“And one more thing, what do you want to do about this?”
The trainee looks so apologetic that I suspect a medical error relating to our octogenarian patient admitted overnight with confusion. Scrolling through the images on the computer, I see the problem, even without the arrows pointing to an “ill-defined abnormality” around the colon.
I am mildly irritated.
“Who ordered a CT of his abdomen to investigate confusion?” I ask. The trainee winces.
“He had a pan-scan but we didn’t order it.”
A “pan-scan” is a head-to-toe scan, also known in the vernacular as a fishing expedition. It is the poor cousin of a targeted scan, performed to confirm or exclude a clinical diagnosis.
A chat with the patient’s son quickly reveals the cause of confusion to be underlying dementia and a medication mishap (sedatives mistaken for antihypertensives) and the patient makes a good recovery but alas the “fishing expedition” hauls a net of incidental findings.
Besides the possible bowel abnormality, his prostate looks craggy and the pancreas odd – and this is just the abdominal scan. The chest CT discovers a “lesion”, its benign description paired with the admonition to do a follow-up scan – thus destroying any chance of it being left alone.
“Please tell me the brain CT is pristine,” I plead.
“It shows an old stroke the patient can’t recall.”
At the bedside, my happy observation that his confusion has resolved is drowned out by questions from concerned relatives about “all those cancers”.
Wary of easily harming an elderly patient, I counsel:
“In the absence of any symptoms, I’d avoid the risks of a colonoscopy.
“A prostate biopsy has significant consequences, even if it shows cancer.
“An MRI brain will rattle him and won’t provide more useful information.”
The family thinks:
“They’re saying no to everything but we want the best for our dad.”
Their wish to explore the incidental findings makes other specialists troop past. No one is genuinely concerned but in the fragmented way of modern medicine, everyone ends up suggesting something. Surgery says the patient can have a colonoscopy if he wants; urology orders a blood test; respiratory says repeat a scan in a few months; and neurology laments that sometimes, it’s just easier to do an unnecessary MRI.
The only person surprised is the intern, still reconciling textbook medicine to real life.
A whole costly week later, on the eve of discharge, the patient sustains a fall and miraculously gets away with only a bruised ego. He surreptitiously asks me whether he really needs all the outpatient “stuff”, when he best liked my original advice to let him be. But by now I am browbeaten and I just want him in safer surroundings, so I tell him to talk to his kids.
If I were betting, I would say he went on to receive every one of those tests – at great cost, and negligible benefit to him or the system. As for harm, I wouldn’t know because I never saw him again.
There was a time in my training when I needed permission to even request an MRI whose approval was contingent on the adequacy of the request. Those were also the days when radiologists posed questions to clinicians like, “How will it change your management?”
My favourite radiologist would ask interns doing their superiors’ bidding to “come back after examining the patient so you can order the right test”. These days when doctors “talk” to one other through electronic orders and are prone to staying in “our lane”, such sensible questions would probably be deemed impudent.
I thought about these issues as I read a thoughtful article in the Medical Journal of Australia about the harms of incidental findings on imaging.
The authors observe that “incidentalomas” (purely incidental findings on imaging), arise in up to a third of all tests and are as high as 40% in CTs and MRIs. Most findings are benign and clinically unimportant but absent risk stratification based on patient factors, clinical context and better description, patients end up receiving unnecessary – and harmful – interventions.
At an individual level, it may seem beneficial that a scan detected a mass that could later turn cancerous. But on a population basis, a substantial increase in the incidental detection of cancers including those of the kidney, thyroid, prostate and breast has led to no reduction in cancer-related mortality but an uptick in major surgery.
An earlier Australian study found that each year about 11,000 cancers in women and 18,000 cancers in men may be overdiagnosed. This figure sticks in my craw because I regularly see the trauma of receiving a cancer diagnosis even when one’s life is not imperilled by it.
As our universal healthcare system feels the strain, low-value medical care should worry us all.
Some of the authors’ advice is directed towards radiologists. Using more specific reporting, consensus guidelines and explicit follow-up recommendations has been shown to reduce overmanagement by anxious clinicians.
An excellent suggestion is to incorporate “incidentaloma outcome” in the growing number of clinical trials to provide prospective data on the natural history of these findings.
As a clinician who spends a lot of time requesting tests but also grappling with results of scans I never ordered, I am especially drawn to the appeal of two suggestions.
One, the public must be made more aware of the potential for harm. Patients are rarely informed of the frequency of incidental findings on imaging but clinicians have a duty of care to explicitly discuss this. In turn, patients must recognise that demanding a scan (yes, it happens) is not a substitute for a thorough history and examination.
But the best advice by far is simple: avoid unwarranted imaging. If clinicians order fewer scans, we will find fewer incidentalomas. We can reduce harm by acting less reflexively and using more validated decision rules aided by artificial intelligence in time.
Reducing harm from imaging cannot be an idea confined to PowerPoint presentations and academic publications when the consequences for patients are so significant. This is another area set for shared decision-making.
So, the next time your doctor orders a scan, heed the benefits but don’t forget to ask about the harm.
• Ranjana Srivastava is an Australian oncologist, award-winning author and Fulbright scholar. Her latest book is called A Better Death
