1. The authors have retrospectively analyzed records of 37 children with tubercular meningitis hydrocephalus (TBMH) all of whom underwent ventriculoperitoneal shunt surgery, correlating stage of disease at the time of surgery and outcome. They conclude that since children who were shunted in earlier stage did better, shunts should be performed in all children with TBMH as soon as they are diagnosed.

It is well known that stage of the disease at the time of diagnosis is a strong prognostic indicator in TBM because it also determines the time when antitubercular therapy was started. Time of starting antitubercular therapy would thus act as a confounder/effect modifier as it is itself a predictor of outcome. The authors have not taken this into account. Although they mention the duration of antitubercular treatment in the group as a whole, they have not analyzed this with respect to outcome. To delineate the beneficial effect of shunt surgery and stage, one would have to compare outcomes in shunted vs nonshunted children stratified for stage. The better outcome in children who were shunted in earlier stages of TBM was probably because antitubercular therapy was also started earlier in these patients.

2. The authors themselves admit that shunt surgery in children has a high rate of complications. The shunt remains in place throughout the life of the person. In the social mileu of our country, it would be a stigma making marriage and jobs difficult. We can recount many children in whom shunt surgery was advised but was refused by the family and on follow up the child was normal or near normal. Such children would have been subjected to unnecessary shunt with its attendant complications and stigma.

3. It is quite clear that hydrocephalus in TBM does arrest in a proportion of patients. What is not clear is (i) What clinical or radiological features in the patient predict arrest of hydrocephalus (ii) whether shunt in such cases improves outcome and (iii) what clinical and radiological features (apart from just stage of disease) predict a favourable response to shunt surgery. I think these are questions that urgently need to be answered. Till then, it may be best to give a trial of medical treatment and shunt surgery be kept for those with TBMH who fail to improve on medical treatment.