Use of translated versions of the MMSE with South Asian elderly patients in the UK.

The elderly population is increasing all over the world, a trend expected to continue well into the next century, particularly in low-income countries (Levkoff et al, 1995). There is an established association between increasing age and cognitive decline (Fillenbaum, 1984) and dementias are common in this age group.

Many South Asian people migrated to the UK in the 1950s and 1960s, mainly as young adults, to meet the demands of a growing labour market. Initially, therefore, older people constituted a relatively small proportion of the UK’s South Asian population. However, this proportion is now expected to grow (Rait et al, 1996).

South Asians in Britain are a heterogeneous group, with different religions, languages and cultures. Their mental health needs have been investigated to a lesser extent than their physical health needs. In this respect they have been disadvantaged by communication difficulties and other barriers to diagnosis, lack of culture-sensitive research, poor access to psychiatric services (Manthorpe & Hettiaratchy, 1993) and the traditional stigma attached to mental illnesses in their communities (Rait et al, 1996).

To diagnose dementia it is necessary to have a valid and reliable tool with which to assess cognitive function. However, sociocultural factors may complicate both the use of these tools and the interpretation of their results (Kabir & Herlitz, 2000). Efforts have been made to improve the validity of instruments used to screen for dementias by producing adapted and translated versions for different cultures. Adapted and translated versions of the Mini-Mental State Examination (MMSE; Folstein et al, 1975) have been developed in five South Asian languages commonly spoken in the UK – Bengali, Gujarati, Hindi, Punjabi and Urdu (Ganguli et al, 1995; Lindesay et al, 1997; Kabir & Herlitz, 2000; Rait et al, 2000). However, their use in routine clinical practice has not been widely reported.

The aim of this study was to determine the exposure of old-age psychiatrists in the West Midlands region of the UK to South Asian patients and their awareness and usage of translated versions primarily of the MMSE but also of any other assessment tools for cognitive impairment.

Method

Data collection was undertaken in an electronic and postal survey between March and April 2007. A brief questionnaire was developed that covered demographic details of respondents, their exposure to ethnic minority patients, and their awareness, previous experience with and views on the usefulness of translated or otherwise modified versions of the MMSE. Initially, the questionnaire was emailed to all the old-age psychiatrists working in the West Midlands. Those who did not respond were then sent a questionnaire by post, with a postage-paid reply envelope.

The West Midlands has a population of approximately 5.3 million. It has the largest Black and minority ethnic (BME) population outside London; 7.3% of residents are South Asian, whereas the national average for England is 4.6% (Office for National Statistics, 2009).

Results

Thirty-nine out of 66 questionnaires were returned completed, giving a response rate of 59%.

Fifteen (38%) respondents were female, 18 (46%) were male and 6 (15%) did not declare their gender. The ethnic background of the respondents was 16 (41%) White, 17 (44%) Asian and 6 (15%) Black.

In view of our response rate of approximately 60%, we tried to establish whether the responders were broadly representative of our original sample. For confidentiality reasons, we were not able to get a demographic breakdown of all the psychiatrists who were sent the questionnaire. We were, however, able to identify clearly the Asian names on our list. There were 28 of these, which represents 42% of the 66, which is similar to the percentage of responders (44%) who identified themselves as Asian. We therefore believe that the Asian responders were probably representative of the original sample.

Three respondents returned questionnaires that did not include information on frequency of contact with South Asian patients. Of the remaining 36 respondents, 3 (8%) had seen no patients from South Asian populations in the past 2 years, 20 (55%) had seen 1–9 patients, 8 (22%) had seen 10–20 and 5 (14%) had seen more than 20.

Eighteen (46%) psychiatrists believed that the number of South Asian patients they had seen underrepresented the morbidity in the population,15 (38%) felt it reflected the true morbidity and 6 (15%) were not sure.

Thirty-two respondents (82%) identified barriers to detecting cognitive deficit in their South Asian patients. The most common barriers mentioned were: language and communication problems, sociocultural issues such as attitude of the carers to mental illness due to the stigma attached, awareness of cognitive disorders, and educational status of the patients and carers. Lack of culturally sensitive tools to detect cognitive deficits was also commonly cited as a barrier. Some respondents questioned the validity or cultural appropriateness of commonly used assessment tools, even when translated. Other identified barriers included lack of education, fear of institutional racism, and poor access to and poor provision of services to these population groups.

