A plethora of retrospective studies have reported that delays in antibiotic administration are associated with higher mortality rates in patients with sepsis. These data have spurred an increased focus on time-to-antibiotics as a metric for best practice guidelines and quality measures. The Centers for Medicare and Medicaid Services Severe Sepsis/Septic Shock Early Management Bundle (SEP-1), for example, requires antibiotics to be given within 3 hours of sepsis onset, whereas the 2018 update to the Surviving Sepsis Campaign Guidelines recommended antibiotic initiation within 1 hour of emergency department (ED) triage time (later modified to 1 hr of sepsis recognition) (1).
Aggressive time-to-antibiotic targets, however, have generated substantial controversy. A major point of contention is that these metrics promote rapid initiation of broad-spectrum antibiotics in any patients presenting with suspected sepsis yet fail to acknowledge the high diagnostic uncertainty associated with this condition. Unlike other emergent conditions that are easy to recognize (e.g., severe trauma) have rapid and accurate biomarkers (i.e., myocardial infarction), or can be quickly and accurately diagnosed by radiographic imaging (i.e., stroke), sepsis is a constellation of signs and symptoms with many potential noninfectious mimics and no diagnostic gold standard. Indeed, up to 20–40% of patients initially treated for sepsis are later found to have a low likelihood of infection (2). Critics of sepsis measures therefore argue that the push for immediate antibiotics risks a culture of “treat first, ask questions later” that risks driving antibiotic resistance, toxicity, and Clostridium difficile infections through excessive or unnecessary antibiotic use. This concern is supported by reports that SEP-1 and other sepsis quality-improvement efforts have been associated with increases in broad-spectrum antibiotic use (3–5).
A second point of contention is the quality of the data supporting the necessity of immediate antibiotics. If the benefit of immediate antibiotics for all patients with suspected sepsis was unequivocal, most clinicians would gladly accept the inevitable trade-off of unnecessary antibiotics in the subset of patients who turn out to be uninfected. To date, however, only one prospective randomized controlled trial has compared differential timing of antibiotics in sepsis by randomizing nearly 2,700 patients to antibiotics in the ambulance versus usual care in the ED (6). This trial found no difference in 28-day mortality between the two groups despite the fact that the intervention arm received antibiotics on average greater than 90 minutes earlier. Importantly, the study by Taylor et al (8) primarily enrolled septic patients without shock, precluding any definitive conclusions about the importance of timely antibiotics in the most severely ill subset of patients.
All other analyses supporting the importance of time-to-antibiotics have been observational studies, which are susceptible to confounding and other important methodologic limitations. One particularly important but often overlooked limitation is that retrospective studies have generally blended together mortality estimates from patients who had very long intervals until antibiotic administration (sometimes on the order of 12–24 hr) with patients with shorter intervals and then reported a single estimate for the average increase in mortality associated with each hour until antibiotics (7). These estimates, unsurprisingly, suggest associations between hourly delays and mortality, but this approach prevents us from knowing whether administering antibiotics at hour 1 or 2 truly makes a difference in patient outcomes compared with antibiotics at hour 5.
A new study published by Taylor et al (8) in this issue of Critical Care Medicine provides important new insight into whether time-to-antibiotics impacts mortality in patients with sepsis and if so, what time intervals really matter. The authors retrospectively analyzed 24,093 encounters among adults with clinical evidence of suspected infection and concurrent organ dysfunction in 12 EDs and separately assessed the impact of recognition delay (time from ED triage to antibiotic order) and administration delay (time from antibiotic order to infusion). Importantly, they assessed these delays in discrete time intervals relative to less than 1 hour and less than 30 minutes, respectively, rather than making the common mistake of blending together mortality estimates from a long interval into an hourly estimate. The primary findings were that 1) recognition delays accounted for the majority of the interval from ED triage to antibiotic infusion (median 2.7 vs 0.6 hr with administration delays) and 2) after detailed adjustment for a rich set of potential confounders, recognition delay was only significantly associated with mortality after 6 hours, whereas an association between administration delays and mortality was observed after 1.5 hours.
The study by Taylor et al (8) has important implications. First, it reinforces the concern that current time-to-antibiotic metrics targeting 1 or even 3 hours for all patients with suspected sepsis is not strongly evidence based and may be overly aggressive. If antibiotics can be safely withheld for up to 6 hours, this has massive implications for clinical care since even a few hours can allow providers to initiate a rapid workup for possible noninfectious etiologies and potentially trial nonantibiotic interventions before deciding whether antibiotics are truly indicated. Allowing additional time may also allow clinicians to choose a more thoughtful and targeted antibiotic regimen as opposed to always defaulting to a combination of broad-spectrum antibiotics. An important exception, however, may be septic shock, as several of the largest and most rigorous observational studies of time-to-antibiotics have suggested that most of the harm associated with delays in antibiotic administration occur in patients in this category (9,10). In the analysis by Taylor et al (8), only delays in antibiotic administration were associated with mortality in patients with septic shock, whereas delays in recognition were not. This, however, could reflect inadequate power from a smaller sample size of septic shock patients or that recognition of septic shock in this healthcare system with a robust sepsis identification pathway was already very fast.
