A diagnostic decision rule for management of children with meningeal signs.

In a previous study we devised a diagnostic decision rule to improve management of children with meningeal signs, suspected of having bacterial meningitis. The decision rule aimed to guide decisions on (1) whether a lumbar puncture is necessary in children with meningeal signs, and (2) which children need hospitalisation and empirical antibiotic treatment for bacterial meningitis. In this study we assessed the validity of this rule in an external population of four (paediatric) hospitals in The Netherlands. The decision rule included two scoring algorithms using symptoms, signs and quickly available blood and cerebrospinal fluid (CSF) laboratory tests. To evaluate the discriminative value of both algorithms, the absolute numbers of correctly diagnosed patients and the area under the receiver operator characteristic curve were estimated, and compared with the results from the original population (n = 360). In a 18 month period, we included 226 children, median age 2.2 years, who visited the emergency department with meningeal signs. Bacterial meningitis was present in 25 (11%). Using the scoring algorithms patients could be categorised in groups of increasing risk of bacterial meningitis. The discriminative values of the clinical and CSF algorithm in this new population were similar to those in the original population. In the total population of 586 children with meningeal signs, the rule selected 205 children (35%) who did not need a lumbar puncture and 366 children who did not need empirical treatment (62%). In conclusion, this diagnostic rule performed well in a new population of children with meningeal signs. This diagnostic decision rule is a valuable tool for the clinician when deciding to treat these children for bacterial meningitis and thus improving their management.