Types of medication errors
Medication errors are most usefully classified by the stage they originate in: prescribing errors, transcription errors, dispensing errors, administration errors and monitoring errors. Classifying by stage rather than by outcome is what makes the taxonomy actionable, because the stage determines which control would have caught it.
Two errors with identical outcomes can need completely different fixes. The outcome tells you how bad it was; the origin tells you what to do.
What are the categories?
Each is defined by where it starts, with the control that addresses it.
Prescribing errors
The wrong drug, dose, route or frequency was ordered, or an order was placed despite an allergy, interaction or contraindication. Caught by clinical decision support at ordering and by pharmacist verification. Not caught by anything downstream — a closed loop executes a wrong order perfectly.
Transcription errors
The order was correct and was copied wrong — onto a chart, into a different system, or from a verbal instruction. Caught by eliminating the copying rather than by checking it, which is why structured order capture removes this category rather than reducing it.
Dispensing errors
The wrong product, strength or quantity left the pharmacy or the cabinet. Caught by barcode verification at dispensing and by packing verification. Look-alike and sound-alike products are the dominant mechanism.
Administration errors
The right dose reached the wrong patient, the wrong dose reached the right patient, or the timing or route was wrong. Caught by the bedside scan — this is the only category barcode administration primarily addresses.
Monitoring errors
The medication was correct and correctly given, and the response was not observed — a level not checked, a side-effect not acted on, a therapy not stopped when it should have been. Caught by nothing in the dispensing chain at all, which is why it is the category most often missing from a medication safety programme.
What about omissions?
Omission — the dose that was never given — deserves separate mention because it is systematically under-counted. An error of commission produces an event; an error of omission produces an absence, and absences are not reported by the people who caused them because they usually do not know.
The reasons are mundane. The patient was off the unit for a scan, the medication was unavailable, the round ran out of time, the patient was asleep and nobody returned. Each is individually reasonable and collectively they represent a substantial share of the doses a hospital intends to give and does not.
An eMAR changes this more than most safety technology, and quietly. A blank on a paper chart is ambiguous; a required reason field on a missed administration turns an absence into a record. That is why omission rates often appear to rise after an eMAR goes live, and it is worth telling a board in advance so the rise is read correctly.
Why classify by stage rather than by severity?
Because severity is largely a function of which drug was involved and how lucky everyone was, and neither of those tells you what to change. The same administration error involving a vitamin and an anticoagulant produces two very different reports and identical corrective actions.
Stage-based classification also stops the most common analytical mistake in this field: attributing an error to the last person who touched it. An administration error caused by two look-alike products stocked adjacently is recorded against a nurse and belongs to procurement and stocking. Classifying by origin rather than by proximity is what surfaces that.
The practical test when reviewing an incident: at which stage did the medication first stop matching what the patient needed? Everything after that point is propagation rather than cause.
Frequently asked
- What are the main types of medication errors?
- Prescribing errors, transcription errors, dispensing errors, administration errors and monitoring errors — classified by the stage they originate in. That classification matters because the stage determines which control would have caught it: a closed loop executes a wrong order perfectly, and barcode administration primarily addresses only the administration category.
- Why are medication omissions under-reported?
- Because an omission produces an absence rather than an event, and the people who caused it usually do not know they did. The causes are mundane — the patient was off the unit, the medication was unavailable, the round ran out of time, the patient was asleep. An eMAR with a required reason field on a missed dose converts the absence into a record, which is why omission rates often appear to rise after go-live.
- Should medication errors be classified by severity or by stage?
- By stage, for corrective action. Severity is largely a function of which drug was involved and how lucky everyone was, and it does not tell you what to change — the same administration error with a vitamin and an anticoagulant produces different reports and identical fixes. Stage classification also prevents attributing an error to the last person who touched it when the cause was stocking or procurement.