Cardiology, Radiology The conventions of cardiac echocardiography are derived from comprehensive studies performed by technicians and cardiologists. Traditionally, patients are positioned supine (i.e., lying on their back), although there is no evidence to support this practice. Clinicians in the emergency department (ED) are sometimes faced with acutely ill patients who cannot tolerate the supine position. For this reason, a study is planned to compare ultrasound studies performed while a patient is supine vs. while a patient is upright. Patients > 18 years of age who present to the ED with chest pain or shortness of breath and consent to the study will be enrolled. Each patient will receive two ultrasound studies, one in the supine position and one in the upright position. The order of administration of the two studies will be at random. Readers will review the images in a blinded manner and assess a number of clinical characteristics. An important endpoint is pericardial effusion (or fluid around the heart). Suppose 50 patients are enrolled in the pilot phase of the study. i. For 28 patients it is determined from both images that there is no fluid around the heart. ii. For 9 patients it is determined from both images that there is fluid around the heart. iii. For 10 patients it is determined from the supine image, but not the upright image, that there is fluid around the heart. iv. For 3 patients it is determined from the upright image, but not the supine image, that there is fluid around the heart. You can use the Distribution Calculators page in SALT to find critical values and/or p-values to answer parts of this question. 2 (a) If we focus on patients where there are different results for the supine and upright images [i.e., groups iii and iv above], what test can be performed to assess whether there is a significant difference in determinations for supine vs. upright images? O Yates-corrected chi-square test for 2 x 2 contingency table O McNemar's test for correlated proportions (normal theory test) Mantel-Haenszel test McNemar's test for correlated proportions (exact test) Correct R (b) Perform the test in (a) and report a p-value (two-tailed). (Use a = 0.05. Let p be the proportion of images where fluid is detected in the supine image but not the upright image out of the images where different results were obtained from the two positions.) State the null and alternative hypotheses. (Enter != for # as needed.) Ho: P = 0.5 correct