OBJECTIVE: To determine the frequency of clinical situations in which the CSF GL and PR influence MD decisions regarding disposition (DISP) and antibiotic therapy (Rx) in patients (pts) undergoing lumbar puncture (LP) in the pediatric ED.

DESIGN: Prospective survey of MDs performing LPs in EDs of 4 academic children's hospitals. Data collected at the time of LP included age, LP indication, history of antibiotic use, CSF results (WBC/differential, RBC, GL, PR, & Gram stain), DISP and Rx. MDs were also asked, ”As an independent factor, did the CSF GL or PR result cause you to do anything differently regarding: 1. your DISP decision (admit/discharge)? 2. your Rx decision (antibiotics/no antibiotics)?“

RESULTS: 714 LPs were reviewed. 69 were excluded because MDs had failed to answer questions re: DISP/Rx; 38 because no CSF GL or PR were obtained. In the 607 study pts, there were no significant differences between pts with normal or abnormal CSF GL for DISP or Rx. Pts with abnormal CSF PR were more likely admitted (79.9% vs. 67%, χ2 = 12.17, p <.001) and treated (90% vs. 80%, χ2 = 10.99, p <.001) than those with normal CSF PR. MDs answered “yes” regarding the CSF GL & PR's influence on DISP in 15 cases (2.47%, 95%CI = 1.39%-4.04%) and “yes” to an influence on Rx in 11 cases (1.81%, 95%CI = 0.91%-3.22%). There were no significant differences between “yes” and “no” cases in terms of% neonates or% normal CSF GL. “Yes” cases were more likely pretreated (6.8% vs. 2.4%,χ2 = 5.085, p =.024) and more likely to have normal CSF PR (77.8% vs. 53%, χ2 = 4.300, p =.038), but were less often admitted (38.9% vs. 74.2%, χ2 = 11.11, p <.001) or treated (66.7% vs. 84.3%,χ2 = 3.977, p =.046).

CONCLUSIONS: In the pediatric ED, CSF GL and PR seldom influence MD DISP or Rx decisions and should not be ordered routinely.