Mapping National Adult Social Care and Care Home Data Sources in Scotland: Baseline understanding & future potential: September 2022

Adult Care Home Data Sources

National Data Collections

This section presents information on the range of national adult care home collections.

Care Inspectorate Register

• Detailed above, there is a legal requirement for all care homes to be registered with the Care Inspectorate (CI).

Care Inspectorate Datastore

• Data file including a list of ALL currently registered care home services in Scotland, updated monthly.
• Each service receives a CS number upon registration. If a service changes ownership, it will be removed from the register and will need to be re-registered, whilst other service changes may be updated in the registration information.
• Publicly available online: CI Datastore. Includes address data, service information and quality ratings based on inspection outcomes.
• No information on staffing models available within the public data.

Care Inspectorate Annual Return Data

• Includes ALL Care Home Services.
• Lengthy, detailed aggregate data by each care home service, unpublished but accessible by Care Inspectorate team to inform inspections and bespoke pieces of work.
• Includes data on staffing, shared with Scottish Social Services Council (SSSC) who publish the official statistics on the social service workforce.
• Includes data on staff vacancies published annually in a joint CI SSSC publication.

Complaints about care services

• CI receives over 4,000 complaints about care services per year, and publish this data annually.
• It holds detailed information about the nature of concerns and complaints received.

Care Inspectorate (other reporting)

• Mandatory reporting of outbreaks of infectious disease, mortality events – reported by service-level, unpublished (during pandemic, mortality, staff absence and outbreak data shared with Scottish Government and now COVID-19 statistics are published by CI).
• Additional enhanced surveillance during pandemic to inform organisational/regulatory response including RAG ratings about staffing (shared with Scottish Government).

SOURCE Social Care Data

• Public Health Scotland data collection.
• Original consultation document available.
• Based on data submitted by Health and Social Care Partnerships (HSCPs) on individuals receiving funding for care home places (including those claiming free personal care allowance/nursing care allowance).
• Misses those funding their own care/not known to HSCPs /those receiving continuing care in a hospital bed / placements funded by an NHS Board.
• Data are submitted with demographics, but require to be CHI seeded from the identifiers shared with SOURCE team. The most recent return (for 2020/21 data) had an average 99.5% match rate to CHI for those people in a Care Home (excluding Orkney who did not return any information).
• Initial data publication did not include individual level data from Glasgow City Health and Social Care Partnership.
• Previous publications in June 2019 on all social care data (PHS Social Care Report), including care homes, and data for 2019/20 and 2020/21 (PHS Insights in Social Care Statistics) published in April 2022.
• Potential for linkage to other national data sources where CHI numbers can be added.

Scottish Care Home Census (adults)

• Annual data collection of care home numbers, registered places, residents and activity (admissions, discharges), submitted by care home staff.
• Registered places and number of residents are as at the census point. Whereas the survey also attempts to capture and summarise activity throughout the year (not on single 'census' point).
• Important source of trend data as collections date back to 2003, albeit format has changed over time. The census has been undertaken in eForms since 2010.
• Data submitted to CI, collected on behalf of Scottish Government, annual publication produced by Public Health Scotland Care Homes Team.
• Last care home census publication is for financial year to March 2022, published September 2022.
• Aggregate information on every care home service including total admissions, discharges, deaths, respite stays, weekly charges.
• Individual data requested on long-stay residents (defined as those anticipated to stay ≥6 weeks).
• Historic individual-resident data 2010-2016 has been linked to CHI^[3], however surname and forename have only been collected since 2013, impacting the linkage to CHI.
• Limited range of population/clinical variables collected.
• Data collection for April 2019-March 2020 suspended due to competing pressures of data collection in response to COVID-19 response and burden to the sector, data collection for April 2020-March 2021 resumed July 2021.

