Narcotic count discrepancies
A narcotic count discrepancy is a mismatch between the recorded quantity of a controlled substance and the quantity physically present. Most are caused by documentation timing rather than by diversion — a waste recorded late, a count taken during handover, an administration charted after the count — but the investigation process has to work as though any of them might not be.
The organisational risk is not the discrepancy. It is the habit of resolving every discrepancy the same way.
What causes discrepancies?
Ordered by frequency, which is roughly the inverse of how much attention each receives.
Documentation timing
The dose was given and charted afterwards, or the waste was documented at the end of a sequence of tasks. The count falls in the gap. This is the majority of discrepancies in most organisations and it is a process artefact rather than a loss.
Shift-change counting
Counts performed at the busiest, least attentive moment of the day, by two people who both want to leave or have just arrived. Miscounts here are ordinary and entirely predictable.
Waste not witnessed at the time
The remainder is discarded and the witness signature is collected later. The record says witnessed; the event was not. This is both a common cause of genuine discrepancies and the primary concealment route for diversion, which is what makes it the important one.
Returns handled informally
A dose removed and not given, put back or into a bin without documentation at the moment. The physical count and the record diverge until someone reconciles them from memory.
Diversion
The least frequent cause and the one every process exists to catch. It usually presents as a pattern rather than as a single discrepancy, which is why individual-discrepancy investigation alone will not find it.
How should a discrepancy be investigated?
With a defined process that runs the same way every time, because the value is in consistency rather than in intensity. An organisation that investigates hard when someone feels suspicious and closes everything else quickly has a process driven by intuition, which is both unfair and ineffective.
Reconstruct the transaction sequence first: removals, administrations, wastes and returns for that drug and that period, in order. Most discrepancies resolve here, visibly, as a timing gap — and the resolution is documented as such rather than as "resolved".
Where it does not resolve, the question is not "who took it" but "which record is missing". That framing keeps the investigation on the evidence and avoids the social cost that stops these investigations from happening at all.
And record the resolution category, not just the closure. An organisation that knows what proportion of its discrepancies were timing, miscount, or unresolved has a signal. One that records everything as "resolved" has destroyed the only data that would show a change.
Why does the resolution habit matter more than the individual case?
Because the concealment mechanism for diversion is usually a normal-looking resolution rather than a hidden transaction. If discrepancies are routinely closed as documentation errors without reconstruction, then a diverted dose produces a discrepancy that is closed as a documentation error — and the process has functioned exactly as designed while the loss continues.
This is the uncomfortable part: the routine dismissal is reasonable on any individual case, because most discrepancies genuinely are documentation errors. The failure is statistical rather than judgmental, and it is invisible from inside any single decision.
The defence is that the reconstruction is always performed and always recorded, however quickly it resolves. It is a small amount of work per discrepancy and it converts a dismissal into a finding.
What removes most of the problem?
Counting at the transaction rather than at the shift. When the quantity updates at each removal, administration, waste and return, there is no window in which the record and the shelf can drift, and a discrepancy points at a specific event with a name and a timestamp rather than at eight hours and a whole team.
Documenting waste at the point of disposal, with the witness present. This is the hardest one operationally and the most valuable.
And reconciling dispensing against administration automatically, so a removal with no corresponding administration is surfaced without anyone comparing two systems by hand. That comparison is the single most useful controlled-substance report available and most organisations cannot produce it routinely, which is worth knowing before it is needed.
Frequently asked
- What causes narcotic count discrepancies?
- Most often documentation timing — a dose charted after the count, or a waste documented at the end of a task sequence. Then shift-change counting performed at the least attentive moment of the day, waste witnessed after the fact rather than at disposal, and returns handled informally. Diversion is the least frequent cause and the one every process exists to catch, and it usually presents as a pattern rather than a single discrepancy.
- How should a narcotic discrepancy be investigated?
- With a defined process that runs identically every time. Reconstruct the transaction sequence — removals, administrations, wastes, returns — for that drug and period. Most discrepancies resolve visibly as a timing gap. Where one does not, the question is which record is missing rather than who took it, which keeps the investigation on evidence. Record the resolution category, not just that it closed.
- Why is routinely closing discrepancies as documentation errors dangerous?
- Because the concealment mechanism for diversion is usually a normal-looking resolution. If discrepancies are routinely closed without reconstruction, a diverted dose produces a discrepancy closed as a documentation error, and the process functions exactly as designed while the loss continues. The failure is statistical rather than judgmental — the dismissal is reasonable on every individual case.
- How do you reduce narcotic count discrepancies?
- Count at the transaction rather than at the shift, so there is no window in which the record and the shelf can drift and a discrepancy points at a specific event with a name and timestamp. Document waste at the point of disposal with the witness present. And reconcile dispensing against administration automatically so a removal with no corresponding administration surfaces without manual comparison.