NHS Referrals are Broken

Cinapsis came about because one of the jobs I had as an ENT registrar was to vet the referrals coming in from GPs into the hospital, into our departments. There were 200 a week, and in that day they were letters — the GP had written them, put them in an envelope, and they'd arrive at the hospital, and I had to go through this pile.

The reason the department wanted me to do that was they were hoping I'd turn some away. But reading the letters, I realised pretty quickly that the horse had bolted. Unless there was something particularly clear in the letter, it was very unlikely I was in a position to reverse that decision — the GP had already had a discussion with the patient, and the patient was already expecting an outpatient appointment. So there was very little I could do other than see them when they came.

And I could tell, even from reading the letters but certainly from seeing the patients, that if only I'd had a chance to talk to the GP when they were seeing the patient, there were often just a couple of things they could have done that would have made the outpatient appointment unnecessary, or at least sped up the process for that patient. The other side of it is that these patients would arrive very anxious — especially a two-week-wait, cancer referral. And why wouldn't they? They'd think there was a good chance they were going to be diagnosed with cancer. Or even if it was more routine, if they're seeing an ENT surgeon, they'd expect they were going to have an operation, and they'd be pretty disappointed when I said, actually, you probably don't need one — let's try this medication first.

So I realised that this gap between primary and secondary care causes so many problems. It makes life harder for the GP, slows down treatment for patients, inconveniences the patient, fills specialist clinics with patients who probably don't need that level of care at that moment, delays treatment for other patients, and just costs a huge amount. It became really clear to me that things could not continue as they are — they were going to fundamentally change. And the way to do it is to join up these parts of the health system. It seemed inevitable that things were going to move in this direction, and I was excited to be part of changing that. I really wanted to lead the change.

Yeah — it gives you time to think. It didn't take me long to get through them, but it was just a waste of time. I thought, this is not adding any value. And then I thought fundamentally about the problem. I'd imagine the GP writing the letter, and how they don't have many levers to pull: if they wanted a specialist opinion, their only lever was to refer — to send the patient physically into another building, maybe in another town or city. It just doesn't make sense.

Because my experience of working with GP trainees was that we'd upskill them very quickly. They'd start with us not knowing much about ENT, and just by calling me up and talking me through the problem with the patient in front of them, they'd be upskilled within a few weeks and independently managing most of the emergencies we'd see. So I had the real personal experience that you can do an awful lot just by connecting clinicians remotely — you can help someone increase their sphere of competence and manage more problems than they could have otherwise. And often it's just reassurance that they're doing the right thing.

What Cinapsis does is make access to specialist advice, if it's required right now, very quick for the GP. We connect GPs to the right specialist in their hospital in less than 30 seconds. They can talk the problem through and come to a joint decision about the next step for that patient. If the patient does need to be seen, we can direct them to the right part of the hospital.

For example, we've reduced the number of patients attending A&E from primary care by 83%. 30% of them stay in primary care; 50% still need to be seen by a specialist but don't need to go to A&E — rather than A&E, they go directly to the specialist clinic they need or directly to the ward. And from the hospital side, they know what's coming their way. It's not just a surprise — you're bleeped, there's somebody in A&E you know nothing about. Now they understand the work coming their way and can plan with the GP: discharge planning, for example. Fine, this patient needs to come in, but what's their home environment like? Have they got support? Questions that are very quick to ask in the moment, but once the patient's arrived it takes real effort to find those things out.

It's very simple, really, in terms of the communication and joining people up — but the impact of that is huge, and the downstream processes we can put in place are really impactful.

Absolutely. You're right that there are a lot of standalone solutions in place. If you're a GP, you might bleep the medical registrar on call, or sometimes there's a hotline. But from a GP's perspective, it's a different way of communicating for different problems: for a medical problem you bleep this number, for urology you call this number, for general surgery another number. So it's very disjointed. And secondly, there's no recording of that interaction — who was called, what patient it was about, what advice was given, what the outcome was, whether the next step happened, whether there's any learning from it, how we change our pathways.

So these conversations do happen, but it's a very unsatisfactory experience. We've done a lot of work on visibility from the GP side. When they engage with Cinapsis, they can see how many people are being called at once, how many are on call, who's answered, whose phone has gone to answerphone. And at a system level you can see all of this interaction. Not only is the call recorded and the outcome recorded, the administrative next steps are put in place. If the outcome is that they need to come into an urgent clinic, the receptionist now knows to expect this patient, and the team in that clinic have access to the conversation that happened — they can see the patient's notes, any images that were exchanged, the ECG at the time.

