In our latest interview series we asked Arron Thind from East Surrey Hospital a few questions about an application he and two partners have developed to automate the process of calculating dosing regimens for four commonly prescribed intravenous antibiotics (gentamicin, amikacin, vancomycin and teicoplanin).
The Antibiotic Calculation and Monitoring System can automatically adjust regimens to account for various patient-specific parameters (indication, kidney function, sex and bodyweight), as well as interpret serum levels for all four intravenous antibiotics, in accordance with trust guidelines.
Pilot studies in both Croydon University Hospital and East Surrey Hospital have demonstrated faster and more accurate prescribing. For Croydon Hospital, the application was calculated to save the trust over £4,000 per year in junior doctor time calculating and prescribing, and over £1,500 per year by preventing drug waste.
Could you tell me about yourself and your background?
I am a junior doctor, currently in my second year. I undertook my medical training at Oxford University, where I also studied a BA in medical sciences, specialising in cardiovascular, respiratory and renal physiology. This is where I initially became interested in health innovation, and alongside my studies I led published research in cutting-edge surgery and novel cancer biomarkers. Later, my passion turned towards digital health whilst working for Your.MD, a symptom-checker app that uses AI technology.
Following graduation, I started as a doctor at Croydon University Hospital, where I applied the skills gained in research to lead quality improvement projects with a junior doctor colleague, Dr Joe Hogan. For our projects, we were awarded three Health Education England awards in quality improvement, leadership and outstanding contribution to the trust. To take our efforts further, Dr Imran Qureshi (a microbiology consultant at Croydon, with a qualification in computer science) trained Joe and I to code in Java programming language, a task we managed whilst balancing our clinical commitments. From doctors with no prior coding experience, we were able to co-develop the Antibiotic Calculation and Monitoring System, with the aim to tackle the widespread dosing errors of intravenous antibiotics identified in our quality improvement project.
Where are you currently with the development of the application?
So far, we have built a fully working system with no third-party assistance. It is an app built by junior doctors, for junior doctors. We’ve had to think hard about the design, and as a result, our pilot studies have demonstrated it to be faster and more efficient than the current method of calculating and prescribing doses.
We want the app to be bespoke for each trust. Currently, it is built in accordance with Croydon’s antibiotic guidelines, as they are in the process using it throughout the trust. However, we are amid discussions with East Surrey Hospital, who would receive a version of the app specific to their local guidelines.
What does the technology do and what are the benefits?
Essentially, it is a computer-based application, coded in Java, to calculate doses of intravenous antibiotics: gentamicin, amikacin, vancomycin and teicoplanin. Dosing these commonly prescribed antibiotics is not as straightforward as, say, penicillin because you have to take into account a patients weight, obesity, renal function, the indication of that antibiotic (as in the type of infection it is prescribed for). To do this you have to follow a series of complex calculations, which, frankly, you just do not have time to do if you have a seriously unwell patient in front of you.
However, failure to follow the guidelines risks overdosing the patient, leading to toxic effects, or underdosing, which is not effective in treating the infection. There is a wealth of literature evidence that both over-and underdosing these antibiotics is associated with significant patient mortality.
Many doctors will be familiar with the difficulties prescribing these antibiotics, but our audit aimed to quantify the extent of the problem. Our team gathered all the information on thousands of antibiotics prescribed over a one-year period. Results highlighted widespread prescription errors. Out of over 6000 gentamicin prescriptions, for example, 61% were prescribed incorrectly, with 17% categorised as serious under- or overdoses. There have been similar results in other trusts.
To rectify this problem, our app allows you simply input all the relevant patient information (such as age, weight, sex, renal function) and the calculations and dose adjustments are done automatically. We tested the app on a pilot group in East Surrey and Croydon, which showed significantly faster prescribing and reduced the error rate to 0%.
Not only could the app save Croydon over £5500/year in saving doctors’ time and preventing drug waste, it also has further cost savings associated with saving the pharmacy department time, whilst reducing costs due to toxicity, mortality and length of hospital stay.
What have been some of the challenges with this project?
The problem with being a clinician wishing to develop digital based innovations is that you are working in an inherently high-pressure environment. So, even after identifying a problem, actually creating a solution can be a low-priority task. For our team, this was a tough battle, since patient safety always comes first.
In addition, there are not enough healthcare professionals trained with the necessary skills to develop healthcare technology. For me, after learning to code, I was presented with a whole new range of possibilities in quality improvement and digital health. Many healthcare professionals, unfortunately, do not get the same opportunities.
One aspect of digital health that I am currently learning is the path of actually delivering the project to the trust. NHS commissioning is local and each trust or organisation operates differently with different needs. Some trusts are also more pro-digital innovation than others. Navigating through this route requires experience and connections.
What advice would you give to someone starting a project similar to yours? What advice would you have given yourself four years ago?
Being a doctor is no longer about just treating patients in the traditional sense. Responsibilities now extend to management, research, audit and quality improvement. In fact, an increasing number of doctors are the founders of digital apps and innovations. As more hospitals undergo digitisation, the benefits digital health technology and innovation can offer will expand exponentially.
Being a healthcare professional, you’re always going to be finding problems in the system; on a daily basis you’ll think ‘wouldn’t it be better if we could do it like this’ or ‘wouldn’t it be better if we had an app or a system that could do this’, and at that very moment you’ve identified a potential solution. My best advice, from what I have learned in quality improvement, is to follow up those ideas. Collect data to prove that there is a problem, and then work on the solution. To not address the issue, you’re missing an opportunity to have a seriously positive impact in healthcare!
My next best advice would be to learn to code, a skill that will enable you to develop and scale your solution. Our team has launched a course, called Code Med, which aims to teach healthcare professionals to code in Java. Currently, we are training our third cohort at Croydon but our expanding our course to Queen Mary’s University London students. We hope that our students will be empowered to use these skills to take their digital healthcare technology ideas to the next level.
What’s next for you?
Based on user feedback on the Antibiotic Calculation and Monitoring System at Croydon, we have extended our work to a mobile app (the Infection Guidelines App), which is an interactive resource that allows for quick and efficient access to microbiology guidance, eliminating the need to read through large amounts of text, as is the case with current local guideline resources. It also includes the calculators for gentamicin, amikacin, teicoplanin and vancomycin. A bespoke version has been developed for Croydon Hospital’s antibiotic guidelines and we are in the process of adapting the mobile app for distribution to other trusts.
My experience thus far has been a huge learning curve and, in the longer term, I intend to develop those skills further to build more advanced, practice-changing software for the NHS. In particular, our team is prototyping ideas for applications that address the paperless NHS aim, in line with the five-year forward view.