Appraising 2-armed randomised controlled trials
It should be noted that the perfect RCT has yet to be undertaken. Any RCT can be criticized for some methodological issue. A key issue in critically appraising studies is that, while it can be thought of in terms of being an academic exercise, the reason why we are doing it is to evaluate what weight we can put on the findings, and how far we can generalize the results from trials into routine practice to inform clinical care. These pragmatic issues always need to be borne in mind when critically appraising study data.
There are numerous checklists available, none of which is perfect. Individual issues may arise that are relevant to one study only. Alternatively, certain disease states may raise methodological issues that affect studies in that field and are specific to that disease state. Hence, no checklist can be entirely comprehensive and can only be regarded as a broad framework to apply. However, at the most general level, three possible scenarios arise when critically appraising the quality of an RCT:
- methodology sound
- methodology suboptimal
- methodology unsound/fatal flaw
The first assessment would be whether a study attained the explicit minimum quality criteria (that is, in terms of the minimum acceptable size, level of blinding [if blinding is possible], length of follow-up, etc).
The following framework forms an initial aide memoir for assessing RCTs, but other issues will arise on a study by study basis.
Appendix 1. Framework for assessing RCTs