Analysis of Long-term Benefits of Intensive Blood Pressure Control —Reply

In Reply Mr McCaw and colleagues point out the potential utility of alternative approaches to analysis of survival data when the intervention effect may vary over time, as appears to be the case for the outcome of probable dementia in SPRINT. The proportional hazards model is known to be robust to certain departures from the assumption of constant proportionality, and when our analysis was prespecified in 2010, we did not expect to observe an increasing separation of the incidence curves late in the follow-up period. We were careful to note the increasing difference in our report of results from our prespecified analysis.
Source: JAMA - Category: General Medicine Source Type: research