Waiting room deterioration between reassessments

1. System failure identification

   A patient triaged as stable deteriorates while waiting. Reassessment is scheduled by acuity band, so a mid-band patient may go an hour without observation during exactly the period their condition changes. The failure is one of scheduling attention, not of clinical knowledge.

2. Data pipeline diagnostics

   Vital signs exist only as discrete timestamped rows. Nothing computes the interval since the last observation, so a patient overdue for reassessment looks identical in the record to one just seen. The record answers what the last reading was, never how long ago it was taken.

3. Predictive computational model

   Derive an observation-overdue ratio per waiting patient: time since last vitals divided by the interval their acuity band requires. Rank the waiting room by that ratio rather than by arrival time. The ratio is comparable across bands, so one ordering serves the whole room.

4. Workflow integration

   Make the ranking the default sort on the waiting room list. It replaces a sort order rather than adding a notification, so it costs no additional attention to read. Changing a default is the cheapest intervention available in a busy department.

Back to the case study database