Emeritus Professor Paul Glasziou AO sat down with Dermatology Republic for a chat about non-drug interventions.
Evidence-based medicine expert and GP Professor Paul Glasziou is technically semi-retired. Which means, of course, that he’s just as busy as ever.
He sat down with Dermatology Republic to talk about one of the projects that he is continuing work on – the RACGP’s Handbook of Non-Drug Interventions (HANDI) resource.
Dermatology Republic: we hadn’t really heard of HANDI before it was mentioned to us earlier in the week. We’ve had a flip through, but what exactly is it?
Professor Glasziou: It’s a compendium of evidence-based non-drug treatments.
And when I say evidence based, what I mean is that we look for at least a couple of randomised trials or a systematic review that supports the evidence for it.
But importantly, it provides the how-to details [of these interventions].
To design it we looked at how pharmacopoeias – drug compendiums – are structured, and we made it fairly similar to that so it felt familiar to GPs, but also contained the sort of information that they would need.
When I was a GP, no such thing existed.
It was one of my dreams to be able to make prescribing non-drug interventions as easy as prescribing a drug intervention, so that you could look up which were the evidence-based ones.
There are hundreds or thousands out there, very few of which actually have a solid evidence base.
We wanted to find the evidence-based ones and then make it very easy to have them in one place, but also to make it easy to know when and how you prescribe those non-drug interventions.
DR: I’d be interested to hear how you select what to put in there – it really seems to run the gamut. I saw wet hair combing for head lice in there, as well as ginger for migraine.
Professor Glasziou: There’s several ways that we scope out potential interventions.
The committee members include several GPs, but also a dietitian, two physiotherapists and occupational therapists, et cetera. And we’re always keeping an eye on the literature.
We each have different means – I use a thing called ACCESSSS from McMaster PLUS, which does keep track of the literature and sends me information. I go through and go, ‘oh, that one looks good, we might consider that for HANDI’, we then put them up to the committee to decide whether this would actually be appropriate for HANDI.
To be appropriate for HANDI, it has to be something that a general practitioner would use.
As you can imagine, there are lots of non-drug interventions that, for example, a physiotherapist or an oncologist might use, but that aren’t something that a GP would be prescribing.
Our focus is really, ‘is this something that a GP is going to see on a regular enough basis that it’s actually something we should put into the Handbook of Non-Drug Interventions?’
There are no strict rules.
We often discuss things that are at the borderline. And one of our principles is, ‘is it really a non-drug intervention?’
The dividing line there is if you would find it in a pharmacopoeia or in the Australian Medicines Handbook, then it’s a drug.
Something that would therefore qualify for HANDI, to give you an example, is Omega-3s for antenatal care, which have been shown to reduce certain outcomes for higher risk women.
We discussed, well, Omega-3s sound a bit like a drug … but because it’s not in the Australian Medicines Handbook for that, then how else are GPs going to know about it?
We tend to favour things like simple procedures, exercises, simple devices like the otovent for treating glue ear, for example.
And they’ve got to be shown to be effective. To give you one example, we included exercise for cancer fatigue, because that’s something that GPs should know about to warn their patients about when they’ve got cancer.
DR: More broadly, you’ve been working in the evidence-based medicine field for a long time. Did HANDI come out of your research in that area?
Professor Glasziou: I’ve been in evidence-based medicine since before the term evidence-based medicine was coined, which goes back to about 1992.
I was always interested in the bridge between evidence and research and practice, but [Canadian physician] Dr David Sackett really inspired me to start working heavily in that area.
I eventually retrained as a GP with a central idea of trying to work out how, in general practice, you might be able to use the research evidence that we have for better taking care of patients.
One of Dave’s principles was to never tell people to do something that you don’t do yourself.
He would always try it out himself and say, “find your own way to do this, but here’s how I am doing this”.
My principle was that I needed to be able to actually do this in my practice first, and then I could tell other people how I’m doing it.
And that’s where the non-drug interventions idea, in a way, was born, because I was editing the Journal of Evidence-Based Medicine, which scanned the world literature to try and find that small number of really important studies that would change practice.
What I noticed was that I would do the drug changes, but that, though I got interested in the non-pharmaceutical changes when the evidence came out, I wasn’t implementing them.
I wondered what the barrier was, and I discovered that it was often the fact that there wasn’t enough detail on the how.
There are so many more barriers to implementing non-drug interventions than there are to implementing drug interventions.
Even I, who was publishing stuff on this is in the evidence-based medicine journal, was not doing it in my practice, so I set about to solve that problem.
