Wait for It...Conservative Initiation of Antimicrobials in the Surgical ICU

Trevor C. Van Schooneveld, MD


January 02, 2013

Editorial Collaboration

Medscape &

Aggressive Versus Conservative Initiation of Antimicrobial Treatment in Critically Ill Surgical Patients With Suspected Intensive-Care-Unit-Acquired Infection: A Quasi-Experimental, Before and After Observational Cohort Study

Hranjec T, Rosenberger LH, Swenson B, et al
Lancet Infect Dis. 2012:17:774-780

Antimicrobial Initiation in Critically Ill Patients

Delays in antibiotic therapy until an infection is clearly documented may be preferable to rapid initiation of antimicrobials at the first signs of infection in critically ill surgical patients. Retrospective reviews have demonstrated that delays in active therapy are associated with worse clinical outcomes. This finding has driven the practice of aggressive antimicrobial initiation in patients exhibiting signs of sepsis.

Despite its wide adoption, whether this practice is beneficial in all circumstances and patient populations has not been established. Nor has the downside of administering antimicrobials to a large number of patients who are later determined to be without infection.

Hranjec and colleagues used a quasi-experimental prospective study to evaluate whether time to antimicrobial initiation altered clinical outcomes in patients hospitalized in a single surgical intensive care unit (ICU) in the United States. Two different strategies of antimicrobial management were practiced over 2 consecutive years. During year 1 (aggressive treatment), antimicrobials were started as soon as infection was suspected and cultures were sent. During year 2 (conservative treatment), antimicrobials were withheld until objective data indicating infection were present (positive blood, urine, or sputum culture), although clinically unstable patients could have antimicrobials initiated immediately if deemed necessary.

The 1483 patients, who were admitted to the ICU primarily for polytrauma and emergency surgery, were well matched between the conservative and aggressive treatment periods. More than 100 patients each year developed more than 235 discrete infections, including pneumonia (34%), bloodstream infection (20%), urinary tract infection (14%), and intra-abdominal infection (11%). No difference was observed between the groups in terms of severity of illness, as measured by APACHE (Acute Physiology and Chronic Health Evaluation) scores, overall or in patients who required pressor support. Time to start of therapy (measured either from onset of fever or return of positive blood cultures) was longer in the conservative therapy group than in the aggressive therapy group (11.1 vs 35.2 hours and 20.9 vs 34.8 hours, respectively; P < .0001 for both).

Despite more rapid initiation of therapy, the aggressive treatment strategy was associated with an increase in hospital mortality, even after adjustment for age, sex, trauma, site of infection, and severity of illness. This held true when the data were analyzed by individual infection (odds ratio [OR], 2.5; 95% confidence interval [CI], 1.5-4.0) or by patient (OR, 4.0; 95% CI, 1.6-9.8). The aggressive strategy was less likely to use appropriate initial antibiotics (62% vs 74%; P = .0095) and to result in longer durations of therapy (17.7 vs 12.5 days; P < .008).

Certain features of this study deserve mention. First, the conservative strategy resulted in a 79% reduction in initiation of antibiotics in patients without defined infection (23% vs 5%). The aggressive strategy encouraged continuing antimicrobial therapy in patients with persistent elevations in white blood cell count and fever, which is the likely explanation for prolonged treatment duration. The time from fever to initiation of therapy was prolonged, even in the aggressive treatment group (11 hours), and the benefits of early initiation of therapy may have already been lost by the time antibiotic therapy was begun. Contradicting this, in patients who required pressor support and were managed conservatively, the mean time to initiation of therapy was 31.8 hours compared with 9.2 hours in patients managed aggressively, yet mortality was lower in the conservative therapy group (26% vs 66%; P = .0004).


Some important weaknesses of the study include the single-center, nonrandomized design; the presence of possible unrecognized or unquantified case mix or treatment differences between years; and findings that are applicable to only to similar populations of surgical ICU patients.

What do we do with the findings of this study? Should we continue to start antibiotics aggressively, or take our time and be more targeted? This study adds to the growing body of literature suggesting that aggressive antimicrobial management strategies (eg, combination therapy and early initiation) may not benefit all types of patients and might actually be harmful in certain circumstances and patient groups. In patients who are in septic shock, therapy should still be initiated aggressively, but in those who are less ill (particularly those in the surgical ICU), this study suggests that we have some time to make informed and targeted decisions with our antimicrobials.