Table 1 CONSULT-ENABLE key findings to inform SWAT design (expanded from Shepherd et al 2022 [12])

SWAT design component and key findings

Selection of host trials

Whilst the DA was considered applicable to all study types, it was thought to be particularly useful for more complex and burdensome studies that have a higher decisional burden for consultees, and less suitable for acute or emergency trials with short recruitment windows.

Consent process

Researchers stressed the need to reduce the informational and consent burden for SWAT participants by avoiding a ‘double consent process’. Requiring an extra layer of consent for the SWAT might deter people from participating in the SWAT and acting as a consultee/legal representative host trial. Given the low-risk nature of the study and the impracticability of gaining consent prospectively prior to randomisation or receiving the intervention, consent to providing data was considered appropriate. This could be achieved through the return of the questionnaire indicating consent to participate. Researchers stressed the importance of harmonising information provided to family members about the host trial and SWAT in order to create a ‘whole package’ and improve clarity.

Level of randomisation

In addition to the usual issues relating to decisions about the appropriate level of randomisation (i.e. cluster or individual), specific factors to consider in trials involving participants with impaired capacity include that more than one researcher may be involved in seeking consent or consultee involvement, and in settings such as care homes there may be one or multiple researchers recruiting at each care home. Therefore, particular attention to allocation processes is needed in order to reduce the burden for host trials and minimise the risk of contamination.

Intervention delivery

The method of delivery of the DA needs to be aligned with the host trial recruitment process as, depending on the context, trials may approach consultees/legal representatives in person or by phone or may post out information with families sometimes having the option to speak to a researcher or just return the consent/declaration form by post with no contact with researchers.

Data collection

As with other study processes described previously, data collection processes need to minimise the additional burden for researchers and participants involved. This must be balanced with the need to collect information not normally collected by the host trial such as personal information from a consultee.