It is important to develop cost-effective and efficient methods of treatment in emergency psychiatry, especially where resources are poor. Alexander et al (Reference Alexander, Tharyan and Adams2004) in their paper comparing two methods of rapid tranquillisation concluded that the injectable haloperidol-promethazine mix is as effective as lorazepam and suggested that in India the former is more cost-effective. We acknowledge the findings of their study but would like to make some observations regarding cost-effectiveness and methodology.

The preferred combination for rapid tranquillisation at the two largest psychiatric centres in India (the National Institute of Mental Health and Neurosciences, Bangalore, and the Central Institute of Psychiatry, Ranchi) (combined monthly out-patient attendance of 49000) is haloperidol with lorazepam rather than haloperidol with promethazine. This is guided by the literature as well as existing practice (Reference McAllister-Williams and Nicol FerrierMcAllister-Williams & Nicol Ferrier, 2002; Reference Hughes and KleespiesHughes & Kleespies, 2003). This combination is about 25% cheaper than the haloperidol-promethazine mix (CIMS, 2004). Since promethazine has both alpha-1 and dopaminergic antagonism its combination with haloperidol is more likely to produce hypotension and neuroleptic malignant syndrome in agitated patients, who are often dehydrated and have electrolyte imbalance. On the other hand lorazepam decreases the required dose of haloperidol. Hence we feel that the combination of haloperidol with promethazine for rapid tranquillisation may not be the most cost-effective or the most efficacious even when resources are poor.

Alexander et al used the Clinical Global Impression Scale to rate aggression and violence. We feel that use of more aggression-specific measures, such as the Overt Aggression Scale (Reference Coccaro, Harvey and Kupsaw-LawrenceCoccaro et al, 1991), which assesses different aspects of aggression and its severity, would have generated more specific results.

Alexander et al also showed that the combination injection produces sedation quicker than intramuscular lorazepam. However, this finding should be viewed with caution because the lorazepam group included more patients with mania, more patients with substance misuse or already on benzodiazepines (who could have developed tolerance to benzodiazepines) and more patients with marked or severe illness (which would necessitate a higher dose of medication to control aggression and violence). Together these factors might have contributed significantly to the results.