General Notes on Evidence-based Veterinary Medicine (EBVM)

These notes were written following the 1st International EBVM Conference (2014) and a good overview blog post of that event entitled EBVM 2014: Building a Community to Advance Evidence-based Veterinary Medicine can be found here at The SkeptVet website.

Evidence-based veterinary medicine (EBVM) must impact on clinical practice, patient care and outcomes and it is important to remember that ‘statistical significance’ is definitely not the same as ‘clinically significant’. EBVM cuts across all disciplines rather than being isolated to individual disciplines. It is a misunderstanding to label evidence-based medicine as ‘algorithmic’ or ‘robotic’. Individualised decision-making still remains very central with clinical decisions made on the basis of not just the best available evidence, but the experience and reasoning of the clinician and patient-related factors including those relating to the pet’s carers. There is a need to eliminate the fear of uncertainty amongst some practitioners while continuing to emphasise the importance of clinical experience.

Some major issues for consideration:

Evidence

There remains a paucity of adequate evidence on which to build EBVM and the limitations of veterinary versus human medicine are implicit in terms of what type of evidence we will ever have. One of the limitations of having much less evidence available for example is our relative inability to account for comorbidities which can result is heterogeneity within study populations. There are also not many placebo-controlled studies and it is hard to know the true significance of the ‘placebo effect’ in veterinary medicine.

When considering the utility of evidence it is important to ask ‘who is the study about’ and therefore can it reliably be extrapolated to my patient. Furthermore a lot of studies use surrogate markers of outcome but clinically relevant outcomes with impact on patients is what is needed. It is also important to remember that association is not the same as causation, i.e. if a patient is started on a treatment and gets better, the treatment and the recovery are associated but the treatment did not necessarily cause the recovery; there are likely many examples of scenarios where association has incorrectly been interpreted as causation.

That said the availability of data of variably higher quality from an ‘evidence critique’ point-of-view is increasing. Going forward it is essential to not just identify but to also prioritise knowledge gaps to allow research to be focused.

The term ‘evidence’ has been hijacked by some operators claiming to have evidence that when scrutinised falls far short of the expected standard. This reminds me of a blog post “Clinically proven – what should it mean?”

More publication of negative studies is needed – but often these are not published due to stakeholder considerations and conflicts of interest.

Accessibility

The aspiration going forward is for EBVM to become part of every clinical staff member’s routine thought processes BUT this can only work if EBVM can be taken into the settings where clinical decisions are made. Time-related barriers need to be overcome and access to evidence made possible in a practical and user-friendly way. This involves multiple interrelated considerations such as:

  • Adequate and appropriate technology
  • Buy-in by ‘Management’ even at an individual practice level
  • Training of ‘evidence seekers’ with respect to principles of EBVM and how to acquire evidence; it is also key to ask evidence seekers in what format they think evidence is best accessed/consumed.
  • Access to evidence (e.g. some veterinary societies are offering access to multiple journals as part of their membership; the Journal of Veterinary Internal Medicine has made its Reviews and Consensus Statements freely available).*

(*Further useful information can be found at these links:

http://www.vetfoamed.com/ebvm-critical-thinking-resources/
http://www.vetfoamed.com/multispecies/)

Implementation

There remains wide variation with respect to the implementation of EBVM. This is likely to be due to several reasons including lack of education about EBVM but also lack of accessibility to evidence in a practical and clinically feasible way. At first glance EBVM may seem like a practice that would add to time and work stress; however with the right tools available it can actually lead to reduced time and work stress by streamlining and simplifying the decision-making process.

The EBVM process

The EBVM process can be thought of as five steps and when it comes to patient care this process should both start and finish in practice:

Ask a specific question (e.g. using PICO format)
Acquire evidence
Appraise the evidence: the evidence should be appraised both for its value (strong-weak, ‘good’-‘bad’) in general terms and in the individual’s circumstances. Appraisal should lead to a conclusion being drawn and the level of confidence in that conclusion should be identified.
Apply the evidence to the clinical scenario
Assess the clinical outcome

What does having evidence achieve in practice?

(1) Allows us to identify practices that are harming our patients and generally provide better patient care; also allows us to have better communications with pet carers and get more informed consent.

(2) Question ‘experts’: the opinion and experience of experts fits in to the evidence pyramid but whether or not it counts for more than the available evidence depends on the quality of the evidence! Ideally experts will be abreast of the evidence and also be practicing EBVM.

(3) Resist drug reps: their presentation of the ‘evidence’ can sometimes be less than ideal and even range from disingenuous to deceitful – not all drugs reps of course!

(4) Choose ‘better’ products and services based on the evidence.

Hierarchy of evidence

One could say that at the moment the current ‘hierarchy of evidence’ by which many clinical staff operate is as follows (Brennan McKenzie):

My opinion (highest)
Expert opinion
Synthetic literature (systematic reviews, good EBVM guidelines, critically appraised topics (CATs))
Primary literature (randomised controlled trials (RCTs), human studies, case reports, pre-clinical data) (lowest)

It should ideally be:

Synthetic literature (systematic reviews, good EBVM guidelines, critically appraised topics (CATs))(highest)
Primary literature (randomised controlled trials (RCTs), human studies, case reports, pre-clinical data)
Expert opinion
My opinion (lowest)

What is the role of the general practitioner?

Be informed about the evidence
Think critically about evidence and uncertainty
Be explicit with clients and colleagues about evidence and uncertainty
Synthesise evidence
Talk to Academia about what you want to know
Talk to Industry about what you need