Editorial Article

The Sale Has Been Made at the Cabinet and Not at the Compounding Suite

Updated November 18, 2026

There is a number in the 2023 ASHP national survey of hospital pharmacy practice that should be quoted in every board paper written about medication safety, and almost never is. Eighty-six per cent of US hospitals use automated dispensing cabinets as their primary method of maintenance-dose distribution. Three point seven per cent use any automation at all to compound sterile preparations.

Both figures come from the same survey of 1,497 general and children’s hospitals, published in the American Journal of Health-System Pharmacy. They describe a market that has finished the easy installation and has not started the hard one.

What the cabinet solved, and what it did not

The cabinet solved a distribution problem. It moved the drug closer to the patient, locked it, recorded who took what, and removed a trolley round from the day. Its spread from 22 per cent of hospitals in 2002 to 86 per cent today is one of the more complete technology diffusions in hospital operations, and the fact that it looks unremarkable now is the measure of how completely it worked.

The compounding suite is a different problem in kind. A cabinet is a storage and verification device deployed per nursing unit against a workflow that already existed. A sterile compounding robot is a manufacturing device deployed once per pharmacy against a process that is regulated, validated, sampled and inspected. The cabinet had to be accepted by nurses. The compounder has to be accepted by a regulator, a quality team, an insurer and a patient-safety committee before a single dose is made on it.

That difference in kind, rather than a difference in enthusiasm, explains why the penetration numbers diverge by a factor of twenty-three.

The proof layer is half-built

What hospitals have bought, consistently, is proof at the point of dispensing. Machine-readable coding verifies doses in 73.6 per cent of hospitals. Autoverification in the electronic health record handles 73.4 per cent. Barcode scanning to verify ingredients during sterile compounding reaches 53.6 per cent, and workflow software that structures the record before a dose is made reaches 31.0 per cent.

Then it stops. Gravimetric verification — the technology that confirms the dose was actually drawn rather than merely scanned — sits at 7.7 per cent. Image or video capture of the compounding process, which is retrospective evidence rather than prospective control, reaches 38.6 per cent. Advanced analytics, including machine learning and predictive models, reaches 5.7 per cent.

Read as a stack, the pattern is coherent: hospitals have bought the technologies that produce a record, and have not bought the technologies that control the process. Barcode scanning tells you what was picked up. Gravimetrics tells you what went into the bag. Only one of those is a control, and it is the one that sits at single-digit adoption.

The regulatory argument is real but it is not the constraint

It is tempting to explain the compounding gap as a compliance problem waiting for a rule change. United States Pharmacopeia Chapter <797> governs sterile compounding conditions and beyond-use dates; Chapter <800> governs hazardous drug handling and exposure control; state boards of pharmacy enforce both, and accreditation surveyors check them. Tighten any of the three and adoption should rise.

The evidence does not support that as the primary constraint. If regulation were the binding limit, the technologies that satisfy a documentation obligation would be the ones with high adoption — and workflow software sits at 31 per cent while barcode verification of ingredients sits at 53.6 per cent, both well above the 3.7 per cent that have automated the process itself. What is missing is not a mandate. It is floor space, capital, dose-range flexibility across a formulary that changes constantly, and the validation burden of proving a robotic process is equivalent to a manual one that a pharmacist has performed for thirty years.

There is a second, quieter constraint: a pharmacy that supplies a hospital cannot stop supplying it. Installing a compounding robot means rebuilding the cleanroom, requalifying the environment, and running parallel processes through a transition. The operational risk of the transition is frequently larger than the business case for the destination, and that is why the projects that proceed are the ones attached to a new build or a consolidation decision rather than a standalone equipment refresh.

The labour market is doing more work than the rulebook

In the same survey year, 9 per cent of full-time hospital pharmacy technician positions and 4.3 per cent of full-time pharmacist positions were vacant. Those are averages across a profession with wide regional variation, and the hospitals with the worst vacancies are not always the ones with the capital to fix them.

