Mirror, mirror, on the wall, whose local care is fairest of all?

How local health equity data can support quality improvement initiatives.

Blog post

Published: 31/10/2016

Equity is a key domain of quality and a recurring theme in many of our indicators, but rarely something focused on in its own right. Professor Richard Cookson at the University of York, with his research team (Miqdad Asaria and Shehzad Ali), helped develop NHS England’s new local health equity indicators for CCGs. He believes these equity indicators can play a useful role in helping NHS and local authority decision makers to evaluate and improve locally-led new models of coordinated care.

The problem with national data on health equity is that nobody owns it. It isn’t any one person’s problem, and it is easy to explain away bad news. Social inequalities in health and health care are influenced by all sorts of complex social, economic and technological factors, and there is no ‘national control group’ telling us what would have happened if national policy had been different. So it is hard to tell whether particular NHS policies are responsible for particular national equity trends. Policy makers are thus able to take credit for good news, and shift blame for bad news, without anyone learning any useful lessons.

Local data is therefore the key to improving understanding. My team at the University of York has looked at pulling this together and now there is a new NHS data set that shows at CCG level the amount of inequality between deprivation groups. The current published indicators focus on inequality in potentially avoidable hospitalisation (‘emergency admissions for ambulatory care sensitive conditions’), though in principle other equity indicators could be added in due course – such as inequality in the supply of GPs and nurse practitioners, inequality in primary care quality, and inequality in hospital waiting times. There are two indicators of inequality in potentially avoidable hospitalisation – one for acute conditions and one for chronic conditions. The latter is particularly useful, since it helps to indicate how well health and social services are working together to manage care for people with complex long-term conditions – one of the most important challenges facing today’s NHS.

Inequality gradients

Local equity is assessed by comparing the level of local inequality with the national picture and with ten similar CCG areas based on deprivation, age profile, ethnic mix and rurality. The diagrams below illustrate this for two CCGs in 2015. In the diagrams, each dot is a neighbourhood, with bigger dots for bigger neighbourhoods. The lines are linear regression lines through the dots, showing social inequality ‘gradients’ – the steeper the gradient, the greater the health inequality. The solid green line shows local inequality, as compared with the dashed red line showing national inequality and the dotted blue line showing inequality within ten similar CCG areas.


Liverpool health equity gradient


Brent health equity gradient

From the diagrams we can see that Liverpool has statistically significantly ‘worse-than-expected’ equity compared with these benchmarks, whereas Brent has significantly ‘better-than-expected’ equity.

What this means

We cannot say for sure that Liverpool CCG is performing badly at tackling social inequality in healthcare access and outcomes for people with long-term conditions. All we can say is that, statistically, inequality in potentially avoidable hospitalisation is worse than expected in Liverpool compared with similar CCG areas and with the country as a whole. This might conceivably be due to a range of special local circumstances outside the immediate control of local health and social services – including special features of the local population, the patterning of diseases and health behaviours, the economy, housing, schools and so on.

Furthermore, we do not yet have a clear set of actions that Liverpool CCG should be taking to remedy the situation. There is a dearth of evidence about effective ways for local health systems to reduce inequality in avoidable hospitalisation.

However, there is at least a mild warning sign for Liverpool CCG. It might be performing badly on equity, and further investigation is warranted including keeping a close eye on how these equity indicators change in future in response to local initiatives to improve the coordination of care for people with long-term conditions.

This indicator is therefore designed to help CCG managers identify areas of concern requiring improvement, and learn how to improve the quality of care and reduce health inequalities. It is not intended to help national policy makers turn the screws on local CCG managers by introducing high powered performance targets or financial incentives. For this reason, NHS England is reluctant to call this a ‘performance indicator’, and CCGs doing badly on this indicator are of course particularly reluctant.

As an academic, I am less squeamish about terminology – so long as it is clearly understood that this indicator is for use in quality assurance, not for designing targets and incentives. My view is that this is a performance indicator, but it shouldn’t be used as a stick to bash CCGs with. Instead it should be used as a ‘tin-opener’, to help find examples of good and bad practice, and as a yardstick for evaluating local change programmes in the NHS.

Testing approaches to care quality improvement

Large numbers of CCGs around the country are busily introducing new models of coordinated care for people with long-term conditions, with the aim of reducing the costly hospital admissions that are currently putting so much pressure on NHS finances. For example, NHS and local authority decision makers in Liverpool have recently introduced a number of local policies to improve the coordination of health and social services and reduce emergency hospitalization, including health centers providing integrated primary and long-term care services, ‘step down’ hospital beds for non-acute care, day shelters for the homeless providing integrated health and welfare services, and the ‘delivering assisted living lifestyles at scale’ (DALLAS) programme that combines tele-monitoring in the home with structured case management and coaching delivered by a multi-professional support team (Devlin et al. 2016).

These equity indicators should be used by Liverpool CCG to evaluate these new models of care – and by other CCGs who are introducing their own new models of care. These indicators provide NHS decision makers with the tools they need to perform simple and quick quasi-experimental evaluations of new models of care in real time.

The benefits to the NHS and patients of communicating and using this information more effectively are potentially huge. As well as increasing fairness, which is a good in itself, there are big cost implications for the system as a whole. Learning how to reduce costly emergency admissions associated with social deprivation has a payoff not just for the poorest members of society but for everyone who uses the NHS or pays taxes to support it.

Review the data

The CCG gradient data for 2015 are now available on the NHS data website and the plan is for these indicators to be updated quarterly. Unfortunately, the data are not yet well communicated on the official My NHS website that is supposed to act as a one-stop-shop for a wide range of NHS indicators, including all of the other indicators in the CCG Improvement and Assessment Framework. Improvements in the presentation and communication of this information would help people to use it more effectively. As a quick fix, my academic colleague Miqdad Asaria has created an online tool for exploring the data for 2015. However, we hope to see improvements in the official NHS presentation of this information over time and are in discussions with NHS England about how this can be done and how the data can be used to evaluate promising local new models of care.