The two barriers were getting an adequate description of the intervention, which was often not in the publication and having the evidence-based ones compiled in one place.
DR: What are some examples where the intervention wasn’t completely described or wasn’t standardised?
Professor Glasziou: Those are two separate issues, but there were lots of ones where it wasn’t completely described.
In fact, Professor Tammy Hoffman and I did an analysis of non-drug interventions in the six major general medical journals over a year and found that less than 50% of them had an adequate description.
We looked on their websites and in appendices that they provided and in supplementary materials, and often we couldn’t find adequate descriptions of the interventions.
To give you a couple of examples, one was the reduction in salt.
There was a long-term follow-up study, the TOHP study, which published its long-term results showing that salt reduction reduced patients’ cardiovascular outcomes in a randomised control trial.
But how do you get patients to reduce salt? It’s really hard, because most of it’s not table salt, and in the description, it just said that they had weekly sessions with a nutritionist.
What does that mean? How long were they? What materials were used?
Eventually we found another publication which described the sessions, but even that didn’t give the materials, it just gave a bit more detail about what those sessions entailed, and it was clearly not practical.
We’ve never put that into HANDI, because we’ve never found a practical version.
We are putting in one now on salt substitution, which is a much easier way to do it.
You use what’s called low-sodium salt, which is a mixture of potassium chloride and sodium chloride.
It tastes pretty similar, but it means it reduces the bad salt, which is sodium chloride, and increases potassium chloride, which actually has a blood pressure-lowering effect.
The George Institute in Sydney has done this large-scale trial in China where they randomised villages to be supplied with the low salt, salt or normal salt, and it showed – I’ve forgotten the exact reduction – like a 25% reduction in stroke and cardiovascular events.
We’ve done a meta-analysis of that, gone “actually, this is fantastic, this is something every GP should know about”, and it’s much simpler to prescribe using low-salt salt to patients with high blood pressure than it is trying to get them to reduce their salt.
DR: And do you go by Dr Sackett’s principle and use it yourself?
Professor Glasziou: Oh yes!
DR: What about the interventions that are hard to standardise?
Professor Glasziou: That’s harder. What we try and do is find some sort of average or range of things.
For some interventions it’s a big discussion about how we should actually describe it in HANDI.
For example, one would be activity for low back pain.
We’ve found various Pilates exercise program, walking programs, etc, which show improvement. What we give people is a range of the options, and sometimes that’s a good thing to have several options.
We give them links to different programs that have been found to be effective, with the caveat that there’s this trade-off between having a really rigorous program [and having a practical program].
You’re looking for the trade-off of something that’s relatively simple and prescribable for patients, versus something that’s effective.
There’s a sort of sweet spot or an optimal spot that you’ve got to find of enough that it’s effective, but not so much that nobody can do it.
DR: So, it’s harder than it might seem?
Professor Glasziou: Some items are really easy to describe, like combing for head lice.
But even that has got its complications, because you want to use three different combs, and you need to put conditioner in to make it easier to pull the combs through.
You don’t just go and buy a comb and comb your kids hair, you’ve actually got to get the progressive combs and use conditioner at the same time to do it. But that’s pretty easy to describe.
There are other ones, like the low back pain where we often take a couple of meetings to really sort out exactly how we’re going to do it, but that’s why we’ve got physiotherapists and nutritionists on our panel.
We try and call on colleagues who understand the area better than us or write to the trialists – we do all sorts of things to be able to get a better description of the intervention.
DR: What other projects are you keeping up with?
Professor Glasziou: I’ve worked a lot in the area of overdiagnosis – that is, conditions that are asymptomatic or don’t cause a problem for people in their lifetime.
There are lots of those, and I see it as one of the growing problems in medicine, with the push for early detection and a push to change definitions of disease.
We’ve got a big, worldwide collaboration now working on that, but a specific and important component to that is creating rules on what is and isn’t a disease.
We produced a checklist about 10 years ago. We’ve now got a plan to get an international consensus on revising that checklist that will include a number of important organisations, hopefully including WHO, that hopefully then push that the use of that checklist so that we prevent extending definitions of diseases unnecessarily.
And the most recent example of that in Australia is the winding back of the definition of gestational diabetes, where women with gestational diabetes had been being diagnosed, but with the trial showing no benefit for them.
The major organisations got together last year to wind back that definition, to pull back the threshold, so that it’ll probably be at least 25% less women being diagnosed with gestational diabetes.
Even though it seems like a small thing, moving the definition from A to B, it’s got such a big implication at the population level.
I see that as the single most important thing that I could help facilitate a better set of rules about internationally.
This interview has been edited for length and clarity.
