First be aware of their likely limitations as published which are summarised well in On Limitations. Bandolier 161 2007;14(7):1-4.
EPIQ at the University of Auckland provide interactive excel spreadsheets to guide critical appraisal of RCTs and other study designs.
Another limitation is that trials usually take about 2 years to be published. With the increasing frequency of pre-registration the never published group becomes easier to evaluate. Good clinical practice in drug trials will increasingly lead to more data being in the public domain.
SERIBCOAA or (Aims. Authors’ conclusions. Methods (as below). Outcomes. Analysis.)
- Selection of patients
- Ethics, informed consent
- Intention to treat, drop-out rate, compliance
- Blinding, bias
- Cost implications of the different treatments
- Outcomes - relevance, standard? Objective?, appropriate to the aims? Open to biases?
- Analysis - pragmatic or explanatory; how are drop-outs etc. handled?
- Aims of study - stated?, the right aims?, achieved?