Table 1 List of inclusion and exclusion criteria for this systematic review. RCT - Randomised Controlled Trial
| Â | Inclusion criteria | Exclusion criteria |
|---|---|---|
Publication type | Peer-reviewed methodological papers whose focus is to compare two or more existing methods under a sufficient simulation study. | Non-peer reviewed articles, books or book chapters, theses or other grey literature such as conference proceedings. |
Focus | The methodological topic of interest is non-compliance to the randomised intervention, which may be by participants in the intervention or control groups. This compliance could be described as all-or-nothing or time varying/partial. Some papers refer to non-adherence, but we are considering these terms to be interchangeable in this paper. | Papers that focus on issues such as missing data or the combination of these issues with non-compliance. Papers whose focus is a novel method rather than comparison of existing methods (e.g., that describe/reference a current method and propose an extension to it or propose a new method). |
Setting | The methods considered are explicitly applied to account for non-compliance in the setting of a superiority RCT. | Papers that focus on an observational setting. Papers that consider a non-inferiority or equivalence setting. Methods based on aggregated data such as meta-analysis. |
Simulation study | A simulation study was defined as ‘sufficient’ based on the following criteria: o The simulation study clearly states its objectives and gives a description of how the simulation was conducted/the nature of the simulated data. o The simulation study compares at least two existing methods that aim to account for non-compliance and estimate a point estimate of the intervention effect. o Existing methods refers to those that have not been proposed in the paper of interest and the authors have referenced previous work when describing the method. o The authors consider several non-compliance scenarios, such as varying the proportion or type of non-compliance. o Amongst performance measures, at least the bias of methods is reported or can be easily deduced. |  |
Publication date | Papers published from databases inception to 30th November 2022. | Â |
Publication language | Papers published in the English language. | Â |