Statistical Reasoning in public Health - Math Problem Example

The children were followed for one year after randomization, and monitored for AOM during this period. 262 children were randomized to the vaccine group, and 150 of these children experienced at least one incident of AOM during the follow-up period. 134 children were randomized to the placebo group, and 83 of these children experience at least one incident of AOM during the follow-up period. (a) Report a 95% confidence interval for the proportion of children experiencing at least one incident of AOM during the follow-up period in: 1) the group randomized to receive AOM 51.2% to 63.1% 2) the group randomized to receive a placebo 53.5% to 69.7% 3) How do these 95% CI’s compare?

(similar range of values? Overlap?) There is an overlap of values. (b) 1) Report a 95% confidence interval for the difference in proportions of children experiencing AOM over the follow-up period. (you may choose the direction of comparison – ie: placebo to vaccine or vice-versa) . Placebo to vaccine: 47.3% to 63.1% 2) Interpret the confidence interval in a sentence. At the 95% confidence level, those who use a placebo have 47.3% to 63.1% chances of experiencing AOM over the follow-up period compared to those who have taken the vaccine. . Yes, the p-value is consistent. (d) 1) Give an estimate of the relative risk of AOM (in the follow-up period) for children in the vaccine group compared to those in placebo. 0.92 2) Interpret this estimated relative risk.

Those who are in the vaccine group are .92 times as likely to suffer from AOM than those in placebo. (e) 1) Give an estimate of the odds ratio of AOM (in the follow-up period) for children in the vaccine group compared to those in placebo. 0.82 or 50:41 2) Interpret this estimated odds ratio. Those who are in the vaccine group are .82 times more likely to suffer from AOM than those in placebo. 3) How does it compare in value to the estimate of relative risk from part d1? They are almost similar. (f) 1) Is this a randomized study? Yes 2) What does this study design suggest when translating the statistical result from part (b) into a substantive/scientific conclusion?

(Note: this is an extension question, we have not covered this in detail, I just want you to think about it) A randomized study minimizes allocation bias and thus makes the statistical result from part (b) a substantive, unbiased, scientific conclusion 2. A study was done to investigate whether there is a relationship between survival of patients with coronary heart disease and pet ownership. A representative sample of 101 patients with CHD was taken. Each of these patients was classified as having a pet or not and by whether they survived one year following their first heart attack.

Of 52 pet owners, 50 survived. Of 49 non-pet owners, 28 survived. Suppose you were interested in doing a statistical analysis of these study results. Answering the follow questions to help you with this goal! (g) Using Stata (or the posted Stata output) , report