I have no idea how to do part 1D on this assignment (the website is calling it 7e38a841bc6f722b44e265058b05abdc_a02ad7f18d4780407dd32225fc9d0). I'm also unclear on the last part of 1c ii (the clear interpretation part). Help would be much appreciated!

Molly Thompson CHS 786 Assignment VII April 24, 2020 13.1 In a clinical trial of patients with respiratory illness, 111 patients from two different clinics were randomized to receive either placebo or an active treatment. Patients were examined at baseline and at four visits during treatment. At each examination, respiratory status (categorized as 1 = good, 0 = poor) was determined. These data are from Koch et al. (1990), and are reported in Davis (1991) and Stokes et al. (1995). The main objective of the analyses is to understand the joint effects of treatment and time on the probability that respiratory status is classified as good. It is also of interest to determine whether the effect of treatment is the same for patients from the two clinics. The raw data are stored in an external file: respir.dat Each row of the data set contains the following eight variables: ID Clinic Treatment Y0 Y1 Y2 Y3 Y4 Note: The respiratory status response variable Yj is coded 1 = good, and 0 = poor, at the jth occasion. The categorical (character) variable Treatment is coded A = Active drug, P = Placebo. The categorical variable Clinic is coded 1 = clinic 1, 2 = clinic 2. 13.1.1 Ignoring the clinic variable, consider a model for the log odds that respiratory status is classified as good, including the main effects of treatment and time (where time is regarded as a categorical variable with five levels), and their interaction. Use generalized estimating equations (GEE), assuming separate pairwise log odds ratios (or separate pairwise correlations, if available software does not permit the withinsubject association to be parameterized in terms of log odds ratios) among the five binary responses. Construct a test of the null hypothesis of no effect of treatment on changes in the log odds that respiratory status is classified as good based on the empirical standard errors. Since we are interested in \"changes\" over time and this was a randomized trial I decided not to include the main effect of treatment in the model. data respir; infile '\\\\tsclient\\E\\Cohort Data\\HW7\ espir.txt'; input ID Clinic Trt$ Y0 Y1 Y2 Y3 Y4; A=0;P=0;Treatment=0; if Trt='A' then A=1; if Trt='P' then P=0; Treatment=A+P; run; data respirLong; set respir; y=y0; time=1; visit=1; y=y1; time=2; visit=2; y=y2; time=3; visit=3; y=y3; time=4; visit=4; output; output; output; output; 1 y=y4; time=5; visit=5; output; drop y0 y1 y2 y3 y4 A P; run; PROC GENMOD descending data = respirLong; class id time; model y= time Treatment*time/dist= bin link = logit; repeated subject = id /withinsubject = time type = un covb corrw; run; 13.1.2 What conclusions do you draw about the effect of treatment on changes in the log odds? Provide results that support your conclusions. The results of the analysis in 13.1.1 showed no difference in the log odds at the baseline measurement (time 1) which suggests that the randomization of participants to treatments was successful in ensuring the two groups were equal at the start. After the baseline measurement, the difference between log odds among the groups was significate at times 2, 3, and 4 (in all cases p 0.0396), and this difference was increasing from time 2 to time 4. However, at time 5 the difference in log odds between the treatment and placebo groups decreased and was no longer statistically significant (p = 0.0704). If time allows try to talk specifically about log odds 13.1.3 Patients in this trial were drawn from two separate clinics. Repeat the analysis for Problem 13.1.1, allowing the effects of treatment (and, possibly, time) to depend on clinic. 2 (a) Is the effect of treatment the same in the two clinics? Present results to support your conclusion. Because which patients go to which clinic could not be randomized, and may have been different in a way that impacted the study outcome, I think it's necessary to include the main effect of clinic when investigating the relationships between clinic, treatment, and time. Consequently I used the following model: PROC GENMOD descending data = respirLong; class id time(ref='1')clinic; model y= time clinic clinic*treatment clinic*time Treatment*time/dist= bin link = logit; repeated subject = id /withinsubject = time type = un covb corrw; run; 3 The clinic*treatment interaction (i.e. the term that describes differences in the treatment effects between the two clinics) was insignificant suggesting that we don't need to allow the effect of treatment to depend on clinic. Similarly, The time*clinic interactions were also insignificant, suggesting that the outcome at the different clinics is not different over time. (b) Find a parsimonious model that describes the effects of clinic, treatment, and time, on the log odds that respiratory status is classified as good. For the model selected, give a clear interpretation of the estimated regression parameters for the final model selected. I removed the two non-significant (not very useful) interaction terms, clinic*treatment and clinic*time, but kept clinic in the model as the two clinics did have different levels of the outcome. My final model was: PROC GENMOD descending data = respirLong; class id time(ref='1')clinic; model y= time clinic time*treatment/dist= bin link = logit; repeated subject = id /withinsubject = time type = un covb corrw; run; 4 13.1.4 For the final model selected in Problem 13.1.3, construct a table of the estimated probabilities that respiratory status is classified as good as a function of both time and treatment group (and, possibly, clinic). What do you conclude from this table? For 13.1.4 - seems straightforward to get from odds to pp and put it in a table. 5