Electronic Letters to:
|
|
eLetters: Electronic letters published:
-
![[Read eLetter]](/icons/misc/sectionDOWN.gif)
Suicide rates in people of Chinese origin in England and Wales?
- Edwin Lee (27 November 2008)
-
Suicide rates in people of South Asian origin in England and Wales
- Veena S Raleigh (10 February 2009)
-
Re: Suicide rates in people of South Asian origin in England and Wales: 1993–2003
- Peter Aspinall (10 February 2009)
-
South Asian Suicides in the UK
- Kwame J McKenzie, Kamaldeep Bhui (30 March 2009)
|
|
|||
|
Edwin Lee, Assistant Professor The Chinese University of Hong Kong
Send letter to journal:
elee{at}cuhk.edu.hk Edwin Lee
|
McKenzie et al deserve praise for their contribution in suicide research of ethnic minorities. They reported lower age-standardised suicide rate for men and young women of South Asian origin and higher suicide rate over 65 years than that of England and Wales.1 Their paper may be helpful in identifying at-risk groups for specific suicide prevention strategy. I propose to include Chinese names in their name recognition software in future studies for several reasons. First, Chinese women were found to have higher average annual age-adjusted death rates for suicide than Asian in the United State and Chinese immigrants men and women had 44% and 95% higher suicide rates than their US-born counterparts, respectively.2 Second, people of Chinese origin constitute a substantial proportion of non-white population with a total of over 247 thousands but there are limited number of studies of suicide-related behaviour in the United Kingdom examined people of Chinese origin.3,4 Third, distinctive epidemiological patterns of suicide were found in China that contrast with the patterns characteristic of Western societies and information of suicide in people of Chinese origin in England and Wales may be useful to understand the effect of migration on suicide.5 REFERENCES 1 McKenzie K, Bhui K, Nanchahal K, Blizard B. Suicide rates in people of South Asian origin in England and Wales: 1993-2003. Br J Psychiatry 2008; 193: 406-9. 2 Singh GK, Miller BA. Health, life expectancy, and mortality patterns among immigrant populations in the United States. Can J Public Health 2004; 95: I14-21. 3 National Statistics. Population of the United Kingdom: by ethnic group, April 2001. National Statistics Online (http://www.statistics.gov.uk/cci/nugget.asp?id=455). 4 Hunt IM, Robinson J, Bickley H, Meehan J, Parsons R, McCann K, Flynn S, Burns J, Shaw J, Kapur N, Appleby L. Suicides in ethnic minorities within 12 months of contact with mental health services. National clinical survey. Br J Psychiatry 2003; 183: 155-60. 5 Ji J, Kleinman A, Becker AE. Suicide in contemporary China: a review of China's distinctive suicide demographics in their sociocultural context. Harv Rev Psychiatry 2001; 9: 1-12. |
|||
|
|
|||
|
Veena S Raleigh, Epidemiologist Healthcare Commission, University of Surrey
Send letter to journal:
veena.raleigh{at}healthcarecommission.org.uk Veena S Raleigh
|
Sir McKenzie et al’s findings (1) of low suicide rates among South Asian men in both 1993-98 and 1999-2003, and of high suicide rates among young South Asian women in 1993-98, are consistent with previously reported findings (2). The difference from previous findings lies in the absence of an excess in young South Asian women in the recent period, 1999-2003, and an excess instead in older women. In the absence of observed numbers of deaths and confidence intervals for the rates, it is not possible to interpret the statistical significance of the findings in Tables 1 and 2 ie which ethnic differences by age, gender and over time are statistically significant. Likewise, although the results were “essentially unchanged” following the sensitivity analysis, it’s unclear which differences remained statistically significant after the 11% inflationary adjustment for potential under-identification of South Asian suicides arising from the use of SANGRA. High suicide and attempted suicide rates among young South Asian women have been a consistent and enduring finding in national and international research over decades (see 2 for references). Research specifically commissioned to examine this issue reported high attempted suicide rates among young South Asian women in London, including the UK- born (3). A recent study found a 2.8 