Objective: To assess the predictive accuracy of serum procalcitonin in distinguishing bacterial aspiration pneumonia from aspiration pneumonitis.
Design: Prospective observational study.
Setting: Intensive care unit of a university-affiliated hospital.
Patients: Sixty-five consecutive patients admitted with pulmonary aspiration and seven control subjects intubated for airway protection.
Interventions: None.
Measurements and main results: Quantitative cultures from bronchoalveolar lavage fluid were conducted on all participants at the time of admission. Serial serum procalcitonin levels were measured on day 1 and day 3 using the procalcitonin enzyme-linked fluorescent assay. There were no differences in the median serum concentrations of procalcitonin between patients with positive bronchoalveolar lavage cultures (n = 32) and patients with negative bronchoalveolar lavage cultures (n = 33) on either day 1 or day 3 postadmission. The areas under the receiver operator characteristic curves were 0.59 (95% confidence interval, 0.47-0.72) and 0.63 (95% confidence interval, 0.5-0.75), respectively (p = .74). However, duration of mechanical ventilation and antibiotic therapy were shorter in those who had a decrease in their procalcitonin levels on day 3 from baseline compared with those who did not (6.7 ± 7.1 days and 11.1 ± 13.5 days, p = .03; and 8.2 ± 2.6 days vs. 12.8 ± 4.6 days; p < .001, respectively). Hospital mortality was associated with radiographic multilobar disease (adjusted odds ratio, 1.14; 95% confidence interval, 1.01-1.31; p = .04) and increasing procalcitonin levels (adjusted odds ratio, 5.63; 95% confidence interval, 1.56-20.29; p = .008).
Conclusion: Serum procalcitonin levels had poor diagnostic value in separating bacterial aspiration pneumonia from aspiration pneumonitis based on quantitative bronchoalveolar lavage culture. However, serial measurements of serum procalcitonin may be helpful in predicting survival from pulmonary aspiration.