Morph's video: SWAT Design Animation

Transcript SWAT Design SWATs often need to be repeated in several host trials. The same intervention can be tested in several host trials running at the same or at different times. Putting the same SWAT into multiple host trials all running at the same time provides faster answers but requires coordination. To avoid research waste, it is important to confirm that the SWAT will add value to the evidence base. Randomisation SWATs can be randomised… or non-randomised. An example of a non randomised SWAT would be a before and after study. A SWAT can be randomised alongside a host trial… or separately from the host trial. If the SWAT is focussed on trial participants, randomisation of each participant is usually best. If the SWAT focuses on site processes, randomisation by site might be more appropriate. The type of randomisation does not need to be the same as the host trial. Outcomes and Analyses SWAT sample sizes are limited by the host trial. They may well contain a subset of the participants. SWATs should therefore be designed to allow results from multiple SWATs to be combined… to ensure there is enough data to answer the SWAT question. When replicating an earlier evaluation, the same outcomes and analysis techniques should be used… to make it possible to combine results in a meta-analysis. An overview of associated costs or cost analysis of the SWAT is also useful. This may include a breakdown of costs… or an additional cost per participant. Costs SWATS do not need to be expensive. For example they may test changes to existing trial documentation or processes. Costs to consider include: Materials and resources required, staff time to design the SWAT, randomise participants, prepare the intervention and to complete the analysis as well as costs to make findings publicly available. By applying these principles, SWATs can be impactful and cost-effective.