Hospital And Laboratory Tests Can Be Wrong

The first thing that everyone (patients and doctors) should remember is that test results can be wrong.

This may sound rather blasphemous. We are all of us encouraged to believe that when a machine spews out a piece of paper containing lists of figures the result must be accurate. When results come back from the laboratory they are regarded with the sort of reverence once accorded to messages on tablets of stone.

But the plain fact is that most laboratory tests are only ninety five per cent accurate.

If a test to detect a disease, whose prevalence is one in 1,000, has a false positive rate of five per cent, there will be 51 positive results if 1,000 people are tested. Of those 51 patients with positive results just one will actually have the disease concerned. The other fifty will appear to have the disease but will, in reality, be healthy. (When considering this remarkable figure you should remember that the prevalence rate for many diseases is much less than 1 in 1,000. A false positive rate of 5% for a disease which affects 1 in 1,000,000 people will mean that 50,001 people have positive results but only one of them actually has the disease concerned.)

Furthermore, that 95% accuracy rate which I have quoted is only reliable when all the equipment in the laboratory is working absolutely perfectly - and this is something that usually happens about once a week. As a general rule ninety per cent accuracy is more likely.

Moreover, all this assumes that all the technicians involved do their jobs perfectly and never make mistakes. As has been shown many times in recent years human error can be extremely significant - and can cause catastrophic results, much illness and many deaths.

And, of course, most patients have more than one test done. If a patient has twenty laboratory tests done (not at all unusual) then there is a good chance that if the patient is perfectly healthy the tests will show at least one abnormality.

When doctors spot an abnormality on a test report they immediately think of disease and then they think of treatment. There is tendency to forget or ignore the condition of the patient.

It is easy to see from all this that every patient going into hospital will have a good chance of being treated for a disease he or she hasn't got.

Even when no treatment is given there is a good chance that when a false diagnosis is made a patient's life may be changed. One recent study showed that out of 93 children who had been diagnosed as having heart disease - and who had lived their lives as `heart patients' - only 17 really had heart disease.

When the first blood test for syphilis was introduced doctors accepted it as accurate. It wasn't until several decades later that doctors found that fifty per cent of all patients whose blood test had shown them to have syphilis didn't have syphilis at all. The lives of many of those patients must have been ruined quite unnecessarily.


False Negatives

Of course, if there is a chance that a laboratory error may mean that a normal sample results in a `false positive' there must also be a chance that a laboratory error may mean that an abnormal sample results in a `false negative' reading.

A `false negative' reading means that a patient who has symptoms and signs of an underlying illness will be falsely reassured that there is nothing wrong with him or her

The doctor (or the patient) may make a diagnosis and then, when the laboratory results come back, assume that the diagnosis must be wrong.

My guess (and this is only a guess because I am not aware of any scientific research on this subject) is that when faced with a patient with clear cut symptoms strongly indicating an underlying problem, and a set of `normal' laboratory results, the vast majority of doctors will allow themselves to be influenced more by the laboratory results than by the patient's condition and their own clinical experience.


What Is Normal?

Next, we should consider the fact that, much as the technicians might dislike and dispute this, I do not believe that everyone fits neatly into the range of `normal values'.

We are all different (for which we should thank God, though if the genetic engineers get their own way and cloning becomes commonplace this will no longer be true) and the whole concept of `normal values' is an entirely artificial one. It should be used only as a guideline, rather than (as it so often is) as though it were written in stone that `normal value = healthy patient' and the corollary, `abnormal value = unhealthy patient'. Who decides what `normal values' are? Were the samples which provided the raw material for the `normal values' taken from healthy young patients? If so then there must be a chance that `normal values' are different for older patients.

There are a infinite number of reasons why a patient could have a result outside `normal values' and yet be perfectly normal and healthy.


Too Many Tests

The problems associated with tests and investigations are made worse by the fact that doctors undoubtedly order far too many tests. Inevitably, this leads to more false positives - and even more unnecessary treatment.

At least two thirds of all the tests and investigations ordered by hospital doctors and general practitioners are unnecessary.

Routine blood and urine tests help doctors make a diagnosis in only around 1% of cases.

And routine sampling for microbiological testing is unnecessary. One survey showed that only in three per cent of cases was the treatment given based on the result of culture and antibiotic sensitivity.

It is important to remember that just about every procedure (however `minor' it might appear to be) carries a physical risk. Even sticking a needle into a patient is potentially hazardous and the more complex the procedure the more likely it is that something will go wrong. One study showed that 14% of all patients who undergo invasive procedures (such as biopsies, catheterizations or bronchoscopies) have at least one complication. Most complications then need treatment and involve a longer stay in hospital. The longer the patient stays in hospital the more tests are done. And on it goes.


Why Doctors Order Too Many Tests

There are several reasons why doctors order too many tests.

