Can the NHS make better use of Data?

In two weeks time it will be exactly one year since I had a life-changing operation. I was born with a Bicuspid Aortic Valve, however I only found this out in 2016 after I had experienced difficulty in ramping up my cycling efforts after the long winter. It had become stenosed (narrowed).

So after lots of tests and consultations I was scheduled for an Aortic Valve Replacement operation in July 2017 (a four hour operation which included stopping my heart). Due to my age I chose to have a mechanical On-X valve (CryoLife Inc.) inserted into my heart rather than a tissue valve. A tissue valve would need replacing in about 15 years time, whereas the mechanical one should outlast me. The consequence of this is that I now have to take anticoagulation therapy for the rest of my life (Warfarin), and I now tick!

Warfarin was originally developed as a rat poison, but it wasn't long until it was being used as an anticoagulant in humans. The big problem with Warfarin is that if not closely monitored then the blood can become so thin that uncontrollable bleeding occurs. Common outcomes include intracerebral bleeding and stroke. You also need to be careful not to use it in conjunction with certain drugs (Ibuprofen, Aspirin), and certain foods high in vitamin K(Broccoli, Kale, Bananas, Liver). You also need to moderate your alcohol intake!

I like to know what is happening and, after almost 30 years in the IT industry, I turned to data to help me.

The key measure in Warfarin use is the International Normalised Ratio (https://en.wikipedia.org/wiki/Prothrombin_time#International_normalized_ratio). This is a value that determines how thin your blood is. Normal human beings exhibit an INR of 0.8 to 1.2. With Prosthetic valves normally an INR of 2-3 is desirable and in some cases as high as 3.5.

I am a cyclist and I average around 150km every week. I mainly cycle in the Scottish countryside, but I do occasionally fall off and bleed! The prospect of not being able to stop bleeding after an incident was a big concern for me and that is why, in conjunction with my surgeon, we chose to have an On-X Valve. Being a more modern valve with advanced materials meant I would only need an INR of 1.5-2.0 (i.e. easier to stop bleeding!).

Getting your INR value is relatively straightforward. It involves taking blood and having it tested in the lab. Immediately after my operation this was happening daily as it takes time to find out what is the right Warfarin dosage for you as an individual. Eventually it does settle down and you end up on weekly and then monthly measurements of your INR. Each time involves a trip to the practice nurse at my GP, and then getting my results the next day, and adjusting medication if required.

Early on, I decided I wanted to take more control of my situation and get better data on how my body reacts to certain foods. To help with this I bought a device from Roche called a Coaguchek XS. This device enables me to measure my INR in the comfort of my home whenever I want. Many in the medical profession express doubts in the ability of these devices to accurately measure INR, but mine has been correlated with my lab results and found to be an excellent match over the year that I have had it.

This device enabled me to discover that it's ok for me to eat the occasional piece of broccoli or have a glass of wine with my meal as they did not materially affect my INR. However, more importantly, it has caught a few occasions where my INR was too low or indeed too high and I was able to make a slight adjustment to my medication to combat that (note: my wife is a doctor, and was consulted on all medication changes).

It is these additional data points which clearly showed to me that INR varies up and down and that there are clearly points in time where my INR are outwith the boundaries set and which I have not been able to capture (I do not test every day as the cost would be prohibitive).

It is these additional data points that make me wonder whether the NHS could make better use of data such as mine (and from similar technologies). Wider adoption of technologies such as Coaguchek, could free up valuable time from GP surgeries and NHS Labs. Linking the data from my device to AI or RPA technologies could enable small changes in medication to be processed in a rapid and more efficient manner, and help prevent strokes and other negative outcomes that cost the NHS a fortune.

Admittedly, any such adoption needs to be in a controlled manner and piloted properly to ensure the correct processes are implemented and patient safety and confidentiality is not compromised, but the technology is there, we just need to find better ways to make use of it.

Now, which hill shall I go and climb tonight?

(Author note: I have no current association with either Roche or Cryolife, although I have consulted with Roche in the past)