TURAS Care Management Tool 'Safety Huddle'

• In response to the COVID-19 pandemic and with care homes reporting significant resource pressures from multiple data requests, the Scottish Government, in collaboration with care homes, the Care Inspectorate, Scottish Care and the Scottish Social Services Council, developed the 'safety huddle tool' (SHT) as a management tool for care homes to monitor the pandemic, is hosted on the TURAS platform and launched in August 2020.
• The SHT was designed for use both during and post COVID-19 to provide collection of pertinent data (e.g. visiting, PPE, Infection Protection and Control (IPC) measures, occupancy, staffing and outbreak levels) and enables both situational awareness and risk assessment for care homes, to support timely escalation and early intervention where appropriate.
• Usage and information compliance input has been positive but varied over time, with electronic, daily data submission on key domains for pandemic response.
• Completion rates for the SHT were high during the pandemic (consistently above 90%).
• Data sharing restricted to care homes themselves (can see own data only), providers (can see data for the homes in their group), local oversight teams, health and social care partnerships, Public Health Scotland, Care Inspectorate and Scottish Government.
• Management information tool, including escalation and involvement of other agencies.
• Domains focused on safety and pandemic response including: occupancy, staffing, testing, outbreaks (confirmed and awaiting confirmation), adequacy of Personal Protective Equipment, Infection Prevention and Control measures, count of deaths (COVID and Non-COVID), transfers to hospital, 1:1 care, end of life care, cognitive impairment, dependency/complexity.
• Expanded to include vaccinations data (resident and staff, collated from data captured within the vaccination programme dataset) and data on visiting practices to support monitoring implementation of evolving guidance.
• The SHT was initially created for monitoring the impact of COVID-19, but as the pandemic response moves to a 'living with COVID' phase, the tool will likely be adapted, and this is subject to consideration via the Care Home Data Review.
• The SHT is under consistent evaluation and review (July 2022 – onwards) to ensure it meets the needs of users/providers and reflects present and future pandemic/endemic status. This aims to determine its future frequency and usage, alongside potential evolution 'post-pandemic' to serve as a supportive improvement tool with additional functions. The July 2022 plan for the SHT review is to look at the SHT in 2 phases:

a) The short term immediate future for the tool –COVID-19 focus and actionable information gathered, especially in relation to winter planning – looking to reduce and revise questions.

b) The long-term strategic view for the future of evolving the SHT (or an alternative tool) – for enhanced usage to support care homes with additional functionalities (determined through the 2022 care home data review).

Inpatient Census: includes Out of Scotland NHS Placements and Hospital Based Complex Clinical Care and Long Stay

• The census' part 1 includes patients with a main diagnosis of a mental health condition, learning disability or addiction who are treated in an NHS Scotland bed. Some patients may be treated in NHS beds in care homes.
• Part 2 patients with a main diagnosis of a mental health condition, learning disability or addiction are NHS Scotland funded but treated either out of Scotland (e.g. by NHS England) or out of NHS (e.g. in a private/voluntary/local authority care home).
• The census also includes a section on hospital-based complex clinical care and long stay hospital patients, although in the 2019 census none of these patients were recorded to be in care homes.
• Annual census snapshot. However, the census was paused in 2020 and 2021 due to the pandemic.
• Although a declining model of care, this population and needs in terms of length of stay, funding and clinical model vary across different NHS Boards.
• Most recent inpatient bed census report to March 2019.

Care Home Data Review 2022

Scottish Government, working with analysts in partner organisations, are conducting a review into the care home data landscape and are considering improvements around the content, quality and frequency of current data collections, and/or alternative methods of data collection. The aims of the review are to ensure a coherent suite of data collections, reduce the burden on data providers and meet the existing and emerging needs of data users. The review is being undertaken as it is recognised that the current evidence base relating to care homes is neither complete nor coherent and does not provide the level of insight required to meet the needs of the sector/policy makers.

Selected Other Sources

These sources have been identified due to their use and relevance as a care home data source.

UK wide Laing & Buisson Care Homes for Older People Market Report – annual, costs £1,345-3,450 for access, print and digital formats (not inclusive of all care home services) [parallel reports for Homecare and Supported Living UK market report and Retirement Housing UK market report].

Knight Frank Care Home Trading Performance Review (not inclusive of all care home services).

Local Data Dashboards

This section presents information on local dashboards of national adult care home collections.

Several NHS boards have developed and use 'care home dashboards'. Both NHS Greater Glasgow and Clyde & NHS Lothian use these internally. They have underlying algorithms based on address matching to identify and include the care home population. Using these data it is possible to look at trends around admissions, attendances and identify which care homes use more unscheduled care than others.

There are data quality issues around over-inclusion of those in neighbouring postcodes and the timeliness of snapshots when looking for longitudinal data (e.g. residents are included based on current residency not historic/temporary care home placement - historic placements may be counted separately or also counted by another health board). Data not available real-time (in National Health Service Greater Glasgow and Clyde).

