Purulence and gram-negative bacilli in tracheal aspirates of mechanically ventilated very low birth weight infants

Abstract

Objective: Tracheal aspirates (TAs) from mechanically ventilated very low birth weight (VLBW) infants are frequently obtained during the evaluation of suspected sepsis, tracheitis, or ventilator-associated pneumonia (VAP). Purulence and bacteria in Gram stain of bronchopulmonary secretions are considered signs of respiratory infection, and medical decisions are made on the assumption that they are predictors of positive bacterial tracheal cultures (TCs). The purpose of this retrospective investigation was to establish the relationship of purulence and bacteria in TA from ventilated VLBW infants with positive TC and to identify its clinical significance.

Study design: One hundred and seventy consecutively born VLBW infants (1996 to 1998) who remained on mechanical ventilation longer than 1 week were studied. Demographic, laboratory, and clinical data were obtained from hospital medical records. Purulence, defined by the number of polymorphonuclear leukocytes (PMNs) per low power field (LPF), was reported as light (<25 PMNs/LPF) or moderate/heavy (>or=25 PMNs/LPF) for every TA.

Results: Purulence was absent in 469 of 646 (72%) TA taken from 170 infants. Light purulence was present in 17% and moderate/heavy purulence in 11%. TCs were positive in 58% of non-purulent, 94% of light, and 100% of moderate/heavy purulent TA. Bacteria on Gram stain were present in 12% of non-purulent, 70% of light purulent, and 83% of moderate/heavy purulent TA. Moderate/heavy purulence in TA was predictive of a positive TC with Gram-negative bacilli (GNB) with 70% sensitivity, 100% specificity, 100% positive predictive value, and 67% negative predictive value. Purulence in TA, as well as GNB airway colonization, became more frequent as mechanical ventilation progressed and was not associated with a particular GNB species. There were 79 infants who never had purulent TA and 91 who, at some time during the hospitalization, did. At the time of first purulent TA, 65 (71%) of 91 infants were asymptomatic. Twenty-six infants (29%) had clinical deterioration for which they underwent sepsis work-up. Three had blood stream infection, 5 VAP, 5 tracheitis, and 13 respiratory complications of non-infectious etiology. Four of five VAP infants died; all others survived.

Conclusion: In VLBW infants, purulence in TA is associated with prolonged endotracheal intubation and is temporally related to GNB airway colonization. At the time of the first purulent TA, the majority of mechanically ventilated VLBW infants are asymptomatic. Only a few symptomatic VLBW infants had nosocomial respiratory infection. Understanding the clinical significance of purulence and GNB in TA from this unique patient population is important for management and prognosis, and it may decrease concern for infection and the associated use of antibiotics.