I need help with 3, the rest is context. 3. As a secondary endpoint, you compare the actual value of CD4 (as a co tinous measure) at the end of the study (1yr) between 2 treatment groups.a. what kind of a test could you use to compare the CD4 count between the 2 groups, of you also want to adjust for covariance (age, gender, baseline, C4, time since diagnosis)b. What are 2 tests (different from your answer to pt a) that you could use to compare the CD4 count between the 2 groups if you did NOT feel you needed to adjust for covariance ( since this is a randomized trial, they should be reasonably well balanced)?c. Many of the patients are were very sick, and quite a few died from AIDS-associated Imfections or TB before 1 year. These patients do not have an observed CD4 count at the end of the study. Can you validly compare the distributions of CD4 counts for the 2different treatment groups, given these missing data? if yes, describe how. if no, explain not

(15 pts) The investigators from the previous question decided to stick with comparing two different ARV's; as the primary endpoint, they want to compare what proportion of people in each treatment group as CD4 count below 200 by the end of the study. The following table shows their observed data: CD4 = 200 at end of study Total 245 ARV 1 50 195 255 ARV 2 90 165 Total 140 360 500 What is the ARV1-to-ARV2 odds ratio for this data? What is the ARV1-to-ARV2 relative risk (also known as the Risk Ratio)? Odds Ratio: (50)(165)/(195)(90)=0.4700854701 Relative Risk: {(50)((50 + 195)}/ {(90)/(90 + 165)}=0.5782312926 b. If you don't know how the sampling was designed (based on ARV, based on CD4 count, or based on taking a group without regard to ARV or CD4 count), is it better to use the odds ratio or the relative risk? Why (briefly)? It is better to use the Relative Risk ratio as the study was conducted to predicated on whether or not the CD4 count falls below 200 by the conclusion of the study, thus it is regarded as a prospective study. Relative risk will be utilized because it is necessary to obtain results. Furthermore, the relative risk can only be calculated when total data is available. C. What is the null hypothesis, expressed in terms of whichever measure you chose in part b? Null Hypothesis: The relative risk ratio = 1 Alternative Hypothesis: The relative risk ratio # 1 I. Fill in the following table with the expected counts, under the null hypothesis that there is no association between ARV regimen and probability of CD4= 200 at end of study Total 68.6 ARV 2 176.4 71.4 245 Total 140 183.6 255 360 500 Using your answers to part (d), compute the chi-square statistic. The Chi Square Statistic value derived is X?=13.73406505