HNL Helps Distinguish Infections
Fjaertoft G, Foucard T, Xu S, et al. Human neutrophil lipocalin (HNL) as a diagnostic tool in children with acute infections: a study of the kinetics. Acta Pediatrica costarricense. 2005;94:661-666.
In pediatrics, the clinician is often faced with the diagnostic challenge of differentiating a bacterial infectious process from a viral infection. History, physical exam, and laboratory data make the distinction. In this article, the authors’ purpose was to assess the kinetics of HNL with viral and bacterial infections. Further, they assess the response of HNL when the infection is treated with antibiotics. The response of HNL is compared with that of C-reactive protein.
In the study, 92 patients with a median age of 26 months were hospitalized because they required systemic antibiotics or because of the severity of their medical condition. Upon admission and on hospital days one, two, and three, the C-reactive protein, white blood cell count with differential, and HNL were measured. The patients were retrospectively classified into five groups: true bacterial infection (n=28), true viral infection (n=4), suspected bacterial infection (n=18), suspected viral (n=34), and other.
A true bacterial infection required bacterial isolation from blood, urine, or cerebrospinal fluid culture, or radiographic demonstration of pneumonia. Patients were classified as having a suspected bacterial infection if they had a nonspecific diagnosis, but an elevated C-reactive protein and erythrocyte sedimentation rate. A true viral infection required isolation of a virus. If a patient did not meet any of the above criteria, the person was classified as having a suspected viral infection. Those patients in the “other” group were diagnosed with Kawasaki disease, Borrelia meningitis, and one undiagnosed patient. The patients were classified using history, exam, and laboratory values including white blood cell count, erythrocyte sedimentation rate, C-reactive protein, and cultures. HNL values were not used in the classification.
The results demonstrated that both C-reactive protein and HNL are elevated with true bacterial infections compared with viral illnesses. Neither C-reactive protein nor HNL were significantly different in true bacterial infections versus suspected bacterial infections. The optimal cut-off for C-reactive protein was 59 mg/L with 93% sensitivity and 68% specificity. The optimal cutoff for HNL was 217 micrograms/L with 90% sensitivity and 74% specificity. In patients with true bacterial infections, HNL was highest at admission and decreased one day after admission. In contrast, the C-reactive protein values were similar on the day of admission and on hospital day one. C-reactive protein decreased significantly on days two and three of hospitalization. After hospital day one, HNL was elevated in only 11% of patients with true bacterial infection in contrast to 83% patients with elevated C-reactive protein.
In summary, HNL may be a useful marker to distinguish bacterial and viral illnesses. In comparison with C-reactive protein, it normalizes more rapidly after appropriate antibiotic therapy is initiated. In the future, HNL may be a useful marker in monitoring the response to antibiotic therapy.
CEDKA in Peds
Lawrence SE, Cummings EA, Gaboury I, et al. Population-based study of incidence and risk factors for cerebral edema in pediatric diabetic ketoacidosis. . 2005;146:688-692.
New onset insulin dependent diabetes mellitus is complicated by diabetic ketoacidosis (DKA) in 15% to 67% of patients. The incidence of cerebral edema in diabetic ketoacidosis (CEDKA) has been reported as 0.4-3.1. In the article, the authors seek to determine the incidence, outcome, and risk factors for cerebral edema in DKA in patients younger than 16.
The study was case-controlled with an active Canadian surveillance study. The authors surveyed pediatricians for a two-year period. During this time in Canada, all physicians were requested to submit reports monthly on patients with CEDKA younger than 16.