A pilot evaluation of a rapid screening test for nosocomial infection at a neurosurgical intensive care unit in Sweden

Abstract

Background When postneurosurgery patients develop fever, there are no convenient methods to immediately indicate the site of infection. The choice of empirical antibiotic therapy is evidently different in nosocomial meningitis compared with ventilator-associated pneumonia or urinary tract infection. Conventional bacterial cultures run a risk of being false negative due to antibiotic prophylaxis, and direct microscopic analysis of cerebrospinal fluid (CSF) from such patients has limited diagnostic value. Because of the substantial mortality associated with nosocomial meningitis, broad spectrum antibiotics in high dosage are, therefore, commonly administered. Neutrophils as a part of the innate immunity system, trap and kill bacteria by neutrophil extracellular traps (NETs). Neutrophil extracellular traps are composed of extracellular DNA which is released to the CSF during bacterial meningitis. Using a combination of sulphated-glucosaminoglycan and aniline dyes, a measurement method was developed that reacts to extracellular host DNA and changes color within 1 minute in proportion to the amount of NETs in the body fluid tested. The present study aimed to evaluate the sensitivity and specificity of the "rapid NETs test"in identifying the site of infection. Methods We performed analysis on the left-over CSF samples (n = 199) that were collected routinely at neurosurgical intensive care unit. No samples were included after patients were transferred to the ward. Besides CSF, bronchoalveolar fluid, and urine samples were collected in febrile patients. The "rapid NETs test"was used for simultaneous analysis of the fresh left-over samples. The final diagnosis was settled at discharge. Results A total of 75 (64%) patients had received empirical antibiotic therapy against nosocomial meningitis, whereas only 19 patients of these cases (16%) had a verified diagnosis. The rapid NETs test could distinguish a verified meningitis (n = 19) with 89.5% sensitivity and 92.5% specificity. The test also identified ventilator-associated pneumonia (n = 32) with 93.8% sensitivity and 86.8% specificity. However, the test was not reliable in identifying nosocomial urinary tract infection (sensitivity and specificity, 78.6% and 55.2%, respectively). Conclusions The rapid NETs test indicated the site of infection in febrile patients postneurosurgery with clinically relevant sensitivity and specificity, which might show its potential to minimize the unnecessary use of antibiotics.