It's fundamentally the same as two people talking, but the experience of it, the governance, the learning — and from a systems point of view, you can use Cinapsis to create new clinics. All it takes is: we've created this clinic, and it comes up as an option for the GP. The pathways are much more flexible and can be disseminated much quicker through our platform. What you can't do with a bleep and a switchboard operator is just harder to do.

That's a big question, and you're right, there are a lot of moving parts. If you're starting anything new like this, this would be my advice. Having something to show is a great starting point. With a problem like the one Cinapsis is solving, what we started with was a minimum viable product. We started with the communication piece — we can manage your on-call rotas, we can connect people to one or many people in one go. And people can see the value of that: instead of me calling switchboard, I can see if there's anyone on call and the system helps me understand how to connect with all these different services.

So you think of it from the end user's point of view. Who are the end users? The GPs, and the consultants or medical team on call. You build it from that point of view, understanding their pain points and how you can address them. And once you've shown them one pathway, you can say, imagine if we added all your pathways, and you could monitor the flow of patients in real time, and connect that data to population health data, to hospital episode statistics, tie it into your financial planning. So you start from end users addressing one problem, and then you build the system perspective from there — rather than starting from "we need to save the NHS millions of pounds a year."

If I think of successes, the biggest one is that Cinapsis is used by clinicians I respect a huge amount, and they're saying it's making their lives easier — making work easier, making caring for their patients easier, so they can deliver better care. I draw so much satisfaction from that. I love the individual anecdotes: they'll tell me about a particular patient who didn't have to come to A&E, or who received treatment quicker. That's really gratifying.

In terms of the impact on a system, for COVID we responded really quickly. There was an understanding that there'd be huge demand on urgent care services, and the aim was to keep patients out of hospital — even if they had COVID, unless they needed to come in. We rolled out very quickly to the paramedic service, community nurses, inpatient mental health wards. In less than two weeks we rolled it out, and that reduced the number of patients with suspected COVID having to come to hospital by 30%. If you think of that over a whole county, and the reduction in exposure for clinicians on the front line, that's a huge impact.

And the fact that Cinapsis is now used for paediatrics, orthopaedics, frailty, acute medicine, dermatology — we've realised the vision we had at the beginning, that you can have a single point of access for all these different pathways. In total we have more than 50 pathways, and GPs can very quickly say, whatever the problem is, I want advice about this, and they get connected. Plus we have a business intelligence part of it, so we can demonstrate this information in real time across the population, and it can all be monitored.

There are loads of challenges, obviously. Running any company, and a tech company, is hard work. It's the hardest thing I've ever done — much, much harder than surgical training. It's about having a long-term view, an idea of directional travel, and attention to detail. We're handling a huge amount of patient data now, so the things we have to have in place to protect against hacking, cybersecurity, data sharing, GDPR compliance — all the bread and butter of a health technology company — I had to learn all about that and build the team to address it. But it's fun. If you like learning about stuff, all those challenges are fun, and I really do like learning about things.

I think having the end in mind — thinking, where is this going? And approaching things from first principles. There are lots of reasons not to do things, and lots of reasons to think things will never change. But what I did was think about how crazy it was that GPs had to write letters, that the exchange between GPs and specialists was so infrequent and so disjointed. That just felt so fundamentally unsustainable that things had to change. So being very clear about the drivers and the principles — why things are changing.

I also think reading is a very good habit. It's kind of cliché now — everybody says you should read — but I didn't do that until about a year leading up to starting Cinapsis, when I read more than I'd read before. I had almost a two-hour commute on public transport in London, so I had more than three hours a day to read. And it's kind of free experience: these people write down their experience and you get it for the price of the book, without having to expend the energy yourself.

The other thing is getting people to work with you. You have to be somebody that people want to work with. I'm all for flexible working and supporting people to meet their goals, and focusing on how to help them. That's rewarding for me — our tech team now is exceptional, and that's because we've allowed them to work in a way that suits them.

One of my favourite books is Thinking, Fast and Slow — that's a really good book. And a good one for anybody thinking about health tech is The Innovator's Prescription by Clayton Christensen. He's written a lot about disruptive innovation, and this book describes how the health service is set up and the problems with the way it's currently constructed. Some of the things NHS England have put in place, you can see are influenced by this book — I'm not sure if directly, but I'd be surprised if it didn't inform some of the strategic decisions being made.