The following article requires a subscription:



(Format: HTML, PDF)

Objective: To evaluate the usefulness of the Gram-specific probe-based quantitative polymerase chain reaction test for rapid detection and differentiation of Gram-negative and Gram-positive bacterial bloodstream infection in preterm infants.

Design: Cross-sectional study.

Setting: University-affiliated Level III neonatal intensive care unit.

Patients: Preterm infants with clinical features suggestive of late-onset infection.

Interventions: None.

Measurements and Main Results: In addition to the full sepsis screen, 0.5 mL of EDTA blood was collected aseptically for Gram-specific quantitative polymerase chain reaction evaluation. The results were analyzed with respect to outcomes of bacterial culture in blood and other body fluids, including peritoneal and cerebrospinal fluids. The diagnostic utilities of the quantitative polymerase chain reaction were determined. A total of 218 suspected infection episodes were investigated, of which 42 episodes were culture positive and 176 were culture negative. For Gram-negative infection, the quantitative polymerase chain reaction test correctly identified 19 of 22 episodes, and the sensitivity and specificity were 86.4% and 99.0%, respectively. For Gram-positive infection, the test correctly identified 14/19 episodes, and the sensitivity and specificity were 73.7% and 98.5%. The remaining one episode was Candida albicans septicemia. None of the episodes with positive quantitative polymerase chain reaction test were classified into the wrong Gram stain category. More importantly, despite negative blood culture in five infants suffering from intra-abdominal sepsis (peritonitis [n = 4] and hepatosplenic abscess [n = 1]), the quantitative polymerase chain reaction test could detect the Gram-specific category of causative organisms in blood.

Conclusions: The Gram-specific quantitative polymerase chain reaction test is reliable and highly specific for rapid identification and differentiation of Gram-negative and Gram-positive bloodstream and intra-abdominal infections. The result could be made available within 5 hrs after the specimen reaches the laboratory. A positive test is able to "rule in" bacterial bloodstream infection before blood culture results become available, and serves as a guide to predict the virulence of the causative organism according to its Gram-specific category so that critical patients can be targeted for intensive treatment.

(C) 2009 by the Society of Critical Care Medicine and Lippincott Williams & Wilkins