No one wants to be sued, but I think fear of legal action over missing a diagnosis means that a lot of doctors send patients for all kinds of tests – the more high-tech, the better – which may actually be harmful.
The first principle in medicine is to do no harm. And over the years, as I have gained more medical wisdom, I have come to realise that my role isn’t ‘just’ about diagnosing people, it’s about balancing risks.
This is especially true when it comes to organising scans.
I worked night shifts over the Easter four-day bank holiday and I organised lots of CT scans for patients, which led to a change in their treatment, and I would say were obviously the right thing to do.
One elderly man had a scan which showed he had a perforated duodenum (a part of the bowel), so he went straight to theatre.
And a young woman with shortness of breath had a CT scan of her lungs which revealed a massive lung clot. We successfully treated this with a clot-busting drug and saved her life.
But I also organised X-rays and CT scans which were clear, meaning I was able to reassure my patients.
The problem here – and what troubles me – is that I exposed them to unnecessary radiation.
In the UK, over seven million CT scans are done annually
Scans such as X-rays and CTs work by using ionising radiation – essentially, high-energy waves which pass through the body to create images
Being honest, I didn’t fully explain the risk of the scans to them, mainly because I had underestimated the risks.
I now know this thanks to reading a new study in the JAMA Internal Medicine.
Scans such as X-rays and CTs work by using ionising radiation – essentially, high-energy waves which pass through the body to create images.
But that same radiation can alter your DNA in the wrong way, causing mutations and setting the scene for cancer to develop years down the line.
The study was a sobering read: researchers calculated that the 93 million CT scans carried out in the US in 2023 could be responsible for over 100,000 future cancers. This is around 5 per cent of all new cancer cases.
And while this is US data, it’s highly relevant to UK practice, especially as scan rates continue to creep upwards.
In the UK, over seven million CT scans are done annually. This means each year, on average, one in ten people will be getting a CT.
The University of California researchers found that the scans which caused the most radiation and raised cancer risk were those of the abdomen, pelvis or chest –ones we often organise in A&E.
As I have gained more medical wisdom, I have come to realise that my role isn’t ‘just’ about diagnosing people, it’s about balancing risks
On average, for every 930 CT scans performed, one unlucky patient developed a cancer (such as lung, colon, breast, bladder or leukaemia) which they otherwise would not have got – due to the radiation.
The risk to children was higher, even from scans without much radiation exposure, such as those of the head for trauma.
The issue for all of us doctors who organise scans is to think what are the risks versus the benefits of them, rather than ordering them as a knee-jerk response.
Take head trauma: a huge study, published in 2023 in The Lancet Oncology, which monitored more than 650,000 young people from nine European countries, all of whom had their first head or neck CT before the age of 22.
Over 15 years of follow-up, the researchers found a clear link between radiation exposure from CT and subsequent brain cancer – not just in children who had had multiple scans, but even for those who’d had just one.
They found there was roughly one extra case of brain cancer in every 10,000 children who had had a single head CT scan. One in 10,000 might sound small, but when you factor in how many low-risk head injuries are scanned every week in A&E, the cumulative harm becomes all too real.
We must stop kidding ourselves that scanning ‘just to be sure’ is harmless.
The evidence is stacking up and it’s pointing to a simple truth: while CTs are sometimes essential, they are never benign.
Every CT scan must be questioned, justified and, wherever possible, avoided.
For example, many inexperienced or over-cautious doctors over-order chest CT scans looking for blood clots because of a knee-jerk reaction to shortness of breath and abnormal blood results.
But when the scan is requested without thinking about other more likely causes which don’t need a CT scan, such as a chest infection, this over-investigation puts patients in avoidable long-term risks of scan-induced cancer.
Although CT scans are the greatest risk because of the amount of radiation used, X-rays and mammography scans (also a form of X-ray) can also be a risk.
For example, it is estimated that for every 14,000 women who take part in breast screening, one will develop cancer because of the radiation exposure.
The radiation and risk from mammography is tiny, but a small increase in risk in millions of patients increases the numbers of people overall getting cancer. (Reassuringly ultrasound and MRI scans carry no risks of inducing cancers.)
But as a doctor, I know that while breast screening does induce some cancers, overall it saves more lives than causing cancer.
Back to my Easter night shifts.
I saw two patients who came in thinking they should get scans because 111 told them to come to A&E. But I chose not to send them for scans.
One was a 19-year-old with a head injury and the other was a 32-year-old who was 34 weeks’ pregnant and had chest pain.
After my assessment, I felt both had a very low risk of having a serious illness and I thought more harm would be done (this was even before I’d seen the new study) so I discharged them.
I wasn’t able to give them a 100 per cent reassurance (even with a scan I couldn’t do that), but I told them that this was the correct course of action and they should come back if they felt worse.
It is a complex decision for doctors to decide if a patient needs a CT scan. But my advice to you as patients is to ask them to imagine that they are not advising you as a patient, but as one of their loved ones.
Only then can they really think about the benefits and risks rather than making decisions to over-use cancer-inducing scans because of a fear of missing a diagnosis.
@drrobgalloway
