BALTIMORE – Screening based on a new chest x-ray infiltrate or fever correctly identified more cases of microbiologically confirmed cases of ventilator-associated pneumonia than a traditional screening tool, in a study of patients in a surgical intensive care unit reported at the annual meeting of the Surgical Infection Society.
In the study, the Clinical Pulmonary Infection Score (CPIS) clinical score (using the threshold of 6) would have missed almost 80% of the cases of microbiologically confirmed VAP. The finding of a new chest x-ray infiltrate was highly sensitive for diagnosing VAP, identifying most cases of VAP, followed by fever as the next most sensitive variable. Each had a sensitivity of about 90%, said Dr. Fredric Pieracci of the department of surgery, Denver Health Medical Center, University of Colorado.
Dr. Fredric Pieracci
Another notable finding of the study was that the presence of organisms on gram stain in the early VAP window (within 5 days of intubation) was highly sensitive for diagnosing VAP, he added.
VAP is the most common nosocomial infection in intubated, critically ill surgical patients and is the most common reason antibiotics are prescribed in the surgical intensive care unit (SICU), he said. Screening criteria for VAP vary widely, but every algorithm includes some variation of the CPIS, with a score that ranges from 0 to 12. Although the CPIS screening tool, which uses variables that include tracheal secretions and chest x-ray results, has come under scrutiny, it is commonly used, with a result over 6 used as the threshold for both obtaining a lower respiratory tract culture and initiating empiric treatment.
The study analyzed the results of 1,013 bronchoalveolar lavage cultures from 497 SICU patients aged 18-88 years, over a 3-year period (2009-2012). Most of the patients (81%) were males and 71% were trauma patients; cultures were obtained a median of 8 days after intubation (range, 1-109 days), and patients had a median of two cultures. VAP was defined microbiologically as at least 105 CFU/mL if no antibiotics had been given within the previous 72 hours; or at least 104 CFU/mL if antibiotics had been given within the previous 72 hours. CPIS scores were calculated retrospectively.
Of the 1,013 cultures, 438 (43%) met the VAP criteria, and 310 of the 497 patients (62%) had at least one episode of VAP.
Most of the CPIS clinical scores were 4, 5, or 6. When the likelihood of VAP was analyzed, CPIS clinical scores from 1 to 9 all correlated with about a 40% chance of VAP, Dr. Pieracci said. The median CPIS clinical score was 5 for those diagnosed with VAP as well as those not diagnosed with VAP, based on the microbiologic criteria.
The sensitivity of the CPIS clinical score, when the threshold of greater than 6 was used, was only 21%, so by using the CPIS, "we would have missed almost 80% of the VAP cases in this group of patients," he pointed out.
Every case of VAP had at least one of the following: fever, a new chest x-ray infiltrate, or the presence of organisms on gram stain.
Of the individual components of the CPIS, the most sensitive for diagnosing VAP were the new finding on chest x-ray (a sensitivity of 91.1%), and fever (a sensitivity of 89.0%).
When the gram stain results were added to the CPIS clinical score, there was a marginal improvement in sensitivity, "but it was still a very poor screening tool," he said. However, the presence of organisms on the gram stain "was highly sensitive for diagnosing VAP, with a reasonably high negative predictive value" in the early VAP window, when cultures were sent within 5 days of intubation, he said.
The results indicate that the CPIS clinical score in the study "had poor discriminative ability for diagnosing VAP in all the clinical scenarios we tested," and it had a sensitivity that was acceptable only at a threshold lower than 6, Dr. Pieracci said. "Based on these data, we recommend abandoning the CPIS clinical score as a screening tool for VAP," and instead, adopting the three criteria and withholding antibiotic therapy in patients with no organisms on gram stain in the early VAP window.
"What we’ve adopted and are now studying is a screening algorithm that differentiates between the early and late period, and is based on either fever or new chest x-ray finding," Dr. Pieracci said. If the gram stain is negative in the early VAP window, then withholding empiric antibiotics is recommended; this is the only scenario identified in which empiric antibiotics could be safely withheld, he added.