Background: Early administration of antimicrobials is one of the most effective interventions to reduce sepsis mortality. Despite this, delays in antibiotic initiation occur, particularly in patients presenting without hypotension. We sought to understand patient characteristics and presenting symptoms associated with antibiotic delays in this population.
Methods: Cohort study of adult patients hospitalized with community-onset sepsis from 11/2020 to 5/2024 at 67 hospitals participating in the Hospital Medicine Safety Consortium sepsis initiative (HMS-Sepsis). Data, including presenting symptoms, were abstracted by professional abstractors. We included patients presenting with sepsis (infection and acute organ dysfunction) and excluded patients with positive COVID or influenza testing. We focused on patients without hypotension (hours 0-2). Timely antibiotics were defined as antibiotics given within 5 hours of hospital arrival—a relaxed cut-off compared to the Surviving Sepsis Guidelines to account for triage and sepsis recognition time. Multivariable logistic regression models adjusted for hospital variation were fit on predetermined covariates to identify patient characteristics and symptoms associated with timely antibiotics. Association of timely antibiotics with 30-day mortality was also evaluated.
Results: Of 23,915 patients in the HMS-Sepsis registry presenting with sepsis without hypotension, 17,732 (74.1%) received timely antibiotics [Figure 1]. For patients receiving timely vs. delayed antibiotics, median time to antibiotic order was 1.7 (IQR 0.9, 2.6) vs 6.0 (5.7, 10.0) hours and median time to antibiotic delivery was 2.4 hours (IQR 1.6, 3.4) vs 7.6 hours (5.9, 12.2), respectively. Patients were more likely to receive timely antibiotics if they presented with subjective fever (adjusted odds ratio 1.37 [95% CI 1.28, 1.47], p<.0001), urinary symptoms (1.18 [1.10,1.28], p<.0001), or respiratory symptoms (1.08 [1.00,1.16], p=0.04), and less likely to receive timely antibiotics if they presented with GI symptoms (0.81 [0.76,0.87], p<.0001), had chronic liver disease (0.72 [0.59, 0.89], p=0.002), chronic kidney disease (0.86 [0.79, 0.95], p=0.002), congestive heart failure (0.90 [0.83, 0.98], p=0.02), or metastatic solid malignancy (0.59 [0.49, 0.73], p<.0001) [Table 1]. Patients with vital sign abnormalities within 2 hours of presentation (hypo/ hyperthermia, tachycardia, tachypnea, hypoxia) were more likely to receive timely antibiotics. Adjusted mortality at 30-days was reduced for those receiving timely antibiotics (0.79 [0.72, 0.86], p<.0001).
Conclusions: Patient symptoms, comorbidities, and vital sign abnormalities influenced the likelihood of receiving timely antibiotics, and 30-day mortality was reduced for those receiving timely antibiotics. Focusing on improved timeliness of antibiotics in patient populations more likely to experience delays represents an opportunity for quality improvement.
