2. A validation study is to be performed in a local hospital to check the accuracy of assessment of hospital-acquired infection (INF) following coronary bypass surgery (coronary-artery bypass graft, or CABG). In a given year, the hospital performs 1100 CABG procedures. A Centers for Disease Control and Prevention (CDC) algorithm is currently used to categorize subjects as having INF. To validate the algorithm, all CDC+ subjects (N = 100) and a random sample of CDC subjects (N = 1000) will be ascertained by an infectiousdisease (ID) fellow and a detailed investigation will be performed, including a chart review and documentation of antibiotic use. Assume the ID-fellows determination is correct. Suppose 100 CDC+ subjects are ascertained, of whom the ID fellow conrms 80 (a). What is the PV + of the CDC algorithm? (Hint: In this problem, Sympton= CDC+, Disease=Conrmed by ID-fellow.)

(b). What is the PV of the CDC algorithm?

(c). What is the sensitivity of the CDC algorithm? (Hint: You want to use Bayes Rule by switching the Sympton and Disease in Equation 3.9, i.e., Pr[Sensitivity] = Pr[Sympton|Disease] = Pr[Disease|Sympton]Pr[Sympton] Pr[Disease|Sympton]Pr[Sympton] + Pr[Disease|No sympton]Pr[No sympton].)

(d). What is the specicity of the CDC algorithm?