fold higher suicide rate among South Asian women aged 25-39 in contact with mental health services (4). Given the evidence overall, any decline in suicide rates in this group over the past decade would therefore be welcome. However, as this finding goes counter to the evidence to date, it should be kept under review to ensure it is a real trend and not a data artefact, given the caveats associated with analyses based on software-assigned ethnicity, many of which are acknowledged in the paper. The constraints to inclusion of ethnicity at death registration were established by ONS in its review of death certification some years ago. Given the growing need for epidemiological monitoring of mortality rates and trends by ethnicity and cause of death, ONS, the Department of Health and the Information Centre should consider alternative approaches for making this data available eg through data linkage, as undertaken in Scotland and recently by ONS for deriving infant mortality rates by ethnic group (5). This would provide sound, comprehensive epidemiological data with self-assigned ethnicity coding of numerators and population denominators on a consistent and comparable basis, thereby avoiding the potential mismatch between numerators and denominators in the use of name- recognition software. It would also obviate the need for researchers to have access to names, which is frequently not possible for data protection reasons. In the interim, given the growing use of such proxies for epidemiological purposes, there is a strong case for these national agencies to undertake a systematic review of the available name- recognition software programs, to establish their robustness for epidemiological analyses using national datasets and across the spectrum of morbidity and mortality. This would also be in keeping with the statutory responsibility of these national agencies for ensuring the availability of comprehensive national data to support equality monitoring. NB: It may be helpful for other readers to clarify that in Tables 1 and 2, columns 2 and 3 refer to E&W and South Asians for 1993-1998 respectively, and columns 4 and 5 refer to E&W and South Asians for 1999- 2003 respectively. Dr Veena S Raleigh, Fellow in Information Policy, Healthcare Commission Reader, Postgraduate Medical School, University of Surrey 1. McKenzie K, Bhui K, Nanchahal K, Blizard B. Suicide rates in people of South Asian origin in England and Wales: 1993-2003. British Journal of Psychiatry 2008; 193: 406-409. 2. Raleigh VS. Suicide patterns and trends in people of Indian subcontinent and Caribbean origin in England and Wales. Ethnicity and Health 1996; 1: 55-63. 3. Bhugra D, Desai M, Baldwin DS. Attempted suicide in west London I. Rates across ethnic communities. Psychological Medicine; 1999; 29:1125 -30; 4. Bhui K, McKenzie K. Rates and risk factors by ethnic group for suicides within a year of contact with mental health services in England and Wales. Psychiatr Serv 59: 414-420, April 2008. 5. Infant mortality by ethnic group, England and Wales, 2005. Office for National Statistics, 2008. http://www.statistics.gov.uk/StatBase/Product.asp?vlnk=15111&Pos=1&ColRank=1&Rank=272 |
|||
|
|
|||
|
Peter Aspinall, Reader in Population Health Centre for Health Services Studies, University of Kent
Send letter to journal:
peter.petjasp{at}googlemail.com Peter Aspinall
|
Dear Sir, A notable finding in McKenzie et al’s study (1) of suicide rates in people of South Asian origin is that the high relative rates in younger Asian women reported in previous research studies are found in the 1993- 1998 dataset but not that for 1999-2003, which shows high relative rates for Asian women over 65. In discussing their results the investigators acknowledge potential problems with the study’s methodology, including the numerator (how well the SANGRA name recognition algorithm ascertains individuals of South Asian origin in more recent samples) and denominator (the validity of a linear interpolation of numbers over their period). However, perhaps cautions are required with respect to the overall robustness of the SANGRA algorithm and the issue of numerator/denominator compatibility: the numerator uses an operational definition of ethnicity (derived from name information) while the denominator is based on self- assignment by individuals to census categories. These