Background: The Mini-mental State Examination (MMSE) is a globally used instrument for cognitive screening, which nevertheless has a bias with respect to education and language. The Indo-U.S. Cross National Epidemiology Study developed a modified version of MMSE, the Hindi Mental State Examination (HMSE), to counter this bias in India among rural and illiterate elderly. Thus, two parallel tests are available for screening the population. This study was conducted to explore and compare the applicability of MMSE and HMSE when scanning for cognitive impairment among urban elderly people.

Methods: The sample consisted of 40 subjects (20 illiterate and 20 literate) aged 60 years and above drawn from the urban community who met the inclusion criteria. A systematically translated Hindi version of MMSE (HVMMSE) was administered to both groups. After one month, HMSE was administered to the same groups. χ2 with Yate's correction, percentage, rank order correlation and qualitative analysis were used to analyze data.

Results: All illiterate subjects scored below the cut-off on translated HVMMSE while only four of them scored below the cut-off on HMSE. Among literate subjects, the translated HVMMSE and HMSE classified three subjects and one subject respectively as having possible cognitive impairment among urban elderlies.

Conclusion: The two instruments are not in agreement with regard to classifying elderly people as having possible cognitive impairment or not. This disagreement is more pronounced in the case of illiterate subjects. The study emphasizes the need to develop a fair cognitive screening instrument for elderly people in India.