Management guide

Stockouts and expiry: measuring and cutting the cost of clinical inventory

A hospital that buys too little runs out and cancels. One that buys too much accumulates expiry and throws money away. The usual answer is to look for a middle point, but no middle point exists when the real problem is lack of visibility. This guide explains how to measure both costs and how to reduce them at the same time.

Updated August 2026 · 8 min read

The two opposing costs of clinical inventory

Stockouts and expiry look like opposite problems and are usually handled separately, by different teams with different budgets. In practice they share one cause: nobody knows precisely what exists, where it is and when it lapses.

When information is missing, the operation protects itself with stock. That extra stock covers the stockout and produces the expiry. Adjusting inventory levels therefore moves the problem from one side to the other without solving it.

Stockouts are paid in cancelled care, changes of technique and emergency purchases. Expiry is paid in discarded product. Both are the same information cost, expressed in two currencies.

What a stockout really is

A stockout is not a zero in the central warehouse. It is the item not being where it is needed, when it is needed. A hospital can hold sufficient stock and still run out, because it sits in another service, on an unrecorded cart or in a box nobody found.

Measuring stockouts at warehouse level systematically understates the problem. The useful measurement happens at the point of use: how many times somebody looked for something and it was not available.

True stockout: no stock anywhere in the hospital
Location stockout: stock exists, but not at the point of use
Information stockout: stock exists at the point of use, but the system said otherwise
Expiry stockout: stock exists, but it is not usable

Why expiry is not a purchasing problem

The instinctive reaction to expiry is to review how much was bought. That is almost always the wrong question.

A product expires because nobody consumed it before an identical one with a later date. That happens when the system does not distinguish lots: if inventory says twelve units without specifying which expires first, the person takes whichever is nearest to hand.

Expiry is, at bottom, a problem of unit identity. It is solved by making each piece's date visible at the moment of withdrawal, not by buying less.

What to measure and how to calculate it

Without these numbers, any improvement is an impression. With them, the discussion stops being about perceptions and starts being about causes.

Point of use stockout rate: failed withdrawals over attempted withdrawals, by service and by family
Inventory accuracy: absolute gap between system and physical count, over the total counted
Expired value: cost of units written off for expiry, separated from damage and loss
Coverage: days of consumption the current stock covers, by reference rather than on average
Turnover by reference: to identify stock that never moves and is the natural candidate to expire
Inventory ageing: share of stock with fewer than ninety days to expiry

Measure by family and by service, never on average. A healthy average hides one service in permanent stockout and another with frozen inventory.

What reduces both costs at once

The mechanisms that work share one trait: they increase the precision of the data instead of increasing stock.

Identify by unit, with lot and expiry, so the system can prioritise what to consume first
Record consumption at the point of use, so replenishment answers real consumption rather than an estimate
Apply nearest expiry consumption, with the date visible at the moment of withdrawal
Warn in advance about lots that will expire, with enough margin to redistribute or return
Redistribute between services before buying, which requires seeing the whole hospital's stock rather than each store separately
Replenish from consumption rather than from the calendar, which is the difference between covering demand and covering habit

The mistake of optimising only one of the two

A programme chasing only expiry ends up cutting stock until stockouts appear in the operating room, and that cost never shows up in the waste report.

A programme chasing only stockouts ends up raising stock until expiry grows, and that cost surfaces months later, when nobody links it to the decision that caused it.

The only way to improve both indicators together is to reduce uncertainty. That is where unit level traceability stops being a regulatory requirement and becomes a financial decision.

How radio frequency traceability works

Frequently asked questions

What is an acceptable level of expiry?

It depends on the family. For high turnover, low cost supplies a small percentage is tolerable. For high value devices, a single expired unit can exceed a full year of management savings. That is why the indicator must be measured in value and by family, not in units over the total.

How do you calculate the cost of a stockout?

By adding what was spent for not having it: emergency purchase at a premium, lost operating room time, rescheduling and, where applicable, switching to a more expensive alternative. It almost always exceeds the cost of the missing unit by a wide margin.

Does lowering safety stock reduce expiry?

It reduces expiry and raises stockouts in the same proportion, unless data precision improves in parallel. Safety stock is the compensation for not knowing, so it can only come down once more is known.

FEFO or FIFO?

FEFO, first expired first out. FIFO orders by entry date, which does not always match the expiry date, because a lot received later may expire sooner. Applying FEFO requires knowing the expiry of every unit at the point of withdrawal.

How often should inventory be counted?

The question changes when inventory is permanent. With automatic reading, counting stops being a periodic event and becomes continuous, and the formal inventory shrinks to a control check instead of being the year's only source of truth.