Industries / Health / Medical supply distribution
Hospital supply rooms
Hospital supply rooms are where the stock a distributor delivers is held, counted and refilled close to the wards, clinics and operating rooms that use it.
7
Milestones achieved, from 1980 to 2025.
2
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1980
Carts swapped on schedule
In an exchange cart system a hospital unit has two identical carts of supplies: at set times a full cart from the central store replaces the one in use, which goes back to be counted and refilled. J. F. Coarse and P. Pierpaoli described such a system run across a whole institution in 1980, as an extension of the hospital's unit-dose drug system, and in 1981 a two-unit study by B. L. Tilley, weighing cost, labor and staff acceptance, found that exchange carts would benefit the entire hospital.
Credit
J. F. Coarse and P. Pierpaoli (Hospital Pharmacy, 1980); B. L. Tilley (Hospitals, 1981)
1990
Supplies delivered to each department
In 1990 Baxter, which a 1985 merger had made the largest single source of hospital products, launched ValueLink: it delivered products just in time, in ready-to-use packages, straight to the departments that used them, such as operating rooms and laboratories, so that a hospital could run down its own stores. Hospitals signed multi-year contracts, Baxter promised a 100% fill rate, and at Vanderbilt University Hospital Baxter was moving its own staff into the hospital. Baxter reported 53 hospitals in the program at the end of 1993, averaging 450 to 500 beds, up from 22 in 1991.
LessonA 1990 study commissioned by the Health Industry Distributors Association found that this "stockless" supply saved money but met many barriers, and one hospital that modeled its own costs in 1991 concluded it was not right for them.
Credit
Baxter International; described by James E. Short and N. Venkatraman of MIT (Sloan Management Review, 1992)
2011
Two bins and a tag
In a two-bin system each item is kept in two compartments: staff take from the front one, and when it is empty they post its tag to order a refill while they use the back one. Hôtel-Dieu d'Amos, a 96-bed hospital in Quebec, had run a par system, in which each item is topped up to a set quantity based on average use; from February 2011 it used Logi-D's radio-tagged two-bin system for about 1,400 kinds of small item, some 8,000 objects, and stopped the daily counting rounds, RFID Journal reported. The hospital had not measured the time saved. A 2010 case study of a nursing unit by Ygal Bendavid, Harold Boeck and Richard Philippe, based on interviews, observation and a simulation, found that the radio-tagged two-bin design cut the stock held on the unit and freed staff time, for less outlay than radio-tagged cabinets.
Credit
Logi-D; Hôtel-Dieu d'Amos; Bendavid, Boeck and Philippe (Business Process Management Journal, 2010)
2017
Opened, then thrown away
Researchers at the University of California, San Francisco (UCSF) watched 58 neurosurgical operations in August 2015 and recorded every disposable supply that was opened for the case but not used. The unused supplies averaged $653 per case, 13.1% of the case's surgical supply cost, and the authors estimated about $968 per case and $2.9 million a year for their neurosurgery department.
LessonThe share wasted depended on the surgeon and the type of case, not on how long the operation took or on the surgeon's years of training.
Credit
C. C. Zygourakis and colleagues, UCSF (Journal of Neurosurgery, 2017)
2017
A $275 million system dropped
In September 2013 the US Department of Veterans Affairs (VA) awarded Shipcom Wireless a $275 million, five-year contract for Catamaran, a point-of-use system that records supplies where they are used. VA let the contract expire in February 2017 because Shipcom failed to meet its needs; only 22 VA medical centers had installed it, and all of them moved back to VA's older Generic Inventory Package. At 11 of those centers the VA Inspector General found errors in the recorded stock of 208 of the 330 supply items it checked (63%): 63 items were below their reorder point and 148 above their normal stock level.
LessonData moved into the new system only as accurately as it had been kept in the old one, which VA's logistics director called "garbage in, garbage out".
Credit
VA Office of Inspector General, report 17-05246-98 (May 2019)
2020
Scanning across six English trusts
In 2016 England's Department of Health gave £12 million to six hospital trusts, the Scan4Safety demonstrator sites, to scan GS1 barcodes (the global product-code standard) on patients, products and places at the point of care, so that stock is recorded when it arrives and again when it is used. A 2020 GS1 UK report, based on audits by the Department of Health and Social Care during the 2016–2018 program, counted recurrent inventory savings of nearly £5 million and £9 million of one-off stock reductions across the six; at Salisbury, reorganizing the stores revealed that 6% of the orthopaedic implants on the shelves were already out of date.
LessonAn interview study at one demonstrator site found that the gains for tracing devices depend on how many items are tracked in stock and on staff scanning consistently at the point of care.
Credit
Department of Health and Social Care; the six demonstrator trusts, among them Leeds and Salisbury; GS1 UK
2025
Kanban on a busy ward
At Hamad General Hospital, a 600-bed teaching hospital in Doha, Qatar, nurses had judged supply needs by eye; audits found one inpatient unit spending QAR 31,000 (US$8,500) a week on consumables and 10–20% of consumables thrown away because they had expired. A team rebuilt the unit's storeroom as a kanban, a pull system in which an emptied bin triggers a refill of a set quantity. Over three years weekly consumable costs fell to QAR 16,000 (US$4,300), stockouts fell to near zero and staff compliance with the system rose from 76% to 95%, by the team's own measurements.
