https://www.infirmiere-canadienne.com/blogs/ic-contenu/2016/01/06/terminologie-101-tests-de-depistage-et-valeurs-pre
Jan 06, 2016, By: Maher M. El-Masri, RN, PhD
Predictive values: Calculations that determine the likelihood that a screening test result will accurately predict the actual disease status of the screened individual
Source: Webb, P., & Bain, C. (2011). Essential Epidemiology: An Introduction for Students and Health Professionals. (2nd ed.). Cambridge, UK: Cambridge University Press
The sensitivity and specificity of a screening test are measures of how well the test does in classifying people as ill or not (addressed in the October and November 2015 issues). Clinicians use these measures to judge whether the test is worth using. Of even greater interest to clinicians, however, is knowing how well a test predicts whether or not the patient sitting in front of them has the disease in question. For this practical purpose, they need to know the positive predictive value (PPV) and the negative predictive value (NPV) of the test. The PPV refers to the likelihood that a patient with a positive test result actually has the disease; the NPV refers to the likelihood that a patient with a negative test result is truly disease free.
To illustrate how PPV and NPV are calculated and interpreted, let us say that 100 adults are screened for colon cancer using fecal occult blood testing. Twenty-five of these individuals have positive results and 75 have negative results. Subsequent testing shows that five of the 25 people with positive results do not in fact have colon cancer, while four of the 75 with negative results do have the disease. These data indicate that we have 20 true positive (TP) cases, five false positive (FP) cases, 71 true negative (TN) cases, and four false negative (FN) cases. The PPV is calculated by dividing the TP count by the sum of the TP and FP counts (TP/[TP+FP]). In our scenario, the PPV indicates that a patient who receives a positive test result has an 80 per cent chance of actually having colon cancer. The NPV is calculated by dividing the TN count by the sum of the TN and FN counts (TN/[TN+FN]). In our scenario, the NPV indicates that a patient who receives a negative test result has about a 95 per cent chance of not having the disease.
It is important to note that the usefulness of a screening test in predicting whether or not an individual has a disease is inversely affected by the prevalence of the disease. The reason? Tests for rare diseases are more likely to yield FP results, which will lower the PPV of the tests (there will be few TP results, so the FP results in the denominator will have a disproportionate effect on the calculation). In fact, even tests with excellent sensitivity and specificity scores will have a low PPV if the disease is uncommon. Thus, it is not prudent to screen for rare diseases.
Formula for screening-related calculations
TP/(TP+FN) = Sensitivity
TN/(TN+FP) = Specificity
TP/(TP+FP) = Positive predictive value
TN/(TN+FN) = Negative predictive value
NurseONE.ca resources on this topic
MyiLibrary
- Maltby, J., Williams, G., McGarry, J., & Day, L. (2010). Research Methods for Nursing and Healthcare.
- Supino, P. G., & Borer, J. S. (Eds.). (2012). Principles of Research Methodology: A Guide for Clinical Investigators.
Maher M. El-Masri, RN, PhD, is a full professor and research chair in the faculty of nursing, University of Windsor, in Windsor, Ont.
#research#nursing-research