The factors that influence Covid-19 infections, deaths and the likelihood of survival are complex. The latest analysis from the LCP Covid-19 Tracker suggests that deprivation, age and population density are all associated not only with Covid-19 infection rates, but also with mortality rates.

Crucially, we have found that higher infection rates alone do not necessarily lead to higher mortality rates.

Identifying the factors that impact the risk of death or life-changing consequences due to Covid-19, and targeting susceptible groups as a priority (both by vaccinating early and by working with local communities to increase vaccine uptake) will be key to reducing the morbidity and mortality from Covid-19 throughout the UK. Our results highlight the need to consider a wide range of variables when planning for the pandemic, but also point to public health level approaches that could improve the nation’s health once the worst of the pandemic has passed.

Data and approach

The LCP Covid-19 Tracker aims to provide comprehensive estimates of the scale of Covid-19 infections across England at the Lower Tier Local Authority level (District, Borough or City Councils, LTLAs) by combining data from Public Health England (PHE) and the Office for National Statistics (ONS) and applying actuarial techniques. Lower Tier Local Authorities are the smallest geographical areas for which these datasets are available, and there are 315 in England. A full description of our methods can be found here.

For this analysis, we also incorporated three additional datasets: cumulative death data from PHE, where a Covid-19 death is one that occurs within 28 days of a positive Covid-19 test; the median age of each LTLA; and data from the Index of Multiple Deprivation (IMD). The IMD is a composite measure of the relative deprivation of local areas within the UK, taking into account factors such as income, education, health, crime and housing. The Government publishes IMD scores by Lower Layer Super Output Area (LSOA), which is at postcode level, and there are around 35,000 LSOAs across the country. A higher score corresponds to a higher level of deprivation. We generated IMD estimates for each LTLA by taking a population-weighted average of the IMD scores of the LSOAs within each LTLA. We additionally integrated median age and IMD scores, which allowed us to explore and estimate the cumulative Covid-19 infection rates across LTLAs in England from 1 July 2020 to 10 January 2021, and cumulative Covid-19 death rates and fatality rates across LTLAs in England from 1 July 2020 to 31 January 2021 according to deprivation status, population density and age.

Associations with cumulative Covid-19 deaths

A reasonable assumption might be that the higher the number of infections, the higher the number of deaths in any given area.  We found that the death rates in the ten LTLAs with the highest cumulative infection rates were double those of the LTLAs with the lowest cumulative infection rates.  However, the ten LTLAs with the highest cumulative infection rates were not the same as the ten LTLAs with the highest cumulative death rates, meaning that higher infection rates did not necessarily lead to more deaths. This indicates that factors other than rates of infection impact the outcome of the disease.

Deprivation is a key driver of both infection and mortality

Our previous analysis found a strong association between deprivation at LTLA level and cumulative Covid-19 infections during the second wave.  A similarly strong association has been found when looking at cumulative Covid-19 deaths, suggesting that death rates are 60% higher in the most deprived 10% of local authorities, compared to the least deprived 10%.  In the most deprived 10% over the last six months, there have been 134 deaths per 100,000 people, compared to 87 deaths per 100,000 in the least deprived 10%.

The ten LTLAs with the highest cumulative death rates had a higher combined IMD score (on average 25% higher) than the ten LTLAs with the lowest cumulative death rates.  This is consistent with the recently developed QCovid risk prediction tool, which recognised deprivation and ethnicity as risk factors for severe Covid-19 infections.  The QCovid tool also allows for other factors that may increase people’s likelihood of death to be taken into account, such as body mass index, and physical and mental health conditions.  As a result of this analysis, a further 1.7m people have been advised to shield and 800,000 people added to the priority list for vaccinations.  Modelling such as this is consistent with our findings, that all factors should be considered in order to identify and protect the most vulnerable.

