To the Editor: The standard recommendation of giving a high dose of dexamethasone in severe typhoid fever, especially when complicated by neurological findings, lacks any systematic review or meta-analysis. The minimal randomised evidence from two trials [1, 2] which was later disputed by another trial [3] in a tropical disease with a high incidence and a high rate of complications, is debatable. These trials excluded the very high-risk patients who died within 6 h, even though the trial specifically aimed to study high-risk patients. In fact, 68 patients had been randomized in the trial, but only 38 were analysed. For 2 of the 68 cases, it is not even stated which arm they were randomized to. For the other 66 patients, the death difference was not statistically significant (10 ⁄ 33 vs. 18 ⁄ 33, P = 0.057). These data are over 25 y old, and this regimen has not been redemonstrated to be effective in any recent or updated trials.

It is also relevant to discuss the maximum dose of steroid. It is a point of discussion among the clinicians when dexamethasone in the dosage of 120 mg is prescribed to a 40 kg weight teenager and 15 vials of dexamethasone are used. If the usual mechanism postulated in treating the complication of typhoid fever involves inhibition of the acute inflammatory cascade, can this steroid regimen be used in other gram-negative sepsis?

It is a challenge for future trials as the culture results are unknown at the time of presentation. Robust clinical trials with a large number of participants are needed to further strengthen the recommendations. In an ideal condition, patient tailored treatment is needed, and evaluation of the baseline cortisol level is pre-emptive to guide dexamethasone-based therapy in these patients.