Antibiotics in Sepsis: The New Frontier*
In this issue of Critical Care Medicine, Seymour et al (11) ask and study a very important question: When does the risk of delay in antibiotics really start: In the emergency department (ED) or when sepsis patients (or their caregivers) first seek medical attention by calling for assistance in transport to the hospital? The “time zero” often used as the baseline in studies evaluating the relationship between antibiotic delay and outcomes is that of recognition—in the ED—of sepsis and septic shock (12). This “time zero” is, without doubt, an arbitrary reference point when, in actuality, the episode of sepsis certainly started before that time. Seymour et al (11) performed a sophisticated analysis based on their previously published methodology that evaluated the relationship between several prehospital time periods and mortality from sepsis. Of importance, a multivariable analysis, which was adjusted for potential confounders, demonstrated a statistically significant relationship between “total medical contact” and each hour delay in antibiotic administration. In addition, the authors establish a more granular view of the delay between a call to 911 and arrival in the ED by dividing that period into response delay and prehospital delay. “Total system delay” was divided into response delay, prehospital delay, and ED delay. The importance of these new terms rests in the fact that they may provide targets for future quality improvement studies.
What from this study can we apply to our practice related to antibiotic administration? Most importantly, this study confirms that any delay in antibiotic administration to a septic patient is associated with a higher risk of death. This study extends that delay to the furthest point yet identified: the time that the ambulance arrives at the door of a patient home or skilled nursing facility to pick up a patient with sepsis. Clearly, the true “time zero” begins even before that moment, but this study lays the ground for establishing a new, earlier time zero than had previously been used when a call for help is made. If a delay in antibiotic administration is associated with increased mortality when the timing starts at the ED door, then it stands to reason that the association between antibiotics delay and unfavorable outcome starts well before that moment. This may provide new targets for both clinical trials and quality improvement efforts. No longer can we focus only on improving process within the walls of the hospital. If the results of this study are confirmed by further prospective trials, attention must be focused on prehospital delays.
The authors conducted several important sensitivity analyses, all of which confirmed the initial results. Even when different criteria were used to define infection and organ dysfunction (qualified Sepsis-Related Organ Failure Assessment vs Sepsis-Related Organ Failure Assessment vs Surviving Sepsis Campaign criteria, including lactate), the results did not change. Therefore, whether or not an individual or institution applies the recently published definitions of sepsis, the results of this study remain valid.