COVID-19 local area health risk rating

Weekly update on which local authorities face the greatest health risk from COVID-19

5 May 2020

By Ben Franklin

11 minute read

Please note CPP has now paused the weekly covid-19 health update.

Over the last month we have run a weekly Covid-19 health risk ranking to explore which local authorities are most at risk of high mortality in the face of the virus. This revealed high levels of risk in the Midlands, North East and North West of England and pinpointed places such as Middlesbrough and Walsall that were of particular risk – both due to high case load and high level of underlying vulnerability related to healthy life expectancy, mortality from treatable conditions and the quality of care in the area. The ONS has since revealed the links between high levels of deprivation and higher mortality from Covid-19 and their most recent analysis showed how people working in particular low paid occupations, such as care workers and security guards, have suffered a higher death rate during this epidemic. They also revealed that ethnic minority groups have higher mortality which may be one reason why deaths have been so high in London. With the ONS now producing such detailed, timely and relevant analysis, we think it is most useful for people to keep updated on the risks of Covid-19 through their coronavirus portal and so we are pressing pause on the CPP index for now.

Please note though, that we do not think the health risks of Covid-19 have passed and the factors linked to high levels of place-based vulnerability remain highly relevant. Over the long run, supporting place-based population health through greater investment in public health and the social determinants of health will be vital for building up resilience in those communities hit hardest by the virus.

About CPP's weekly update

Different local areas and health systems will experience the crisis differently. Not only will the number of cases differ substantially from place to place so that the timing of the peak will be localised, but different areas will be better or worse equipped to deal with the surge. Variation in health outcomes will be driven by multiple and complex factors, including the existing “stock” of health in the population and the quality of the health and adult social care systems locally.[1]

Over the duration of the crisis, local areas that experience the largest number of cases and have the least systemic resilience are most likely to see the highest mortality rates. Excess mortality will not just be the result of the COVID-19 virus, but from other conditions that stretched health services will be less able to treat. In this context, CPP has developed a risk rating to predict where the current surge in case numbers could have the biggest impact on the health of local populations.

View past risk rankings:

The risk rating

Indicators and weighting

The risk rating is based on the indicators below. Data for the indicators is assigned to an upper tier local authority (UTLA) and its relevant CCG(s). To normalise the data, each of the 149 local authorities is ranked against each indicator, ranking 1 if it has the highest level of risk (i.e. highest cases per 100,000 people) and 149 if it has the lowest level of risk (i.e. lowest cases per 100,000).

In determining weightings for the indicators, we seek to strike a balance between accounting for the systemic vulnerability of local areas on the one hand, and the present reality of rising COVID-19 cases on the other. Since we are particularly interested in providing a forward look at health systems most at risk for the duration of the crisis, the systemic indicators make up the largest proportion of the risk rating – accounting for 75% of the overall rating (25% weighting for the prevalence of COVID-19).

We partially base the weights for each of the four systemic indicators, on their relative ability to explain variation in local authority age-standardised mortality rates (ASMR).[2] The assumption being that if all places have the same number of COVID-19 cases, mortality will be higher in those areas that already suffer from higher death rates.

Table 1: the indicators

Indicator name and weighting

Description

Sources

COVID-19 cases (25% weighting)

This is the cumulative number of cases per 100,000 people. Data is available in real time on a day by day basis.

Government COVID-19 dashboard:

https://coronavirus.data.gov.uk/#local-authorities

Health Life expectancy (20% weighting)

The average number of years a newborn can expect to live in good health. Data available for 2016-18.

ONS: Life expectancy (LE), healthy life expectancy (HLE) and disability-free life expectancy (DFLE) at birth and age 65 by sex, UK, 2016 to 2018: https://www.ons.gov.uk/peoplep...

The size of the older population (20% weighting)

The % of people over the age of 65 in a local area in 2018.

Population estimates – local authority based by 5-year age bands. Accessed via Nomis: https://www.nomisweb.co.uk/dat...

Treatable mortality (20% weighting)

Age-standardised treatable mortality rates by local authority in England and Wales, 2016-2018. Per 100,000 people.

Avoidable mortality by local authority in England and Wales https://www.ons.gov.uk/peoplep...

Quality of care (15% weighting)

The average “overall” rating for adult social care providers in a local area. CPP derived variable from the ratings provided by the CQC.[3]

CQC: State of Care 2018/19: Ratings data: https://www.cqc.org.uk/sites/d...

Rationale behind the indicators

The objective is to capture both the prevalence of the virus, as well as the strength of local characteristics and institutions that can support local health resilience in the face of the surge.

To capture the prevalence of the virus in a local area, we use the laboratory confirmed case load per 100,000 people. This has several limitations including that it only measures the number of cases within hospitals and depends on the resourcing and ability of NHS staff to administer tests for patients with symptoms which may itself differ by area and trust. But it is arguably the best official and timely data currently available.

