Hypothesis: 4 days of antibiotic therapy, as compared to 8 days, is equally effective and results in decreased antibiotic exposure among surgical ICU patients with early VAP.

The prevalence of multi-drug resistant (MDR) pathogens in intensive care units (ICUs) worldwide has reached epidemic proportions. In some cases, the choice of potential therapy is limited or even non-existent. Antibiotic prescription, through selection pressure, represents the main mechanism by which resistance emerges. Limitations in the development of new antibiotics underscores the importance of adherence to the principles of antibiotic stewardship.

Ventilator associated pneumonia (VAP) is the most common serious infection in mechanically ventilated, critically ill patients. Approximately one half of antibiotic prescription in the ICU is related to VAP, including prophylactic, empiric, and definitive therapy. The development of evidence-based algorithms for the rational use of antibiotics in the management of patients with both suspected and confirmed VAP is pivotal to decreasing the emergence of MDR pathogens.

Shortening the duration of antimicrobial therapy for VAP represents one strategy to curtail the emergence of MDR pathogens. Although current guidelines recommend a treatment course of 8-14 days, both clinical and microbiologic resolution (MR) of infection typically occur much sooner [10, 11]. In one study of ICU patients ventilated for > 5 days who developed VAP, 8 days of antimicrobial therapy was equally as effective as 14 days, provided VAP was not caused by a non-lactose fermenting gram negative bacillus. Favorable results following shorter courses of therapy for VAP have been observed, albeit in small, uncontrolled series.

One subset of patients for whom a decreased duration of antimicrobial therapy may be particularly effective is those who develop VAP ≤ 5 days after intubation (early VAP). Early VAP comprises approximately one half of cases of pneumonia diagnosed in the ICU. Furthermore, as compared to patients who develop late VAP, patients who develop early VAP are more likely to be infected with community-acquired pathogens sensitive to narrow spectrum antibiotics. Finally, nearly all cases of early VAP caused by sensitive pathogens demonstrate MR after relatively short (3-5 days) courses of therapy.

Inclusion Criteria:

1. Surgical patient
2. VAP, defined as clinical suspicion plus a bronchoalveolar lavage (BAL) culture showing ≥105 cfu/mL of at least one pathogen. The quantitative microbiology threshold will be lowered to ≥104 cfu/mL if the patient was being treated with antibiotics to which the pathogen is sensitive at the time of the BAL. Clinical suspicion of VAP is defined as at least one point for ≥ 2 variables in the Clinical Pulmonary Infection Score (CPIS, described below).
3. Ventilated ≤ 5 days at the time that the BAL was obtained.
4. Hospital LOS ≤ 5 days at the time that the BAL was obtained.

Exclusion Criteria:

1. Age < 18 years.

2. Prior episode of VAP for the index admission (the patient may have had prior BALs sent for culture, but these cannot have met the above mentioned diagnostic criteria for VAP).

3. VAP caused by a MDR pathogen: Early VAP is rarely caused by a MDR pathogen; in a recent analysis of our surgical ICU, 94% of cases of early VAP were caused by a highly sensitive pathogen (MSSA 39%, H flu 35%, S. pneumo 16%, E. coli 9%) (Pieracci in press). Patients with early VAP caused by the following MDR pathogens will be excluded: Methicillin-resistant Staphylococcus aureus (MRSA), Vancomycin-intermediate Staphylococcus aureus (VISA), pseudomonas aeruginosa, Vancomycin-resistant enterococcus (VRE), Acinetobacter baumannii, Stenotrophomonas maltophilia, and extended-spectrum beta lactamase producing gram negative bacilli.

4. Antibiotic therapy for ≥ 5 of the last 10 days preceding the BAL.

5. Septic shock, defined as evidence of tissue hypoperfusion after adequate volume expansion, due to infection, and requiring ≥ 1 vasopressor.

6. Current or recent (within 30 days) use of immunosuppressive medications.

7. Length of stay ≥ 48 hours in a transferring facility.

8. Inpatient hospitalization within 30 days of admission.

9. Pregnancy or lactation.

10. Legal arrest or incarceration.

11. Moribund state in which death is imminent.

    -

Patients were following during their hospital stay for 30 days or until discharge, whichever event occurred first. Adverse event monitoring occurred daily throughout this time.

There was an independent Data Safety Monitoring Board (DSMB) that met every 6mths to review data. The PI and study coordinator met daily to review unanticipated problems.