COMMENTARY Dexamethasone for Management of Neonatal Meningitis Praveen Kumar & Gagan Mahajan Received: 17 January 2013 / Accepted: 17 January 2013 / Published online: 26 January 2013 # Dr. K C Chaudhuri Foundation 2013 Bacterial meningitis is associated with significant mortality and devastating neurological sequelae. The pathophysiology of neurological lesions in meningitis correlates with the severity of inflammation in cerebrospinal fluid (CSF), which can be alleviated by systemic corticosteroids. Ran- domized clinical trials (RCTs) to assess the efficacy and safety of adjunctive dexamethasone in bacterial meningitis, however, have shown conflicting results. A 2010 update of Cochrane review identified 24 trials with 4041 participants and concluded that corticosteroids significantly reduced hearing loss and neurological sequelae in high-income countries, but did not reduce overall mortality. There were no beneficial effects in low-income countries [1]. A recent systematic review included 44 publications, and reached similar conclusions [2]. An individual patient data meta- analysis showed that dexamethasone did not reduce deaths or composite endpoints of death, neurologic sequelae, or hearing loss [3]. Studies from Pakistan and India have shown either no benefit or increased mortality with dexa- methasone [4, 5]. The different results in developing world may relate to late diagnosis, pre-treatment with antibiotics before patients reach referral hospitals, different organisms or frequent malnourishment. Meningitis is more common in neonatal period than at any other time in life. The true incidence of neonatal bacte- rial meningitis may be underestimated, particularly in resource-poor settings because of non-performance of lum- bar puncture in all cases of neonatal sepsis, lack of micro- biological support and difficulties in interpretation of the cerebrospinal fluid findings in the newborn [6]. Mortality from neonatal meningitis in developing countries is estimated to be 40– 58 % against 10 % in developed countries [6]. The world wide risk of at least one persisting major or minor sequelae is 20 % (IQR 12 to 35 %) with large geographic variations. The natural corollary to above information is whether steroids would work in neonatal meningitis. The situation in the newborn is different in two major ways from that of pediatric or adult meningitis—first, the organism profile is different and secondly, the brain is still developing. There are two questions to be addressed. First, are steroids effec- tive in neonatal meningitis? The second, even if cortico- steroids are effective, are they safe for the developing brain? In this issue, Mathur et al. report a much desired RCT of dexamethasone in a relatively large number of neonates with meningitis [7]. They show a dramatic reduction in mortality and hearing loss among survivors. As a possible mecha- nism, they also demonstrate significant reduction in CSF cell count, proteins and some cytokines after 24±6 h of dexamethasone treatment. However, the study has several limitations which restrict the conclusions which can be drawn. On the methodological front, the drawbacks of non-blinding are obvious even though mortality is a hard outcome and the laboratory was apparently blinded. Appli- cation of blocking in the random allocation sequence in a non-blinded study provides an obvious route to revelation of the sequence and a chance for selection bias. It is remarkable that none of the 101 lumbar punctures was traumatic. In the literature, the incidence of traumatic taps in neonatal lumbar punctures has varied between 15–35 %. Neonates who had received antibiotics for less than 24 h before enrollment were included in the trial. On the other hand, it has been shown that dexamethasone has effective anti-inflammatory activity if given before or simultaneously with the first dose of antibiotic but not if given 1 h later. The paper does not provide information whether the number of babies who received prior antibiotics were equal in each group. The P. Kumar (*) : G. Mahajan Department of Pediatrics, Post Graduate Institute of Medical Education and Research, Chandigarh, India e-mail: drpkumarpgi@gmail.com Indian J Pediatr (February 2013) 80(2):155–156 DOI 10.1007/s12098-013-0975-1