When a drug goes into shortage, a hospital pharmacy team doesn’t simply wait for stock to arrive. They run a demanding daily process: checking inventory across the hospital using both software and manual counts, assessing what future supply looks like, working out whether approved substitutes are available, and then deciding which mitigation steps to take, all while weighing cost and disruption to patient care.
One piece of information sits at the center of those decisions: the estimated release date (ERD), the date a manufacturer expects to release its next batch of a drug into the supply chain. During a shortage, that date is often the single best signal a pharmacy has for when it will next be able to restock. So the team leans on it to decide whether they can hold steady or whether they need to act now.
Researchers at Northeastern University did something that, remarkably, hadn’t been done before. They collected real ERD data alongside real shipment-arrival data from a hospital system over an eight-month period, and measured how accurate those release dates actually were.
What they found
The results were sobering. Across nearly 200 tracked cases, shipments arrived significantly later than the estimated release date promised, and the variation was wide, confirming what pharmacy staff had long suspected: the dates simply don’t reliably predict when product will show up.
It got worse the deeper they looked. Almost half of the tracked cases saw the release date revised two or more times before the shipment actually arrived, some as many as ten or twelve times. And crucially, an updated date was no more accurate than the one it replaced. The changes behaved essentially at random: when the researchers modeled the timing of updates, they found the intervals were memoryless, meaning a past update told you nothing about when, or by how much, the next one would come. When dates did change, they were pushed later about 71% of the time, by an average of more than two and a half weeks.
Why this is really a decision-velocity problem
Here’s the trap, and why speed matters so much. If the release date looks far off, a pharmacy may proactively launch mitigation measures, conserving stock, sourcing substitutes, changing clinical practice. But if the shipment then arrives early, all that effort was wasted: time, money, and labor spent unwinding steps that turned out to be unnecessary.
If the release date looks near, the team may decide to hold current practice steady and wait. But if the shipment arrives late, they’re forced to scramble, implementing widespread mitigation fast, at higher cost and far greater stress on staff than if they’d started earlier.
Either way, the decision was reasonable given the information available. The information was simply wrong. And when the underlying signal is unreliable, the team loses the one thing that protects patients: the ability to move early and with confidence. Slow, hesitant, second-guessed decisions are the real cost of bad data, not the shortage itself.
This reframes the whole problem. The shortage is largely outside the hospital’s control, from patient demand to manufacturing to shipping times, all external. The one thing a hospital can control is the speed and quality of its own response. And that response is only as fast and as confident as the data feeding it allows.
The real lesson
It’s tempting to conclude that the release dates are useless and should be ignored. But that’s not quite right either. For a few specific product types, the dates were meaningfully more reliable, and throwing them out entirely would discard genuine signal. The smarter path is to weight information by how trustworthy it actually is, to pull in additional sources where they exist, and to make decisions that account for uncertainty rather than pretending it away.
That is a much harder thing to do by hand, across hundreds of products, every single day, under stress. It requires one trusted view of reality: seeing all the relevant data at once, including inventory, consumption, contracts, substitutes, and supplier reliability, and understanding how much confidence to place in each piece before acting. Without that, teams don’t just decide badly. They decide slowly, and slowness during a shortage is its own kind of harm.
Where this points
This is exactly the gap a decision system is built to close. The value isn’t in generating one more number for an already-overloaded team to interpret. It’s in taking fragmented data, from any source, any format, any quality, and turning it into trusted, actionable decisions: surfacing where the signal is weak, showing the reasoning and the alternatives behind each recommendation, and letting teams act with confidence and no hesitation.
That is the thinking behind Optivian, MUUTAA’s enterprise supply chain decision system. It starts with the data you already have, however messy, and turns it into one trusted view, one aligned recommendation, and decisions you can defend, with the why, the evidence, and the impact attached. Shortages will keep happening; that much is outside anyone’s control. What healthcare organizations can improve is their decision velocity: how quickly and how confidently they move when it counts. And that starts with knowing which data to trust.
This article draws on: Chicoine, N. & Griffin, J. (2025), “The Unreliability of Estimated Release Dates in Hospital Drug Shortage Management: A Case Study of Hospital Pharmacy Operations During the COVID-19 Pandemic,” medRxiv, doi:10.1101/2025.07.10.25331166.