Official statistics on health inequalities

Data and statistics about health inequalities can tell us which groups are experiencing poorer health outcomes and why, and how inequalities are improving or worsening over time. They can also give us vital information on the impact of events such as the COVID-19 pandemic and the cost-of-living crisis, guide policies and targeted interventions to tackle inequalities, and indicate the effectiveness of these interventions.

The main official statistics on health inequalities were previously thought to be those looking at the relationship between socioeconomic circumstances and outcomes such as life expectancy, healthy life expectancy and avoidable mortality. More recently though, statistics producers have started to focus on broader factors, such as the social determinants of health inequalities and their outcomes.

For example, the Department of Health (Northern Ireland) publishes an annual report on the indicators underpinning the Making Life Better strategy. This strategic framework provides direction for policies and actions to improve the health and wellbeing of the people of Northern Ireland and reduce inequalities in health. Meanwhile, Public Health Scotland’s ScotPHO online profiles tool contains a Care and Wellbeing Profile focused on indicators relating to population health, inequalities and wider determinants of health. The indicators in the profile are structured around the Marmot framework, and support the aims of the Scottish Government’s Care and Wellbeing Portfolio.

To fully understand health inequalities, it is important to have demographic data, such as data on ethnicity and deprivation level, available for analysis. We often hear that users of statistics would like more information about demographics to explore inequalities, for example in our assessment of A&E statistics in England last year. The need for more-granular data can present challenges, for example if demographic information is not collected in an administrative dataset or if analysis of multiple variables results in small sample sizes.

Producers have been using novel approaches, such as data linkage and the use of rolling datasets, to tackle some of these challenges. One example is the experimental statistics release looking at inequalities in mortality involving common physical health conditions in England. This was a collaborative piece of work carried out between ONS and the Department of Health and Social Care, with input from clinical experts from across the NHS. The analysis involved linking census data to death registrations, and enabled ONS to publish age-standardised mortality rates broken down by sociodemographic characteristics such as sex and ethnicity.

Another example of a novel approach to analysis is the most recent National Survey for Wales. Statisticians in the Welsh Government pooled three years of data so that they could carry out analyses on protected characteristics such as disability, ethnic group, marital status, religion and sexual orientation. This would not have been possible to do with data from a single year. New systems are also being developed within the NHS in Wales to better capture and link patient characteristics. This will allow for a greater analysis and understanding of inequalities within health and social care in Wales.

The annex contains more-detailed summaries of current work and future plans written by producers from each of the four countries of the UK.