Van de Velde [Van de Velde 2018] provides an overview of the results of 66 randomised head-to-head controlled trials (RCTs) exploring which technical features make a DSS more effective in changing user behaviours or improving patient outcomes. Table 6 below summarises the findings. Note that these findings relate primarily to decision support embedded in electronic care systems, rather than the other forms of decision support also delivered by the Right Decision Service.

Table 6: Technical factors that improve the effectiveness of decision support, with size of the evidence base and amount of improvement.

Redrawn from Van de Velde et al [2018]

* See Table 7

~ For example, patient information leaflets or giving them the same advice

This systematic review identified several other design features that may lead to improved effectiveness, but require more investigation before we can be sure of this:

• Requiring the user to respond to advice
• Providing an action button with the advice (e.g. to prescribe the suggested drug or order the suggested test)
• Making the advice more explicit
• Requiring the user to enter additional data to obtain the advice
• Giving the advice before the decision is taken (though it seems so obvious that it seems frivolous to mount an RCT to investigate it)
• Citing the source of the evidence or guideline on which the advice is based.