The gold standard fallacy

Almost like clockwork, once we finish medical school, we doctors are asked by those around us “what is the best treatment for…[insert obscure and usually embarrassing personal condition here] ”.

Initially, I responded with the clinical guidelines that I had so meticulously memorised during medical school.

But over time, I realised that the best treatment for a certain condition or symptom wasn’t always as straightforward as looking up a guideline. What my medical training had actually (somewhat) prepared me to navigate was a grey area of labyrinth-esque decision trees and caveats, all qualified with a range of confidence intervals, exceptions and side-effects.

The thing is, the science that underpins what we know as ‘the best treatments’, is always changing. Not only that, but the ways in which science is changing, are changing.

To top it off, the more we understand just how unique each of us truly are, the more customised we expect our treatment plans to become.

Let me explain.

‍

Science is not dogma.

Science, by definition, is an evolving body of knowledge.

The guidelines that I memorised were based on science; evidence-based treatments, collated and moderated by peer-review and held to the highest level of scrutiny before being deemed safe enough to be labelled a recommendation. These guidelines are taught as lore and become the gold-standard upon which many well-meaning healthcare providers rely to make safe and helpful decisions (1).

Science, by definition, is an evolving body of knowledge(2). While inspiring, this can be problematic, because keeping up to date with the evolving body of evidence means that our gold-standards that we, as healthcare providers, so vehemently defend, are prone to bias. Finding our knowledge, or worse, our practice, to be out of date can be very disempowering.

‍

New ways of gathering evidence are becoming available.

As doctors, we’re taught that the best way to get information about a patient’s health is to visit them at their home. We’re also taught that “a careful history will lead to the diagnosis 80% of the time”(3).

To this end, being able to understand how patients are tracking, across a period of time in finer and finer granularity, is helping inform our understanding of how each treatment really impacts a patient’s quality of life. Data points such as biometrics from wearables, or patient-reported outcomes (PROs), are being recognised not only as valid but as highly valuable (4). To what extent this vast and novel body of data will impact the stack rank of treatments in our current guidelines is yet to be ascertained, but does point to further risk of bias.

‍

Precision medicine, an elusive frontier.

The more data we’re able to collect and connect, the more complicated this decision-making process can become, for doctors and patients alike. New data points such as microbiomes, environmental factors, social determinants of health and other lifestyle factors all have the potential to significantly alter which treatments you receive in the future, as well as our understandings of how these treatments might impact you and your symptoms.

So you see, ascertaining what is best for you, isn’t as straightforward as it seems.

‍

At Human Health, we’re working to accelerate the world’s transition to precision medicine. We plan to learn from the world's doctors and their patients aboutwhich treatments are working for different individuals - and we'll publish this information for free for the medical and patient community.

‍

References

1. Feder G, Eccles M, Grol R, Griffiths C, Grimshaw J. Using clinical guidelines  BMJ  1999;  318 :728 doi:10.1136/bmj.318.7185.728
2. The difference between science and dogma. Nature, 400, 697 (1999). https://doi.org/10.1038/23307
3. Hampton JR, Harrison MJ, Mitchell JR, et al. Relative contributions of history-taking, physical examination, and laboratory investigation to diagnosis and management of medical outpatients. BMJ 1975;2:486–9.
4. Yetisen AK et al. Wearbles in Medicine. Advanced Materials, 30, 33. 2018; https://doi.org/10.1002/adma.201706910

‍

‍