Estimating the size of the self-funding population in the community, England: 2022 to 2023

• From 1 March 2022 to 28 February 2023, an estimated 23.0% of people using community care services were self-funders (83,844) and 77.0% were state-funded (279,916), which is not a significant change in the proportion of self-funders since last year.

• The South East remained the region with the highest proportion of self-funders in community care services (33.2%), which is statistically significantly higher than North East, which had the lowest proportion of self-funders (16.2%).

• Community care services that had no self-funded users accounted for 30.8% (2,708) of all community services; in contrast, 5.8% (510) of community services had all self-funded users.

• Services providing community care for older people (aged 65 years or older), had the highest proportion of self-funders (28.4%), which was statistically significantly higher than services providing community care for younger adults (aged 18 to 64 years), which had the lowest proportion of self-funders (6.8%), the same pattern as last year.

• Services providing domiciliary care continued to have a statistically significantly higher proportion of self-funders (23.5%) than services which only provide non-domiciliary care - for example, extra care housing and supported living services (10.7%).

• These estimates are based on community care services that are regulated by the Care Quality Commission (CQC) and were invited to complete a provider information return (PIR), and therefore do not cover all people receiving community care in England.

The methods used to derive these estimates follow those of our Care homes and estimating the self-funding population, England article, with the following exceptions.

Community care services were defined as care providers whose primary inspection category in the Care Quality Commission's (CQC's) Care directory was "Community based adult social care service", and the service type was not "Shared lives." Community care services could include domiciliary care services, extra care housing services and supported living services (see CQC service types).

In our Care homes and estimating the self-funding population article, the estimates are weighted using the variable "care home beds" to derive the population of care home residents. There is no equivalent variable on the provider information return (PIR) that can be used to derive a community care population. There is also no suitable alternative data source to use for the actual population size of people who use community care services. Therefore, we took the weighting population to be all community care services that were sent a PIR and could be linked to the Care directory in the relevant period, regardless of whether they provided a response.

There were 12,953 services in the Care directory that met the above definition of a community care service in the relevant period (May 2022 to April 2023, two months after the data collection period to account for de-registration and long-term changes to care providers). Of these, 9,365 community care services were sent a link to an online PIR form. There were 8,014 community care services which responded (85.6%). CQC removed 204 responses affected by system errors, for a sample of 7,810 (83.4%). 

Of all invited community services (9,365), 8,781 could be linked to the Care directory in the relevant period. Out of the 7,810 valid responses, 7,532 were from community care services which were still active two months after they submitted the PIR. These responses were retained, resulting in a response rate of 80.4% before data cleaning.

The cleaning and editing steps are described in Section 5: Data cleaning of our Care homes and estimating the self-funding population methodology, except that there is no equivalent to the "care home beds" variable, and therefore there is no "occupancy" check. After validating and editing the data, the final sample was 5,776 community care services. This is 44.6% of all community care services meeting our definition of community care that were registered in the Care directory in the relevant period (12,953 services), and 65.8% of the population used for weighting (8,781).

Weighting methods have been used to create an annual estimate of the population of self-funders from incomplete data on community care services in England, with region used to calibrate the weighting. As discussed, the population used to weight the estimates is the number of community care services sent a PIR and linked to the Care directory (8,781), rather than all community care services in the Care directory during the relevant period (12,953). See Section 7: Weighting the data of our Care homes and estimating the self-funding population methodology for a description of how the weighting was carried out. However, please note that only the region variable was used for calibration.

We compared the distributions of the final cleaned sample, the community care services that were sent a PIR but did not respond or were dropped during cleaning, and the community care services that were not invited to complete a PIR (see Table 1 for a comparison of region). The biggest difference was seen between the final sample and non-invited sample, where there were 13.9% more providers running single care services (see Table 9 in the accompanying dataset for the full comparison).

Notes:

1. Percentages may not add to 100% because of rounding.
   
2. The region variable was derived from the May 2023 National Statistics Postcode Lookup File (NSPL), which is linked to the Care directory.
   
3. For more information on cleaning and potential bias in the sample, please refer to our Care homes and estimating the self-funding population, England: 2019 to 2020 methodology.

Data quality

There are fewer cleaning stages than in residential care as there is no equivalent to the occupancy check in our Care homes and estimating the self-funding population 2022 to 2023 article. The CQC also has limited data validation process during data collection. There was low coverage of the actual community care service population (44.6%). Many responses were returned with an error (2,408, or 30.8%). Of these, 72.9% could not be cleaned. Therefore, 26.4% of all responses were dropped because of errors.

The responses that were dropped when filtering for community care using the Care directory (232, or 3.0%) reflect where the service is now defined differently in the Care directory in the relevant period (May 2022 to April 2023) compared with when the community PIR form was sent. By linking to the Care directory two months after the data collection period, some providers, which previously had a primary inspection category of "Community based adult social care service", may now be "Residential adult social care service", or their service type may now be "Shared lives". Therefore, these will not be included in the analysis. This is the same process we use for our residential care release, to avoid issues with de-registration and short-term changes. However, the community care service population is less stable than the care home population, and this step may have a higher impact on data quality for the community care statistics compared with residential care.

Since we weight our figures up to the number of community providers invited to complete a PIR (8,781) rather than the actual population (12,953), our figures are likely to be an under-estimate of the actual number of people who use community care services in England.

Please note, up to July 2022, community services deemed to be under pressure were not sent a PIR because two or more staff, or people using the service, have suspected or confirmed coronavirus (COVID-19). After this, the criterion was no longer applied. However, if a service was under pressure, they could request to delay the submission until a later time, after being invited in their original scheduled month. This may partly explain why the response rate has improved since last year (80.4% compared with 77.9%).

Therefore, we advise caution in using these estimates because of the data quality issues. These data have been published as a guide only. For further information on difficulties around estimating the number of people who use community care services, please refer to last year's blog post, Who is paying for their own community care? about why it is so difficult to estimate the number of people who self-fund their care in the community.

Feedback

These statistics are designated as Experimental Statistics. The experimental method presented here provides a first look at self-funders using regulated community care services. We welcome feedback to social.care@ons.gov.uk from anyone with an interest in the data, methods and analysis presented.