PrEDICT — Predicting Emergency department Delirium with an Interactive Computer Tablet

What this tool is and how to interpret its evidence

PrEDICT stands for Predicting Emergency department Delirium with an Interactive Computer Tablet. It is a tablet-based serious game developed to flag delirium-related cognitive and motor-performance change in older emergency patients. Participants tap visual targets across four difficulty levels while the application records time and accuracy automatically. The abstract does not give a stable total completion time or the numerical threshold, so neither should be invented. App guidance simplifies delivery, but staff must consider vision, motor ability, language and capacity to use a tablet.

Lee and colleagues conducted a prospective observational study of 203 emergency-department patients aged 70 or older who had not been recognised as delirious in routine care. Critically ill people and those with communication barriers, visual impairment or inability to use the tablet were excluded. Sixteen participants had unrecognised delirium. Overall completion was 196 of 203, and 14 of the 16 delirium cases completed the game. A threshold derived in that same cohort produced sensitivity of 100%, specificity of 59.7% and an area under the curve of 0.86.

PrEDICT is promising and genuinely designed for delirium case-finding, but it remains an emerging component screen. The perfect sensitivity estimate rests on only 16 cases and lacks external validation. Exclusions also matter when judging real-world usability. The page should report these design limitations, link to the primary study and avoid implying that target-tapping performance establishes acute onset, fluctuation or a clinical diagnosis.