User:Tialaramex/Screening

From RationalWiki
Jump to navigation Jump to search

Screening is a statistical approach to detecting diseases or risk factors for disease in a population that lacks symptoms. Lack of statistical training by lay people makes screening a popular topic for arguments driven by emotion rather than rational assessment of the situation, by both proponents and opponents. Where medical tests are part of the diagnostic process this is not screening even if the same exact tests are also used for screening populations.

Considerations[edit]

At the surface screening seems like an unambiguously good idea. If there's a test that could tell you whether you have disease X, why wouldn't you take it? Why wouldn't everybody take it? Indeed, why shouldn't a public health system want everybody to take it? Here are some considerations which must be taken into account before we can say that it's a good idea to have or make use of a particular test in this way:

Type I error[edit]

Tests may erroneously report that a particular patient has the disease or risk factor when they do not. Even if further tests are run which detect the type I error the patient may be distressed by the experience unnecessarily. The rate of type I error should be compared to the incidence of the disease in the population, not the population itself.

Type II error[edit]

Tests may erroneously report that a particular patient does not have the disease or risk factor when in fact they do. This may cause the patient to ignore future symptoms, or make unduly optimistic plans and either way lead to a worse outcome than if they hadn't been screened at all. Again the rate of error should be compared to the incidence of disease.

Overdiagnosis and treatment[edit]

When a test shows that you have a particular risk factor or disease, it is hard for both patients and clinicians to avoid the conclusion that they should do something about it. Indeed, if from the outset you didn't plan to do anything, why screen at all (see below)? So patients may end up consenting to a radical intervention that significantly reduces their quality of life, for example surgical removal of affected tissue, based on an uncertain forecast of the future.

Overhead of the test itself[edit]

Such tests are of course not free. The patient must take time out of their life to attend the screening, perhaps undergo an uncomfortable procedure (even a blood test is quite an ordeal for some people) and may subsequently need counselling about how to react to the results. Trained people must be available to take any samples, and then laboratory technicians are needed to run the actual tests and a specialist may be needed to interpret the results. However this is paid for, resources diverted to screening cannot be used elsewhere.

Demographic bias[edit]

Some demographic groups are more or less than usually inclined to attend screening tests. Without any deliberate mechanism to ensure or prevent take up by some groups the screening will tend to detect more disease in those groups most inclined to attend. If screening is paid for by cutting services elsewhere the net effect may be that those groups least likely to attend actually have overall worse health outcomes even if screening itself is effective.

Treatment availability and suitability[edit]

If there is nothing useful to be done, or if the patient is already certain that they would reject the available treatment, then it is arguably useless to carry out the test and provide patients with information they can't act on.

Proponents[edit]

Proponents of screening in general, or of a particular technique, especially those with a financial stake in the use of the screening test (e.g. practitioners, patent rights holders) may emphasise the benefits of screening for a particular disease and ignore or downplay the costs. They may be predisposed to interpret changes to screening criteria as a "cut" in services, because it will mean less income for them even though the overall health outcomes for the population will be better.

Opponents[edit]

Opponents may exaggerate the concerns listed above to suggest that overall screening (or some particular type of screening) is pointless and either shouldn't be offered or people shouldn't take it up. If they are connected with particular type of "alternative" medicine they may suggest that rather than undertaking screening to detect disease people should use unproven methods of preventing the same or all disease instead.