Reliable and valid methods should have been used to determine the outcomes of treatments.
Look out for outcomes that were not assessed reliably in treatment comparisons. Some outcomes are easy to assess such as births and deaths. Factors such as quality of life are more difficult to assess.
It is preferable to use standardised methods which have previously proven reliable. An example of this would be the Glasgow pain scale which provides a standardised process to assessing pain, replicable by anyone, at any time, and has been rigorously tested and improved to ensure reliability.
It is also important to assess whether the outcomes have clinical significance i.e., will the outcomes measured in the study improve the resolution of a condition or quality of life for the patient?
REMEMBER: Think about whether the study methodology can be proven to be reliable.