Regarding the issue of bias in RCTs, hope you will discuss the increasing use of complex combined and surrogate endpoints. This has led to degradation of information quality in cardiology and probably other fields. Also, run-in periods like they used in the ARNI pivotal trial - they seem de…
Regarding the issue of bias in RCTs, hope you will discuss the increasing use of complex combined and surrogate endpoints. This has led to degradation of information quality in cardiology and probably other fields. Also, run-in periods like they used in the ARNI pivotal trial - they seem designed to "weed out" subjects destined to have a bad reaction to the agent. Is there a legitimate use for them?
Finally, the deepest source of bias is that investigators have a strong professional stake in the outcome of an RCT. This is understandable, given the amount of time and effort invested and the differential professional consequences of a "positive" and "negative" trial, but if we are really acting as scientists, we should not care whether the null hypothesis is true or not, only what the truth is.
Great start, I will look forward to this series.
Regarding the issue of bias in RCTs, hope you will discuss the increasing use of complex combined and surrogate endpoints. This has led to degradation of information quality in cardiology and probably other fields. Also, run-in periods like they used in the ARNI pivotal trial - they seem designed to "weed out" subjects destined to have a bad reaction to the agent. Is there a legitimate use for them?
Finally, the deepest source of bias is that investigators have a strong professional stake in the outcome of an RCT. This is understandable, given the amount of time and effort invested and the differential professional consequences of a "positive" and "negative" trial, but if we are really acting as scientists, we should not care whether the null hypothesis is true or not, only what the truth is.