There are patients who want to know not only whether their lab results are good or bad but also why some tests are more reliable than others. So, sometimes physicians are expected to explain intellectually challenging concepts such as test sensitivity and specificity. I'd be interested to hear from you some plain language definitions of these two concepts. These definitions may be useful not only in the context of face-to-face patient-physician relationship but also in the context of printed or electronic patient education materials. Thanks!
For patients, predictive values may be more meaningful than sensitivity and specificity (they incorporate both sensitivity and specificity as well as pre-test probability).
Positive predictive value (75%) : if you test positive, there is on average 75% chance that you have the disease and 25% that you do not.
Negative predictive value (80%): if you test negative, there is on average 80% chance that you do not have the disease and 20% that you do.
Sensitivity (90%): the test will identify 90% of patients with the disease, it misses 10%.
Specificity (90%): the test will misidentify 10% of healthy people as having the disease.
It is just saying the proportion of people with disease or health condition in question shows the screening test result as positive, this can be called as Sensitivity. Similarly the proportion of people without the disease or health condition in question shows the same screening test result is negative, this can be called as Specificity.
if you put it in other way as test showing the proportion with disease as positive , then it called as prediction of the test as positive.
Sensitivity: if it is there it will be detected
- Tells us if you have the diseases (80% sensitivity means we're sure that the test will detect (true positive) the presence of the disease in 80% of the patients, and the 20% may have a false negative results "they have the disease and the test didn't detect it)
- Here we miss some patient with the disease untreated
Specificity : if it not there it will refute it
- Tells us if you don't have the disease (90% specificity means we're sure that 90% of the tested people are disease free. The remaining 10% means the tested people don't have the disease yet the results says that they do, which is false positive)
- Here we will treat someone without the disease
Very good idea to mention also the consequences of low sensitivity/specificity (not treating real patients vs treating patients without disease)! Thanks!
Merriam-Webster definitions:
Definition of sensitivity
: the quality or state of being sensitive: such as a : the capacity of an organism or sense organ to respond to stimulation : irritability b : the quality or state of being hypersensitive c : the degree to which a radio receiving set responds to incoming waves d : the capacity of being easily hurt e : awareness of the needs and emotions of others a book written with just the right mix of empathy and sensitivity— L. C. Brown
Definition of specificity
: the quality or condition of being specific: such as a : the condition of being peculiar to a particular individual or group of organisms host specificity of a parasite b : the condition of participating in or catalyzing only one or a few chemical reactions the specificity of an enzyme
Explained in layperson's language in my book: Social Epidemiology: Strategies for Public Health Activism (Columbia University Press, 2005), pages 319-323 with charts and explanations but easily understood by lay persons. Happy reading!.
A highly senstive test rarely misses a truly positive result while a highly specific test rarely misses a truly negative result.
But note that because sensitive tests tend to give positive results, some of the results may in actuality be negative. Similarly, because specific tests tend to give negative results, some of the tests may in actuality be positive.
What you want is a test that is both highly senstive and highly specific. Sometimes, this is not attenable and so two tests are conducted- one with a highly sensitive test and one with a highly specific test. That is the logic behind having patients undergo a confirmatory test after an initial positive test. A screening test is usually highly sensitive. If negative, it rules out the disease. If positive, a confirmatory test is done using a highly specific test. If the confirmatory test is also positive, it rules in the disease.
- Badges
- Science topic
- Similar topics
- Psychology