What is Sensitivity and Specificity?
I promise it’s not boring!
Diagnostic tests help us determine the presence or absence of a disease. All tests in medicine have a unique sensitivity and specificity for a disease – this could be a physical exam maneuver, screening tests, and diagnostic tests.
As we know, tests are not perfect. There are 4 outcomes of a test: true positive, false positive, true negative, and false negative.
False negative consequences: delay in diagnosis, delay in treatment.
False positive consequences: more testing, unnecessary treatment, cost, risk, stress/anxiety.
Have I lost you yet? Stay with me.
Case: You’re seeing a 33-year-old female of Asian ethnicity. She presents with joint aches, fatigue, mouth sores, and a malar rash. The differential diagnosis of systemic lupus erythematosus (SLE) crosses your mind. You decide to order ANA (antinuclear antibodies) as part of your work-up. You recognize it is a screening test and it will not definitively diagnose lupus.
What is sensitivity? the ability of a test to correctly identify a patient who truly HAS the disease (1). A test that is 100% sensitive means that all patients with the disease are correctly identified as having the disease (i.e. no false negatives).
What is specificity? the ability of a test to correctly identify a patient as DISEASE-FREE (1). A test that is 100% specific means all healthy patients are correctly identified as healthy (i.e. no false positives).
Back to the case: What is the sensitivity of ANA for lupus? It has a high sensitivity, meaning it is good at correctly identifying patients WITH lupus (low risk of false negative result) (2).
What is the specificity of ANA for lupus? It has a low specificity, meaning it is poor at identifying patients WITHOUT lupus (risk of false positive result) (2).
In this case, your pre-test probability is high, so ordering an ANA is reasonable. The result comes back as positive, with a titre of 1:160, which you know is high. You recognize that given your high pre-test probability, there is a good chance your patient has an autoimmune/rheumatologic process, but requires further workup as ANA can be positive for a number of reasons (e.g. connective tissue disorders, thyroid disease, etc.).
Let’s review a second case: You’re seeing a 50-year-old female with non-specific symptoms including muscle aches, fatigue and dry skin. You again question whether an autoimmune condition like SLE should be on your differential. Your pre-test probability is low in this case and you recognize that ordering an ANA may return as falsely positive, which could create negative consequences (patient anxiety, more testing, use of health care resources). You think twice about ordering an ANA in this case.
Key Take Home Points:
1. The sensitivity and specificity reflect the accuracy of a test, guides us on its appropriateness in a clinical encounter, and helps us interpret the results.
2. SNOUT: a test with high SNsitivity is good at ruling a disease OUT (low false negatives).
3. SPIN: a test with high SPecificity is good at ruling a disease IN (low false positives).
References/Readings:
1. Swift A, Heale R, Twycross A. What are sensitivity and specificity? Evidence-based nursing. 2020;23(1):2-4. (VIEW HERE)
2. Olsen NJ, Karp DR. Finding lupus in the ANA haystack. Lupus science & medicine. 2020;7(1):e000384-e. (VIEW HERE)
3. American College of Rheumatology. Antinuclear Antibodies (ANA). [internet]/ [cited 2023 August 28]. Available from https://rheumatology.org/patients/antinuclear-antibodies-ana
4. Here are more readings about AUC and ROC curves or types of diagnostic test accuracy study design.