[Diagnostic test in emergency departments for bacterial infections in infants younger than 12 months]

To evaluate clinical and analytic numeric data that may help the emergency departments to identify bacterial infections in infants.

A retrospective study of 430 infants with bacterial growth in cultures (culture from blood, 30; urine, 207; stools, 193, and/or cerebrospinal fluid, n 25) was performed. These patients were compared with a control group (n 430), randomly selected from patients aged less than 12 months with negative cultures who were hospitalized with suspected infection. Neonates and surgical patients were excluded from both groups. Statistical analysis was performed using Student's t-test for independent samples, Levene's test for the study of equality of variances, bivariate correlation and one-factor ANOVA, and receiver-operating characteristic (ROC) curves and odds ratios were calculated when statistically significant (p < 0.05) results were obtained. These analyses were performed using the SPSS 10.0 statistical software package.

Of the infants admitted to the pediatric unit, 11.7 % had at least one positive bacterial culture. Temperature (p 0.005), leucocyte count (p 0.003), percentage of neutrophils (p < 0.0001) and C-reactive protein (p < 0.0001) were significantly higher in infants with positive cultures. In invasive infections significant differences were found in sex (more frequent in males) (p 0.03), heart rate (p < 0.0001) and respiratory rate (p 0.003). In the ROC curves, the best diagnostic yield was obtained for C-reactive protein (0.93 for a cutoff value of 29 mg/l, 86 % specificity and 91 % sensitivity).

C-reactive protein is essential for diagnosis of bacterial infection in infants in the emergency department.