ABSTRACT
Objective To identify the clinical and imaging characteristics in neonatal refractory purulent meningitis. Methods Clinical data of 70 cases of neonatal purulent meningitis admitted to the neonatal intensive care unit at Children's Hospital of Fudan University from January, 2009 to December, 2014 were reviewed retrospectively. The patients were divided into refractory group (n=28) and non-refractory group (n=42) according to the course of antimicrobial therapy.The clinical and brain MRI characteristics of neonatal refractory purulent meningitis were analyzed. Parameters were compared between the two groups using Chi-square or Fisher's exact tests, and Wilcoxon tests where appropriate. Risk factors of neonatal refractory purulent meningitis were investigated by univariate and multivariate Logistic regression analysis. Results Among the 70 cases, 31(44.3%) were positive for cerebrospinal fluid (CSF)/blood culture. The positive rate was higher in the refractory group than in the non-refractory group [75.0%(21/28) vs 23.8%(10/42),χ2=17.843, P<0.01]. The most common pathogenic bacteria isolated in the refractory group were Escherichia coli [8 cases (38.1%)] and group B streptococci [5 cases (23.8%)]. Compared to the non-refractory group, patients in the refractory group were more likely to have seizure, higher CSF white blood cell count, higher CSF protein concentration and lower CSF glucose concentration [53.6%(15/28) vs 7.1% (3/42), 965.0 (463.0-2 200.0)×106/L vs 116.5 (61.0-327.5)×106/L, 3 221.1(2 354.3-4 633.5) mg/L vs 1 487.6(988.2-1 924.1) mg/L, and 0.2 (0.1-0.8) mmol/L vs 1.5 (1.2-1.8) mmol/L; all P<0.01]. Multivariate Logistic regression analysis showed that seizure, low CSF glucose concentration on admission, and a positive CSF/blood culture result neonatal refractory purulent meningitis (OR=9.6, 95%CI: 1.2-76.0; OR=15.0, 95%CI: 5.6-63.3; and OR=7.3, 95%CI: 1.5-36.0, respectively). Abnormal brain MRI findings, including intracranial extracerebral space abnormality, ventricular dilatation and periventricular white matter injury, were more common in the refractory group [100.0%(28/28) vs 61.9%(26/42), χ2=13.827 totally; 64.3%(18/28) vs 21.4%(9/42), χ2=13.023 for intracranial extracerebral space abnormality; 60.7%(17/28) vs 19.0%(8/42), χ2=12.704 for ventricular dilation and 28.6%(8/28) vs 2.4%(1/42) for periventricular white matter injury; all P <0.01]. Compared with the non-refractory group, the refractory group had a longer hospital stay [(48.0±17.4) d vs (26.0±10.2) d, t=6.016, P<0.01] and more adverse events [67.9%(19/28) vs 31.0%(13/42), χ2=9.220, P=0.002], including hearing impairment and requirement of neurosurgical intervention [14/18 ears vs 10/46 ears (21.7%), χ2=4.292, P=0.038]. There was no death in both groups during hospitalization. Conclusions Neonates with seizure, low CSF glucose concentration and positive CSF/blood culture results are more likely to have refractory purulent meningitis. Brain MRI abnormalities are more common in neonatal refractory purulent meningitis.