Operations guide

Hospital medication control: from the sealed box to the traced unit

Controlling medication is not counting boxes at month end. It is knowing, at any moment, which units exist, where they are, when they expire and who withdrew each one. This guide covers the groups where risk concentrates and what it takes to control them.

Updated August 2026 · 8 min read

What controlling a medication actually means

Pharmaceutical control has three layers, and usually only the first gets implemented. The first is quantity: how many units there are. The second is identity: which units those are, with which lot and which expiry. The third is accountability: who withdrew each one and for which patient.

An inventory that stops at the first layer allows replenishment but not investigation. When a discrepancy appears, there is no way to tell whether it was a recording error, waste, undocumented consumption or something more serious.

Quantity: how much sits at each storage point
Identity: which lot and which expiry each unit carries
Accountability: who withdrew it, when and for which patient
Condition: whether cold chain or storage stayed within range

The four groups where risk concentrates

Applying the same level of control to the whole formulary is expensive and unnecessary. The sensible practice is to segment.

Controlled substances: narcotics and psychotropics, subject to named records and per shift counts under most regulations
High cost: oncology, biologics and specialised therapies, where one lost unit equals weeks of consumption of another family
High alert: medication whose administration error causes serious harm, such as anticoagulants, insulins, concentrated electrolytes and sedatives
Emergency: the contents of the crash cart, where availability outweighs every other consideration

All four groups require unit level control. The rest of the formulary can be managed by quantity with no meaningful loss of safety.

The crash cart: where manual control fails most

The crash cart holds a contradiction: it must always be complete and always available, and its verification is a manual task competing with patient care.

The usual procedure is checking the cart per shift or when the seal is broken, against a list. It is slow, it gets postponed and it catches the missing item late. In a sealed cart, moreover, the only way to know whether a medication has expired is to open it.

Radio frequency reading reverses the problem: it verifies the full contents without opening the cart, in seconds, and shows at once what is missing and what is about to expire.

Cold chain and expiry

A medication that went out of temperature range is no longer safe even with intact packaging and a valid expiry date. That is why condition control belongs to inventory control rather than being a separate matter.

With expiry, the problem is rarely missing information: the date is printed on the package. The problem is that the date is not in the system, unit by unit, so nobody can prioritise consumption of the lot closest to lapsing.

When the system knows the expiry of every unit, the consumption rule stops depending on the discipline of whoever opens the fridge.

How to measure and cut the cost of expiry and stockouts

What to record at every event

A useful record is one that lets the history be reconstructed without interviewing anyone. That standard defines the minimum data set.

Reception: product, lot, expiry, quantity, origin and owner
Storage: exact location and, where applicable, temperature
Withdrawal: who, when, which unit and for which patient or procedure
Return: unit returned, condition and destination
Write off: reason, unit, lot and owner, separating expiry from damage and from loss

How to implement it

The most common mistake is starting with the entire formulary. The route that works starts with one contained group, in one service, treating the clinical team's real workflow as a design constraint rather than something to correct later.

The second mistake is leaving recording as an extra task. If control requires someone to write down something they did not write down before, the system coexists with a parallel spreadsheet from week one.

Pick one group and one service, not the whole hospital
Clean up master data before installing any equipment
Capture the event at the point of use, with access identified per person
Integrate with prescribing and with the ERP, so the record serves pharmacy, costing and the patient chart
Measure the result in shortages avoided and units expired, not in events recorded

Frequently asked questions

What are high alert medications?

Those whose administration error has a high probability of causing serious patient harm, even when used correctly under normal conditions. Lists vary by institution but usually include anticoagulants, insulins, concentrated electrolytes, opioids and sedatives.

How are narcotics and psychotropics controlled?

With named records, locked custody, per shift counts and documented reconciliation. The specific requirements are set by each country's regulation. A cabinet with per person identified access and automatic recording covers the requirement without depending on a manual logbook.

Does RFID work on small medication packages?

Yes, with the right tag. The real constraint is not size but content: liquids and metallised packaging degrade UHF reading, so tag validation is part of the project rather than an afterthought.

How does this relate to medication reconciliation?

Reconciliation compares what was prescribed with what was administered. A dispensing record per unit and per patient supplies half the data reconciliation needs, and supplies it as it happens instead of reconstructing it later.

How do you avoid double recording with pharmacy?

By integrating. If the dispensing system and the pharmacy system do not share the same event, someone will key it twice and the figures will diverge. Integration is not an optional improvement, it is the condition for the record to be single.