Using digital solutions to target interventions helping prevent hospital falls

By Rex Parsons, AusHSI PhD Scholar

AusHSI PhD student Rex Parsons

I’m writing this blog post days before submitting my PhD thesis for external review, and having already spent my last day in the office as a PhD student. This is probably the point in the PhD timeline where most are having extreme feelings of relief or are pushing through to the very end. I don’t think I have these feelings. Instead it all seems quite low-key.

I have already done the hard work leading up to my final seminar in late 2023 and have been working in a post-PhD job for more than 6 months (which has successfully distracted me from attaching too much of my self-worth into the 200+ page document that I’m about to submit).

My PhD journey took a bit of a turn from what was described in my earlier blog post. This was a result of a potentially risky situation where my planned PhD studies were relying on data that could only be accessed by our hospital partners, and the process of data acquisition was taking longer than anticipated. I was feeling the pressure to go off script to ensure that I made some progress and kept to schedule, even though Susanna (my primary supervisor) said something to effect of, “you ride motorbikes, since when are you risk-averse?”.

In hindsight, this was very fortunate, as it forced me to generate interesting ideas for studies to complete in the meantime. Trial and error uncovered a couple of dud ideas, but also found one that stuck. I ended up doing a simulation study to evaluate a proposed method that could possibly align clinical prediction models towards value-based care, which aims to achieve better patient outcomes at lower costs to the healthcare payer. Decision support systems are often used to optimise assignment of interventions to patients to reduce unnecessary treatments, focusing resources on patients that are likely to see the most benefit from the intervention (and not giving them to patients that are at low risk).

The simulation study was a success and snowballed into another idea, to make some software. The software replicates the simulation approach and could be interpreted to estimate whether a hypothetical model would be worthwhile. These studies led me to discover an enjoyment in the software development side of statistical programming and contributed to landing a post-PhD job as a Data Scientist.

I’m happy with how my PhD journey turned out. Going off script allowed me to pursue things that I found interesting and gain skills that I wanted. A big part of my PhD’s success was a great supervisory team and colleagues at AusHSI who support me going off script (even if I did change my mind during the PhD itself).

As a partnering member of the Digital Health Cooperative Research Centre, AusHSI manages several research projects in the area of digital disruption in health environments. Hear Rex talk about his PhD work with AusHSI and the DHCRC with Peter Birch in this Talking HealthTech episode (44:03 min).