Predictive Value

By | March 1, 2015

No test regardless of the sensitivity and specificity should ever be used without taking a good history and performing a careful physical exam.

No test regardless of the sensitivity and specificity should ever be used without taking a good history and performing a careful physical exam.

No test regardless of the sensitivity and specificity should ever be used without taking a good history and performing a careful physical exam. At Schlesinger Pain Centers we know that the reason for this is that the incidence[1] of any disease varies with the patient population and the predictive value of the test is greatly influenced by the prevalence[2] of the disease in the population being studied.  An absurdly obvious case would be the diagnosis of pregnancy, which should rarely if ever be seriously entertained in male patients. Let us first define predictive value and then I will give you another example, which is slightly less far fetched. Predictive value comes in two flavors, positive and negative. The positive predictive value of a test is a measure of how much you can trust a positive result and is defined as the number of true positives divided by the total number of positives, both positive and negative. In the same way the negative predictive value of a test would be a measure of how much you could trust a negative result and is defined as the number of true negatives divided by the total number of negatives, both true and false.

[1] Incidence is the number of new cases of a disease occurring in a defined population in a given period of time. It is more often given as a percentage to take out the effect of changes in population size. For example the incidence of breast cancer in the United States is about 232,000 new cases per year with an incidence rate of 125 new cases per 100,000 women.

[2] Prevalence is the number of cases of a disease found within a given population at a given time. Prevalence like incidence can be expressed as a ratio of the number of cases divided by the population size. Prevalence is influenced by incidence, the mean survival time of patients with the disease as well as the cure rate for the condition. The prevalence of breast cancer in the U.S. is about 2,975,000 cases or about 1600 per 100,000 women.


Leave Your Comment