Case Study: The Cisplatin and Carboplatin Shortage

When cancer drug inventory became a governed planning problem

In 2023, two of the most important generic oncology drugs in the United States became scarce: cisplatin and carboplatin. These platinum-based chemotherapies are used across multiple cancers and treatment regimens. For many patients, they are not interchangeable convenience items. They are part of a treatment path that may be curative, life-extending, or clinically time-sensitive.

The shortage began with cisplatin after manufacturing quality issues disrupted supply. As cisplatin availability tightened, demand shifted toward carboplatin, creating a second shortage. A failure at one fragile point in the generic sterile injectable market became a national cancer-care planning event.

Hospitals did not merely need to know that a shortage existed. They needed to convert a shortage signal into defensible action.

That required a much richer planning loop than ordinary inventory management.

First, hospitals needed a real demand basis. Historical utilization was not enough. Demand had to be understood by drug, disease area, regimen, cycle, site, clinical intent, patient appointment, and timing. A vial needed for a curative-intent patient next week is not the same planning priority as a vial used in a setting where an alternative regimen is clinically acceptable.

Second, hospitals needed to know usable future supply, not just visible inventory. On-hand stock, wholesaler allocations, open purchase orders, expected deliveries, imported product, vial sizes, dose forms, local stock by site, and already-committed supply all had to be interpreted together. Supply that exists somewhere in the network is not necessarily supply that can be promised to a patient.

Third, hospitals needed explicit prioritization rules. A shortage of 500 vials does not automatically mean "buy 500 vials." It may mean conserve, substitute, reserve, transfer, defer, split, round doses, escalate, or refer. In the platinum shortage, many of those decisions depended on clinical intent: preserve scarce supply for patients with the greatest expected benefit, especially where treatment was curative or where alternatives were inferior.

Fourth, hospitals needed available-to-promise logic for care delivery. The real question was not "how much do we have?" It was "which future treatments can we safely promise?" That requires matching expected supply against future clinical commitments and identifying where the promise breaks: by patient cohort, by site, by regimen, by date, and by policy.

Fifth, hospitals needed governed recommendations. A recommendation to substitute, delay, conserve, or reserve supply should not silently become an operational commitment. It should pass through the right authority: oncology pharmacy, treating clinicians, pharmacy and therapeutics leadership, service-line leaders, or crisis allocation governance, depending on the consequence.

And lastly, sixth, hospitals needed traceability. Every consequential decision needed a record: what was known at the time, what supply was available or expected, what demand was already committed, which clinical rule was applied, which alternatives were considered, who approved the action, and what future commitments were affected.

Where traditional enterprise systems show their limits

The gap is not that Workday, Oracle, SAP, or other enterprise platforms cannot manage inventory, purchasing, order promising, or supply planning. They can. The gap is that high-consequence shortages are not ordinary replenishment or fulfilment problems. They sit between ERP, pharmacy systems, EHR schedules, clinical protocols, supplier allocations, committee decisions, and patient-level commitments.

An ERP can tell a hospital what it has on hand. A planning suite can help balance demand and supply. An ATP engine can help promise against available supply. But the cisplatin/carboplatin crisis required something more specific: a governed planning record that connects scarce supply to clinical consequence, feasible alternatives, decision rights, and audit-ready rationale.

That is the distinction.

In ordinary planning, the system asks:

Can we fulfill demand?

In high-consequence planning, the system must ask:

What is at risk, why, what can we still promise, and what should we do first?

Seeing a shortage versus managing one

The cisplatin and carboplatin crisis was not only a manufacturing failure. It was a planning-stress event. It exposed the difference between seeing a shortage and managing one. Managing it required a disciplined loop from observation to forecast, from forecast to plan, from plan to governed action, and from action to traceable commitment.

Those are the capabilities Vista is built to provide.

Vista is designed for high-consequence supply chains where planning decisions must be explainable, governed, and auditable. It brings together messy source data, deterministic calculations, usable future supply, ATP/CTP logic, risk and recommendation traceability, approval workflows, commitment impact, and exportable proof packets.

Vista does not replace expert judgement. It gives expert judgement a governed planning system to work inside.