Intervention the term intervention[T.Intervention] to the left of the results indicates that intervention=1 is the numerator of the hazard ratio. In this case, the hazard ratio is 0.57 which indicates the rate of death and/or myocardial infarct diagnosis in the intervention group is about half of that in the control group.
The p-value is 0.502 and it is higher than 0.05. Therefore, it indicates that there is no statistically significant difference in the hazard ratio …show more content…
The survival in the control group at 273 days is about 0.62.
The absolute rate reduction is the cumulative incidence in the control group minus the cumulative incidence in the intervention group : 0.38-0.23= -0.15
2a. According to the data, compare with ‘smoker’ and ‘high cholesterol’, ‘overweight (body mass index >25)’ is the most likely to be a confounder. The reason is this variable has a large unequally distributed between two treatment groups (intervention group is 48% and control group is 56%), which means, it is associated with the exposure.
2b. If we perform this unplanned subgroup analysis, the people in smoker subgroup going to have higher risk than non-smoker subgroup, because the confounder may associated with exposure. In addition, because this is an unplanned subgroup analysis, there are unequally distribution between smoker subgroup and non-smoker subgroup, and study not randomize. Two statistical issues will be cause. The subgroup analyses can increase type 1error and type 2. Type 1 error is defined as the texts detect an effect that is not existent. Type 2 error is defined as the text failing detects an effect that is