Barcode medication administration — and why it should not be the first thing you buy.
Barcode medication administration software checks a scanned patient wristband and a scanned medication package against the active order before a dose is given, and writes the result to the medication record. It is the only widely deployed mechanism that verifies the physical dose against the physical patient — and it only works when the dose it scans carries an identity that something put there.
A scan performed on a hand-assembled package confirms that a nurse scanned a package. It does not confirm what is inside it.
What does BCMA reliably catch?
Wrong patient, which is the strongest claim it makes and the most valuable one. That error is otherwise close to uninterceptable: the nurse is confident, the drug is correct for somebody, and nothing about the moment looks unusual. Replacing a recognition judgement with a comparison between two identifiers is the whole mechanism, and the comparison does not become less careful at the end of a long shift.
Wrong drug, where packaging is what confused the human. Look-alike and sound-alike products are indistinguishable to a tired person and entirely distinguishable to a scanner.
Discontinued medication and duplicate doses, because the check runs against the live order rather than against a printed chart that reflects the order set as of the last reprint.
What it does not catch: a wrong order — a closed loop executes a verified mistake faithfully. Route, because the package identifies the drug and not where it went. And whether the patient actually swallowed it, since administration is an act by a nurse rather than an event in a patient.
What has to be true before it works?
Doses that carry a readable barcode linked to the order
The single decider. A dose split in a med room, repackaged by hand, decanted or crushed has lost its identity, and a scan performed on whatever is left reads a container rather than the medication being given. This is why packaging sits upstream of BCMA in any sensible sequence.
Wristbands that scan reliably, every time
Faded thermal printing, bands under dressings or oedema, bands on a limb with a line in it. In the units where identification matters most, this fails most — and a band that will not read produces a copied barcode kept in a pocket, which defeats the control entirely.
Wireless coverage in the rooms, not the corridors
Surveys are frequently done where it is convenient to walk. The scan happens at the bedside, often in a side room with thick walls, and a dead spot there converts into batch scanning at the station within a fortnight.
A device per person, that logs in fast
Login time repeated at every patient is real time taken from a round. Ten seconds per patient across a bay, twice a shift, is the difference between a workflow that holds and one that gets optimised by the person carrying it.
An alert set that is worth reading
A system that warns on most administrations trains dismissal, and the habit does not distinguish the important alert from the four hundred that preceded it. Fewer alerts is a safety intervention here, not a convenience one.
Why is scan compliance the wrong metric?
Because it measures that a scan happened, and the safety comes from what was scanned and where. A ward that scans every dose from a pre-pulled batch at the nurses’ station has near-perfect compliance and no bedside verification — and it has removed precisely the control that catches the wrong-patient error, which was the strongest reason to deploy BCMA at all.
The metric also creates its own pressure. Once a scan rate is reported upward, the fastest route to improving it is to scan more rather than to scan better, and batch scanning is exactly what “scan more” looks like when the compliant path is slower than the round allows.
This is worse than doing nothing, and the reason is worth stating precisely: before enforcement, a low scan rate was an accurate signal that verification was not happening. After enforcement, a high scan rate is an inaccurate signal that it is. The instrumentation has been destroyed by the act of managing it.
Better measures, all of which are harder to satisfy without doing the right thing: the proportion of scans occurring inside the patient’s room, the proportion of doses with no readable barcode at all, and the alert override rate by drug. The middle one is usually the most revealing and the least often measured, because it points at packaging rather than at nurses.
So what should you buy first?
Whatever makes the dose carry its identity. In most hospitals that is packaging — because the proportion of doses that arrive without a usable barcode is the ceiling on every scan rate, and no amount of engagement lifts a ceiling.
This is an argument against buying the visible thing first, and it runs against how these projects are usually scoped, because barcode administration is the part with a business case attached and the part a board recognises. The predictable outcome of that sequence is high scan compliance and an unchanged incident rate — which is the most demoralising result available in this category, because it looks like success until somebody audits it.
Sequenced forward instead: structure the orders, make packaging carry identity, close dispensing, then deploy the scan — which now checks something real — and then take on the segment after discharge, which needs the record the first four produced.
Frequently asked
- What is barcode medication administration software?
- Barcode medication administration (BCMA) software checks a scanned patient wristband and a scanned medication package against the patient’s active order at the point of administration, and writes the result to the medication administration record. It is the administration stage of a closed medication loop, and the only widely deployed mechanism that verifies the physical dose against the physical patient at the moment it matters.
- What do you need in place before deploying BCMA?
- Doses that carry a readable barcode linked to the verified order, a wristband that scans reliably, wireless coverage inside patient rooms rather than corridors, and a device per staff member. The first is the one that decides the outcome: BCMA deployed on hand-assembled or repackaged doses verifies that a nurse scanned a package, which is a much smaller thing than it appears on a slide.
- Why do BCMA deployments plateau?
- Because the remaining misses are usually physical rather than behavioural — doses with no readable barcode after splitting or repackaging, wristbands that will not read, slow logins and wireless dead spots, and alerts that fire so often that dismissal becomes a habit. None of those appear in a compliance report, which only shows that a scan did not happen.
- Is a high BCMA scan rate proof of safe administration?
- No, and it is a weaker signal than it looks. A ward that scans a pre-pulled batch of doses at the nurses’ station has near-perfect compliance and no bedside verification at all. More useful measures are the proportion of scans occurring inside the patient’s room, the proportion of doses with no readable barcode, and the alert override rate by drug — all three are harder to satisfy without doing the right thing.
- Does OneDose replace our existing eMAR?
- No. Where a hospital runs its own eMAR, OneDose writes administration events back to it rather than replacing it. Replacing a medication record that clinicians already trust and know is a large, disruptive project with a poor return, and the value here is in what feeds the record rather than in the record itself.
- What does a barcode need to encode?
- Ideally the specific verified order the dose was packed against, not just a product code. A product code lets the scan catch a wrong-drug error and nothing else; a reference to the order lets it confirm this dose, for this patient, against this order. Ask any vendor what their barcode encodes and what the scan compares it against — the answers separate a closed loop from a well-organised open one.