A challenge working on digital prevention services is the tension of providing healthcare that works for individuals on a scale that works for the country.
Remote access to healthcare professionals
The highlight of my week was a talk by East Suffolk and North Essex NHS Foundation Trust (ESNEFT).
Working with a tech partner, the Trust has created an online channel for people with obesity to access care. Importantly this includes speaking to qualified healthcare professionals (doctors, psychologists, and specialist nurses) at the same level of care as in‑person services. The Trust has a dedicated clinician responsible for the service and online patients. The digital channel offers direct referral into pathways that are affected by obesity (for example, sleep apnoea and musculoskeletal). The service has some other features like diet and exercise guidance.
But the important thing about the design of the service is integration. The users of the service are essentially remote patients, the level of care is identical to patients who attend appointments in‑person. They can move between online and in‑person care seamlessly.
This has clear benefits. It saves patients the time, cost and inconvenience of travelling for an appointment. A benefit for everyone, but for people with BMI of 60 or higher, it makes a huge difference. It also means clinicians can see more people.
The Trust has also joined pathways that are usually standalone. It combines bariatric surgery and GLP‑1 weight‑loss medication within the same service. This allows the best solution for the patient to be provided. For example, bariatric surgery might be best for someone who is young with relatively low comorbidity, while GLP‑1 might be better for someone more elderly with longer established conditions.
The service feels transformational because it lines‑up incentives for the Integrated Care Board (ICB), the Trust, and patients. As part of the service, the ICB gets help with providing GLP‑1s (something they’re now required to fund under national policy) with reduced burden on GPs. The Trust can offer joined‑up support and reduce burden on hospital staff. Patients get choice and faster access.
For me, working from the centre, with a focus on scale, this level of integration between physical and digital feels distant. Even a one‑time hand‑off from our services to local services is difficult. Truly integrated digital and in‑person care is years away. But I am convinced this is what we need to strive for.
Asking people population level questions
One of our services is looking at digitising the lung cancer screening eligibility assessment – something that currently happens over the phone for most people.
The triage consists of a few questions that calculate the user’s likelihood of having lung cancer. If they meet the threshold then they’re offered a CT scan.
The risk calculation is based on population level statistics. Things like ethnicity and level of social deprivation are factors used to calculate the risk in the model we’re likely to use (there are other models that simply ask if the user has ever smoked and their age).
The evidence and logic of asking population level questions is clear. But the implementation is another thing. For example, in the offline service, a user’s level of deprivation is often estimated using proxy questions, such as asking them what level of education they have. It’s pretty startling to be asked this question, especially when it’s not obvious how it relates to their health. In an over the phone service, this can be handled. With a digital service the motivation for asking this question is harder to explain. These population risk questions appear hostile to users. It’s possible they will impact completion, there’s a compromise to consider around accuracy of prediction vs the number of people who complete the triage.
I think the exact model we use to calculate risk is still up for grabs. Ideally we can use one that accurately calculates risk without feeling discriminatory. The team are also exploring if we can use different questions to get to the same calculation and ways to explain why the question is being asked, which offers some mitigation.
Long term we hope to use data about the user to avoid asking as many questions as possible.
Incentives and integration
Both examples show the challenge. Individual Trusts control their local system, employ clinicians and change how pathways operate, but national services have to work across a complex set of unrelated systems.
Trusts also have a relationship with their patients and access to data that lets them personalise care. We don’t have access to the data we need to personalise services yet. In the interim, we’re building services that ask intrusive questions.
We need to look forward with our services. Which of them can deliver value as digital alone? Which ones need integration with in‑person services? For the services that need integration – what do local organisations need to make partnership worthwhile?