Table 1: The ten LTLAs with the highest cumulative Covid-19 death rates over the period 1st July 2020 – 10th January 2021

Table 2: The ten LTLAs with the lowest cumulative Covid-19 death rates over the period 1st July 2020 – 10th January 2021

The association between population density and cumulative Covid-19 mortality is less clear

Areas with high infection rates might also be expected to have high death rates, but this is not always the case.  In our previous analysis, we found that more urban areas had higher cumulative infection rates, but not necessarily higher death rates. We found a more mixed picture, with the ten most densely populated LTLAs having a 50% lower average cumulative death rate compared with the ten most sparsely populated LTLAs. 

However, this is not universal. When analysing the results by population decile, the cumulative death rate is variable, with the most populated 20% having a death rate of 194 deaths per 100,000 and the least populated 20% a death rate of 186 deaths per 100,000.

This variability further highlights the need to look at other factors that influence the severity of Covid-19.

Age is not just a number

Age is very strongly correlated with both infection rates and death rates, albeit in opposite directions. On average, the median age of the LTLAs differed by 20 years (31 years and 51 years). The ten LTLAs with the lowest median age had more than double the level of infections, however, the death rate was 40% higher in the ten LTLAs with the highest median age. This highlights the impact of age and confirms its importance when prioritising vaccination.

Associations with infection fatality rate

Another vital statistic to consider is the likelihood of death across populations that become infected with Covid-19; as we estimate infections, we term this ‘infection fatality rate.’ As expected, this is widely variable across the country and should also be taken into account when distributing health resources.

Don’t just rely on infection rates

A common view is that, once infection rates fall in a particular area, lockdown restraints can be relaxed.  However, it should be remembered that even if the chance of becoming infected with Covid-19 is lower, the chances of survival for those infected may be starkly different.  Our analysis found that the four regions in the UK with the highest level of infections were also the four regions with the lowest infection fatality rates.  Conversely, the region with the lowest infection rate had the third highest infection fatality rate.  

There is a large amount of variability within regions but the difference remains significant at LTLA level.  The ten LTLAs with the highest fatality rates had an average fatality rate of 3.8%, six times higher than that of the ten LTLAs with the lowest fatality rates, at 0.61%.  But at the same time, the infection rate in the ten LTLAs with the highest fatality rates was about half that of those with the lowest fatality rates. 

Table 3: Ten LTLAs with the highest infection fatality rates over the period 1st July 2020 – 10th January 2021

Table 4: Ten LTLAs with the lowest infection fatality rates over the period 1st July 2020 – 10th January 2021

Density and deprivation are not the only factors influencing mortality variations across populations

Where we are seeing more deaths in more deprived areas, our analysis seems to show a higher fatality rate in less deprived and more sparsely populated areas.  Caution should be exercised when reviewing the impact, as the large variation in numbers of infections and deaths can cause an exaggerated effect on infection fatality rate for areas with very low infection and death rates.

However, this finding is likely to be linked to the demographics of the groups of people that live in urban or rural areas.  The ten LTLAs with the highest population density had an average median age of 34, while the ten LTLAs with the lowest population density had a median age of 50.

Age remains the biggest predictor of survival across populations

The three areas with the highest infection fatality rate also had the highest median age, indicating a strong correlation between survival and age.  This is a significant finding, indicating that the current vaccination roll out by age is the correct response.  The trend continues across all LTLAs; there is a 3.5-fold increase in infection fatality between the ten LTLAs with the lowest median age and the ten with the highest median age. 

However, age clearly isn’t the only determinant – Broadland and Harrogate both have a median age of 48 but Broadland has the fifth highest fatality rate while Harrogate has the fourth lowest.

Future of healthcare

The infection fatality rates show that a number of factors impact the health outcomes of the population, and these should be considered not only in the pandemic response, but also in future healthcare planning.  The more deprived areas of England have experienced the highest levels of Covid-19 infection and the highest numbers of deaths.  This stark inequality is likely to affect other aspects of health, which, as we’ve seen from this pandemic, can have a significant impact on the economy. 

The recently published ONS Health Index and Health Index Explorer highlighted the fact that social, economic and health outcomes are complex and intertwined, and underline the need for holistic planning in order to reduce the inequalities the pandemic has revealed.  Metrics and interactive tools such as these can help identify priority areas for investment to repair and improve health in the UK as the pandemic recedes.