Neighbourhood Caremaps

Gloucestershire County Council and NHS Gloucestershire, 2021-2025

Background

Gloucestershire's commissioners wanted to understand the condition of the County's home care market as COVID restrictions eased. In particular, they wanted to explain waiting-times for home care and excess demand for hospital and care home services, and to plan remedial action.

The Council's brief indicated that home care provision was not sufficient in some places. About half of the County's population lives in rural parts, while nearly three-quarters of its care providers, and their staff, were located in two large, central towns. The Council's brokers reported persistent delays finding care for people in places distant from the main towns, especially on the County's rural periphery.

The Council's community care framework defined the County's home care market using its six local authority Districts, with fee-rates distinguished by rurality. The framework paid lower fees in the central towns and a single, higher fee-rate in the four surrounding predominantly rural districts.

Gloucestershire chose Sphere for our experience in home care commissioning and statistical modelling.

Approach

We formed a hypothesis that scarcity of home care might be due to rurality and rurality might therefore account significantly for excess demand hospital in hospitals and care homes.

To test the hypothesis, we needed statistical measures of rurality and scarcity in home care, and geographical units smaller than a district to isolate the relationship between them. This is because the rural districts each contain towns with urban population densities; and the urban districts each contain rural areas with low population densities.

We chose "lower-layer super-output areas" (LSOAs). A local authority district typically contains between 50 and 100 LSOAs. Their rurality can be estimated from the density of their population and, where an LSOA's population is sparse, its proximity to urban areas. LSOAs are also a commonplace unit for socio-economic data that measure other local factors that affect home care markets, especially demography, deprivation, employment, income and unpaid care.

We chose waiting-times as our measure of scarcity. Waiting-times are a common measure of unmet need in health and care. They are easily measured in LSOAs because each delay is linked to the home address of the person waiting.

Waiting-times are imperfect measures of scarcity because people wait for care for reasons other than scarcity, for example, because of delays in administration. Scarcity does, however, explain excessive waiting-times, which we defined as people who wait a fortnight or more for care. The Council's data showed that delays in administration rarely accounted for delays this long, making lack of capacity the likely cause.

Findings

  1. In most of the County the supply of home care sufficed. We found persistent scarcity in about a fifth of LSOAs.
  2. The correlation of scarcity and population-density was weak. Rural LSOAs accounted for less than half of excess waiting.
  3. Some areas where people waited for care neighboured areas of abundance. The LSOA with the longest waiting-times neighboured LSOAs with the highest density of home care agencies in the County.
  4. Consultation with providers found varied causes of scarcity. Local factors always played a part. Examples included with district fee-rates that did not fund local pay-rates; long distances between customers and care agencies (or physical barriers between them, like rivers); real or perceived crime rates, deterring evening care visits; and unpredictable demand, deterring agencies from investing in a permanent local presence.
  5. A sub-measure of provider-density (the number of home care agencies serving each LSOA per day) showed high concurrency of providers in some areas, with consequent inefficiencies.
  6. A sub-measure of customers' proximity to their home care agency found that Gloucestershire's agencies operate mostly inside a six-mile radius from their premises, which is typical of the English home care market. We found a legacy of emergency COVID measures that stretched some providers' operating territories: some agencies served customers 40 miles distant.

Engaging the market

Our findings suggested that the Council's framework for home care would generate more capacity, without proportionate growth in costs, if it replaced Districts with LSOAs as its unit of price and contract for home care.

Smaller geographical units permit fee-rates and other commercial incentives that better reflect local conditions. As an organising principle, LSOAs would also cause care providers to operate nearer their customers, reducing travel costs and releasing time for care.

We tested our findings with the County's home care providers in consultation events around the County, including areas known for scarcity. We asked about their practical experience of the statistical phenomena; asked them to propose solutions; and asked them to tell us about other factors that limit their capacity that we had not discovered.

For the most part, providers confirmed our findings. They brought our attention to care management and commercial practices that deserved improvement.

Recommendations

From the results of our model and providers' feedback, we recommended that the Council define a graded fee structure by LSOAs, with attention to local factors beyond rurality, especially local employment rates, physical obstacles, crime rates and the proximity of the nearest home care agency.

We added a recommendation, found in the Council's original brief, to introduce block contracts. We said block contracts should be used as an exceptional, temporary stimulus in areas of intractable scarcity on three conditions:

  1. The normal fee-rate for an area does not suffice for a provider to establish a permanent local presence.
  2. The contract is temporary, has a clear objective and rules for termination. For example, the block covers a provider's costs only while they build custom. It ends when the provider has sufficient business to sustain the service without guaranteed income.
  3. The contract applies to only an LSOA or clusters of LSOAs with persistent waiting-times for home care.

Hyperlocal commissioning

The Council adopted our recommendations and created a "Hyperlocal Commissioning" project in its transformation programme to execute them.

