The expected Treatment A deaths are at month 1, at month 2, and at the three tied deaths at month 5. ThusThis is not half of the five deaths because right censoring changes the treatment proportions in successive risk sets. For Treatment B, and . One observed-to-expected relative-risk estimate is therefore
Articles by others on the same topic
There are currently no matching articles.