Deep Vein Thrombosis Standard anticoagulant therapy (to prevent blood clots) requires frequent laboratory monitoring to prevent internal bleeding.
A new procedure using rivaroxaban (riva) was tested because it does not require frequent monitoring. A randomized trial (Einstein-PE Investigators 2012) was carried out, with standard therapy being randomly assigned to half of 4832 patients and riva randomly assigned to the other half.
A bad result was recurrence of a blood clot in a vein. Fifty of the 2416 patients on standard therapy had a bad outcome, and 44 of the 2416 patients on riva had a bad outcome.
a. Test the hypothesis that the proportions of bad results are different for riva and standard therapy patients. Use a significance level of 0.05, and show all four steps.
b. Using methods learned in Chapter 7, estimate the difference between the two population proportions using a 95% confidence interval, and comment on how it can be used to evaluate the null hypothesis in part a.