Brendan’s first video gave an insight into our research and the fallibility of clinical trials due to problems such as confounding (Clarke et al, 2013). Now, we delve deeper to further explore the reasons why trials cannot provide all the answers and how mechanisms can help.
Problems with ‘Clean’ Test Populations
Clinical trials are often carried out on ‘clean’ test populations. Whilst this is efficient in various ways by helping to make trial evidence universally applicable, it also presents a number of issues. A common criticism of population testing is that it houses inherent biases regarding factors such as age, gender and ethnic background. Somewhat neglected, however, is the failure of trial populations to account for differences at an individual-level.
What’s the point in testing a drug on an ‘ideal’ population, if it is then wrongly prescribed to an individual patient? After all, they may share very little physiological characteristics with that given trial population. We may view this is a large-scale extrapolation problem, whereby the process of subsequently treating an individual patient based on a population-level trial misses the point – by mistreating the patient.
People are diverse
As Trisha Greenhalgh recently stated in her comment on the EBM Manifesto, “the world is messy” and as Phyllis similarly points out, “people are diverse” – in ‘real-life’, from the biochemical level to the social level, there are numerous complex factors that make us different. We have different genetic information, different social backgrounds, different drug interactions and thus, based on these mechanisms, we all have slightly different needs.
Moreover, the video’s example of ACE inhibitors works well to demonstrate that specific patient preferences are hugely important when prescribing drugs to them, as this decision can detrimentally impact their very livelihood. Interestingly, we are slowly uncovering the mechanism of action to understand how antihypertensives induce coughing.
Using Evidence of Mechanisms
So how do we account for individual needs? Take a guess… With evidence of mechanisms.
We must support statistical evidence from population trials with mechanistic understandings of how and why a particular drug or treatment works the way it does – i.e. integrating population-level evidence with individual-level mechanisms and physiological demands. By ‘adding’ mechanisms, we can better accommodate individual patient preferences so that clinicians can make better healthcare decisions as a result.
As a last side note, this might sound like an advert for personalised medicine or genomics, but EBM+ certainly has much broader scope. For instance, whilst genome research may help to discover how we meet the high demands of individual patient care, myriad contributing factors – such as a person’s diet, pre-existing conditions or occupation – are also important but often ignored.
Therefore, EBM+ urges researchers and doctors to see the bigger picture – to think globally but act locally – so that EBM realises the needs of individual patients.