Care Home Data Review: workshop summary

Issues relating to gaps in data

Variable gaps

There was discussion around specific variables that are needed but not collected (or of such poor quality in their current state that they are not of any use).

Staffing and workforce

• timely and accurate data is needed on staff recruitment, turnover & retention. This is needed to determine safe levels of staffing to meet the needs of residents (it was noted that there is data on turnover and retention available relative to other social care sectors)
• poor data on movement of staff
• there is currently a poor response rate on data relating to staff leaving social care (e.g. the type of employment people move to)
• data needed on use of agency staff, the ratio of permanent staff : agency staff, and the impact of agency staff on quality of care
• staffing numbers in relation to occupancy are needed (i.e. headcount per resident)
• data needed to help with the safer staffing legislation
• more data needed on staffing and workforce patterns and trends
• real time data on staffing cover is needed (together with qualification & skills levels)
• data on staffing absence have been removed from Turas - would be helpful to have these back

Vacancies (as in bed vacancies, not staffing vacancies) and bed occupancy

• real time occupancy levels are needed, but data is often affected by poor timeliness
• reasons for vacancies (in placements and care homes)
• status of beds. 'Beds pending' , 'beds held for assessment', available beds, occupied beds, total beds, temp closed beds etc.
• need for data on 'viable vacancies', not just empty rooms that can’t be used
• occupancy and availability trends

• a single number is not enough, but staff need more context behind bed vacancies. Some care homes have many vacancies but don't have the staff to admit to these, or care homes limit the number of admissions per day
• interim and respite care availability
• data split by care home / placement category (i.e. residential / nursing / etc.)

Resident data

Participants commented that there was a need for 'more resident focussed data'.

Resident needs

Information on resident needs is not currently captured.  Data on the level of need people have and the support they require, both before and after they are admitted. The ability to track how quickly resident needs change from point of admission (i.e. residential to nursing care) would also be helpful.

Data on dependency and high levels of care is required, especially if acute. This would help ascertain safe levels of staffing to meet resident needs.

Types of bed

Data on the numbers of residents in different types of bed (e.g. specialist beds) is needed.  

Types of care / placement types

Data is needed on the types of care required, provided and received . Also more data is needed on placement types.

Resident health conditions, wellbeing & complexity

Accurate data is needed on the number of residents with specific health conditions and complexity:

• Dementia. Particularly mentioned was the need for accurate data on dementia. The care home census has dementia status but this is just a snap shot. Both the Care Home census and PHS Source social care data collection under-report the expected numbers based on prevalence. More detailed data is also needed – by severity and giving information on how long residents with dementia stay in a care home

• Medication levels

• Resident wellbeing. Details around residents’ wellbeing are needed, to flag early concerns around Adult Support & Protection (ASP) and provide support to care homes. Other specific data needed include:

  • falls trends
  • prevalence of pressure ulcers (in particular “pressure acquired wounds above grade 2”)
  • catheter-associated urinary tract infections (CAUTI)
  • dysphagia

Home address of residents

To determine home LA area, and give data on the number of 'out of area residents' in local care homes. There’s a need for a national picture of out of area placements by HSCP to HSCP

Outcomes and experience data

Both resident and care provider experience.

Funding

• types of funding (self-funding, non-self-funding)
• who is on contract rate
• details of the charges made to self-funders

Demographic data

For example, understanding the changing demography of care home residents.

Other more detailed resident data

• are people living in care homes (and the staff that care for them) receiving the support that they need from healthcare teams (e.g. GPs, Nurses, Allied Health Professionals)?
• how many people have / have not had the opportunity to develop an Anticipatory Care Plan?

Finance

Lack of robust finance data (e.g. expenditure by client group is no longer classed as a national statistic). Data has to be apportioned, as it is not available directly from LAs’ ledgers.

Data is needed on sustainability – care home costs vs income from funded places. It was noted that “sustainability info is being gathered at a local level by HSCPs”.

Discharge data

Discharge data is very patchy - wasting time contacting GP and hospital trying to verify sparse information. This is worse on rushed discharges.

A&E attendance

There is poor recording of Accident & Emergency (A&E) attendance codes. Difficult to identify any pattern of A&E attendance by care home residents. The reason for A&E attendance is particularly hard to determine.  A&E data only records injury type, not reason.

Health / outbreak data & reasons for care home closure

There is a need for better data on outbreaks (including, but not limited to covid, i.e. flu, norovirus etc.), including those causing care home closure. Currently there’s only data on lab-confirmed cases, so – especially now that testing is becoming less common – it’s hard to determine the true size of an outbreak. There is a need for data on symptomatic residents.

Other variable gaps

• how many different GP practices are providing care to a single care home?
• data on volunteers in care homes
• data on access and completion rates of staff training/Continuing Professional Development (CPD, e.g. e-learning). The ability to track the impact of this training and development on quality of care

Forecasts and predictions

Data currently used is retrospective. Several participants said that it would be useful to know the predicted future demand for Care Homes and for what purpose (e.g. residential, interim, respite etc.)

Longitudinal / care pathway data and outcomes

Useful to have data on each client’s care pathway. This would include records of the services they received before they were admitted to hospital or care home (e.g. home care and/or telecare); where people were admitted from (e.g. community or health), and the type of admission (planned / emergency). Linking this to data on the services received after admission; how long they stay; and where they are discharged to will give a much fuller picture of the full experience of a client through the care system. There is currently limited ability for addressing key questions about client outcomes.

A comment suggesting this would be possible:

Timeseries data

There was discussion around wanting to know how care home usage is 'changing over time'. Current challenges include changing definitions over time (meaning data is not comparable year to year), low data quality, and lack of specific variables focussing on care home usage are all possible reasons.

Suggested solutions to data gaps

Specific variable gaps

Solutions follow on from the variable gaps detailed above.

Data linkage

Many of the suggested solutions here relate to data linkage, including:

• linking PHS Source social care data to Hospital and Primary Care data
• linking individual resident data to dementia registers (via CHI numbers)
  • date stamping dementia data to facilitate pathway analysis
  • there has been a previous (one off) exercise to link the dementia register to Source social care data and the Care Home Census. It would be good if this was done more regularly and with data on severity
• linkage between residents' profiles, medical records, hospital admission/discharge
• “dataset easily linkable to the new Adult Support and Protection (ASP) dataset”
• regular collection at an individual level and linked to other service use data enables pathways analysis

In order to do this:

• “include CHI with all data returns to allow linkage and analysis”
• add a flag to other datasets to indicate that a person is in a care home. This will enable care home records to be linked to other data - linkage to hospital records for example
• “a robust online platform is required to link data. This will reduce cost and time to access and process data. This will also reduce data duplicity and redundancy problem”

The solution to aspire to is a “shared health and care record”.