A typical Triscribe week
Moving forward our clinical focus, more product growth and solving the knucklehead stuff
Most weeks, I write about general topics and themes. Maybe pick up on some news item and consider the wider implications. Triscribe works to the principles I describe in these overview posts. But the power and success of our software is driven by details.
So in this post I want to summarise the areas we are working on right now. I hope it gives you a flavour of the way Triscribe continually learns and strengthens analytics and AI to better suit the needs of frontline NHS clinicians.
Clinical focus
We have been building analytics to help hospitals track opioid deprescribing. Early versions are now being reviewed by clinical users.
There is a basic data challenge. Hospitals cannot easily access primary care prescribing records. The only way to know if a patient is taking prescribed opioids on admission is through an optional checkbox in the EPMA system. No surprise that this is used inconsistently.
The simple solution to this is data linkage. Until that happens, the Triscribe team is looking at the data we have. We aim to infer the best answer available. Any ideas or suggestions are welcome.
We are also supporting a small practical research project around Anticholinergic burden. Triscribe already has analytics in this area. Our customer’s pharmacy falls team is looking at this data now. We will move forward when we have their feedback on needs and priorities.
Product growth
Every day we work on building customer feedback into our software. The primary focus at the moment is a list of requests from a new customer. The work is almost done and we are opening up to feedback from clinicians.
Our new customer brought us access to Medchart and Ascribe for the first time. The work has been focused on the Medchart EPMA system. Over the past week, we have started extracting data from the Ascribe stock system. We now have the basic movements and balances aligned so we can study this data daily.
The most exciting conversation this week was about a new data source. If available, we could include data from closed loop prescribing for the first time. All part of aiming to provide a complete view of NHS hospital medication data.
Knucklehead stuff
We have another hidden aspect in our work. NHS staff work with the consequences of long-term underinvestment in IT. Just as much as in other areas. That means legacy systems with a wide variety of data quirks and challenges.
Over time, Triscribe handles these things for our users. I call this the knucklehead stuff.
This past week that involved three data access and availability issues for two different customers. Two of these are fixed (one after several months of effort.)
In addition, we were working to refine our data extractor and reduce the time it needs to run on NHS systems. Another regular task was uploading the latest update to dm+d. These things happen in the background. Mostly automated. That means we constantly improve the user experience.
Work with Triscribe to make this better
Looking back over that, it’s a pretty typical snapshot. I’m afraid Triscribe doesn’t do Eureka moments or moonshots. Steady accumulation of improvements and learning is the way we work.
We think that’s the best way digital tech can help the NHS. We would love to work with clinicians, clinical needs and data from more hospitals and more systems. If you are interested in finding out which analytics we could build for your hospital, get in touch: contact@triscribe.net
Also a good week - as it was nice to connect with Matt and Richard at Christchurch Hospital, New Zealand - sharing our stories and working on ways to collaborate.