escalation strategies, ultimately promoting judicious antibiotic use whilst maintaining patient safety.
Source: JAMA
Image Credit: iStock
References:
Gupta AB, Heath M, Walzl E et al. (2025) Antibiotic De-Escalation in Adults Hospitalized for Community-Onset Sepsis. JAMA Intern Med.