In the face of rising cases, the first protective factor is the level of good health in the local population for which we use healthy life expectancy—it is well known that the virus is more dangerous for those people who have underlying health conditions. But healthy people have also become critically ill and the virus appears to have worse health impacts for older people (even if healthy) than younger people.[4] This implies that local areas with a higher proportion of older people are more vulnerable if the virus spreads.

Once seriously ill, it is the job of the NHS to nurse people back to health and it has been on the frontline in the fight against COVID-19. To measure the strength of the healthcare system we use the mortality rate for treatable conditions. Treatable mortality refers to causes of death that can be mainly avoided through timely and effective healthcare interventions.[5]

Finally, the adult social care sector continues to support the most vulnerable groups during this crisis—older people with long term conditions. Rising cases amongst social care professionals clearly poses a risk to care home residents and those with care needs in the community—both directly by potential transmission of the virus to older people and indirectly—by reducing the number of people able to provide care and support. Better resourced and managed adult social care systems should be better placed to cope with these big safety and capacity issues. Failure to cope will result in increased loss of life in the community and greater pressure on the NHS.

What does the risk rating tell us about local area vulnerabilities?

The ranking underlines the point that when considering the severity of risk posed by COVID-19 to the health of local populations, it is vital to go beyond the number of confirmed cases. This will be crucial to directing resources at those healthcare systems which are going to struggle the most in the face of high demand (i.e. prioritising where to substantially increase the number of beds, equipment and personnel). While all healthcare systems are stretched, the risk rating reveals those areas that are likely to find it particularly challenging and where more stringent capacity building is required.

As well as showing which areas will require additional NHS capacity, the risk rating also reveals the degree to which the adult social care sector has the capacity to meet the challenge. In Barnsley for instance, which has a moderate-to-high case load, the adult social care sector is likely to be particularly vulnerable given the prevalence of low ratings from the CQC—29 care providers are rated as requiring improvement and 2 are inadequate. While much of the political and public policy focus has rightly been on the NHS, finding ways to support enhanced capacity in the adult social care sector in places like Barnsley, will be another crucial way of coping with the rise in cases and keeping the most vulnerable people safe.

The risk rating is also useful because it identifies the places where public health restrictions are so important to preserve life. For instance, in Walsall, healthy life expectancy is just 56.4 years of age—significantly below the age set by government for self-isolation (age 70 and older). Ensuring that those who have a pre-existing health condition self-isolate, irrespective of age, and social distancing is fully observed will be critical to avoiding greater pressure on Walsall’s health and social care sector and saving lives.

Finally, the analysis reveals that while places with older populations currently tend to have a lower number of cases, this is far from universally true.[6]Cumbria, for instance, has a very high number of cases (306 per 100,000, as of 22 April) and a high proportion of people over the age of 65 (24.1%). As the crisis continues, it will be important that places with a high proportion of older people are sheltered from the virus and that where cases are already prevalent, self-isolation will remain vital.

Conclusion

The risk rating is intended to kick start discussion and action on taking broad systemic health system vulnerability into account when considering the risks posed by COVID-19 to the health of local populations. Until now, most data tracking and subsequent policy and resource allocation has focussed on the overall case load and hospital capacity (i.e. number of beds). But as we reach the peak, local vulnerability to the virus will depend on much more than this including the existing stock of health in the population, the overall strengths and weaknesses of the health and adult social care systems and the age structure of the population.

Notes

[1] The value of health as an asset has been long championed by the Health Foundation. See for instance, The nation's health as an asset: Building the evidence on the social and economic value of health. https://www.health.org.uk/publications/the-nations-health-as-an-asset

[2] We ran a multiple regression model of local authority ASMR with healthy life expectancy (HLE), treatable mortality, the proportion of people aged over 65 and the quality of care providers as the independent variables. The coefficients for HLE, treatable mortality and 65+ population were all similarly strong and statistically significant so they are assigned equal weights in the ranking. The quality of care was insignificant, so this is assigned a lower weight. Model was a good fit for the ASMR data, R2=0.8

[3] The ranking system is based on assigning numbers to the ranks. 1-4 for inadequate, requires improvement, good, and outstanding (i.e. 1=inadequate, 4 =outstanding). This is done for all adult social care providers in the local area and based on their overall score. The total is then used to calculate the average care provider score for the local authority.

[4] Latest data from the Intensive Care National Audit and Research Centre (ICNARC) finds that 93% of those critically ill from COVID-19 were able to live without assistance in daily activities. https://t.co/mdq8GhRQEV?amp=1

[5] For more on the definition and data behind treatable mortality rates see: https://www.ons.gov.uk/peoplepopulationandcommunity/healthandsocialcare/causesofdeath/bulletins/avoidablemortalityinenglandandwales/latest

[6] Based on a simple scatterplot, there is a negative relationship between the proportion of people aged 65+ and the number of COVID-19 cases. R2=0.4.