We thank Prasad for his comments and agree that there is some evidence for clinical benefit of steroids in terms of survival, functional deficit and symptoms.1 Our comment that steroids should be withheld before excluding an underlying infection in patients with meningitis may be misleading in suggesting that steroids are not useful for bacterial and tuberculous meningitis, because there is some evidence of benefit in these conditions.1 The evidence needs to be reexamined. We also need to individualize treatment.
The conclusion by McGee and Hirschmann that steroids are beneficial and safe for a wide variety of infections1 should not be taken at face value. There are infections (e.g., bacterial meningitis, severe typhoid fever and tetanus) for which the clinical benefit of steroid treatment has not been convincingly shown for all patients. Although some investigators have seen improved outcomes for bacterial meningitis,2,3 others have found no benefit.4,5 Furthermore, the observed clinical benefit of steroids for typhoid fever and tetanus was found in studies that involved only patients with more severe disease.6,7
The application of evidence-based recommendations on treatment should be appropriate to the specific clinical context. The studies of bacterial meningitis included only patients who had supporting evidence of bacterial infection (i.e., cloudy cerebrospinal fluid, bacteria seen on gram stain or white blood cell count > 1000 × 109/L).4,6 Our patient was a 48-year-old woman with meningitis of unknown cause, and clinical assessment of cerebrospinal fluid suggested that the cause was nonbacterial. Furthermore, a risk–benefit analysis suggests that witholding treatment is preferred so as not to aggravate unidentified infections (e.g., fungal) with steroid treatment.
We subsequently diagnosed Vogt–Koyanagi–Harada disease. Retrospectively, we find little justification for empirical steroid treatment, given that this disease is not known to cause death or neurologic disability.