matters are brought into focus in the derivation of denominators. The investigators use the counts for the 1991 categories ‘Bangladeshi’, ‘Indian’, ‘Pakistani’ and 2001 categories ‘Asian or Asian British: Bangladeshi, Indian, Pakistani’. They also include the 2001 category ‘White and Asian’ (numbering around 190,000 in the census) on the grounds that people in it ‘…could be identified by SANGRA if any of their names were of South Asian origin’. We have no systematic data on how offspring of these inter-ethnic unions are named, although qualitative research has revealed the complexity of the process (2). Inclusion of the ‘White and Asian’ category also introduces heterogeneity into the South Asian collectivity. Evidence from the ONS Longitudinal Study (LS) for members having a 1991 and 2001 ethnic group showed that half (49.0%) of the 993 ‘White & Asian’ persons identified as ‘White’ in 1991 and just 9.5% as one of the three South Asian groups (3). Similarly, in recent research half in the ‘White and Asian’ group prioritised ‘White’ when asked to name just one racial/ethnic group that contributes most strongly to their identity. Our collective identities affect our ability to make an individual life and have relevance in the context of suicide risk. The investigators exclude ‘Other Asian’ (the free-text ‘Any other Asian background’ under the ‘Asian or Asian British’ label, numbering around 240,000 in the 2001 Census) from the denominator ‘because the majority of this group are of Middle Eastern or Sri Lankan origin’. Although around 1 in 4 were born in Sri Lanka and 1 in 6 in the Middle East, 37% had a region of birth in South Asia and 31% in the UK (4). Given that the focus is on ethnicity rather than country of birth, the LS data is, again, informative: of members with a 1991 and 2001 ethnic group, 42% of 1,285 ‘Other Asian’ persons identified as Indian, Pakistani or Bangladeshi in 1991. In this study none from the ‘Other Asian’ group are counted in the denominator. Finally, the investigators point out that SANGRA was validated against real data. However, the key dataset was London and Midlands hospital inpatient admission data from the mid- to late-90s, a period during which the quality of ethnic coding was very poor, the team itself admitting that further studies are needed to confirm whether SANGRA is able to produce valid results across Britain (5). Beyond the parsimonious way in which the statistical data is presented (with no measure of the precision of the rate estimates), the collective effect of potential problems with numerator/denominator compatibility and concerns about SANGRA’s performance is a factor which needs to be considered in making a judgement whether to accept these findings as the accurate contemporary evidence needed to shape specific prevention strategies. Peter J Aspinall Reader in Population Health Centre for Health Services Studies University of Kent CANTERBURY Kent CT2 7NF 30 January 2009 REFERENCES 1. McKenzie K, Bhui K, Nanchahal K, Blizard B. Suicide rates in people of South Asian origin in England and Wales: 1993-2003. British Journal of Psychiatry 2008; 193: 406-409. 2. Edwards R, Caballero C. What’s in a name? An exploration of the significance of personal naming of ‘mixed’ children for parents from different racial, ethnic and faith backgrounds. Sociological Review 2008; 56(1): 39-60. 3. Platt L, Simpson L, Akinwale B. Stability and change in ethnic groups in England and Wales. Population Trends 2005; 121: 35-46. 4. Gardener D, Connolly H. Who are the ‘Other’ ethnic groups? London: Office for National Statistics, 2005. 5. Nanchahal K, Mangtani P, Alston M, dos Santos Silva I. Development and validation of a computerised South Asian Names and Recognition Algorithm (SANGRA) for use in British health-related studies. Journal Public Health Medicine 2001; 23: 278-85. Declaration of interest I have no interest to declare (with respect to fees and grants from, employment by, consultancy for, shared ownership in, or any close relationship with, and organisation whose interests, financial or otherwise, may be affected by publication of the letter) in respect of my authorship. |
|||
|
|
|||
|
Kwame J McKenzie, Medical Director for Diversity CAMH and University of Toronto, Kamaldeep Bhui
Send letter to journal:
kwame_mckenzie{at}camh.net Kwame J McKenzie, et al.
