Why medical tests don’t always have the answer

When someone comes to see you with a worry about their health, it helps if you can guess what they would like you to do. Often, they’re hoping that you’ll send them for a test. Or even several tests. By “tests”, people rarely mean the simple investigations you can do in clinic: blood sugar levels, oxygen saturations, a urine dipstick. These never seem to inspire as much confidence as X-rays or blood tests, or anything involving a consent form. While many people hate hospitals – anxiety, fuss, too hot, too busy – some greet the prospect of an X-ray with beaming relief. Nothing could be more popular – except for a CT scan. You can almost see people thinking: “At last – answers! Results!” There is no doubt medical investigations are potentially invaluable. So why are we always withholding them, a patient with backache asked me. Why not do some tests, if we have so many available?

No medical system in the world is more committed to testing than America’s (because nowhere in the world is more litigious). But recently The American Board of Internal Medicine Foundation issued a list of 45 investigations it considered unnecessary. It states, very clearly: do not do ECGs on patients unless they have chest pain or significant risk factors for heart disease; do not perform MRI or CT scans on patients with straightforward back pain; do not do CT scans on patients with uncomplicated headaches, or on patients who have had a single black-out, or faint; and not everyone needs a chest X-ray before they have an operation.

All of these are already best medical practice in the UK, but it is odd and momentarily counter-intuitive to see a list of do nots so clearly spelled out. The temptation is to ask, why not? Surely 46 tests will find 46 problems which medicine can go on and solve?

Unfortunately, they probably won’t. A test is more likely to provide an answer if you’re asking it a question. And you need to have a plan for what you’re going to do with the result. Indiscriminate investigation has many drawbacks. It is expensive and time-consuming and it can get things wrong. It may find a problem that would have resolved on its own in time. The psychological consequences are obvious: the more tests you do, the more worry you generate.

Then there is the drawback of looking for one thing and finding another. If you scrutinise a leaf, it is rarely perfect: it may have blotches, or a ragged edge, or a trail where a snail has gone over it. But none of these necessarily caused the plant to die. It is the same with scans. If you MRI the spine you will often find prolapsed discs. If you X-ray people’s joints there will invariably be signs of arthritis; the bowel is likely to have diverticulae, and the gallbladder is often filled with gallstones. These are all “findings”, but they can be symptomless and harmless.

But once you know an anomaly is there, it’s hard not to see it as the cause of all trouble, and even harder not to start treating it. And once you’ve started on treatment it is difficult to go backwards (especially if your treatment involves an operation). This is without even considering the potential harm caused by the tests themselves. Most medical imaging has risks, exposing you to radiation, or dye, or powerful magnets. An X-ray of your spine delivers as much radiation as 120 chest X-rays; a CT scan of your chest as much as 400. That is a lot, especially if your findings are not going to change what you do next. Which is why tests need to follow the same rule as the rest of medicine: first, do no harm.

Sophie Harrison is a hospital doctor in South Yorkshire

Copyright The Financial Times Limited 2017. All rights reserved. You may share using our article tools. Please don't cut articles from FT.com and redistribute by email or post to the web.