Short-Course Empiric Antibiotic Therapy for Patients With Pulmonary Infiltrates in the Intensive Care Unit: A Proposed Solution for Indiscriminate Antibiotic Prescription

SHORT-COURSE EMPIRIC ANTIBIOTIC THERAPY FOR PATIENTS WITH PULMONARY INFILTRATES IN THE INTENSIVE CARE UNIT: A PROPOSED SOLUTION FOR INDISCRIMINATE ANTIBIOTIC PRESCRIPTION [Singh N, et al. Am J Respir Crit Care Med 2000;162:505]: The authors address the issue of antibiotic abuse in the intensive care unit (ICU) for suspected pneumonia, which is often attributed to noninfectious causes. There were 81 patients with pulmonary infiltrates and low risk for pneumonia by CPIS (clinical pulmonary infection scores) <6. The CPIS is based on temperature, WBC, purulence of respiratory secretions, hypoxemia, roentgenogram findings, and gram stain. Treatment was IV ciprofloxacin for 39 patients, and 42 controls were given antibiotics according to physician preference. The Cipro TM group had antibiotics discontinued at 3 days if the CPIS was <6, and ciprofloxacin was continued or another antibiotic was started if the score was >6. The results showed that recipients of Cipro had a shorter ICU stay (9.4 vs. 14.7 days), fewer antibiotic days (a mean of 3 vs. 9.8 days), lower incidence of antibiotic resistance, and reduced antibiotic costs (a mean of $235 vs. $640/patient); there was a trend toward lower mortality (13% vs. 31%) that was not statistically significant. The authors conclude that discontinuation of antibiotics in this low-risk group has substantial benefits in terms of cost, outcome, and resistance patterns. Comment. Previous reports with bacteriologic studies using the protected bronchoscopy brush indicate that less than half of ventilated patients in ICUs with pulmonary infiltrates on roentgenogram with fever and purulent secretions have confirmed bacterial infections. The obvious concern is antibiotic abuse in the ICU. One tactic, which is popular in Europe and used occasionally in centers in the United States, is to confirm or refute the presence of infection with quantitative cultures of BAL specimens or with the protected brush catheter. An alternative tactic is shown in the present study.