Over-diagnosis: the view from inside primary care

Of all the teenagers with flu-like symptoms, GPs need to find the one with a lymphoma in his chest. Joey Yee

OVER-DIAGNOSIS EPIDEMIC - We kick off the second week of this series with Jenny Doust looking at some drivers of over-diagnosis in general practice.

It’s easy to dismiss general practice as being about minor illnesses requiring little clinical acumen. We see patients with the common symptoms of life – coughs, fevers, backache and abdominal pain. Most of these are transitory and, for the large proportion, medical intervention makes little difference.

The diagnostic skills of a general practitioner, however, need to be as acute as those of House MD. Among all the children we general practitioners see with diarrhoea, we need to be able to pick the one with Crohn’s disease requiring urgent surgery. Among all the women who complain of feeling tired, we need to isolate the one who has life-threatening Addison’s disease. And among all the teenagers with flu-like symptoms, we need to find the one with a lymphoma in his chest.

These are all patients I have seen during my time as a general practitioner and they remind me constantly to take every person who walks through my door seriously.

Missing diagnosis

The most common reason general practitioners are sued is because of missed diagnoses. Missed diagnoses also invoke a strong sense of professional failure. So how can general practitioners manage in this sea of uncertainty?

One way is to perform more tests. This is also popular with patients, who perceive that tests ensure nothing serious is missed. What is not well understood by patients (and sometimes also by clinicians) is the potential harm from testing.

The most obvious harm is the cost and resources required; we would quickly overwhelm the health system if we performed an MRI on every patient with back pain. A strong system of primary care results in a health-care system that’s both more efficient and less costly because primary-care physicians are skilled at filtering those with severe disease needing further tests, from those with self-limiting illnesses.

But even in Australia, with its highly trained general practice workforce, this skill is often under-appreciated.

Harms of testing

Less well understood are the harms due to the inherent inaccuracy of testing. By the laws of statistics, when the likelihood of a disease is very low, most positive test results will be false positives. We see this most clearly in screening programs, where most abnormal screening results will not have the disease on follow-up testing. This causes only minor harm if there is a follow-up test. But it is of much greater consequence if tests are assumed to be accurate and too much faith is put in the results.

The greatest harm from the increased use of testing, however, is not costs, resources or false positives. Rather, it’s the problem of over-diagnosis.

Clinicians and patients both believe that finding a disease earlier in its process means it will be more successfully treated. But there’s increasing evidence that finding disease early or at a milder stage has paradoxical harmful effects, even reducing survival and quality of life.

Wider availability of more sophisticated tests results in “incidentalomas”, incidental findings that would not have otherwise been diagnosed. The detection of thyroid cancers, for instance, has more than doubled in the past 30 years. But most of these diagnoses are incidental findings from imaging.

If the earlier detection of these cancers improved prognosis, you would expect to see a decline in the mortality rate from thyroid cancer. But the mortality rate has sadly been constant in this time.

Looser boundaries

Even more problematic is the widening of the definition of diseases and the lowering of disease thresholds. Examples of expanding definitions include chronic kidney disease, diabetes, and even the diagnosis of cancers, such as breast and bladder cancer.

We assume patients who are now included in the wider definition benefit from treatment as much as those using the old definition. In fact, we are likely to believe that they will benefit more as we have caught the disease early. But all medical treatments cause harm. And when patients with milder disease are treated, it becomes more likely that the potential harm of medical treatment will outweigh the benefits.

Wider disease definitions become self-reinforcing – we find more and so we test more. And because we have now included milder patients in our disease group, we are misled by the perceived improvement in patient survival and the reduction in disease complications in our new patient group.

What now?

Right now, we are in the midst of a perfect storm – a population that is increasingly anxious about health, doctors who don’t want to miss a diagnosis, a pharmaceutical industry that profits from widening the definitions of disease and a health system that rewards over-testing and fails to acknowledge the harms that can result from it.

We have set up a group to look for solutions to this problem, including a conference in 2013, but the answers are far from clear. Given the pervasive nature of over-diagnosis, solutions will need to involve all layers of the health-care system, including policy makers and key clinical opinion leaders.

As a general practice physician, I am hoping we will be able to spend less time labelling patients with an ever increasing number of “diseases” and to spend more time working with them on solving their health problems.

Have you or someone you know been over-diagnosed? To share your story, email the series editor.

This is part six of our series on over-diagnosis, click on the links below to read other articles:

Part one: Preventing over-diagnosis: how to stop harming the healthy

Part two: Over-diagnosis and breast cancer screening: a case study

Part three: The perils of pre-diseases: forgetfulness, mild cognitive impairment and pre-dementia

Part four: How genetic testing is swelling the ranks of the ‘worried well’

Part five: PSA screening and prostate cancer over-diagnosis

Part seven: Moving the diagnostic goalposts: medicalising ADHD

Part eight: The ethics of over-diagnosis: risk and responsibility in medicine

Expert Database

Want to write?

Write an article and join a growing community of more than 57,600 academics and researchers from 2,183 institutions.

Register now