Flags in National Data Sources

This section presents information on the flags relating to adult care home clients in other national data collections.

CHI Institution Flag

• Coded variable to identify institutional addresses including schools, universities, prisons and care homes.
• Applied based on discretion of those manually coding change of address forms at Practitioner Services Division received from Primary Care.
• Inconsistent definitions of care homes are being used and, in some cases, based on the judgement of coder whether address is an Institution and not based on Care Inspectorate list of registered services. Not all adult care home services are flagged as care homes.
• Two care home codes which are reported as 'residential' or 'nursing' flag (codes 93 and 98 respectively although do not accurately denote as these classifications not used in Scotland).
• Decision to label as care home based on whether care home services receives GP capitation.
• Accuracy varies between NHS Boards – sensitivity 55-89% between Fife and Tayside – incomplete for whole care home population.
• Positive predictive value is high – so likely address is a care home if flagged as a care home, but this classification does include some sheltered housing and residential addresses erroneously.
• Reliant on GP change of address notifications thus time-lag in an individual moving-in to a care home and this being reflected in nationally available data. The length of the time lag is unclear, likely to be 1-2 months on assumption of need for regular prescription, may be improved from January 2021 as weekly updates are being made to CHI.
• Caution when using in specialist care home populations (e.g. substance misuse) as manual review work suggests individuals may have changed their address subsequent to receiving care in these settings, but their GP record in CHI is not necessarily updated to reflect this, thus risk of over-inclusiveness of younger adults no longer resident in specialist care homes.

Other flags in national datasets

• Prescribing Information System Care Home flag – community prescribing dataset uses CHI Institution Flag. The CHI Institution Flag is applied to indicate care home residency if the 'paid date' of the prescription is within the period in which an individual is resident in a care home according to CHI. Paid date is the last day of the month in which dispensing of the prescription was completed, rather than dispensed date (when the prescription was dispensed). There may be a lag in data accuracy for those receiving weekly dispensing as the paid date will be when the prescription completes rather than starts.
• Scottish Patients at Risk of Admission and Readmission dataset – uses CHI Institution flag.
• Unscheduled Care Datamart – uses CHI Institution flag.

Hospital admission/discharge data

• SMR01/01E/04 and other national data do not contain address data fields beyond admission postcode.
• Coded fields of 'admitted/transferred from' and 'discharged/transferred to' available with meaning described in SMR Data Manual and SMR Crib Sheets.
• Variable accuracy as wide range of options and often completed by administrative staff or clinicians without appreciation of role in data intelligence/national statistics.
• Multiple options available depending on clinical system being used, with 'private residence' frequently selected.
• Recent Public Health Scotland Hospital Discharges publication suggests 30% of discharges to care homes not identified using this approach and 3% of non-care home addresses incorrectly flagged using these codes.
• Admission from codes not formally evaluated for accuracy in identifying hospital admissions from care homes, but concerns around data quality have been expressed previously in Assessments of SMR01 data.

National Records of Scotland death registration data

• Death certificate data includes address fields of location of death.
• Identifies those dying in a care home based on Institution Location Codes.
• Multiple codes which can denote care home [J joint user hospitals, K contractual hospital, R homes for the elderly, S other homes, T miscellaneous premises, U care home and V private nursing home].
• Note PHS Institution locations do not align directly with CI list of registered services, particularly where a service name has changed over time. Not all adult care homes registered with the Care Inspectorate have an Institution code.
• Most standalone hospices in Scotland have a V location code, meaning that summaries of deaths occurring in care homes will include some hospice deaths.
• However, NRS use a consistent methodology in classifying locations across their published National Statistics on deaths.
• Additional effort used to identify deaths of care home residents in hospital during pandemic, based on internal use of additional data fields – data not routinely available/published due to issues identifying care homes accurately.

Scottish Ambulance Service (SAS) & Patient Transport Service (PTS) Data

• SAS for emergency transport, PTS for scheduled journeys.
• Both include fields about where individuals are collected from and taken to.
• These are based on address lookups to improve accuracy and quality.
• It is therefore possible to use these data to extract address information which is of high quality, but requires access to identifiable information.

Delayed Discharge Data

• Includes variable to indicate if individual's stay in hospital is delayed due to waiting for placement (note this category is broader than care homes and includes intermediate care which may be provided in other designated locations than care homes).
• Annual detailed publication with monthly updates published by Public Health Scotland.

