Saturday, February 24, 2007

Have you had the test?

When we hear about celebrities being ill or admitted to hospital we hear phrases like ‘undergoing tests’ and ‘tests showed that…’. ‘Tests’ are obviously very important and a lot hangs on the results. ‘Tests’ are safe and always right. Aren’t they?

We’ve all seen patients where the test gets it wrong; the normal x-ray in the patient with bone metastases or the dangerously high potassium level due to a dodgy blood sample in a normal patient. In these circumstances we know the limitations and problems with the test and can repeat the test, do a different test, or even ignore the result.

I’m sure that all medical students and junior doctors have had some rules about diagnostic tests drummed into them by their seniors:

If you don’t know why you’re doing a test you shouldn’t be doing it

If you don’t know what you’re going to do about the test result when you get it you shouldn’t be doing it

If the result of the test isn’t going to change or aid in the patient’s management you shouldn’t be doing it

Treat the patient not the result
(all tests can produce false positive or false negative results)

This avoids patients having unnecessary tests and unnecessary or wrong treatment based on incorrect results. Obviously unnecessary treatment could be risky with surgical complications or drug side effects but some tests carry their own risks, for example radiation exposure (x-rays, CT scans etc), bleeding (liver biopsy) or other injury (colonoscopy, diagnostic procedures under anaesthetic).

We try to take these problems with diagnostic tests into account but sometimes doctors as well as patients can become over reliant on them for diagnosis or reassurance (or medico-legal reasons). Many of the diagnostic tests we do have had their reliability studied in depth before they are recommended for everyday use because it’s important to understand how often and in what circumstances a test is likely to give a false result. Newer tests are still being studied and developed and we’re often cautious about them. The histopathologists out there will know about the latest immunohistochemical stains which show early promise at differentiating benign from malignant or being highly specific for a certain type of tumour. When more research is done they are often not quite as good as we initially thought, although they may still be useful.

In 2003 the RCPath produced some draft guidelines: Who can request a test? This document then seemed to vanish but over three years later it is finally being revised into a more definitive form. The draft version states:

the test must not only be ordered appropriately, but also… the result must go back to someone who can take appropriate action.’ (sounds just like what your consultant told you when you were a house officer).

On a related note they’ve also released this report: Evaluating and introducing new diagnostic tests: the need for a national strategy. This document talks about some of the things I’ve mentioned above about the reliability of tests and summarises the issues that must be looked at when introducing new tests, the first of which is:

Is the scientific evidence for the validity of the new test sufficiently robust?'

The report also goes on to state:

There is increasing pressure to introduce new tests precipitously, without rigorous evaluation of their true utility, when related to emotive topics (such as cancer) or when there are possible medicolegal threats.

It is difficult for staff to evaluate recently introduced tests objectively and consistently across the NHS; published evidence demonstrates that papers on the diagnostic accuracy of new tests are often of poor quality and to make a good decision is difficult.

The RCPath might be a bit late with these documents as increasing numbers of ‘diagnostic’ or ‘screening’ tests are being offered to anybody with the money to pay, as well as the time and stress resilience to cope with subsequent investigation of their incidentalomas.

Shinga has written about some of the allergy tests available and finds, via some dedicated literature searching, that they aren’t as reliable as some suggest. Despite the amount of work that should go into studying tests, Shinga has found that tests can be let loose on the public without being properly validated, and worrying they can be requested by anybody – regardless of the risks.


At 16:23, Blogger Shinga said...

I had no idea that this was such a hot topic for the RCPath - primarily because I'd heard that whole-body scanning has more or less tanked in the US and I didn't know that the 'scanners' had set up home over here.

The reliance on some of these tests in the vaccuum of evidence to support them is absurd. I was called a fool this week for mentioning that I consider vaccination to be A Good Thing. The denouncer of my mental and cognitive status declared incredulity that I didn't know that children only developed measles because they have sub-clinical deficiencies of vitamin A (it had to be sub-clinical - isn't it always).

OTOH - I was fascinated to read an article about whole-body scanning some time ago that argued that it seems as if a surprising number of people have incidentalomas in the form of non-progressive tumours and that it would benefit cancer research if we understood why some tumours progress and others don't. (I don't at all believe that this is a justification for whole-body scanning - it was just an interesting nod in a research direction.)

Regards - Shinga

At 11:31, Blogger Dr K said...

Hi Shinga

The RCPath aren't too interested in the scanning but it's a good example of the kind of test available. They are more interested in tests which come under the umbrella of things like immunology, clinical biochemistry, haematology, genetics, microbiology and histopathology. The availability to anybody of unvalidated tests is a worry which they don't really stress maybe as much as they should. They have published a guide to allergy and allergy tests which you might find interesting. It's at:

At 21:17, Blogger The Angry Medic said...

Whoa, interesting post. And here I thought tests were merely plot devices in medical TV shows...

At 17:41, Blogger Dr K said...

Wait til you get onto the wards and they start asking you what tests you might do on the patients, you say full blood count, U+E (because that's what they always do) and then you get asked why you want to do those and you don't know. It's happened to all of us!


Post a Comment

<< Home