Types of Medication Errors by Process Stage

A wrong drug ordered at 2am and a correctly-ordered drug infused at the wrong rate are both “medication errors,” but they broke in different places in the medication-use process. Sorting errors by where they happened — not by how much harm resulted — is a separate classification system from the harm-severity scale most hospitals already run, and the two answer different questions for a medication-safety program.

Written for hospital pharmacy, patient-safety and quality staff who categorize reported medication events and need to route each one to the right improvement work, not just log it.

Two classification systems, not one

It’s easy to conflate “how severe was this error” with “what kind of error was this” because a single incident report often gets scored on both dimensions at once. They are genuinely separate axes:

Severity classification asks how much harm occurred or nearly occurred — the NCC MERP-style A-to-I harm gradient covered in our guide on patient safety event severity classification. It determines investigation depth and reporting obligation.
Type classification — the subject of this guide — asks where in the medication-use process the error originated: prescribing, transcribing, dispensing, administration, or monitoring. It determines which team owns the fix and which intervention actually targets the failure point.

A single event gets scored on both: a wrong-dose order that reaches the patient and causes temporary harm is simultaneously a prescribing-type error and a temporary-harm-severity event. Neither classification substitutes for the other, and a medication-safety program that only tracks severity has no way to tell whether its errors are clustering in ordering, order verification, dispensing, administration, or monitoring — which is exactly the information needed to decide where to spend improvement effort.

The medication-use process: five stages where an error can originate

ASHP’s Guidelines on Medication-Use Evaluation frame medication-use evaluation as a method that examines the full pathway a medication order travels — prescribing, preparation, dispensing, administration, and monitoring. Type classification uses this same pathway as its organizing structure, with an order-verification/transcribing step commonly broken out as its own category in taxonomies built for error reporting specifically (rather than the broader evaluation framing ASHP uses), because that step remains a distinct failure point even in largely electronic systems — a correctly-written order can still be entered, verified, or communicated incorrectly downstream of the prescriber.

The sections below work through each stage: what counts as an error there, and an illustrative example. The examples in this guide are illustrative composites written to show the mechanism of each error type — they are not real incident reports, and no specific institution, patient, or clinician is described.

1. Prescribing errors

A prescribing error occurs at the point the medication order is written or entered — the wrong drug, dose, route, frequency, or formulation is selected, or a relevant patient factor (allergy, renal function, interacting medication) is missed at the decision point. This is the stage targeted by clinical decision support, allergy/interaction checking in CPOE, and structured order sets.

Illustrative example: a prescriber orders a medication at an adult dose for a pediatric patient because the order was entered on a weight-based order set without a hard weight-entry requirement.

2. Transcribing / order-verification errors

This stage covers the order after it leaves the prescriber but before it reaches dispensing — historically the point where a handwritten order was copied onto a medication administration record, and in modern systems the point where a pharmacist verifies an electronic order or where an order is communicated between systems that don’t share a single record. A correct order can still be miscommunicated, misread, or altered here.

Illustrative example: a verbal order is read back and confirmed correctly, but the person documenting it selects the wrong drug from a look-alike pair in a dropdown menu.

3. Dispensing errors

A dispensing error happens in pharmacy: the wrong drug, strength, or formulation is selected and labeled, a compounding or preparation step is done incorrectly, or the wrong patient’s medication is filled. This is the stage that barcode verification at the point of dispensing, automated dispensing cabinet configuration, and look-alike/sound-alike (LASA) list controls are built to catch.

Illustrative example: two vials with similar packaging and adjacent storage bins are confused during a high-volume filling period, and the wrong vial is loaded into a dispensing cabinet pocket.

4. Administration errors

An administration error occurs at the bedside: wrong patient, wrong drug, wrong dose, wrong route, wrong time, or wrong rate at the point the medication is actually given. This is the largest category in most reporting systems, in part because it is the last of several checkpoints and the one closest to the patient. Barcode medication administration (BCMA) and independent double-checks for high-alert drugs are built specifically for this stage.

Illustrative example: an infusion pump is programmed at the wrong rate because the concentration on the bag label and the concentration selected in the pump’s drug library don’t match.

