Rules of thumb for positive and negative test results

Abstract

Background: It is well known that tests are not 100% accurate at classifying individuals. The actual condition of an individual(e.g. diseased, or not diseased) does not coincide with her test result (positive, or negative). Nevertheless, it is often presupposed, as a rule of thumb, that individuals with negative results can be "ruled out" if screening test is highly sensitive and "ruled in" if screening test is highly specific. This has led to the mnemonic SNNOUT (sensitive negative out) and SPPIN (specific positive in).

Method: Probabilistic analysis of SNNOUT and SPPIN.

Result: SNNOUT and SPPIN are incorrect. We devise the correct and easily applicable rules of thumb.

Conclusion: The correct rules of thumb could be of great help to doctors and patients.

Keywords: predictive value; rules of thumb; sensitivity; specificity.