Abstract
Mortality rate increased in the period after 1 January 2020 because of the Sars-Cov-2 (coronavirus disease 2019, COVID-19) pandemic. A significant proportion of those deaths occurred within residential care homes who were mandated to put in place stringent preventative measures including vaccinations, regular testing and visitor restrictions, while maintaining access to front-line healthcare. Our question was, by how much did these measures mitigate this increase in mortality rate?
The Office of National Statistics (ONS) annually publish deaths, by age and sex, for each small geographic entity - the lower layer super output area (LSOA). A baseline of national average deaths per population in 2017-2019, by age group and sex, was calculated. This was then applied to local populations to calculate values of expected deaths and, when divided by the actual deaths, to create a standardised mortality rate (SMR). The change in standardised mortality rate (CSMR) was calculated as % change in SMR 2020-2022 compared with SMR 2017-2019. Excess deaths were then calculated on the basis of the assumption that CSMR would be 0% without the pandemic. The link between LSOA social deprivation index of multiple deprivation (IMD) score and CSMR was established by simple linear regression for each age group. The Care Quality Commission publish annually a register of residential care homes (RCH) which includes the post code location, which can be linked to an LSOA, and the number of beds split according to nursing care (CH) or purely residential homes (RH). Linking presence of RCH beds in LSOAs to outcome was evaluated in two ways, (1) by the amount with no RCH beds plus three tertiles of RCH bed number as the percent of older population (≥ 65 years) and (2) by the type of beds, those with RH only, CH only, or both RH and CH. CSMR was calculated for each of these cohorts. As RCH are mostly occupied by people aged ≥ 80 years, to estimate the impact of restrictions in care homes compared with the general community, the difference in CSMR between LSOAs with 'no RCHs' and 'with RCH' with baseline 0% CSMR were used to calculate the change in excess deaths.
Overall CSMR was 8.4%, (age group < 40 years was 5.7%, 40-64 years 13.7%, 65-79 years 11.3%, and ≥ 80 years 5.9%). This reflected 128,385 excess deaths in 2020-2022 compared with 2017-2019 (by age group < 40 years, 2106; 40-64 years: 26,120; 65-79 years: 49,301; and ≥ 80 years: 50,857). Social disadvantage had the most effect on CSMR in the age 80+ years group; in this group, the lowest five deciles (50%) of LSOAs by IMD score had CSMR of 4.5%, with the CSMR then increasing linearly up to 16% in the top IMD decile. In the age group of 80+ years, the 22,357 LSOAS with 'no RCH' had CSMR of 10.0% (as a result of 35,791 excess deaths), while in the 10,484 LSOAs 'with RCH' the CSMR was 3.3%, as a result of 17,840 excess deaths. In those LSOAs with only residential homes, the CSMR was 6.4%, and in those with only care homes (i.e. including nursing support), the CSMR was -0.2%. The average IMD score in LSOAs with RCH was 21.3, whereas without RCH, the average IMD at 21.8 was slightly higher, suggesting that social deprivation difference was not a factor in explaining these outcomes. Modelling if 'no RCH' CSMR had applied to the LSOAs with RCHs, there might have been 24,968 (+140%) additional deaths. If the CSMR of LSOAs with RCH had been applied to those with no RCH, 32,815 deaths might have been avoided.
We conclude on the basis of the available evidence that precautions put in place for RCH residents significantly mitigated the risk of death following a COVID-19 infection, especially so if they were in nursing homes. This suggests that the sacrifice made by family members in avoiding visits to RCHs did reduce the mortality and that rapid access to first line healthcare provided in nursing homes mitigated the consequences for disruption in normal healthcare provision.