Sepsis is a common, life-threatening organ dysfunction driven by a dysregulated host response to infection. Outcomes have improved over the years in line with a focus on intravenous fluids, appropriate antimicrobials, and other supportive measures, but for septic shock, mortality remains at 30% to 50%.
The 2017 Surviving Sepsis Campaign guidelines endorse a focus on the rapidity of treatment once sepsis has been identified, with a strong recommendation for the administration of antimicrobial drugs within 1 hour; however, the quality of the supporting evidence is evaluated as moderate.
Data from multiple observational studies have addressed the question of whether early antibiotic administration is associated with improved survival. In a retrospective analysis of patients with sepsis admitted to intensive care units (ICUs) from 1989 to 2004, Kumar and colleagues found an average 7.6% decrease in survival with every hour delay in receiving antibiotics after the onset of hypotension.
Delay in antibiotic administration was associated with increased mortality only after adjusting for sepsis severity score, location prior to ICU admission, and geographic region, and there was an hour-by-hour effect gradient.
In general, retrospective time-to-intervention studies are fraught with confounding elements because a treatment delay may be caused by factors that affect mortality independently. Confounding factors include patient complexity, hospital arrival time/time of clinical deterioration, quality of the medical facility.
Recently Alam et al. presented the results of the PHANTASi trial in Lancet Respiratory Medicine. This study is the first prospective, randomized, controlled trial comparing the administration of prehospital antibiotics with standard care in The Netherlands.
Participants were adults with the temperature above 38 °C or below 36 °C in addition to either heart rate >90 beats per minute or respiratory rate >20 per minute who were recruited by specifically training emergency medical services (EMS) personnel.
Eligible individuals were randomly assigned to receive either open-label ceftriaxone in the ambulance (following collection of one blood culture sample) or usual care with intravenous fluids and supplemental oxygen. The primary outcome was all-cause mortality at 28 days with an intention-to-treat analysis.
A total of 2698 patients were enrolled, with 1535 in the intervention group and 1137 in the control group. No difference was found in primary outcome, with a mortality rate of 8% in both groups (relative risk 0.95, 95% CI, 0.74-1.24).
Important features of the study that influence interpretation are the low proportion of patients with septic shock (3.3% in the usual-care group and 4.3% in the intervention group) and an acknowledged protocol violation where overenthusiasm for early antibiotics so that, “some EMS personnel purposefully opened the envelopes until they found an envelope instructing randomisation to the intervention group.”
Researchers had a high percentage of patients with severe sepsis but only 4% septic shock. About 9% were admitted to the ICU. The study is not powered to make any conclusions on septic shock. But we think that the study has enough [patients with] severe sepsis to conclude that early antibiotics are not always indicated.
It was noted that more gram-positive bacteria were isolated from blood cultures in the intervention group than in the usual-care group, suggesting a higher rate of contamination. Also, fewer positive urine cultures were found after admission likely due to the effect of the first dose of antibiotics. Antibiotics were not given during admission for 9% in each group, mainly because viral infection was suspected.
The concept of early antibiotic administration in treating infection has evident biological plausibility; however, time sensitivity to outcome appears to be most relevant in septic shock. Further prehospital studies may help to reveal the presence of a benefit for speed at the expense of precision in this sicker patient group.