❓️ How can we use an epidemiological approach in the clinic?
❓️ How much can you rely on test results while ignoring the clinical signs of disease?
Case Example:
A 45-year-old male with no significant medical history presented with sudden-onset chest pain. His doctor suspected a rare disease (prevalence: 0.01%) and ordered a new, high-quality test (Sensitivity = 90%, Specificity = 80%).
Sensitivity = 90%: among 100 diseased individuals, the test will be positive in 90 cases.
Specificity = 80%: among 100 healthy individuals, the test will be negative in 80 cases.
💥 The test result was positive.
❓️ What should be the next step?
❓️ How much can you rely on this test, knowing that most medical tests usually have even lower sensitivity and specificity?
As you can see, there are two types of positive results (true or false). For a proper clinical approach, you must consider both and calculate the precision (positive predictive value):
Precision = 8100 / (8100 + 17,998,200) × 100 ≈ 0.045%
This means you can rely on it with only about 0.045% precision.
✅️ This example shows that the problem with this test is not its material or quality, but rather the disease it measures. For rare diseases, such tests are not useful. The best approach is to save resources and rely on clinical expertise.
