Background Guideline implementation through electronic point-of-care alerting has been shown to be effective. The best displays of these alerts have not been well studied.
Objectives To assess the effect of wording and presentation of electronic alerts on insulin ordering by providers caring for out-of-control diabetics on 2 or more oral hypoglycemic agents.
Methods An electronic message to the provider caring for an out-of-control diabetic was displayed. Prior to randomisation, a generic message was presented to all providers. Health centres were then randomised to 1 of 4 specific messages recommending insulin. Messages differed by wording (active/passive voice) and presentation (black text/red and black text). A 2-arm RCT was then performed where health centres were randomised to a complete absence of any alert or to one of the specific messages.
Results The generic alert triggered 56,878 times. Providers prescribed insulin 5.11 % of the time. During the 4-arm RCT, the alerts triggered 11,744, 11,826, 11,742, and 11,554 times and insulin was prescribed 5.17%, 5.01%, 5.11% and 5.20% respectively. There were no statistically significant differences amongst the 4 rates (p = 0.67) nor was there a statistically significant difference between insulin ordering with the generic message compared to insulin ordering with the more specific messages (p = 0.62). For the 2-arm RCT, insulin ordering remained at 5% for the specific messages and was 0% where there was no alert (P<0.05)
Lessons for Guideline Developers, Adaptors, Implementers, and/or Users Extra effort to craft wording and develop eye-catching electronic alerts may not be worthwhile. Guideline adherence was improved by an electronic alert.
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.