If you change the cutoff for defining disease, what happens to sensitivity and specificity?

Prepare for your Epidemiology Test with our engaging content, including flashcards and multiple choice questions. Each question is accompanied by hints and explanations. Boost your readiness and confidence now!

Multiple Choice

If you change the cutoff for defining disease, what happens to sensitivity and specificity?

Explanation:
The main idea being tested is how changing the cutoff for a positive result shifts the test’s sensitivity and specificity. Sensitivity is the fraction of actual diseased individuals correctly identified, and specificity is the fraction of non-diseased individuals correctly identified. When you lower the threshold for calling someone diseased, you catch more of the true positives, so sensitivity increases, but you also misclassify more healthy people as diseased, so specificity decreases. If you raise the threshold, you miss more diseased cases (lower sensitivity) but correctly exclude more non-diseased people (higher specificity). Because the cutoff affects who is labeled positive and who is labeled negative, both sensitivity and specificity change with the cutoff.

The main idea being tested is how changing the cutoff for a positive result shifts the test’s sensitivity and specificity. Sensitivity is the fraction of actual diseased individuals correctly identified, and specificity is the fraction of non-diseased individuals correctly identified. When you lower the threshold for calling someone diseased, you catch more of the true positives, so sensitivity increases, but you also misclassify more healthy people as diseased, so specificity decreases. If you raise the threshold, you miss more diseased cases (lower sensitivity) but correctly exclude more non-diseased people (higher specificity). Because the cutoff affects who is labeled positive and who is labeled negative, both sensitivity and specificity change with the cutoff.

Subscribe

Get the latest from Passetra

You can unsubscribe at any time. Read our privacy policy