1. Tests are sometimes ordered because they provide doctors with some protection from the risk of litigation. A doctor who has bits of paper containing test results to substantiate and support his selected treatment programme will probably be safe from legal action. That is sad but it is a fact of modern life. In countries which have lots of lawyers (you know who you are) the fear of unjustified litigation (fed more by greed than anything else) is very real. But in the long run it is the patients who suffer. Only the lawyers benefit. (As usual).

2. Unnecessary tests are often ordered to buy doctors time when they don't know what is going on. Doctors who are bewildered or uncertain may simple hide behind technology and order more and more tests in the hope that this scatter gun approach may prove effective. The condition of the patient becomes almost irrelevant as the doctor searches endlessly for the ever elusive diagnosis.

3. Some doctors practice `hot' or aggressive medicine and insist on collecting all the possible evidence they can even when the diagnosis is not in doubt. This type of doctor will insist on performing an endoscopy so that he can actually see a stomach ulcer even though the patient's history (and, maybe, an X ray) have made the diagnosis certain.

4. Doctors sometimes order tests in order to impress their patients, their colleagues, their students or themselves. The more esoteric the test the greater the status associated with its use.

5. It is fairly common these days for doctors to order unnecessary tests because they are planning to write a paper for a scientific journal and need lots of data to fill up the pages and build up their reputations. Most of the papers published in scientific journals are of little or no clinical value and so yet again patients are investigated unnecessarily.

6. Doctors always undergo the first part of their training in hospital. And in teaching hospitals students and young doctors are taught to over-investigate. Consultants frequently berate students who haven't ordered all the possible tests. No attempt is made to decide which tests are likely to prove particularly helpful. No thought is given to the cost of ordering so many tests. And no thought is given to the patient who will spend hours wandering from laboratory to X ray department and who will end up, in consequence, exsanguinated, radiated and exhausted. The pointlessness of all this was well illustrated by one survey which showed that when bacteriological swabs are taken from patients in hospital there is a two to one chance that the patient will have been discharged before the results of the culture become available.

It is, of course, possible that the odd patient may get better before the test results become available but the only sensible conclusion I can draw from this survey is that in most cases the doctor in charge of the patient regards the test that he has ordered as irrelevant to the patient's well being.

7. Finally, doctors frequently perform tests out of habit. So for example, patients who are in hospital will often have daily tests done even though a daily change in a specific test is unlikely to lead to a daily change in the treatment recommended. When this happens diagnosis and investigation have been divorced from treatment. The real risk lies in the fact that the more often tests are performed the more likely it is that there will be one or more results showing a new and abnormal result - and that the new, abnormal result will inspire doctors to introduce another element to the treatment programme.


X Rays

Ever since Rontgen first discovered X rays at the end of the nineteenth century radiology has played an increasingly important part in the average doctor's investigative armamentarium. Go into hospital for a fairly routine operation and there is an excellent chance that they will take at lest one X ray - even if it is only a routine X ray of your chest.

Since 1968 medical journals around the world have frequently carried papers pointing out that most of X rays are unnecessary, extremely expensive, potentially hazardous and unlikely to contribute anything to a doctor's knowledge of a patient's illness, to the way that the patient is treated or to his or her long term health.

Doctors frequently order for X rays as a comforting ritual rather than for any specific purpose. And, of course, many X rays are ordered purely for legal protection.

It has been repeatedly estimated that in most `developed' countries something like nine out of ten X rays are unnecessary. (In the US the figure is probably even higher than this.)

I had good personal experience of the possible danger of an X ray a few years ago. When an investigative X ray showed that I had a strangely shaped kidney the two radiologists who had studied the X ray plates wanted me to see a surgeon. Because of the X ray they thought I had renal cancer and urgently needed major surgery. After I insisted on another opinion a third radiologist immediately spotted that my abnormal kidney wasn't pathologically abnormal at all. I just happened to have a mis-shaped kidney.


Too Many Doctors Are Dependent On Tests

Many doctors have become totally dependent on tests of one sort or another. They feel incapable of making any sort of clinical decision unless they have done a vast range of tests. And they put more faith in the results they obtain from the laboratory or the X ray department than they do in their own clinical skills.

Whenever a doctor plans to do tests or investigations of any sort ask him how the tests or investigations will affect the way he treats you.

If he tells you that the test isn't going to affect the treatment (or lack of it) then the test may not be worth having.

It is, of course, always worth remembering that tests can cause serious health problems - and can kill.


The Bottom Line

The bottom line is that you should remember that tests aren't as reliable, as useful or as necessary as most people think they are.

If you feel ill when tests have shown that there is nothing wrong with you then you should listen to your body and ignore what may well turn out to be `false negative' tests.

If you feel well when tests have shown that you have a problem which needs treating just remember that there is a chance that you are feeling well because there isn't anything wrong with you - and that it is the laboratory test equipment rather than your body which needs treatment.


Copyright Vernon Coleman 2003


back