Credit
K. J. Logrono, B. S. M. Zu'bi and R. Siddiqui (BMJ Open Quality, 2025)
Watch for
Cameras that count shelves
Gartner, a research firm, says advances in AI and computer vision make it "increasingly feasible" for hospital supply rooms to sense their own stock, predict demand and start replenishment with minimal human involvement, and advises supply chain chiefs to prepare to replace manual counting. Roboflow, a computer vision company, sells camera counting for nursing-unit supply rooms, par carts and procedure areas.
Achieved whena hospital publishes camera counts compared with hand counts in working supply rooms over several months, checked by someone outside the vendor.
Claimed by
Gartner, August 2026
NHS stock system savings
The Scan4Safety program's website says NHS Supply Chain's inventory management system (IMS) program had captured £15.3 million in savings across the first 19 providers to adopt it by April 2026, and that stock value at Manchester University NHS Foundation Trust fell from £3.2 million to £700,000 after March 2023.
Achieved whenan audit by the Department of Health and Social Care or the National Audit Office, like the audits of 2016–2018, confirms the figures.
Claimed by
the Scan4Safety program and GS1 UK, 2026
Open
01
Expired stock, counted in public
A standard sweep method and a public table of the share of on-hand medical and surgical supplies that are past their expiry date, by product category, in the supply rooms of at least 10 hospitals or clinics in more than one country. Use AHRMM's measure: expired products divided by all on-hand products that carry an expiry date.
The briefCredit · unclaimed02
Pack codes checked against GUDID
A public audit of the barcodes on medical device and supply packs as they sit in hospital and clinic supply rooms. At three or more sites, scan every barcode on at least 1,000 packs with a phone, note which code carries the device identifier (the product number in a UDI), and check that identifier, the lot and the expiry date against the printed text and against the pack's record in GUDID, the FDA's public device database. Publish the share of packs whose codes scan, the share whose identifier is in GUDID, and the share whose GUDID record matches the pack, such as its brand and the quantity in the package.
The briefCredit · unclaimed
Open 01 · the brief
Expired stock, counted in public
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A standard sweep method and a public table of the share of on-hand medical and surgical supplies that are past their expiry date, by product category, in the supply rooms of at least 10 hospitals or clinics in more than one country. Use AHRMM's measure: expired products divided by all on-hand products that carry an expiry date.
Why it is still open
The figures that exist come one hospital at a time. Salisbury found 6% of its orthopaedic implants out of date, Hamad General Hospital found 10–20% of a unit's consumables discarded for expiry, and a September 2026 VA Inspector General review in Augusta, Georgia, found expired supplies in its stock. AHRMM, the American Hospital Association's supply chain society, defines the measure but publishes no values, and searches found no multi-hospital count.
Built with what exists
AHRMM's measure; phones running a barcode reader that decodes GS1 DataMatrix codes, the square codes that carry a device's identifier, lot and expiry date; a vision-language AI model, such as Claude or Gemini, to read printed dates on packs without a code; a shared spreadsheet.
First step
Agree the method with one hospital's materials manager and sweep one supply room twice, once by hand and once with the phone method, and compare the counts.
Done looks like
At least 10 sites and 20,000 items checked with the same method, the share expired for each category with its confidence interval, and a second person recounting a random 5% of items at each site.
Open 02 · the brief
Pack codes checked against GUDID
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public audit of the barcodes on medical device and supply packs as they sit in hospital and clinic supply rooms. At three or more sites, scan every barcode on at least 1,000 packs with a phone, note which code carries the device identifier (the product number in a UDI), and check that identifier, the lot and the expiry date against the printed text and against the pack's record in GUDID, the FDA's public device database. Publish the share of packs whose codes scan, the share whose identifier is in GUDID, and the share whose GUDID record matches the pack, such as its brand and the quantity in the package.
Why it is still open
Stock is recorded at the point of use only when the code on the pack scans and leads to the right record. The BUILD project's final report (June 2019) found that hospitals regard GUDID data as "not fully validated", with "substantial gaps", and named a missing database record, a degraded label and operator error as reasons a scan fails. In a 2019 survey by AHRMM, the American Hospital Association's supply chain society, 52% of providers cited several barcodes on one pack as a barrier to scanning and 48% unreadable barcodes. These are reports of what staff meet, not counts. The nearest tool, hospital-udi-scan, an open Android app posted on GitHub in August 2026, scans the codes on hospital stock and looks each identifier up in China's national device database, marking it found, to be checked or missing, but publishes no counts. No published audit was found that matches packs on hospital shelves against GUDID.
Built with what exists
Phones running a reader that decodes GS1 DataMatrix and GS1-128 codes, such as the open-source ZXing library; the AccessGUDID API, which returns each identifier's brand, package level and quantity; a vision-language model such as Claude or Gemini to read the printed lot and expiry text for comparison; a spreadsheet. The same sweep can feed "Expired stock, counted in public".
First step
With one materials manager's agreement, scan 100 packs in one supply room and look each identifier up in AccessGUDID.
Done looks like
At least 1,000 packs at three or more sites, each with its scan result, GUDID match and any mismatch by type (no record, wrong package quantity, lot or date that differs from the print); a second person rescans 5% of packs and agrees on at least 95%.