[Meningococcal sepsis in pediatrics. Parameters associated with poor outcome].

BACKGROUND

Mortality due to meningococcal sepsis continues to be extremely high. Patients with a poor prognosis require aggressive therapy and should be identified early.

OBJECTIVE

To investigate the clinical and biological factors associated with poor outcome.

PATIENTS AND METHOD: Seventy-one children aged 2 months to 13 years with meningococcal sepsis were studied. Inclusion criteria were meningococcus isolation in cultures or characteristic clinical features with purpuric exanthema.

METHODS

A correlational descriptive study was performed. In all patients we evaluated the Pediatric Risk of Mortality (PRISM), the Glasgow Scale for Meningococcal Sepsis (GSMS), polymorphonuclear (PMN) count and prolactin (PRL), leptin (LPT) and C-reactive protein (CRP) levels.

RESULTS

Fourteen children (19.7 %) died. Death was associated with multiple organ dysfunction syndrome (MODS) (p = 0.0001), high GSMS and PRISM scores (p = 0.0001) and to a lesser extent with shock (p = 0.01). In patients who died, the determinations showing greatest alteration at admission were PRL levels (p = 0.0009) and PMN count (p = 0.0005). CRP levels were not associated with differences in mortality but were high in patients with shock (p = 0.008). Children with high body weight percentiles were at greater risk of death and showed higher levels of PRL, PCT (p = 0.006) and LPT (p = 0.006), without differences in GSMS or PRISM scores. Age did not influence mortality or PRL levels but did influence GMSM and PRISM scores and PMN and CRP levels. These differences disappeared after the age of 2-3 years. In patients with MODS or shock, the only differences found were reduced PMN count (p = 0.0001) and elevated PRL levels (p = 0.0001).

CONCLUSIONS

In meningococcal sepsis, death is more frequent in children with high body weight percentiles. Moreover, these children present elevated PRL and LPT levels, although whether these variables act independently remains to be elucidated.