Public Health Scotland

• Internal method available to identify care home residents within data and link records to CHI – via CHI Linkage (CHILI) team – methodology not in public domain.
• Additional methods developed using clinical details fields to identify care home COVID tests.

NHS24

• Record address of callers (based on address lookup for quality/consistency).
• Since end August 2020 – are asking if call is about a care home resident, so building a 'care home flag' which is being reported on daily basis to Public Health Scotland.

Alternative Methods

This section presents information on other methods for identifying flags of national adult care home collections.

Address matching

Address matching has been the traditional method used (for research and evaluation studies) to identify care home addresses and thus residents living in care homes. However, address matching for care homes is problematic as they do not have unique postcodes and care home names are not presented in standard formats – often containing room number, house within the home, care provider/company name, care home name and street name or a combination thereof. They often contain colloquialisms about how a home is referred-to locally, or may be recorded in clinical datasets based on what the home was previously called under earlier management. To explain, these are all address records for the same care home):

Abbotsford, Kinglassie, KY5 0XT

Abbotsford, Kinglassie, KY5 OXT

Abbotsford, Kinglassie

Abbotsford Care, Lochgelly KY5 0XT

Abbotsford, Mina Crescent, Kinglassie KY5 oXT

Abbots Ford, Meany Cres, Kinglassie, KY5 OXT

Care Home, Kinglassie, KY500RC

Nursing Home, Kinglassie KY5 0XT

Abbotsford Residential Home, KY5 0XT

Access to Master CHI has allowed detailed examination of care home addresses and development of techniques/algorithms to assist - within appropriate Safe Haven environments. However, while these have been able to improve identification of care home residency, they have not allowed allocation to a specific care home service, without manual allocation.

The newest and most promising development is the use of Unique Property Reference Number (UPRN). This is an Ordnance Survey identifier for all properties across the UK. Public Health Scotland and the Scottish Centre for Administrative Data Research have undertaken work to match UPRN to Master CHI to improve identification of care homes and, most advantageously, identify which care home an individual resides in.

This work has been facilitated by the Improvement Service matching algorithm. It emphasises the need for dedicated work to improve care home addressing within national datasets (helped by more systematic use of address lookups) and the over-representation of care home addresses among those which cannot be matched by algorithmic/automated means.

The Improvement Service receive regular updates from the Care Inspectorate regarding their services through One Scotland Gazeteer. Discussions are underway around adoption of UPRN within the Care Inspectorate data, as this would be transformative for addressing changes of service and could sit alongside CSNumber (to preserve insights about specific services), to allow understanding of service changes moving forwards. Ideally, all services should use a looked-up address which can label records with UPRN, but this would require investment to facilitate. Need to undertake specific work to resolve/manage care home addresses with multiple UPRNs and ensure UPRNs for care home services are accurate and contemporaneous.

Priority Development Areas in Adult Care Homes Data Capture

This section presents information on the development areas of national adult care home collections. As collated by members of the 2021 Scottish Government Data, Analysis and Research Steering Group for Care Homes and Social Care in 2021.

• Reliable identification of who lives in a care home – likely to require national dataset in which all admissions, discharges and deaths are recorded. Incorporating CHI as a linkable identifier which could be used to join to other national datasets.
• Changing role of care homes – there is no national data collection on the provision of intermediate care and step-up/step-down models of care to support individuals on a temporary basis. NHS England has an annual audit of Intermediate Care which helps provide useful evidence on role of services in wider care landscape.
• Short stays: it would be useful to collect individual-level data on respite, short stays and emergency care home stays. This would help in understanding the level of need, access, timeliness, funding and utility of such services and how these link to other models of care and support.
• Links to primary care data – to enable assessment of frailty (using electronic Frailty Index) within the population and record multimorbidity based on conditions not recorded in acute hospital datasets, but which are important markers of chronic disease, disability and functional impairment (e.g. arthritis, visual and hearing loss etc.).
• Outcomes and experiences for those living in care homes – moving towards a much more nuanced and detailed form of data capture but one which could tell us some of the most valuable data about our care home services. Will require bespoke measures and appropriate methods of data collection and synthesis, sensitive and appropriate to the environment. Adult Social Care Outcomes Tool (ASCOT) as one existing approach and exemplar of what can be possible.