From dog sledding in Norway to technology-related errors, the director of the Centre for Health Systems and Safety Research at the Australian Institute of Health Innovation is always on the go.
Recently inducted as a fellow of the Australian Academy of Health and Medical Sciences, Macquarie University’s Professor Johanna Westbrook is riding the crest of a wave in her already storied career.
HSD editor Cate Swannell sat down with her to talk all things health systems and safety research, as well as her youthful ancient history nerdiness and dog sledding in Norway.
Congratulations on your induction.
Thank you. It is an added bonus when you get acknowledged by your peers. To be in the realm of such esteemed colleagues, it does make you sit up and think ‘oh, maybe I’m not a bad researcher, after all’.
Never in doubt, surely?
There’s something about being a researcher, under scrutiny. You’re always constantly trying to improve and so, yes, it’s always nice to be reassured that you’re doing a good job.
What are you currently working on?
We’re doing a lot of really exciting things in evaluating large-scale clinical information systems.
I’ve got a particular interest in medication safety which we’ve recently finished a big study in, so that was very exciting.
We did the first randomised trial of a large-scale IT system in the world to demonstrate whether the system made a difference to prescribing errors. We looked short- and long-term and found some really interesting results.
In the short term, things didn’t improve and some medication errors actually increased. Twelve months later, things did improve, but not everything.
Then we got to dive deep into the data that we collected and [we found] technology-related errors, which are new a new source of prescribing errors that don’t occur in paper-based systems. Now our goal is really to understand what are the mechanisms that drive those types of errors and what are the implications for changing systems.
We’re turning that into one-page bulletins in a new series we call the Health Innovation Series, to actually let people in the system know there are things you can do to optimise your electronic medication systems, to reduce some of these errors.
For me that’s really gratifying – going on that whole long journey that you go on with research, from doing a randomised clinical trial to then digging deep, and then trying to work out how do we get this out, and really make a difference.
Technology-related errors – are we talking about software problems, input problems?
It can be the way the system is designed, or the way people use the technology. For example, it can be people making incorrect selections from dropdown menus, which you obviously can’t do with paper.
But it’s also understanding things like how people make decisions and the way in which the architecture of a system will lead people down a certain path. We know from behavioural economics that what’s on the top is most likely what people choose.
If you have drugs arranged in alphabetical order, or if you have, for example, IV at the top, and people aren’t quite sure what to do, they’ll see that as the default and so they’ll think they might just leave it there.
So there are those types of errors. But then there are more detailed things, like these systems often have those calculators which are great for helping people, particularly in paediatrics, to calculate a dose, but these calculators also have rules within them. For example, they have rounding rules, where you either round up or down automatically.
That’s often because tablets come in a certain size. Now, most of the time, that’s fine, but with very young children who might have very low doses of things, going up or rounding down may result in a significant under or overdose. As a user, if you’re not attuned to this sort of backend rule going on, you may not check it as closely.
So we’re trying to highlight those things to get organisations to check, for example, what are the rounding rules and do you have a lot of kids in your hospital and might this be a problem.
Is it a software problem or a user problem?
There are often combinations of these things. It’s never possible to design a perfect system for every type of individuality. It’s a combination of using design and understanding how people can make errors.
You go to a vendor and they’ll show you this perfect system and you’re assessing it and it looks great. But then the systems go into complex work environments, where users are being interrupted, and the system has to potentially integrate with legacy systems that don’t do that very well.
And that’s really where our research comes in. We’re not interested in assessing the perfect system. We’re actually interested in what happens when you put it in the messy world. And then how do we design systems that are adaptive and can support clinical workflows as they are in the real world?
When you walk into a hospital, what do you see? What are you looking for?
That’s a big question! There’s this term – work as imagined and work as done. We imagine how work is done in this nice linear process. Often, clinical systems are designed for this linear process and everything gets done step by step.
And then you walk into that environment and you realise that’s not the case. There are all these interruptions that people work around because the system doesn’t quite do what you want it to do. You’ve got competing demands.
Part of what I’m interested in is the impact of interruptions on clinical work. There are clearly relationships between how often clinicians are interrupted and how many errors they make.
What keeps you awake at night?
Wondering where the next grant is coming from! It is getting more and more competitive.
My concern is also for the next generation of researchers. How do I get them up to the level where they are competitive? What opportunities can we create for them to make sure they can continue this great work?
The other thing [I worry about] is getting the results out as fast as possible – doing rigorous research and trying to make sure that we get it through the scientific process and the peer review process, which seems to be taking longer and longer.
We’re not only expected to do a lot of robust science stuff, but also, how do we ensure that our results make a difference? And that’s fantastic, but it’s also quite challenging.
You’re also renowned for your research into unprofessional behaviour in medicine. Are there any low-hanging fruit we can pick to make things better for junior doctors, for example?
What the available evidence would suggest is that if there are mechanisms by which they can report their experiences, and there are then mechanisms by which that information is taken forward and fed back [to the perpetrator] in a non-judgmental way – there is evidence the majority of people will change their behaviours in response to feedback.
But we’re not very good at giving people feedback about unprofessional behaviours in a non-judgmental way.
A lot of the professional accountability programs are based on this idea. You can report it through a system. A peer messenger has an informal cup-of-coffee conversation with the perpetrator, saying, “You may not have realised this behaviour was interpreted this way,” to get them to reflect.
None of this is part of the HR process. It’s not recorded. It’s not formal in that sense. Then if you get persistent sort of behaviours, the idea is that then somebody with this individual says, “What’s going on?”
Only after that process, then you can get stuck into those disciplinary type of processes where usually no one benefits.
So, there is some evidence that those sort of structured programs with sufficient resources and commitment can work. If you have this sort of reporting system available, you’d be sending a message across the organisation that we think this is important — we are going to act if you report these types of things.
At St. Vincent’s we found staff were using over 2500 reports, but they could also report positive behaviours, and 50% of the reports [were] positive things about other staff members.
We underplay the value of also giving positive feedback to people. It’s about people feeling valued and so when [they are asked to do extra things] they are much more willing to do them.
Wildcard question: what’s the most fun you’ve ever had?
I loved ancient history when I was at school, that was another passion, so the first time I went to Egypt. And back then you could still go down into the tombs in the pyramids. To actually be there, to see all the colours as they were when they were first applied, that was amazing.
I also went dog sledding in Norway in winter. Absolutely incredible. If you have never tried it, you must put it on your bucket list.