Hospital Admission Notification Service (HANS) Pilot

How might we provide notifications to social care providers when someone they provide care for is unexpectedly admitted to hospital?

What was the problem?

When people who receive domiciliary (home) care are unexpectedly admitted to hospital, the people that provide their care may not be informed. This can lead to carers experiencing ‘failed visits’: they would arrive at a person’s home to provide the care they need, only to find that they are not there. They then potentially have to wait at the empty home until the person has been located, which could involve ringing round hospitals, or in some cases, calling the police to break down the front door.

To better understand the scale and impact of the problem, with the support of NHS England/DHSC’s Digitising Social Care Programme, we ran a discovery phase which included structured research interviews with people working across the homecare sector - including senior managers at national care providers and carers working in individual local branches.

It quickly became clear that the problem of ‘failed visits’ due to people being unexpectedly admitted to hospital was a real one, which while happening relatively infrequently, could have a big impact. We found that:

• Carers have close relationships and rapport with people they care for, and experience high levels of frustration, stress and anxiety when they were not informed that they have been admitted to hospital. Not all carers are paid for travel time, or for visits that don’t take place.
• Registered managers could spend hours phoning round different hospitals and wards, in attempts to locate their clients
• Local authorities who fund adult social care provision spend time retrospectively reconciling care contracts which were unable to be delivered or dynamically reallocated
• National-level care provider managers were increasingly concerned about the financial and workforce implications of failed visits, and the knock-on effect of wasted social care capacity on discharge and other priority pinch points within the system

User research interview with Care Worker

User research interview with Register Manager at a Home Care Provider

We validated these findings by engaging with other key health and care stakeholders, including local authorities, NHS staff at national, ICB and hospital-level, and companies that provide digital software to homecare providers. These groups all identified potential benefits of developing a solution. For example,

• NHS ward staff would spend less time on the phone fielding calls from care providers trying to locate them
• Automatic links with social care rostering software could allow staff’s capacity to be more dynamically reallocated
• if local authorities had more real-time information about admissions they could manage their contracts with care providers more dynamically
• aggregated data about the relationship between home care and admissions may allow the development of better national models, or AI processing to improve policy decisions and identify risks pre-emptively

Why is this an issue?

Once we understood that failed visits were adversely impacting homecare providers (and others), we began to quantitatively model the ‘size of the problem’ in more detail.

Since no specific dataset exists, we analysed data sources covering the UK homecare sector and hospital admissions, to develop a model to estimate the scale of the problem across England. Our modelling suggested that each year there are approximately 40,000 instances of people receiving homecare being admitted to hospital without their care provider knowing.

Further modelling (informed by our user research) showed that the annual consequences of this problem include:

• Up to 100,000 hours of homecare staff hours wasted - that could be better spent managing, planning, or providing care
• Up to 114 tonnes of avoidable CO2 emissions due to carers’ unnecessary journeys

Our hypothesis

What did we do?

Our design goals for finding a solution were to minimise the number of new processes that staff would have to go through; and utilise existing systems and platforms wherever possible.

We eventually designed a potential solution based on a publish-subscribe model, using NHS numbers to allow care providers to “subscribe” to messages relating to a person they provide care for. Existing hospital admission messages could be routed to provide the “publish” message when a person is admitted. This solution met our design goals by using existing processes and platforms. It also fits with existing systems and architecture, utilising

• existing messages generated by hospitals on admission as a trigger
• the existing national event management service (NEMS) to link publishers and subscribers
• a route to provide notifications to care providers either via API to their clinical or rostering system, or secure NHSmail message.

It also had the benefit of minimising the data that needed to be shared between organisations, and allowed multiple subscribers (with a legal basis to receive it) to be notified at once.

These elements would work together as a service called the ‘Hospital Activity Notifications Service’, or ‘HANS’ for short.

To validate the technical feasibility of this solution, we built a prototype system which ingests HL7v2 messages from a hospital, filters and processes them through a message queue, and sends notifications to a registered care provider. This is end-to-end tested with dummy data, but has been built to the standards required to stand it up as a live pilot, in order to test the business feasibility and to validate our modelling work.

We also did substantial assurance work related to business processes, pilot governance, information governance, clinical safety, cyber security and evaluation frameworks.

Throughout this project we have worked in the open, running regular engagement and show and tell sessions with stakeholders including NHS hospitals, ICBs, Local Authorities, Care Providers and NHS England.

We have also published the prototype code for a Subscriber Management API and Management Interface.

What could happen next?

With the support of the Cabinet Office Evaluation Task Force, we have developed a rigourous evaluation framework, and are considering the feasibility of running a limited-time pilot to test the business feasibility and validate our benefit modelling.

Our user research and stakeholder engagement has also identified numerous other exciting additional potential applications and benefits of HANS.

More trigger events

Whilst our focus was on emergency admissions, we have learned from social care providers that notifications across the care pathway would be helpful. For instance, notifications around ambulance conveyance would give care providers even earlier information about an unexpected admission, and information about wards and discharge would support better discharge planning, improving patient flow.

More audiences

If implemented, we found there are audiences outside of adult social care that may also benefit from notifications. For instance, pharmacy staff could save time in preparing weekly prescriptions for individuals that will not need them because they are in hospital. We have also heard about potential applications in community mental health and the third sector. We could even empower individuals, via the NHS App, to manage their own preferences for who is notified if they are ever admitted to hospital.

More information

If you would like to talk to us about this project, you can contact us at england.innovation.lab@nhs.net