Expanded access programs occupy an awkward place in most sponsor organisations. They are medically urgent, reputationally visible and often small enough that nobody builds infrastructure for them. A handful of agreements, a handful of patients, a handful of sites. The paperwork gets done quickly because the clinical need is real.
What follows is a set of obligations that look a great deal like clinical trial obligations, carried by a team that does not have clinical trial systems. Safety events must be reported within defined windows. Product must be accounted for. Records must be retained past the life of the program. And supply commitments, which are the part most often underestimated, can bind the sponsor to continue providing product to enrolled patients well beyond the point where the program stops taking new ones.
This guide covers what those agreements typically oblige, where the tracking fails, and how to run an access program with the same obligation discipline applied to a trial, without building a parallel trial organisation to do it.
What the agreement commits both sides to
The obligation set is narrower than a clinical trial agreement but denser per clause, and it is split between the sponsor and the treating institution in ways that need to be explicit in the document.
- Supply and continuity, the sponsor commits to provide product for defined patients, often with language about continued supply for enrolled patients if the program changes or ends.
- Safety reporting, the physician reports adverse events, with serious events carrying short defined windows and a named recipient on the sponsor side.
- Treatment plan adherence, the physician agrees to follow the treatment protocol or plan, including dosing, monitoring and eligibility, with deviations reported.
- Product accountability, receipt, storage, dispensing records and return or destruction of unused product, documented to a standard the sponsor can present.
- Publication and promotion limits, restrictions on presenting program data and on any characterisation of the product as approved or proven, which protects both parties.
The asymmetry worth noticing is that the sponsor’s obligations are mostly forward-looking and open-ended, while the physician’s are mostly event-driven and short-cycle. Both need tracking, but they need different tracking. Forward supply commitments belong on a long-horizon register alongside other continuing duties; event-driven safety reporting needs a short clock with escalation, the same pattern that governs clinical trial agreement obligations.
Many expanded access agreements commit the sponsor to continue supplying enrolled patients after the program closes to new enrolment, sometimes until commercial availability or an alternative is in place. That is a multi-year forward obligation created by a short document, and it is routinely lost when the program team moves on.
Where access program tracking fails
The failure pattern is consistent and has little to do with the quality of the agreements themselves. It has to do with the mismatch between how long the obligations run and how long the program structure exists.
- Site-by-site obligations treated as one set. Each treating institution is a separate counterparty with its own executed terms, its own reporting contacts and its own dates. Managing them as a single program-level obligation hides per-site gaps.
- Safety reporting paths that assume trial infrastructure. The physician is often outside any trial network, so a reporting route that works for investigators does not reach them. The obligation exists, the mechanism does not.
- Supply commitments not carried into planning. Manufacturing and supply planning works from commercial and trial demand. A continuing access commitment that lives only in a contract file does not appear in that forecast until it becomes a shortage.
- Program closure treated as administrative. Closing to enrolment is not closing the obligations. Without a defined closure process, retention, continuing supply and final accountability duties are left with no owner.
- Records dispersed across functions. Agreements in legal, safety reports in pharmacovigilance, shipment records in supply, and no consolidated view of whether a given site is fully compliant.
Running an access program with obligation discipline
Step 1, Template the agreement, then track the deviations
Use a standard form per jurisdiction so the obligation set is predictable, and treat every negotiated change as a flagged deviation that must be reflected in the obligation register for that site. Uncontrolled variation across sites is what makes program-level reporting impossible later.
Step 2, Build a per-site obligation set at execution
At the point each site agreement is signed, generate its obligation records: reporting contacts and windows, accountability cycle, supply commitment scope and any site-specific terms. Doing this at execution rather than at first audit is the difference between a register that reflects reality and one assembled from memory.
Step 3, Name owners across four functions
Medical affairs, pharmacovigilance, clinical supply and regulatory will each hold obligations. Assign each to a person and record the deputy. Programs that assign obligations to functions consistently find that the ones sitting between two functions are the ones that go unperformed.
Step 4, Connect supply commitments to planning
Push continuing supply obligations into the demand picture used by manufacturing planning, with the patient count and expected duration. A commitment that is visible in the contract but invisible in the plan is a supply interruption waiting for a schedule change.
Step 5, Define closure as an obligation event
Write down what closure requires: notice to sites, transition arrangements for enrolled patients, final accountability and reconciliation, destruction authorisations, retention custody, and formal closure of each site obligation set. Then execute it as a checklist rather than a wind-down conversation.
Access program commitments, owned and monitored
See per-site obligation sets, safety reporting clocks and forward supply duties in one register.
Why small programs deserve real infrastructure
The argument against putting an access program into a managed obligation register is usually scale. A dozen sites and thirty patients does not feel like it warrants system configuration. The counter-argument is duration and consequence. These programs frequently run longer than planned, the obligations survive the team that created them, and a missed safety report or an interrupted supply to a patient with no alternative carries consequences that no amount of process savings offsets. The marginal cost of tracking a small program inside an existing obligation platform is close to zero once the platform exists, which is the practical case for using the same infrastructure that already carries life sciences obligation management across the rest of the portfolio rather than standing up something separate.
Obligation set by counterparty and clock
The table separates sponsor-side from site-side duties and pairs each with its monitoring pattern. Specific windows and retention periods depend on the governing regulation and the executed agreement, so treat the timing column as a structure to fill in rather than a standard.
| Obligation | Owed by | Clock type | Typical owner | Survives closure |
|---|---|---|---|---|
| Serious adverse event reporting | Treating site | Event-triggered, short window | Pharmacovigilance | For events during treatment |
| Continued supply to enrolled patients | Sponsor | Forward, open-ended | Clinical supply | Yes, often for years |
| Product accountability records | Treating site | Recurring per shipment | Clinical supply | Retention only |
| Treatment plan adherence and deviations | Treating site | Event-triggered | Medical affairs | No |
| Record retention | Both parties | Expiry-driven, long horizon | Regulatory | Yes, by definition |
Expanded access is one of the few areas where the contracting is straightforward and the aftercare is not. The agreements are short, the templates are stable, and the negotiation is rarely contentious. Everything that goes wrong goes wrong after signature, in the gap between a program that ends and obligations that do not. Treating those obligations as a tracked register from the first executed site agreement costs very little and removes almost all of that gap. The programs that run cleanly are not the ones with the most resource behind them; they are the ones where every commitment had a name against it from the day the first site was signed, and where closure was executed as a defined sequence rather than allowed to happen by attrition.
For related guidance, see Clinical supply depot agreements, tracked end to end, Safety data exchange agreement tracking for PV teams and Comparator drug sourcing agreements: control datasheet.
