Interview

Interview Series: Malcolm Pradhan, Chief Medical Officer, Alcidion

In our latest interview in the series we hear from Malcolm Pradhan from Alcidion, where we explore decision support and utilising data, automated workflows and alerts to save clinicians time.

Can you tell me about yourself and your organisation?

My background was originally in medicine, in Adelaide, I went to medical school and discovered I was interested in emerging computer technology; that was in the early to mid-80s. When working in hospitals I was struck at how many things could be improved with technology.

I started to get interested in decision support and what was called AI back then. I decided to study the field of medical informatics and I applied to do a PhD programme at Stanford in California. At Stanford I studied medical AI, and how computers can help clinicians make decisions under uncertainty whilst taking into account patient preferences and the complexity of clinical decision making.

I then came back to Australia after the PhD and went into academia. I began consulting in the area of patient safety. In the 90s, there was a growing movement around understanding preventable errors in healthcare that cause patient harm. As I was working in patient safety and the Australian Patient Safety Foundation, I realised that a lot of the problems were very similar; I thought computer technology could really help in making healthcare safer and more efficient. So myself and a colleague, Ray Blight who had been head of a government health sector, co-founded Alcidion in the early 2000s to improve patient safety. Unfortunately we didn’t realise at the time that funding models meant hospitals were reimbursed for errors! In fact that was the case until quite recently. Fortunately financial drivers have changed around the world to support patient safety — hospitals are no longer rewarded for errors.

Our first product was a clinical system for emergency departments and other high pressure areas to help doctors and nurses manage clinical results and respond to emerging patient risks in real-time. From the beginning we focused a lot on user interfaces, visualisation and on clinical usability. Our mission hasn’t changed, we still focus on making healthcare safe and improving efficiency for clinicians by making smart technology that clinicians want to use. We listed on the ASX stock exchange back in 2016. Then in 2018 we acquired MKM Health and Patientrack.

Could you tell me about the system?

What we’ve created is a platform that takes in data from all sorts of different systems in hospitals in real-time, it converts all this data into an open standards format called FHIR (Fast Healthcare Interoperability Resources). The platform has a decision support engine that monitors each patient’s data to detect and highlight emerging risks to clinicians in real-time so they can act to mitigate problems for the patient.

Strangely, real time decision support is a pretty new idea in healthcare. Most current health IT products are essentially big database systems that store and retrieve data; they require doctors and nurses to remember what’s happening for each patient, and then manually look up data in the IT systems. The healthcare industry expects doctors and nurses to perform heroics every day just to get their job done! We believe IT systems should play a more active role in helping clinicians to look after their patients. Smart technology should help with the memory tasks so they can spend more time on the difficult judgement stuff.

One of the reasons real-time decision support isn’t commonplace in healthcare is that it requires a different technology approach than the traditional database system. To actually monitor for patient safety problems and run predictive algorithms in real-time the IT system has to be designed from the ground up to stream data and safely run algorithms. The system we have built, is designed to do this.

For example, let’s say a patient gets a lab test result that shows their kidney function is worsening. In response to this new lab result could trigger many algorithms to help clinicians manage this problem. There are algorithms to work out what might be causing the problem based on the patient’s conditions and medications, some to calculate risk to work out how urgently the clinical team need to respond to the new problem, some researchers have developed predictive algorithms to work out further risks for the patient, and other algorithms recalculate how long the patient may stay in hospital. The system can run algorithms developed by hospitals and researchers, not just from Alcidion.

Our view is that the sustainability of healthcare systems relies on hundreds, if not thousands, of algorithms monitoring patients in the hospital and in the community to make sure that clinicians can focus on those with the greatest need.

There’s a lot of talk about ‘AI’ and predictive algorithms, but before that there are so many problems in healthcare that can be addressed using real-time decision support. Some errors and delays are so common everyone thinks they are just part of healthcare. We believe that many of these problems can be avoided or reduced with smart technology. Our challenge to our customers and potential customers is ‘what do you want to engineer out of the healthcare system?’ ‘What problems do you want to remove?’ And not regard them as just part of the healthcare system.

Where is the software being used?

The new version of the system is being used in Australia, in New South Wales and Australian Capital Territory, it is being used in New Zealand, and we are getting ready to install it in the UK at Dartford and Gravesham NHS Trust.