One thing we often hear, especially regarding health and health care decisions, is ‘question everything’, and ‘do your own research’. Of course it’s reasonable to question health care options, but do you know what question to ask, do you know how to find out the answer, and do you know how to interpret the answer that you find?
Often we are looking for a simplified answer, but realistically, as knowledge and expertise increases, we realise that for most questions of depth, science does not simply provide us with a ‘yes or no’, but the answer is far more nuanced than that.
Evidence based medicine is the process of learning from science; the process of systematically reviewing, appraising and using clinical research to ensure that the best possible clinical care is delivered to patients.
We look to evidence based medicine to provide us with the best answers available, in order that we can practice medicine, where possible, according to a principle of treating where there is evidence of benefit, and not treating if there is evidence of harm.
The interpretation of science using evidence based medicine is a speciality in its own right. Asking the right questions, looking for answers, and summarising the evidence in medicine takes training and expertise. The skills required include having an understanding of the tenets of evidence based medicine, and being able to use this as a tool. In acquiring these skills one learns not only how to weigh up evidence, but also that subtle but vital nuances can be lost without an adequate depth of knowledge.
Oversimplification leads to error, and without training, guidance or expertise, oversimplification (and therefore error) is almost inevitable.
The first thing needed when exploring medical evidence is to identify the specific question you are trying to answer, this can be more complex and involved that initially thought, and requires a degree of background knowledge in its own right. From there, you need to look at all the available published studies, to judge their quality and reliability, and to synthesise and summarise what may appear to be conflicting information.
You also need to recognise your own biases and pre-existing opinions on a subject when approaching the question, and you need to understand that there is a hierarchy of evidence that can give an initial guide to how important or relevant a particular study is.
Here is one example of an approach to evidence based medicine, and how it can be used to formulate clinical recommendations as outlined by the NHMRC (Natonal Health and Medical Research Council)
Determine the evidence base for the subject. Identify all available published material on the subject, and then judge this evidence base on its strength, size of effect, and relevance.
Five separate components of the evidence base should be assessed;
- A score known as the ‘level of evidence’ (see later), the number of studies, their size, power, and quality, and any risk of bias.
- The consistency of the study results.
- The potential clinical impact of the proposed recommendation.
- The ability to generalise the body of evidence to the target population.
- The applicability of the body of evidence to the specific healthcare context.
The first two components give a picture of the internal validity of the study data, whilst the last two components consider external factors that may influence how generalizable and applicable the research is to a different demographic or health care context.
The level of evidence is a score given to each study to help with the process of questioning the reliability and internal validity of the evidence.
This is a hierarchy, with the most reliable, reproducible, and likely to be correct given the greatest weight when making decisions.
A pyramid can represent the level of evidence hierarchy, with the most reliable forms of evidence at the top.
If a systematic review is included in the evidence base, then each study included in that review must also be rated according to strength, size and relevance.
Each of the five separate components of the evidence base can then rated (given a score) according to a body of evidence matrix as illustrated below.
All included studies and their component scores are then summarised and synthesised. An example of how a summary is generated for the evidence supporting hand hygiene and the prevention of infection can be found here.
This generates the most accurate answer for the initial question, but even within that, the certainty of the answer can be graded according to how strong the evidence base has been.
Answering questions in medicine, and reviewing the research, is not as simple as hopping onto the internet and reading the abstracts and conclusions of a select few papers.
Answering questions in medicine, and determining what the best evidence is, and knowing how to use this to guide clinical decisions takes expertise and specialist knowledge.
So yes, ‘question everything’. But part of questioning means also accepting that you may not have the skills or experience to be able to find the most accurate answer yourself, and accepting that the experts spend their lives gaining the skills so that they can guide us, and accepting that when the experts say that there is overwhelming evidence for or against something, that is not a conclusion that they come to lightly.