Epidemiological Findings on the Prevalence of Mental Disorder in India

Sir, I read with great interest the article on epidemiological findings on the prevalence of mental disorders in India (Ganguli,2000). As we do not have epidemilogical studies with sound methodology metaanalysis of the existing studies will definitely help to provide baseline data of varios mental disorders in India. In the metaanalysis the author has included studies from 1967 to 1993. Regrettably, the author has left an epidemiological study conducted in Kerala in the year 1995 (Shaji et al.,1995). This study satisfy the three inclusion cnteria listed in the metaanalysis viz. delineation of sample and collection of demographic data, identification of cases by social workers with good inter-rater reliability coefficient of 0.97 and confirmation of psychiatric diagnosis by two independent psychiatrists according to ICD-10 Diagnostic Criteria. This study has also mentioned the sensitivity and specificity of screening test as 98.7 and 97 5 respectively which shows that this type of survey is statistically and clinically useful. In this survey authors have given prevalence rate of five priority psychiatric disorders including organic psychoses, affective disorder, schizophrenia, epilepsy and mental retardation. Inclusion of this recent epidemiological study in the metaanalysis would have given the most recent baseline data of mental disorders in India. Reddy and Chandrasekhar (1998) have reported a metaanalysis' of various epidemiological studies fulfilling the required selection criteria during the period from 1967 to 1995. I would like to point some discrepancies in these two metaanalyses. The national prevalence rate for all mental disorders reported by Ganguli (73/1000) is much higher than the rate reported by Reddy and Chandrasekhar (58.2/1000) with the difference reflecting mainly in the rural prevalence rate (70.5/1000 versus 48.9/1000). The 1998 metaanalysis reports urban morbidity which is 1.5 times higher than the rural rate. Another major difference was in the national rate of affective disorder which was 34/1000 in 1998 metaanalysis against the national rate of 12.3/1000 in Ganguli's. Unfortunately there are no baseline data on organic psychoses, alcohol/drug addiction and epilepsy in the author's metaanalysis despite that these are reported in many epidemiological studies. It would have been more informative if the author has compared his results with the previous metaanalyses and discussed the reasons for these divergent findings.