Every automated dispensing cabinet keeps a log of its overrides, the times a nurse pulled a medication before a pharmacist reviewed the order. In the imagined version of the process, this pathway is rare, urgent, and justified, an emergency valve for the moments when waiting for review would itself be dangerous. Then you open the log and find the overrides running at hundreds a month, and clustered not around emergencies but around the exact hours when pharmacy review is slowest. The emergency valve has quietly become the ward’s standard workaround for turnaround time.
This is one of the most instructive documents in a hospital, because it is a running, timestamped record of a safety barrier being removed on purpose, by good people, for understandable reasons. Each override is a dose given with a verification step voluntarily skipped. Nobody is being reckless. They are adapting to a real constraint, a pharmacist review that does not keep pace with the ward’s rhythm, and the override is the pressure valve that lets the day continue. The log is not a list of rule-breakers. It is a map of where the process and reality have come apart.
Which is why the instinctive response is the wrong one. Faced with a high override rate, the reflex is to tighten override discipline: more warnings, more justification fields, more reminders that overriding is a serious act. All of that aims at the symptom and leaves the cause untouched, so the workaround simply reappears in a slightly more effortful form, and the barrier stays down. The fix that actually moves the number is upstream and unglamorous. Make pharmacy verification keep pace at the hours the log points to, and the reason to override mostly disappears, which means the barrier stays up without anyone having to be disciplined into keeping it there.
There is a second gift in the log, once you stop reading it as a disciplinary record. The override rate is a ready-made measure. Change the turnaround, watch the rate fall, and you have closed the loop that most improvements never close, with a number you did not have to invent because the cabinet was keeping it all along.
The wider habit is worth carrying everywhere. When you find a workaround running at scale, resist the urge to stamp it out and ask first what condition made it rational. People do not remove safety barriers for fun. They remove them because the barrier, as designed, has started to cost more than it appears to be worth in the middle of a real shift. Fix that, and the workaround retires itself. Attack the workaround alone, and you have declared a war of attrition against your own staff, on ground of their choosing, that you will lose one busy night at a time.
This article accompanies the book Before Harm Happens.