By addressing our findings, first using provisions in its existing home care contracts and then by creating a hyperlocal lot in a new community care framework, the Council generated new home care capacity. The new capacity did not generate the proportionate growth in costs found in alternative methods of stimulus, like global fee-uplifts and permanent block contracts. This was because hyperlocal commissioning improved capacity partly through efficiency, for example, by focussing care providers nearer customers, and so reducing travel.

The Council began to pilot hyperlocal commissioning four months after our appointment.

Between April 2022 and December 2024, the Council doubled the number of people that it supported with care at home, reducing waiting-times for care and delays leaving hospital.

Read the Department of Health and Social Care case study of hyperlocal commissioning in Gloucestershire in its 2023 Market Sustainability Insights Report.

Living maps

Our economic model for hyperlocal commissioning combined forty sources of data from the public domain, the Council's care management and finance systems and the NHS.

The volume and complexity of data in this kind of model sometimes results in their being used only for their initial purpose, with no more value extracted from them.

In our experience, this is especially common when a consultant produces the model using its own skills and systems. We wanted to avoid this and leave Gloucestershire with home care commissioning strategy and routine information to deliver it.

Sphere created the hyperlocal model using data-integration tools, and a geospatial data-warehouse, mostly using Microsoft technologies that the Council already used. This gave Gloucestershire the ability to refresh the model frequently and reliably, and to use it for other purposes among them live, searchable, online maps.

  • The Integrated Brokerage Team, to find home care using the rules and prices in the hyperlocal home care framework and care home beds with real-time capacity and prices
  • Commissioners and contract managers to develop new commission strategies and monitor care markets
  • The County's Proud to Care team, who help care providers recruit care workers, to analyse local labour markets

Our data-warehouse supplied information for other purposes, including:

  • Market research for the national Fair Cost of Care programme
  • Long-term trends and near-term forecasts in the balance of home care and residential care, for the Council's Market Sustainability and Improvement Plan
  • Automated reporting of the 2022 Hospital Discharge Funds
  • Analysis of disability-related expenses and disregards for a new charging policy
  • Commissioning plans to replace COVID discharge-to-assess care home beds with more reablement at home

Neighbourhood care for the future: careplaces

Besides the rich supply of relevant statistical information about LSOAs, they are also convenient proxies for neighbourhoods. The next stage of hyperlocal commissioning is "ordinary life" support.

Visit careplaces.ai to see more about our latest work on neighbourhood care and ordinary life services.

Deep text search: help for unpaid carers

Edinburgh Health and Social Care Partnership, 2024

Background

During our analysis of Edinburgh's sub-acute hospital and care home system for their Older People's Pathway, we found less residential respite than we expected. Statistics about Scotland's 32 Health and Social Care Partnerships confirmed this, although they showed a healthier supply of respite in the home.

We also heard feedback from carer representatives on the City's Integration Joint Board; the trade association Scottish Care; Edinburgh MPs in Holyrood and Westminster; members of the City Council; social workers and health professionals.

We believed any proposal to commission more residential respite care would only increase the pressure on the City's stretched care home system, and risk exacerbating the price-inflation that we found in our research for the Older People's Pathway.

We also felt that responding to an apparent shortage of respite beds by commissioning more risked begging the question. We needed first to understand the needs were left unanswered by lack of supply.

Approach

We began by asking not how many more respite care beds the City needed, but what prompted the idea that more was needed. Bed-based respite care is, after all, little different from other bed-based health and care services. It should used only in the absence of less disruptive, less costly alternatives.

We asked, "what are the adverse outcomes when unpaid carers do not get help when they need it?"

For this exercise, we said that a carer needs help at the beginning of any series events that leaves them unable to care and the person they care for is admitted permanently to a care home.

Research Method

Our question has the benefit that it seeks the earliest moment when a crisis that leads to negative outcome --that is, permanent, unplanned admission of the person they care for to a home--and asks what help was needed then. This avoids the assumption that respite, at the moment of crisis, was the right intervention. We wanted our solution to avoid the need for emergency respite, albeit that this might find unmet need for planned residential respite.

The question has the further benefit that it is easy to find the cohort: it is anyone that the Partnership admitted to care home and who has an unpaid carer. Both items are normal elements of the adult social care record in Scotland.

Two items of data needed to find our cohort were less apparent.

  1. The distinction between planned and Unplanned admissions to a care home care is a discretionary item in the adult social care record.
  2. The "event that led to unplanned admission to a care home" is not typically documented as a standard part of the formal social care record. Such events may occur weeks or months before admission, might involve multiple contributing factors, and are often identifiable only in hindsight. In Sphere's experience, this information is rarely recorded systematically in adult social care records anywhere in the UK.

Findings

Coming soon.