5. Monitoring errors

A monitoring error is a failure to order, perform, or act on the follow-up needed to confirm a medication is working as intended and not causing harm — a missed drug level, an ignored abnormal lab value, or a therapeutic response that isn’t checked at the interval the drug requires. Monitoring errors are easy to under-count because nothing happens at the moment of the error; the harm, if any, surfaces later.

Illustrative example: a patient is started on an anticoagulant with a required baseline and follow-up lab schedule, and the follow-up draw is never ordered because responsibility for scheduling it wasn’t assigned to anyone specifically.

Why the process-stage taxonomy matters for improvement work

Type classification is what makes root-cause categorization actionable at the program level, not just the individual-event level. A pharmacy quality committee reviewing a batch of low-harm events over a quarter can only decide where to spend limited improvement capacity if it knows whether those events are clustering at one stage or spread evenly across all five:

A cluster of prescribing errors around a specific drug class points toward clinical decision support or order-set redesign, and is the kind of pattern a medication-use evaluation (MUE) is built to surface.
A cluster of transcribing/verification errors around specific look-alike names points toward tall man lettering, dropdown-menu redesign, or removing verbal orders for high-risk drugs.
A cluster of dispensing errors points toward pharmacy workflow, storage separation, or barcode-verification gaps — the territory covered in our LASA list guide.
A cluster of administration errors points toward BCMA scan-compliance gaps, smart-pump library configuration, or staffing/interruption factors at the bedside.
A cluster of monitoring errors points toward closing a gap in who owns follow-up — often a medication reconciliation or handoff process issue rather than a single clinician’s oversight.

This is the same reason a hospital’s incident-reporting system should capture stage of origin as a structured field rather than leaving it to free text: aggregated by stage, the data tells a safety program which of the five parts of the medication-use process is under-defended, which is a different and complementary question to how severe any single event turned out to be.

High-alert medications — the drug classes most likely to cause serious harm when an error does occur — get layered safeguards across multiple stages precisely because no single stage’s controls are assumed sufficient on their own; see our guide on high-alert medications and safeguards for how those controls map onto specific stages of this same process.

How type and severity combine in practice

Neither axis is optional if a program wants a complete picture. A severity score alone tells a reviewer how urgently to respond to one event and whether it triggers a reporting obligation. A type classification alone tells a program where its process is weakest, but says nothing about how much any individual event mattered. Used together — every medication event tagged with both its process-stage type and its outcome severity — a program can answer both “what happened to this patient” and “where does our next improvement dollar go,” which neither classification answers by itself.

Frequently asked questions

What are the main types of medication errors?

Organized by where in the medication-use process they originate: prescribing errors, transcribing/order-verification errors, dispensing errors, administration errors, and monitoring errors. Some organizations fold transcribing into prescribing or dispensing depending on how their systems are configured; the underlying idea — classify by process stage — stays the same regardless of exactly how many named categories a given taxonomy uses.

Is this the same as the NCC MERP severity scale?

No. NCC MERP’s A-to-I scale (and the general-purpose harm-tier adaptations built on it, covered in our severity classification guide) classifies outcome — how much harm occurred or nearly occurred. The process-stage taxonomy on this page classifies origin — which part of the medication-use process the error came from. They are independent, complementary systems, and a single event is typically scored on both.

What is the most common type of medication error?

Administration errors are consistently the largest reported category in most hospital reporting systems, in part because administration is the last checkpoint before the patient and errors from earlier stages that weren’t caught also surface there. That doesn’t mean administration is where the underlying problem originated — a wrong-drug administration error can be the downstream consequence of a dispensing-stage mix-up that nothing caught in between.

Is a near miss a medication error?

A near miss is an error that was caught before it reached the patient — it’s still classified by the stage where the underlying mistake occurred (a wrong drug caught at the pharmacy-verification step is still a dispensing-type error, just one that didn’t reach the patient), even though its severity score is at the low end of the scale precisely because no harm occurred.

Types of Medication Errors: A Process-Stage Classification

Ask about Types of Medication Errors: A Process-Stage Classification