How to Write a Safety Data Exchange Agreement
Define who reports what, to whom, and inside which clock — before the first case arrives.
What a how-to is not
A how-to is SPEQ’s practitioner method, not a procedure. It does not replace your own SOP, it is not a validated approach, and the judgement calls in it belong to your quality unit.
A safety data exchange agreement sets out how two organisations sharing a product will exchange safety information. It exists because reporting obligations are not divisible: both parties may hold obligations to different authorities for the same case, and each depends on the other for the information to meet them. Agreements written late, or written by lawyers alone, tend to omit the timings that make them work.
- 1
Establish each party’s own obligations first
Map what each party is independently required to report, to which authorities, and in what timeframe. The agreement allocates work between parties; it cannot reduce an obligation either of them holds, and drafting it before this mapping produces a document that allocates the wrong things.
- 2
Define awareness and the exchange clock
State when awareness at one party is deemed to arise and how quickly it must reach the other. Exchange timings must be short enough that the receiving party can still meet its own external deadline — an agreement that consumes the whole window guarantees late submissions.
- 3
Specify the scope of exchange
Set out precisely what is exchanged: cases, literature findings, signals, periodic report inputs, regulatory correspondence, product information changes. Ambiguity here is where information stops flowing without anyone deciding it should.
- 4
Assign responsibility for periodic reports and signals
Say who prepares which periodic report, who contributes what by when, and how signals identified by one party reach the other. Contribution deadlines belong in the agreement, not in a project plan that expires.
- 5
Provide for reconciliation and its cadence
Agree a periodic reconciliation of cases exchanged against cases received, with a documented process for resolving discrepancies. Reconciliation is how a broken interface is discovered before an authority discovers it.
- 6
Set audit rights, change control and termination handling
Include the right to audit, the process for amending the agreement as obligations change, and what happens to data and obligations on termination. A safety agreement with no termination provision leaves cases in flight with no owner.
- !Exchange timings that consume the whole regulatory window, leaving the receiving party unable to submit on time.
- !Scope written so generally that whole categories quietly stop being exchanged.
- !No reconciliation cadence, so a broken interface is found only at inspection.
- !No termination provision, leaving in-flight cases and retained data without an owner.
How to Write a Safety Data Exchange Agreement: frequently asked questions
Common questions on write a safety data exchange agreement.
Can the agreement reduce a party’s reporting obligations?
No. It allocates the work of meeting obligations between parties; it cannot remove an obligation that either party independently holds. That is why mapping each party’s own duties has to come before drafting rather than after.
How short do exchange timelines need to be?
Short enough that the receiving party still has usable time within its own external deadline after receipt, assessment and coding. Working backwards from the shortest applicable external clock is the only reliable way to set them.
Why is reconciliation a contractual term rather than an operational nicety?
Because without an agreed cadence and a discrepancy process, neither party owns finding the gap. Interfaces fail quietly, and the first evidence is usually a case that neither organisation can show it received.