Rationale for early antibiotic therapy

Worldwide, the annual prevalence of sepsis is estimated at 19 million cases [1]. The mortality rates in severe sepsis were reduced to 20% to 30% because of advances in training, recognition, surveillance, monitoring, and rapid initial antibiotic therapy and organ support [1],[2]. The most recent update of the Surviving Sepsis Campaign guidelines was published in 2013 [3]. A consensus committee provided valuable and clear recommendations on treatment of sepsis and septic shock (Table 1). One of the main focuses is the administration of a broad-spectrum antibiotic (Table 2) [3]. It is recommended that antimicrobials be administrated within the first hour of recognition of septic shock and severe sepsis without septic shock [4]–[7]. Moreover, it is recommended that initial anti-infective therapy include one or more drugs that have activity against all likely pathogens and penetrate in adequate concentrations into the target tissue [3]. The selected antibiotic strategy should anticipate the site of infection, medical and culture history, and local microbial susceptibility results, all in an emergency situation [1],[8].

Table 1 Extract of the key recommendations and suggestions of the Surviving Sepsis Campaign guidelines [3]
Full size table
Table 2 Extract from the Surviving Sepsis Care bundles [3]
Full size table

New research results

Appropriate antibiotic therapy

In a systematic review and meta-analysis, Paul and colleagues [9] found a pooled odds ratio of appropriate antibiotic therapy during the first 48 hours for all-cause mortality of 1.60 (95% CI 1.37 to 1.86), corresponding to a number needed to treat of 10 (95% CI 8 to 15). Kumar and colleagues [4] found a time-dependent mortality with a 7.6% decrease in survival for each hour of antibiotic delay in patients with sepsis.

Timing of antimicrobial therapy

Some recently published investigations supported the findings by Kumar and colleagues [4] (Table 3). Ferrer and colleagues [10] presented a retrospective analysis of a large dataset collected prospectively. In total, 28,150 patients with severe sepsis and septic shock from 165 ICUs in Europe, the US, and South America were included. Of them, 17,990 patients received antibiotics after identification of sepsis, and the in-hospital mortality was 29.7%. After 1 hour, hospital mortality steadily increased with a delay in antibiotic timing. The adjusted hospital mortality odds ratios steadily increased from 1.00 to 1.52 as time to antibiotic administration increased from 0 to greater than 6 hours where 0 to 1 hour was the reference group. The probability of mortality increased from 24.6% to 33.1% (P <0.001) [10]. However, critics stated that, owing to a lack of information on antibiotic appropriateness and focus control, this study was limited [1].

Table 3 Comparison of studies investigating antibiotic treatment in patients with sepsis
Full size table

In contrast, Puskarich and colleagues [11] reported results from a large prospective study of emergency department patients with septic shock, which failed to demonstrate an association between timing of antibiotic administration from emergency department triage and hospital mortality. A delay in antibiotics until after shock recognition, as compared with before, was associated with increased mortality; however, if antibiotics are administered after shock recognition, there is no increase in mortality with hourly delays. These findings were in contrast to those from Kumar and colleagues [4] and Ferrer and colleagues [10]. The differences may be explained by a higher severity of illness in the other two studies. For example, Kumar and colleagues [4] investigated ICU patients with septic shock with an overall mortality of 56% in comparison with the emergency department patients in the study by Puskarich and colleagues [11] with 19%. With respect to these findings, the focus on the observed patient cohort seems to be essential.

In a prospective observational multicenter cohort study in 44 German ICUs including 1,011 patients with severe sepsis and septic shock, Bloos and colleagues [12] did not find a linear association between timing of antibiotic therapy and 28-day mortality. However, regardless of timing, 28-day mortality rate was lower in patients with adequate antibiotic therapy than in those with non-adequate antibiotic therapy (30% versus 41%, P <0.001). Bloos and colleagues stated that, owing to differences in the related patient populations, they were not able to confirm the findings of Kumar and colleagues [4].

Given these findings, the concept of early empiric antibiotic therapy has recently been challenged. More and more, the underlying and treated patient population has come into focus.

Escalating versus de-escalating strategy

With respect to the previously reported investigations, Hranjec and colleagues [13] presented data from a 2-year, quasi-experimental before-and-after observational study of hemodynamically stable patients admitted to a surgical ICU. In the first year, patients suspected of having an infection (n =101, aggressive approach, de-escalating strategy) had blood cultures and antimicrobial therapy was started. In the second year, patients suspected of having an infection (n =100, conservative approach, escalating strategy) had an antimicrobial therapy only after objective findings confirmed an infection. The conservative approach was associated with lower all-cause mortality: (13/100) 13% versus (27/101) 27%, P =0.015. The odds ratio for the risk of mortality in the aggressive approach group was 2.5 (95% CI 1.5 to 4.0) in comparison with the conservative group. In this investigation, waiting for objective data to diagnose infection before treatment with antimicrobial agents for suspected infection does not worsen the mortality.

Conclusions

Early broad-spectrum antimicrobial therapy is necessary within the ‘golden hour’ in septic shock and, as a milestone, reduces mortality. However, the use of broad-spectrum antimicrobial agents may be associated with induction of resistance among common pathogens and therefore may be a sticking point. In septic patients without sepsis-associated hypotension, diagnostic measure may be beneficial before an antibiotic therapy starts. New research is urgently needed concerning different strategies that would balance early administration of antibiotics against the potential harmful effects to patients and resistance. New research strategies have to test a de-escalating strategy with restriction of specific broad-spectrum antibiotics in the initial therapy to the most critically ill patients and patients with suspicion of multidrug-resistant pathogens, against an escalating strategy in less critically ill patients. Moreover, further studies are needed to distinguish between more and less critically ill patients, to differentiate patients’ backgrounds, and to determine indicators for patients who profit from an escalating or de-escalating strategy.