Blinded by the Lab Coat: The Dangerous Reality of Medical Bias

Bias plays a huge part in how we are treated in our daily lives by others, and unfortunately, it can even follow us at our most vulnerable times in life: when we are sick and in need of medical assistance. When we realise the effects medical bias can have on people's feelings, trust in doctors, and, biggest of all, their health, we can finally understand what a huge issue we are dealing with.

Bias can stem from any preconceived notions healthcare professionals have. They can drastically alter the attitude or manner of care they give to patients. An example of this is how some healthcare professionals treat patients differently depending on their gender. Women are seen to experience longer delays when being diagnosed with chronic conditions such as cardiovascular disease and endometriosis, and according to the Australian Government Department of Health, Disability and Ageing, two out of every three women experience discrimination in healthcare settings. There are often psychological misattributions in women's health, as clinicians may dismiss physical pain as stress, which can prevent the true cause of their issues from being discovered. This is not only due to doctors' own biases but can also be due to male-centric models in healthcare, as historically, diagnostic frameworks have been built around men, which can exclude women from being promptly diagnosed.

Another group of people who experience significant medical bias throughout their lives is people who are neurodivergent or have mental health conditions. Their physical problems are often treated as being related to their mental health, which can make it difficult for them to receive the proper care they need for genuine physical problems until those conditions progress to more serious stages. Furthermore, standardised testing and research mainly filter out or penalise people with diagnosed neurodivergence or mental health conditions, which means there often just isn't enough data or research for clinicians to know what to look for when diagnosing these patients.

Many people may say that this is all the more reason for more roles in healthcare to be handed over to technology, as it has no personal prejudice and could be less prone to bias against its patients. However, this introduces the risk of automation bias, where the system that you are trusting to guide your healthcare gives invalid advice or ignores real physical symptoms, which can ultimately result in patients being prescribed the wrong medicine because of a misdiagnosis or having a severe problem overlooked until it develops into something more sinister. Automation bias can also arise from algorithmic blind spots that these systems have because of how they are trained. We can see this when diagnosing melanoma, as many tools are predominantly trained to detect it in people with lighter skin tones, meaning they are more likely to make mistakes when diagnosing patients with darker skin tones. These automated tools also depend on the training data supplied to them. If researchers and clinicians have biases, those biases can be reflected in the data they collect, passing those prejudices on to the tools and making them no better at examining patients than the clinicians themselves.

While we expect doctors to remain neutral, we have explored how they are ultimately just humans who can let their internal prejudices seep into their judgment, causing inexcusable mistakes time and time again.