In a clinical study, a new predictive model is developed to identify patients at high risk for a specific disease. If the model has a sensitivity of 85% and a false positive rate of 10%, what can be inferred about the accuracy of the predictors when applied to a population with a prevalence of 20% for the disease?

The model correctly identifies 85 % of the 200 people who actually have the disease, so it finds 170 true positives. Among the 800 people without the disease, a 10 % false‑positive rate produces 80 false alarms, giving 250 people who test positive. Thus the probability that a positive test truly indicates disease (the positive predictive value) is 170 ÷ 250 ≈ 68 %. Overall, the model correctly classifies 170 + 720 = 890 out of 1,000 people, yielding about 89 % overall accuracy. This shows that while the model is fairly good at catching disease, its positive predictions are only about two‑thirds reliable in a population where 20 % have the disease.