The Problems and Controversies with Overdiagnosis
I would like to preface this article by saying that views expressed are my own and do not represent the medical community. I will be discussing topics such as cancer which may be difficult for some readers.
Medicine, as an academic subject, allows for a look into the most vulnerable state of human existence, a privilege not afforded to many. It’s a subject I have (mostly) enjoyed studying; knowing how the body can malfunction and what we can do about it is a daunting, yet incredibly rewarding, feeling. If I must read about the loop of Henle again, I may combust, but I do enjoy the challenge that this degree provides.
However, I would be lying if I said that I can always see the positives, both of studying medicine and of the way that medicine functions beyond theories of cells and organelles. As part of my dissertation, I have been researching overdiagnosis and the consequences thereof. The very definition of overdiagnosis is already a complicated situation, so I’m bringing forth some examples in the hopes of explaining what it means, or at least my interpretation of it.
The 00s saw a rise of thyroid cancer to epidemic proportions in South Korea. However, the death rate due to these cancers did not increase at all, which was unexpected in such high rates of disease in a population. Eventually, the cause of the epidemic was found: ultrasound machines. Multiple cancer screening programs, one of which was for thyroid cancer, had been introduced in 1999. Cancer was being found, diagnosed, and treated in significantly more people. A lot of these people would never have been affected by their cancer and underwent treatment without any benefit to them (Ahn et al, 2016).
Does this mean these people were diagnosed incorrectly? Not exactly. Evidently, people can get diagnosed, following all the usual guidelines and diagnostic criteria, with a condition that wouldn’t cause any harm if unnoticed. This is overdiagnosis. It is a correct diagnosis, just not a beneficial one.
It can be difficult or even impossible to tell whether a cancer picked up on screening has the potential to become harmful. Two people can have the same test results and yet one will go on to develop serious and life-altering disease, whereas the other would never have had symptoms. Which means that both people must undergo the psychological stress of a cancer diagnosis, as well as the side effects from treatment, to ensure that one of them will have benefitted at the end.
Don’t get me wrong, I think advances in medicine are always good. If someone can be treated, they should be treated, after they consent with all the information available to them, especially if there is a chance the treatment will do more harm than the disease itself. If someone who can’t be treated can be given answers and support with a diagnosis, then this beneficial too. Early cancer detection saves countless lives, and it is an essential part of cancer research.
Overdiagnosis was initially termed due to these types of cancer cases, which could only be seen retrospectively and on a population level. These diagnoses occurred because medical advances let us detect and treat something early, to the benefit of the individual. We only know these cases are overdiagnosis in retrospect, which is valuable information that has been used to change guidance and policy, for example the changes to prostate and breast cancer screening in the UK. Check out the Cancer Research webpages about overdiagnosis (linked below) if you’d like to know more about screening in the UK and the justifications thereof.
There is another, more insidious, side to overdiagnosis. That is one where people could potentially benefit from a rise in diagnoses and treatments.
The Diagnostic and Statistical Manual of Mental Disorders, or DSM, contains guidelines to diagnose psychiatric and behavioural disorders. Its most recent edition (DSM-5), which was revised by a panel of experts, has been speculated to inflate the diagnostic criteria for certain disorders (Kudlow P, 2013), allowing more diagnoses to be made in borderline cases, often based on symptoms which are part of the continuum of “normal” human behaviours. 70% of the expert panel who developed the DSM-5 are tied to drug companies (Cosgrove, Krimsky, 2012).
I chose this example because it is controversial and has been contested. My personal thoughts on whether the DSM-5 has manufactured overdiagnosis in mental disorders to please big pharma are irrelevant. Feel free to investigate this and write me an essay, I could do with some inspiration.
What I would like to point out is that weighing up benefits and harms of a subjective diagnosis, like many in the DSM, is not an easy task. Two people with the exact same symptoms and lifestyle can get completely different outcomes from a diagnosis. One person may feel tied down and defined by their disorder, another may feel liberated and finally have answers they were looking for. Defining overdiagnosis based on net benefit is complicated in a lot of clinical contexts, especially if two identical profiles can have vastly different outcomes due to their own reaction to a diagnosis.
Many medics chose this course because we “want to help people”. It’s cliché, but it is still often quoted in interviews. Realising that sometimes there’s no real benefit to helping someone by providing a diagnosis can feel like a punch in the gut, especially as you won’t know that it wasn’t beneficial until much later, if ever.
Uncertainty is inevitable in medicine. You won’t always get it right. The systems in place and guidelines written won’t always get it right. All you can do is be there for the people in your care, supporting them in the best way you can.
Ahn HS, Kim HJ, Kim KH, Lee YS, Han SJ, Kim Y, Ko MJ, Brito JP. Thyroid Cancer Screening in South Korea Increases Detection of Papillary Cancers with No Impact on Other Subtypes or Thyroid Cancer Mortality. Thyroid. 2016 Nov;26(11):1535-1540. doi: 10.1089/thy.2016.0075. Epub 2016 Oct 18. PMID: 27627550.
Kudlow P. The perils of diagnostic inflation. CMAJ. 2013 Jan 8;185(1):E25-6. doi: 10.1503/cmaj.109-4371. Epub 2012 Dec 10. PMID: 23229001; PMCID: PMC3537802.
Cosgrove L, Krimsky S (2012) A Comparison of DSM-IV and DSM-5 Panel Members’ Financial Associations with Industry: A Pernicious Problem Persists. PLoS Med 9(3): e1001190.