Ethnicity of the deceased person: further background

From the initial request to implementation

In July 2009, the Director of Public Health Science at NHS Health Scotland wrote to the Registrar General for Scotland. He explained that a working group, which he chaired, was producing a research programme for ethnicity and health in Scotland. Its report would express the group’s view that the absence of an ethnicity code on the death certificate was a major obstacle to assessing the relative mortality rates of different ethnic groups, and one which was not adequately overcome either by probability linkage of death records to the census, or by using country of birth as a proxy. Therefore, the working group would like to recommend that ethnic identity should be recorded on the death certificate, using the same classification as for the then forthcoming 2011 Census.

On 6 August 2009, the Director of Public Health Science met the Registrar General. They discussed other possible ways of achieving the same end:

• getting the Community Health Index (CHI) number recorded on the Medical Certificate of the Cause of Death, and getting ethnicity recorded on the CHI. They felt that this would achieve the same effect but it would take a long time to get anything like complete coverage of the population;
• using the Scottish Longitudinal Study. But its sample (for example about 1 in 20 of the 2001 Census population) was not large enough to provide information about relatively small minority ethnic groups;
• ad-hoc data linkage between the death registration and the most recent census (as had been done for work on the cardiac health of people from the Indian sub-continent). They felt that this would be effective, but very laborious.

Having considered the options, the Registrar General agreed that, in principle, the best way forward would be to use the registration computer system to collect the information when a death was registered, which would make it available for statistical purposes, without it being recorded on the death certificate. However, he noted that a number of points had to be resolved, including:

• the legal specification of the new format of the information to be provided when a death is registered;
• the IT changes that would be needed to record the information;
• informing registrars about the change, and the reasons for it. The forthcoming working group report would be the trigger for this, and the process would start at the Registrar General’s annual meeting with representatives of all 32 registration authorities;
• the methodological work necessary to ensure that the ethnicity information to be collected is specified correctly – based on the categorisation prepared for the 2011 Census, but possibly adjusted to make coding simple;
• possible consultation with minority ethnic groups.

The working group’s report was published on 9 November 2009, and is available in the Health in our multi-ethnic Scotland - future research priorities publication on the Health Scotland website.

In the Foreword, the Scottish Government’s Chief Medical Office stated that the report’s recommendations had his full support. He also wrote that:

“The reason we know less than we should is largely because we do not collect data on ethnic identity on the death certificate and efforts to do so for health records have been patchy so far. I strongly support the Working Group’s conclusion that our top priority must be to improve the quality of our data collection. Without this, our ability to make those important links between ethnicity and health will be severely limited. I will be working with colleagues at the Scottish Government to help bring about the changes that are needed.”

Over the next year or so, the Registrar General’s staff developed proposals for the collection, when a death is registered, of information about the deceased person’s ethnic group.

Consultation

In June 2011, the Registrar General consulted a range of interested parties (including local authorities, NHS boards, and representatives of ethnic minority organisations, different parts of the medical profession, funeral directors and burial and cremation authorities) about the proposed approach:

• the provision of data would be voluntary;
• the information would not appear in the entry in the Register of Deaths, or in any Extract (the ‘death certificate’) that was produced;
• the person who was registering the death would be asked for consent to the sharing of the data with the NHS, and to the use of the data for research and to produce statistics.

Over 30 responses to the consultation were received. No-one opposed the proposals, but some had questions or raised possible issues.

Implementation

In December 2011, National Records of Scotland (NRS) circular letter R4/2011 informed registrars that they should collect this information for deaths registered with effect from 1 January 2012, and about the new procedure which should be followed more information can be found in the How the information should be collected section of this website. Key points are as follows:

• the registrar should ask whether the informant is willing to provide the Registrar General with information about the deceased person’s ethnic group, and describe briefly the purposes for which the information might be used;
• if the informant agrees to provide the information, the registrar should ask him/her to pick the most appropriate ethnic group from a list of categories used in 2011 Census outputs;
• otherwise, the registrar should record either ‘not willing to provide’ or ‘ethnic group not known’ (whichever is appropriate);
• the information is held in the registration computer system, and may be used to produce statistics. However, it does not appear in the ‘death certificate’ (this is the Extract of the entry in the Register of Deaths).

The introduction of the collection of this information was reported in the BMJ, in the news section of the BMJ website.