That vacancy rate is the most honest way to frame the investment case, because it removes the comparison that usually stalls a purchase. The choice is rarely between a robot and an equally staffed manual process. It is between a robot and a service line that cannot be reliably staffed at 2am. When a health system consolidates twenty-one hospital pharmacies into a central distribution model — as Ballad Health did across Northeast Tennessee and Southwest Virginia — the saving is measured in six to eight technician hours a day per site that no longer have to be recruited, trained and retained.

What changed in the last twelve months

Two things, and both are about the software layer rather than the mechanism.

BD launched the Incada Connected Care Platform — a cloud environment that unifies device data — alongside a next-generation Pyxis Pro dispensing solution, and signed a first-of-its-kind US pharmacy automation partnership with Henry Ford Health around its Rowa Vmax storage and retrieval robot. Omnicell launched Titan XT at the ASHP 2025 Midyear Clinical Meeting and described the roadmap as a step toward autonomous medication management that extends beyond the pharmacy into nursing care areas.

Neither is a breakthrough in motion control. Both are attempts to change the unit of sale from a device to a fleet. That matters for the compounding suite specifically, because the argument that has failed to sell a robot — it makes doses more accurately — may succeed as an argument that it makes doses visible inside an enterprise system that already holds the cabinet data, the infusion data and the diversion data.

The Henry Ford arrangement is the more interesting of the two, because a health system co-developing applications on a vendor’s robot is not buying capital equipment. It is buying into a roadmap, and it gives the vendor a North American reference for a machine that has been strongest in European and Japanese central-pharmacy installations.

The number to watch

Sterile compounding automation is the smallest of the five equipment families in the pharmacy automation market at USD 500 million in 2025, and the fastest growing at 10.10 per cent, reaching USD 1,080 million by 2033. That growth rate is not a forecast of enthusiasm. It is the arithmetic of starting from 3.7 per cent.

The single indicator that would confirm the transition has begun is not a vendor announcement or a guideline revision. It is the ASHP survey’s compounding line moving from 3.7 per cent to 8 per cent. At that point the reference installations exist, the validation templates exist, and the second wave of buyers stops treating the project as a first-of-its-kind risk. Until then, the honest description of this market is that it has solved distribution and has not yet started preparation.

This analysis draws on the ASHP National Survey of Pharmacy Practice in Hospital Settings: Operations and Technology — 2023 (published in the American Journal of Health-System Pharmacy, covering 1,497 hospitals), the 2025 Pharmacy500 IV workflow management report, Omnicell’s fiscal 2025 results, BD’s fiscal 2025 results and its December 2025 segment disclosure, and the Pheonix Research report Global Pharmacy Automation System Market, Size, Share and Forecast 2026–2033.

Frequently Asked Questions

What share of US hospitals automate sterile compounding?

3.7 per cent of US hospitals use any automation for compounding sterile preparations, and 3.4 per cent use a robotic IV compounder for non-hazardous products, according to the 2023 ASHP national survey of 1,497 general and children's hospitals. Barcode scanning to verify ingredients, by contrast, is used in 53.6 per cent of hospitals and compounded sterile preparation workflow software in 31.0 per cent. Hospitals have bought the technologies that produce a record but not the ones that control the process.

Why did dispensing cabinets spread so much faster than compounding automation?

86.1 per cent of US hospitals use automated dispensing cabinets as their primary method of maintenance-dose distribution, against 27.0 per cent with a high-density carousel, 6.5 per cent with a unit-dose robot, and 2.0 per cent using a robot as their primary central-pharmacy distribution method. The cabinet solved distribution and spread quickly because it was deployed per nursing unit against a workflow that already existed.

What is gravimetric verification and how widely is it used?

Gravimetric verification — confirming that the dose was actually drawn rather than merely scanned — is used in 7.7 per cent of hospitals, against 73.6 per cent using machine-readable coding to verify doses during dispensing. Only one of those technologies controls the process rather than documenting it, and it is the one at single-digit adoption.

How large is the pharmacy workforce gap driving automation?

In the 2023 survey year, 9 per cent of full-time hospital pharmacy technician positions and 4.3 per cent of full-time pharmacist positions were vacant. The investment case is therefore rarely a comparison between a robot and an equally staffed manual process; it is a comparison between a robot and a service line that cannot be reliably staffed.