A second implication is that the time interval from antibiotic order to infusion is an important metric worth tracking. This supports recent recommendations from the Infectious Diseases Society of America (IDSA) and others to consider reporting the time interval between each STAT antibiotic order for sepsis to infusion (11,12). IDSA has articulated the potential advantages of this approach include its objectivity (in contrast to efforts to identify sepsis time zero via medical record review that are subject to interrater reliability issues ) and its emphasis on improving processes for rapid antibiotic delivery (since it is incontrovertible that once a clinician does decide a patient needs an antibiotic, it should be infused as quickly as possible). Unlike the diagnostic challenges present in time-to-recognition, improving time between antibiotic order to infusion represents low-hanging fruit that can be readily addressed by institutions through simple interventions such as improving access to antibiotics in high use areas such as EDs and ICUs with automated dispensing cabinets, creation of “code sepsis” teams that provide prompt delivery of antibiotics and fluids for resuscitation, and defaulting antibiotic orders originating from these high-acuity areas to a “STAT” priority (14). In the analysis by Taylor et al (8), antibiotics were generally administered fairly rapidly, yet other healthcare systems have reported longer delays that are also associated with higher mortality (15). The timely antibiotic administration in the current analysis may reflect the intense quality-improvement efforts that have occurred within this particular healthcare institution.
So what does the future hold for time-to-antibiotic metrics? As important and compelling as the data are from the analysis by Taylor et al (8), we need to confirm their generalizability in diverse hospitals and populations. Studies focusing on patients with septic shock are needed to better differentiate discrete time frames from bundled interventions as this is the patient subset with the highest likelihood of benefit. In addition, there is a need for future analyses to better define the physiologic criteria, aside from shock, that should trigger immediate treatment when infection is a possibility. This is important because sepsis includes a heterogeneous mix of clinical signs, but not all of them may be equally time sensitive with respect to antibiotics. For example, it is doubtful that suspected infection with hyperbilirubinemia or acute kidney injury alone needs to be treated with the same urgency as hypotension, yet most time-to-antibiotic analyses do not differentiate between various types of organ dysfunction in patients with sepsis.
Ultimately, the study by Taylor et al (8) is an important step toward developing more evidence-based sepsis quality metrics. This in turn could have substantial public health benefit by promoting a more rational approach to empiric antibiotic prescribing for patients with suspected sepsis and focusing resources on the processes of care that are most likely to improve outcomes.
1. Levy MM, Evans LE, Rhodes A. The surviving sepsis campaign bundle: 2018 update. Crit Care Med. 2018; 46:997–1000
2. Klein Klouwenberg PM, Cremer OL, van Vught LA, et al. Likelihood of infection in patients with presumed sepsis at the time of intensive care unit admission: A cohort study. Crit Care. 2015; 19:319
3. Pakyz AL, Orndahl CM, Johns A, et al. Impact of the Centers for Medicare and Medicaid Services sepsis core measure on antibiotic use. Clin Infect Dis. 2020ciaa45
4. Hiensch R, Poeran J, Saunders-Hao P, et al. Impact of an electronic sepsis initiative on antibiotic use and health care facility-onset Clostridium difficile infection rates. Am J Infect Control. 2017; 45:1091–1100
5. Miller J, Hall B, Wilson K, et al. Impact of SEP-1 on broad-spectrum combination antibiotic therapy in the emergency department. Am J Emerg Med. 2020; 38:2570–2573
6. Alam N, Oskam E, Stassen PM, et al.; PHANTASi Trial Investigators and the ORCA (Onderzoeks Consortium Acute Geneeskunde) Research Consortium the Netherlands. Prehospital antibiotics in the ambulance for sepsis: A multicentre, open label, randomised trial. Lancet Respir Med. 2018; 6:40–50
7. Weinberger J, Rhee C, Klompas M. A critical analysis of the literature on time-to-antibiotics
in suspected sepsis. J Infect Dis. 2020; 222:S110–S118
8. Taylor SP, Anderson WE, Beam K, et al. The Association Between Antibiotic Delay Intervals and Hospital Mortality Among Patients Treated in the Emergency Department for Suspected Sepsis. Crit Care Med. 2021; 49:741–747
9. Seymour CW, Gesten F, Prescott HC, et al. Time to treatment and mortality during mandated emergency care for sepsis. N Engl J Med. 2017; 376:2235–2244
10. Liu VX, Fielding-Singh V, Greene JD, et al. The timing of early antibiotics and hospital mortality in sepsis. Am J Respir Crit Care Med. 2017; 196:856–863
11. Rhee C, Chiotos K, Cosgrove SE, et al. Infectious Diseases Society of America position paper: Recommended revisions to the national severe sepsis and septic shock early management bundle (SEP-1) sepsis quality measure. Clin Infect Dis. 2020ciaa059
12. Klompas M, Rhee C. Antibiotic order-to-infusion time for patients with septic shock: A potential new quality metric. Crit Care Med. 2019; 47:1467–1470
13. Rhee C, Brown SR, Jones TM, et al. Variability in determining sepsis time zero and bundle compliance rates for the centers for medicare and medicaid services SEP-1 measure. Infect Control Hosp Epidemiol. 2018; 39:994–996
14. Strich JR, Heil EL, Masur H. Considerations for empiric antimicrobial therapy in sepsis and septic shock in an era of antimicrobial resistance. J Infect Dis. 2020; 222:S119–S131
15. Kashiouris MG, Zemore Z, Kimball Z, et al. Supply chain delays in antimicrobial administration after the initial clinician order and mortality in patients with sepsis. Crit Care Med. 2019; 47:1388–1395