Abstract
We are pleased that our initial perspective article in the BJD have sparked discussions. As indicated in our initial paper, our main critical point is the exponential growth in publications relating to drug survival as an efficacy assessment based on non-randomized registry data (Figure 1). Indeed, we ourselves have performed such drug survival analyses.1 While we agree that methods such as propensity score matching could even out some of the between-drug variations in patient characteristics, propensity score matching is a rare sight in drug survival studies.
This article is protected by copyright. All rights reserved.
http://ift.tt/2wiZsJO
Δεν υπάρχουν σχόλια:
Δημοσίευση σχολίου