THE MYTH OF EVIDENCE BASED MEDICINE (IN PRIMARY CARE IN AUSTRALIA)
Let me be clear from the outset: I’m grateful for modern medicine. Without it, myself and many members of my family would be dead.
I’m also grateful for doctors. The majority are very well motivated. Many are excellent at what they do.
With all that said, the more I have worked in healthcare (in Australia), the more concerned I have become about the lack of standardised practice, particularly in primary care. The idea that medicine is practiced based on consistent, recent evidence is, unfortunately, not as true as patients need it to be.
If medicine were practiced based on evidence, we could expect to find a high degree of consistency in the way that different conditions are treated. Sadly, the opposite often tends to be true.
Let me offer some examples…
1) In addition to my day job, I have pleasure of serving on the board of the Hyperemesis Education and Research (HER) Foundation, an organisation that supports women with Hyperemesis Gravidarum (HG), severe nausea and vomiting in pregnancy. As part of that board I get to hear stories from women and their partners from around the world. The variation in treatment protocols in Australia alone is staggering.
I have spoken with women who have had doctors tell them:
“This is just morning sickness. You will get over it.”
“The condition is in your head. Pull yourself together.”
“You just don’t want your baby enough.”
“You can’t take anti-emetic (anti-nausea) drugs, they are not safe in pregnancy.”
“Taking anti-emetic medications in pregnancy is not ideal, but many other women have done it and the evidence currently doesn’t show that it will harm your baby.”
“If you’re sick, you have to go to the Emergency Department.”
“I’m taking you off these anti-emetic medications they gave you in the Emergency Department because you don’t need them.”
“Come in for proactive intravenous hydration as an outpatient whenever you like.”
Now clearly these statements, all made in the last few years, can’t all be true at the same time. They represent the huge range in care for a single condition. How much more then is variation an issue across other conditions?
What concerns me, however, is not just the huge variation in practice and patient experience. It’s the fact that when presented with good practice in the form of evidence-based treatment protocols, these are frequently rejected by health professionals, particularly Australian GPs. I’ve experienced this personally on many occasions as both a patient an expert in the treatment of conditions such as hyperemesis gravidarum.
“You don’t know what you are talking about” and “that’s not the way I do it” have been the most common reactions.
It’s not possible that every health practitioner both does what they want in terms of treatment, and delivers best practice care. There just can’t be that many versions of best practice.
2) Recently I have worked with the board of one of Australia’s leading allied health professions. It’s fair to say that there is a significant amount of concern at board level about the variability in practice.
The worrying reality is that this profession, like most other allied health professions in Australia, collects no data about activity, types of intervention used, variation in practice, clinical outcomes or patient reported outcomes.
Where it goes from worrying to scary is that there is a reluctance to start gathering this data, precisely because of the variability in care that it will almost certainly show.
Once again we have a primary care health profession where practitioners are free to practice with complete discretion, and without any compulsion to follow standardised, best practice treatment protocols.
There is much evidence that shows standardised, consistent practice delivers net better outcomes for patients, regardless of whether that practice represents absolute best practice. We must work hard to create a culture in which standardised practice is not only possible, but the norm. Patients deserve better than the crap-shoot that is today’s primary care in Australia.
Evidence shows that the average time taken for new medical discoveries to be adopted by the majority of clinicians is 17 years. Yes, you read that correctly. 17 years.
The following diagram, courtesy of ePatientDave, shows some more detail:
This further demonstrates that the adoption of clinical evidence is complex, non-linear process. Clinicians are not neutral observers, rationally weighing all available evidence and immediately moving away from treatments that are sub-optimal. Rather, clinicians have complex attachments to existing treatment protocols with rational, emotional, commercial and many other factors playing a part in the decision to change approach.
Evidence-Based?
I work with a number of Primary Health Networks (PHNs, the groups that support GPs in a given region) in Australia, the majority of whom are in the process of implementing “HealthPathways”, a web-based reference resource for GPs, offering standardised treatment approaches for many common conditions.
Let’s overlook the fact that each of the 31 PHNs is independently developing its own pathways (with a little cooperation), driven by a small number of GPs in each region defining what they see as good practice, based on their interpretation of the available evidence.
Whilst useful for reference information, the uptake of HealthPathways has been poor. The key reason for this is arguably that the culturally supported notion of complete autonomy for individual clinical judgement has never been addressed. We still train GPs that, having assessed the evidence for themselves, they are the ultimate authority.
Add to this the idea that the hours required for Continuing Professional Development (CPD) are sufficient to keep up with even a fraction of the available evidence is frankly laughable, even for the most diligent of GPs.
In this environment of autonomy, evidence that counters the way in which a GP may have been treating a particular condition for their whole career can be a real threat. The idea that new evidence is incorporated into practice without a significant emotional attachment to years of prior decision making is unrealistic. In my experience, new evidence is often adopted or rejected through the lens of confirmation bias, i.e. it is very difficult to bring myself to believe that the interventions I have been using for years might have been unhelpful.
In conclusion, whilst we talk about evidence-based medicine, there is still a long way to go in order to realise this reality, particularly in primary care in Australia.
A good start would be to acknowledge that complete autonomy for clinicians in determining treatment protocols delivers an unacceptable level of variation in patient outcomes and experience. It is time for change.
