It's 3am. A nurse has been on her feet for nine hours. She's managing twelve patients, two of whom need close monitoring. She hasn't eaten since 10pm. Her phone buzzes - a reminder that her mandatory infection control refresher is due.
She opens the module on the ward computer. Thirty slides. Auto-advancing audio. A quiz at the end.
Who designed this for her?
The Alert-Person Assumption
Almost all eLearning is designed for someone who is alert, seated, undistracted, and has a block of uninterrupted time. In healthcare, that person doesn't exist. The workforce that needs training the most - frontline clinical staff - has the least capacity to engage with it.
Night shift workers operate with reduced cognitive function. This isn't opinion. Decades of research on circadian rhythm disruption confirm that alertness, working memory, and decision-making all degrade significantly during overnight hours. Between 2am and 6am, cognitive performance can drop to levels comparable with mild intoxication.
We wouldn't design training for someone who's been drinking. But we routinely design it for people who are equally impaired by fatigue.
What Cognitive Load Means at 3am
Cognitive load theory matters more at 3am than at 3pm. During the day, a learner can compensate for poor instructional design through effort. They can re-read a confusing paragraph. They can hold multiple concepts in working memory while making connections. They have cognitive reserves to spend.
At 3am, those reserves are depleted. Every unnecessary word, every decorative image, every ambiguous instruction takes a disproportionate toll. The learner doesn't have the capacity to work around bad design. They need the design to do the work for them.
This means:
Fewer words. Strip every sentence to its essential meaning. If it can be cut without losing the point, cut it.
Clearer structure. One idea per screen. Obvious visual hierarchy. No hunting for the key information.
Simpler interactions. If the interaction requires the learner to remember something from three screens ago, it's too complex for a fatigued brain.
Shorter sessions. Five minutes maximum. Ideally two to three. The module should be completable in the time it takes for a system to load or a patient to settle.
The Decision Architecture
The most dangerous training failure isn't forgetting a fact. It's making the wrong decision under pressure. And decisions degrade faster than recall when fatigue sets in.
This is why scenario-based training matters more for night shift workers than for anyone else. They don't need to memorise the escalation pathway. They need to practise using it when they're tired and the situation is ambiguous.
The Standard
Healthcare training should be designed for the worst conditions the learner will face, not the best. If it doesn't work at 3am, it doesn't work.
Nic