Antibiotic Delay Increases Mortality in Ventilator- Associated Pneumonia

ABSTRACT & COMMENTARY

Synopsis: Delay in initiating or adding appropriate antibiotic coverage while awaiting results of bronchoalveolar lavage cultures is associated with higher mortality in patients developing pneumonia during mechanical ventilation. Gram-negative organisms resistant to third-generation cephalosporins and methicillin-resistant S. aureus were the most commonly missed pathogens in pre-BAL antibiotic coverage.

Source: Kollef ME, Ward S. Chest 1998;113(2): 412-420.

Kollef and ward identified 130 consecutive patients during a recent year who underwent fiberoptic bronchoscopy with bronchoalveolar lavage (BAL) in evaluation of presumed ventilator-associated pneumonia (VAP) in a university teaching hospital. The effect of BAL culture results on antibiotic management was examined. Forty-one patients had not received any antibiotics in the 48 hours preceding BAL, 17 had been started on antibiotics during the previous 48 hours, and the remaining 72 had been receiving antibiotics for at least 48 hours prior to the BAL. Sixty patients had culture results that indicated a potential pathogen accounting for VAP. These infected patients were classified as receiving appropriate antibiotics if the presumed pathogen was sensitive to the prescribed antibiotics, or inappropriate if there was a pathogen that was not covered.

Based on the BAL results and clinical course, the antibiotic regimen was unchanged in 51 patients, antibiotics were started or changed in 51, and antibiotics were discontinued in 28. The groups were similar in admitting diagnoses, age, sex, premorbid lifestyle, and severity of illness (APACHE II). Overall, 40% of the studied patients expired. The mortality of the patients who were continued on current treatment was 33%, while those in the group requiring a change experienced a mortality of 61%. Patients having antibiotics stopped following a negative BAL culture died at a rate of 14%. Patients with resistant organisms identified died at a rate of 59%, while those with sensitive organisms experienced a 31% mortality. Death specifically from VAP was 24% in those requiring a change in therapy, 7.8% in those continued on treatment, and only 3.6% in patients in whom antibiotics were stopped (sterile BAL).

The BAL fluid from most patients requiring a change in antibiotic regimen grew gram-negative organisms resistant to a previously administered cephalosporin (72% of changes) or methicillin-resistant Staphylococcus aureus (MRSA) (27%). Patients requiring a change in therapy were more likely to have bacteremia, septic shock, or an empyema. There was no difference in the duration of mechanical ventilation or in overall hospital stay. In this study, no additional hospital days could be attributed to the development of VAP. The clinical recommendation of Kollef and Ward is to treat VAP empirically with appropriate antibiotics rather than wait for BAL culture results. Antibiotic coverage in patients already receiving antibiotics should include treatment for resistant gram-negative bacteria and MRSA.

Comment by Charles G. Durbin, Jr., MD, FCCM

This is an important study identifying that the development of VAP in patients already on antibiotic treatment results in a worse outcome. As expected, resistant organisms predominated and were associated with a worse outcome. Unfortunately, the recommendations of Kollef and Ward remain speculative. Patients with VAP requiring an antibiotic change were indeed more ill than those not receiving antibiotics or those receiving the appropriate antibiotics at the time of culture. Because those not receiving and not started on antibiotics were included with those not requiring a change, these "healthy" patients (not having VAP) contaminated the unchanged group, making the difference reach significance. The antibiotic treatment recommendations (cover for resistant organisms while awaiting BAL results) make sense but are not proven in this study.