Guidance on terminology

There is no universally accepted definition of medical screening but there is general agreement that the activity contains three elements:

It is a process of selection with the purpose of identifying those individuals who are at a sufficiently high risk of a specific disorder to warrant further investigation or sometimes direct preventive action. It is usually a preliminary process to offering a diagnostic test and, if required, preventive action.

It is systematically offered to a population of people who have not sought medical attention on account of symptoms of the disease for which screening is being conducted. It is normally initiated by medical authorities and not by a patient's request for help on account of a specific complaint.

Its purpose is to benefit the individuals being screened. On this basis, mass testing activities such as surveillance for HIV infection or pre-employment examinations to test fitness for work would not be classified as medical screening.

To encapsulate these elements the following definition has been widely adopted:

Screening is the systematic application of a test or enquiry to identify individuals at sufficient risk of a specific disorder to benefit from further investigation or direct preventive action, among persons who have not sought medical attention on account of symptoms of that disorder.

One aim of this journal is to encourage the use of a common screening nomenclature without being too prescriptive or restrictive – a delicate balance, but by making the semantics explicit the issues may be clearer and better understood.

There are a number of alternative terms that are used for the same measures of screening performance.

Detection rate (DR) and sensitivity are synonyms (the proportion of affected individuals with a positive test result). An advantage of DR is that it avoids confusion as ‘sensitivity’ has a different meaning in analytical biochemistry (the minimum detectable amount in an assay). DR can be used in a different sense in cancer screening – as the number of screen-positive individuals divided by the number screened. This is better described as the prevalence of screen-positive cancers in the population.

False positive rate (FPR) (the proportion of unaffected individuals with positive results) is the complement of specificity (the proportion of unaffected individuals with negative results) so, for example, if the FPR is 5% the specificity is 95%. The advantage of using the term FPR is that (a) it is more easily understood and remembered; (b) it focuses attention on the group who will be offered further medical intervention; and (c) a 10% FPR, for example, is twice as bad as one of 5%, whereas corresponding specificity values of 90% and 95% conceal the difference.

The odds of being affected given a positive result (OAPR) provide the same information as the positive predictive value (PPV). The OAPR is the ratio of the number of affected to unaffected individuals among those with positive results (affected positive: unaffected positive). PPV is the number of affected individuals with positive results divided by the number of individuals with positive results, both affected and unaffected (affected positive/(affected positive + unaffected positive)). The advantage of the OAPR over the PPV is that the OAPR conveys a clearer impression of the performance of the test when either is high. For example, if the OAPRs for two tests are 20:1 and 50:1, the PPVs of 95% and 98%, respectively, tend to conceal the large difference.

The negative predictive value or NPV (unaffected negative/(unaffected negative + affected negative)) is sometimes used in publications. It is, however, of little or no clinical value because it defines a group of unaffected individuals for whom there is no need for action to be taken.

In screening for cancer and certain other diseases, there are difficulties in estimating the DR and FPR because the presence or absence of disease is not easily established. The more detailed the investigation, the more cases of disease will be found. The denominator at any point in time may be unknown. A DR cannot be determined, although the FPR can usually be estimated from the overall positive rate. Alternative measures for the DR are used, such as the ratio of the screen-positive cancer prevalence at the first screening examination to the annual incidence in the absence of screening, which indicates the number of years it would have taken for cancers detected by the screening examination to have presented clinically. Another measure that seeks to estimate the DR is the number of screen-detected cancers divided by the sum of this number and the number of cancers discovered between screening examinations (interval cases). This measure, which has also been called the ‘detection rate’, will be a function of the interval duration.

One source of confusion in cancer screening is that the cancer DR is sometimes used to describe the prevalence of detected cancers at a screening examination (perhaps better described as the screen-positive cancer prevalence, as above) instead of the proportion of all cancers present that are positive.

In other forms of screening, such as antenatal screening for congenital malformations, the estimation of DR and FPR is straightforward because the denominator can be determined.