Eleven consultants (28%) had used translated versions of the MMSE and 7 of these (64%) found them to be useful. Overall, 28 (72%) felt that clinicians should use the translated versions of standardised tools such as the MMSE when assessing patients from minority groups; only 6 (15%) could see no advantage in their use (5 did not respond to this item).

Effects of respondent ethnicity on the results

We were interested to see whether respondents’ own ethnicity had any effect on their responses. The questionnaire results for Asian and White respondents are compiled according to ethnicity in Table 1. A larger proportion of White respondents (44%) than Asian respondents (23%) thought that their exposure to ethnic minority patients was a true reflection of the morbidity in this population. More White respondents than Asian thought that there were barriers in detecting cognitive deficits in this population (93% v. 77%). Interestingly, however, only 7% of the White respondents had used translated/modified versions of screening tools, compared with 23% of Asian respondents. We were, however, surprised that a majority of the respondents who used the translated versions did not find these useful (Asians 75% and Whites 100%). This, though, contrasts with answers to the next question, which indicate that the majority felt medical staff at both primary and secondary care level would benefit from using modified/translated versions of the MMSE: 88% of the White and 71% of the Asian respondents answered yes to this question. The pattern among Black respondents was even more surprising: all six of them had used the translated version and found them useful but only a third considered their use as beneficial at both primary and secondary care level.

Table 1

Questionnaire results, by ethnic group

Questions	Response options	No. (%) of respondents
		Asian (n = 17)	White (n = 16)
1. How many patients within South Asian minorities with cognitive deficits have you seen in past 2 years?	<10	9 (53%)	11 (69%)
	10–20	1 (6%)	4 (25%)
	>20	5 (29%)	0 (0%)
	No response	2 (12%)	1 (6%)
2. Does it reflect the true level of morbidity in this group?	Yes	4 (24%)	7 (44%)
	No	9 (53%)	7 (44%)
	Not sure	4 (24%)	2 (13%)
3. Are there any barriers to detecting cognitive deficits in this group?	Yes	13 (77%)	15 (93%)
	No	4 (23%)	1 (7%)
4. Have you used any translated/modified versions of screening tools for these patients?	Yes	4 (23%)	1 (7%)
	No	13 (77%)	15 (93%)
5. Did you find these useful?	Yes	1 (25%)	0 (0%)
	No	3 (75%)	1 (100%)
6. Would medical staff in both primary and secondary care benefit from using translated/modified versions?	Yes	12 (71%)	14 (88%)
	No	2 (12%)	0 (0%)
	No response	3 (17%)	2 (12%)

Discussion

This research is the first we know of to have looked at the use of an existing tool for the cognitive assessment of minority ethnic elders in clinical settings in the UK.

Our findings indicate that most psychiatrists have seen South Asian patients, although the level of exposure was variable. This may reflect the demographic composition of the individual psychiatrist’s catchment population. However, nearly half of those who responded felt that their exposure to BME patients underrepresented the level of morbidity in that population. This may reflect the reported barriers to healthcare for the BME population (Commander et al, 1997).

Some respondents had used translated versions of the MMSE. However, the majority believed that translated versions had a wider potential and that both primary and secondary care medical staff would benefit from using them.

We were, however, disappointed to see that many psychiatrists were still unaware of the availability of standardised and validated adapted versions of these screening tools. It is possible that these translated versions, while commonly used for research, have not been made available to service providers for clinical use. Even when psychiatrists were aware of them, these tools were not being routinely used to overcome linguistic barriers, possibly owing to the unavailability of linguistically competent individuals, such as interpreters, in the process of assessment; this also probably accounts for the dissatisfaction of those who used these tools.

Therefore, the observation by Rait et al (2000) that ‘there remains an obvious need for a cross-national approach to improve detection, educate practitioners and improve services for older south Asians in the UK’ still holds true. We do, however, recognise that the use of translated versions of the MMSE is only one measure among many to improve the overall assessment process.