Dr Matthias Toh, Public Health Physician Consultant for Information Management with the National Healthcare Group (NHG), is a medical doctor who uses data analytics in the ongoing fight to keep our ageing population healthy. He has been involved in the creation of longitudinal medical records and the setting up of a chronic diseases registry which gives doctors an overview of all their patients’ medical histories. Dr Toh talks to iN.SG about the convergence of chronic disease, numbers, IT and medicine, and how data analytics can help measure the quality of care and improve clinical outcomes.

Can you give us a brief background of your career as a medical professional and how you got involved with IT?
I did my specialist training in public health, which looks at disease prevention, health promotion and public health issues on a larger scale. It’s more than just one-to-one medicine. For example, during the SARS and H1N1 outbreaks, public health people came into play. We manage health policies and how to improve healthcare systems. I was also drawn to public health because of epidemiology, which studies the trends and causes of diseases in populations.

I’ve always had a head for numbers so I was very interested in the data analytics aspect of healthcare. IT was the tool that helped us translate patient information into useful data to improve healthcare systems. I never had formal IT training so I must give credit to the rest of my team members, particularly the IT professionals who work silently in the background.

How did you get involved in data analytics? What sparked off your interest in this area?
In 2001, I was posted to NHG, which was doing a lot of work on chronic diseases. We were coordinating efforts across Tan Tock Seng Hospital, National University Hospital, Alexandra Hospital and nine National Healthcare Group Polyclinics. We had patient records, but we needed a system to measure what we were doing. We tried to do it manually, but it was very tedious. We were still doing it on paper, going to the Medical Records Office to pick up case sheets and the relevant clinical data. At that time, the IT systems were not able to stratify our patients very well or to link up all the different chronic diseases to provide a bigger picture. NHG was prepared to build an IT infrastructure to improve care delivery. The timing was perfect. We received funding to build a Diabetes registry. I was asked to work on the project from a clinician’s point of view – to decide what was needed from the medical perspective and translate it into IT language so that we could build the algorithms.

In subsequent phases, we added Hypertension, Stroke, Asthma and Osteoporosis to the chronic diseases registry.

Can you describe some of the healthcare challenges that analytics can help to solve?
Instead of every clinician having to remember in his head the various care components that should be carried out for a patient, the system can prompt the clinician, which helps save time and reduce duplication. The patient doesn’t have to pay for unnecessary tests and procedures, and hopefully we can improve the quality of care for our patients with clinical decision prompts.

Often, clinicians are busy looking after the individual patients, and most of them will not know for certain the overall quality of care they are providing, other than  a “gut feel” about it, because they don’t have the hard data to quantify the quality of care. If they can have access to the data, they will be able to identify areas that can be further improved. This brings about systemic improvements over time.

At the polyclinics, we see more and more people who are in their 80s and above. We are dealing with more conditions than we used to in the past, as the risk factors accumulate. Some people take over 10 types of medication. With the chronic diseases registry, our physicians will be able to trace most of the conditions that these patients have, if they are recorded somewhere in the system.

Tell us more about what you are currently doing with analytics in healthcare.
We are conducting ongoing analysis, and we share the information with clinical work teams so that they can digest it and identify which clinical care processes need further improvement. Right now, we’re working on diabetic foot data. Diabetics are susceptible to foot conditions or amputation due to poor circulation. They have annual foot screenings to address this. We’re trying to port this data into the registry in a more granular way, so the onus won’t be on the patient to tell the doctor what happened during the screening.

At the Institute of Mental Health, patients staying in the hospital’s long term care wards are also aging and besides their mental health conditions also suffer from chronic medical conditions.  Traditionally, IMH would send these patients out to polyclinics to provide medical care whilst IMH focuses on the mental health needs of these patients.  IMH has decided to enhance their care for these patients by bringing in other medical professionals. We are also working with IMH now as part of this effort to take holistic care of all the patient’s medical and mental health needs.

What are some important skillsets that you need to “do” data analytics?
Public health training was essentially what helped me – where the mainstays are epidemiology and statistical analysis. It taught me how to appreciate the epidemiology of diseases and analyse data in a systematic manner.

How do these skillsets complement the domain expertise that you already have, to solve healthcare-related problems through data analytics.
The tools we create are meant for all clinicians. I’ve discovered one unintended benefit though. When I show the patients a prompt on the screen to, say, do an eye test when they thought they had already done one recently, they realise I’m not forcing them to go for an unnecessary test.

When I show them the trends in their condition, it has a visual impact. It helps them to understand what their targets are. It gives them ownership, and helps me with patient engagement during the consultation. You see a patient three or four times a year – the rest of the time, they are taking care of themselves, so you want them to feel empowered.