Background Paediatricians are concerned that children who present with a non-blanching rash may have meningococcal disease (MCD). An algorithm to help identify which children with a non-blanching rash have MCD has been proposed in recent NICE guidance.
Aim To evaluate the NICE algorithm on previously collected data from children presenting with non-blanching rash.
Methods The algorithm was applied retrospectively to seven cohorts of children who had presented to paediatricians with non-blanching rashes in six paediatric units between 1999 and 2002.
Results The cohorts included 391 children (median age 2 years; range 1 month-15 years). 53 (13.5%) of these children had confirmed or probable MCD. Clinicians had empirically treated 171 (44%) children with antibiotics. This included 52 of the 53 children with MCD. The NICE algorithm suggested that 223 children (57%) should be treated with antibiotics. This included 51 of the 53 children with MCD. Two children with possible MCD were not identified by the algorithm; one child had no clinical features of MCD and no fever, but a CRP of 146 mg/l; the other child also had no clinical features but a neutrophil count of 0.89 x 109/l. Both children were given 5 days of intravenous antibiotics by their treating paediatricians, although neither had meningococcus confirmed on blood culture or PCR.
Conclusion The NICE algorithm would have lead to more children with non-blanching rash being given empirical antibiotics than actually happened (57% vs 44%). The algorithm identified 51/53 children with confirmed or probable MCD which was similar to the clinicians (52/53). Including CRP measurement on all children (not just those with fever) and treating children with low white cell counts (not just raised counts) might improve the sensitivity of the algorithm, but may lead to even more children being given antibiotics.