The online version of this Research paper by D J W McMinn and colleagues (BMJ 2012;344:e3319, doi:10.1136/bmj.e3319) suggested that the National Joint Registry provided raw patient level data to the authors. In fact, the dataset was requested from the NJR via an employee of Smith and Nephew at the request of the first author, and passed on to the authors for analysis. The paper also stated that the dataset included linkage to hospital episode statistics (PEDW in Wales). This dataset in fact was provided solely by the NJR and contained precise information on time to implant revision and time to death.
Cite this as: BMJ 2012;345:e6282