|
Authors reply Kwame McKenzie BM MRCPsych Professor of Psychiatry University of Toronto Senior Scientist Social Equity and Health Research Centre for Addictions and Mental Health, 455 Spadina Avenue Toronto, Ontario Canada M5S 2G8 Kwame_mckenzie@camh.net Kamaldeep Bhui, MD FRCPsych Professor of Cultural Psychiatry & Epidemiology Centre for Psychiatry, Barts & The London School of Medicine & Dentistry Queen Mary University of London & Hon. Consultant Psychiatrist East London Foundation Trust. Our paper is the first to report findings at variance with previous studies and we welcome the opportunity to discuss the findings and subject them to scientific scrutiny (1). The findings of a decreased rate of suicide in South Asian men has not been challenged. It is reassuring that the experimental methods of SANGRA do not produce unexpected findings for this group. The comparison with our study of the national confidential enquiry should be made cautiously, as that study included suicides among people in contact with services rather than from all deaths to ONS (2). We would also suggest self-harm rates are not a proxy for comparative suicide rates. Dr Aspinall makes important comments about ethnicity classification. There are no data that investigate self-assigned versus ascribed ethnic identity, and variations of this relationship across geographical areas of the UK, over time, or the patterns of transmission of ethnic identity through the generations. There are often unpleasant trade-offs when using descriptors of ethnicity and culture from survey research (3). Ethnicity is not a measure of cultural identity (3). Perhaps nested within self- reported ethnic categories we need more complex models of identity that take account of acculturation, social stratification and their interaction (4). This may help more precisely to disentangle specific influences on health. Unfortunately, the concepts and methods to do this are still being developed. The information on the denominators so far is useful but incomplete to forge a new study design or recommend specific changes in routine data sets. ; for example, we would need a breakdown of self reported ethnicity in the Asian Other and White and Asian categories by sex and age. Adding more ethnic categories which are imprecisely measured, or for which the difference between self-rated and ascribed may vary over time and place, may lead to more random misclassification; therefore more ethnic categories may not always be helpful or explain any more precisely which specific ethnic identity groups are at greater or lesser risk. The finding of high rates of suicide in young South Asian women in the UK are based mainly on papers sampling groups born in Southern Asia – using the same methodology would miss the 50% of South Asians currently in the UK (1). Of the two studies that used different methodologies, one used names to ascertain South Asian suicides but the methodology was not validated or described so that it could be replicated, and the other studied part of London, though we know that there are significant differences in South Asian suicide rates by geographical location. (5). The main purpose of the study was to improve the accuracy of the estimate of suicide rates in all South Asian people living in the UK, irrespective of place of birth. We know that over 50% of the South Asians population were born in the UK and future studies need a way of accurately including them in rate calculations. We believe that ethnicity assigned on death certificates is likely to be the most useful way forward. We concur with the view that there is a need to assess trends over time. We agree that there are caveats because of the SANGRA program. However, we do not think that the program would have worked less well in the 1999-2003 cohorts than the 1993-1998. That is would work so differentially for men and women, or that artefact can explain the findings in both the older and the younger female population. We would welcome attempts to replicate these findings using different methodologies and show the raw data to facilitate this and the calculation of other statistics such as confidence intervals (table 1) . However, in the meantime, this is the first contemporary attempt to ascertain the suicide rates for the whole of the UK South Asian population. The findings for older women should be viewed with concern and the drop in younger women should be replicated and followed over time. Policy has to be based on the best evidence available rather than the best evidence that is sought and might one day become available. References 1) McKenzie K, Bhui K, Nanchahal K, Blizard B.Suicide rates in people of South Asian origin in England and Wales: 1993-2003. Br J Psychiatry. 2008 Nov;193(5):406-9. 2) Bhui K, McKenzie K, Rates and risk factors by ethnic group for suicides within a year of contact with mental health services. Psychiatric Services 59: 4; 414-420 3) McKenzie K, Crowcroft NS. Describing race, ethnicity and culture in medical research BMJ 1996; 312: 1504. 4) K Bhui1, Y Khatib1, R Viner2, E Klineberg1, C Clark1, J Head1,2, S Stansfeld1 Cultural identity, clothing and common mental disorder: a prospective school-based study of white British and Bangladeshi adolescents Journal of Epidemiology and Community Health 2008;62:435-441; doi:10.1136/jech.2007.063149 5) Neeleman, J. & Wessley, S. (1999) Ethnic minority suicide: a small area geographical study in south London. Psychological Medicine 1999, 29, 429-436. Table 1 Sangra South Asian Suicides 1993-2003 1993-1998 1999-2003 Male 15-24 136 72 25-34 145 158 35-44 121 109 45-54 54 72 55-64 46 39 65-74 18 18 over75 14 10 Total 534 478 Female 15-24 60 29 25-34 75 55 35-44 54 29 45-54 28 25 55-64 12 15 65-74 13 14 over75 6 11 Total 248 178 Authors reply |
|||