This month we talk to DHIN member and Senior Research Fellow, Dr Aldo Saavedra.
Can you please tell us about a day in your life?
A typical day starts early with a nice bowl of porridge or fruit salad after either a run, watering the garden or bowling at the cricket nets. I am currently a Senior Research Fellow with the Faculty of Health Sciences at the University of Sydney. My work deals with the application of data analytics on routinely collected health data as well as developing a prototype data sharing platform with eHealth NSW to identify the barriers and gains to sharing de-identified data across health districts in NSW. I’m mainly based at Sydney University with significant amount of time spent at teaching hospitals and institutes such as the Adult and Children’s Hospitals at Westmead, Westmead Institute of Medical Research (WIMR) and eHealth NSW.
I have worked in health for the last two years. During that time, I have had the great opportunity to learn about the health system in Australia, and meet passionate professionals by undertaking a number of projects to inform clinical outcomes and processes. The projects have shown me the commonality of data across districts and the barriers, social and technical, that negatively impacts the creation of an ecosystem to foster a common vision and cooperation across universities, institutes, LHDs and government.
My original background is in particle physics where I was a founding member of an international collaboration called ATLAS. The collaboration measured proton on proton collisions at the Large Hadron Collider. In the collisions, fundamental particles thought to have existed at the beginning of the universe should be created if they exist. To discover the Higgs boson, collisions with signatures compatible with the Higgs boson were analysed. We synthesised data to understand what the signature would look like in our study and determine what other sources, like other particles already discovered, could possibly fake such a signature. These were extensively studied to either suppress them or take them into account in the experimental data.
One of the major challenges for me, as I changed fields, was that systematic approach I was used to applying had to be more flexible. In health the ground truth for an analysis is difficult to define. It is hard to select consistent variables that will elucidate a relationship between factors and outcomes. Apart from having a limited ability to control for factors, relationships may be temporal and the fields collected to inform such an outcome may hold very little information. It is an interesting and worthwhile challenge.
What does digital health mean to you and your work?
Digital health to me means the effective application of the various aspects of digital technology within health care. The digitisation of information drastically lowers the barrier to the manipulation of data. It is now feasible to link disparate sources of data and inform any aspect of the health with the ability of feedback to health professionals and patients.
Even though it is feasible, health represents a large ecosystem that is very diverse. The lack of interoperability and standards between systems is a great barrier for any work. It stifles the uptake of new relevant technology and affects any downstream use of data. From uncovering an insight that can be applied in a support decision system to studying the impact how such information can have on clinical practice and patient outcome.
For my work, the promise is there but a lot of work needs to take place before sophisticated algorithms can be confidently applied on data and that people accept them for routine use.
What advice would you have for someone with an interest in digital health? Any useful resources or networks they should know about?
My advice for someone with a data analytics background with an interest in digital health is to first think:
- At what level do you want your impact to be felt? For example:
- Do you want optimise or revolutionise the operation of a medical unit or a hospital?
- Do you want to improve patient outcome or gain a better understanding of contributing factors towards the prevention/identification/treatment of a condition?
- Do you want to identify and enhance how information is presented and used by clinical staff in their day-to-day decisions?
- Do you want to shape policy?
- Think of what you would like to do:
- Develop tools with intuitive interfaces that clinicians love
- Help health experts and data analytics experts talk to each other
- Develop statistical and machine learning algorithms to answer a particular question
- Contribute to the development of a data architecture and standards that facilitates the use of data.
My second bit of advice is to be resilient and make sure you talk to people and do your homework. Reading journals, blogs and attending public events such as meet-ups, HISA meetings, DHIN events, etc. to get an idea of what the challenges, current approaches and with whom and in which area you would like to work in.
Thank you to Aldo for his time.