Skip navigation

Counterintuitive is bad news (Editorial)


There is an ancient truth that if you want to make your job look really important you surround it with barbed wire of obscurity. Baffle the bastards. Make it look as difficult as you can, and watch the plaudits flood in.

Why is Bandolier so toxic? Because Bandolier is unable to remember how to calculate sensitivity or specificity without looking it up on the CEBM website ( http://cebm.jr2.ox.ac.uk/ ). They are thus defined as counterintuitive (nice rationalisation here for post middle-age deterioration). You all know that Bandolier feels much the same way about odds ratios, which are about as much use to the clinician as a wet muffler. But there at least we have NNTs to keep us warm.

So what's the moral? The moral is that methods of working which are muddling should be superseded by methods which are comprehensible. Down with sensitivity and specificity, and up with likelihood ratios and common sense.

Not just Bandolier


It's not just Bandolier who finds these things difficult. A survey of US physicians showed that a whopping 96% do not use formal methods of assessing diagnostic tests or their results - not even likelihood ratios. There are lots of reasons why not, but probably, like Bandolier , they find them counterintuitive and stodgy, and they can't remember what a negative predictive value is.

I want it NOW!


Another reason information on diagnostic tests isn't used is that it isn't available in any easy format, and especially not immediately available when it's needed. But if we can have knowledge that is relevant, valid and usable, and have it fast, then evidence gets used much, much more often than if it's not. That is the lesson from the ' evidence cart ' tested at Oxford's John Radcliffe Hospital.

Waste of space


Bandolier encounters many people who think diagnostic testing is a waste of space. Countering that argument isn't easy. That is partly a reflection of the utility of diagnostic tests being wrapped in the barbed wire of obscurity. Partly it is lack of examples.

When Bandolier asked in a previous issue for good examples of the use of diagnostic tests, or tests which make a difference, the answer was a deafening silence. None of the readers of the 25,000 copies or the 10,000 visitors a day to our Internet site seemed able to provide answers. Bandolier 's own reading in some defined diagnostic areas produces profound feelings of despair that it is so uninformed and uninformative, and shame for occasionally having added to it.

Future of diagnostic testing


That could be a bit of an oxymoron. Yet there's good news, too. There is evidence that we can make better use of our laboratories, and this month's Bandolier contains evidence that we can avoid swamping our labs with unnecessary diagnostic tests. There's also evidence on how to influence physicians effectively to use tests to their best advantage. And there's evidence elsewhere that senior physicians have the experience to order diagnostic tests appropriately.

The bottom line is that we are in a hole. As a famous Balliol politician once said (perhaps his only major contribution) 'When you are in a hole, stop digging!' To treat properly we have to diagnose properly. Most doctors do a pretty reasonable job of that most of the time, in spite of the lack of a solid evidence base to help them.

We have to get back to basics, get some quality research done, and find practical and intuitive ways to help. A big job, urgently needed.

next story in this issue