Log in with your email address username.


[Correspondence] Outcomes in the trial registry should match those in the protocol

We thank Andrew Davidson and colleagues for their reply (April 16, p 1615)1 to our letter on the GAS trial.2 The authors assert that most of the outcomes we considered non-prespecified are in fact prespecified in a protocol that was submitted to The Lancet after trial commencement but not published, and that a summary protocol was subsequently published that did not contain all prespecified outcomes.