Hi Students!
1. I've just posted HW #8 and answers to HW #7.
2. I want to clarify that when I calculated that it costs > $37,000 to prevent a course of oral steroids (which costs about $8) that I was not implying that the only bad thing about the course of oral steroids was its $8 cost. Of course the reason for giving the oral steroids is because the person was having trouble breathing, and perhaps was in danger of needing hospitalization. So preventing that level of symptoms is a goal of treatment. Furthermore, although data were not in the published article, the supplementary appendix did indicate that the group that got salmeterol had slightly lower use of "rescue medication." But here again, the effect size was small: the difference was a mean of 0.90 puffs per day with salmeterol+fluticasone vs 1.09 puffs per day with fluticasone alone. So important take home messages are still: 1) decompose composite outcomes, since the least important one may be driving the results and 2) quantify absolute risk reduction and cost, to avoid confusing clinical and statistical significance and get some idea of value.
3. Attendance. As I've said before. attending section is highly encouraged, but not required. That being said, if you are having any trouble in this course, you should come to class! At our lunch meeting today section leaders mentioned students who did not come to section but emailed them with questions about things that had been (or could have been) covered there. To me this seems like complaining of being hungry after not showing up for the nice dinner your mother made you. Can you see why it would feel like that to those of us working in the kitchen?
Best wishes!
Tom