Why Shared-Control Accountability Needs Explicit Boundaries
In a platform trial, participants assigned to different intervention arms may be compared with participants assigned to a common control group. The design can reduce duplication, but sharing a control does not make accountability disappear or automatically transfer it from one sponsor to another. A safety report, data error, or delayed infrastructure change can affect several comparisons at once, while a product-specific issue may concern only one arm.
The central governance task is to make responsibility traceable: for every important duty, identify who is accountable for ensuring it is performed, who performs it, who must be consulted, and who receives information. This tutorial focuses on sponsor accountability at that level. It does not prescribe a particular allocation of decision-making authority; the earlier material on decision rights and delegation maps addresses how authority and delegated tasks should be recorded.
The legally responsible sponsor, and the precise duties that can be delegated, depend on the jurisdiction and trial structure. A platform may have one sponsor for the whole trial, separate sponsors for different interventions, or a coordinating sponsor alongside other legally responsible parties. Governance documents must describe the actual arrangement rather than assume that a label such as “platform sponsor” resolves every responsibility.
Three Responsibility Domains
A practical allocation begins by separating responsibilities into three domains. These domains are not necessarily separate organizations: one sponsor may hold several roles, and a role may be delivered through qualified service providers. The point is to prevent a shared process from becoming an unowned process.
| Domain | Examples of responsibilities to allocate | Key accountability question |
|---|---|---|
| Shared infrastructure | Master protocol implementation, trial-wide systems, site training, central data processes, common quality controls, and coordination of changes affecting multiple arms | Who ensures the shared process works consistently and that relevant sponsors are informed of failures or changes? |
| Shared-control comparison | Control-arm conduct, control data quality, timing and completeness of control observations, and the production of analysis-ready data used across comparisons | Who is accountable for the control data and procedures relied on by each comparison? |
| Arm-specific obligations | Intervention-specific information, product handling, arm-specific monitoring, and the quality and safety information particular to an intervention | Which sponsor is responsible for the intervention-specific obligation, and how is information with platform-wide implications communicated? |
A shared control is a statistical resource, but also an operational dependency. If a control observation is misclassified or collected under a changed procedure, more than one intervention comparison may be affected. The sponsor arrangements therefore need to identify who oversees control-related procedures and data, how an issue is assessed for each affected comparison, and how the relevant sponsors receive timely information. This does not mean every sponsor must perform every control task. It means the allocation and communication route cannot be left implicit.
Similarly, a shared platform system may be administered by a coordinating organization, but the sponsors relying on it need a defined way to learn about material failures, assess their impact, and take any required action within their own obligations. Operational coordination is not a substitute for an accountable sponsor’s informed oversight.
A Justification for Explicit Allocation
The need for an explicit allocation follows from a simple dependency argument. Suppose an intervention comparison uses a shared control group and a common data system. A defect in the system can alter control records; altered control records can affect the estimated comparison; and affected estimates may matter to more than one intervention sponsor. If the governance arrangement names only the system operator, it identifies who performs a task but not who ensures that the trial obligations are met or who must assess the consequences.
The accountability chain therefore needs three linked elements: an accountable party, an assigned performer, and a route from evidence of performance or failure to the parties who must respond. Omitting any one element creates a predictable gap: an unowned duty, an unmonitored delegation, or an issue that does not reach a sponsor able to act.
This expression is a governance checklist, not a statistical formula. Its components are jointly necessary in practice: naming a sponsor without defining monitoring leaves performance unverified; naming a contractor without naming an accountable party leaves responsibility unclear; and recording both without an escalation route can leave a known failure unresolved.
Worked Example: A Shared Data System Fails
Worked Example: A Shared Data System Fails
Hypothetical situation. A platform compares Intervention A and Intervention B with a common control. A contracted data provider operates the electronic data system for all arms. During a routine check, the platform team discovers that a software update has prevented some control-visit dates from appearing in exported data. The provider operated the system, but the sponsor arrangements do not say who must assess the impact.
Step 1: Identify the affected duties. The immediate issue concerns system performance and the completeness of control data. Because those data contribute to both comparisons, the issue may affect both Intervention A versus control and Intervention B versus control. It is not enough to classify it as a provider’s technical problem.
Step 2: Identify accountability and performance roles. The agreement should identify the party accountable for oversight of the shared system, the provider’s operational duties, and the sponsor or sponsors accountable for the relevant trial obligations. If the platform has multiple legally responsible sponsors, the arrangement must state how each one receives information and oversees its own obligations.
Step 3: Require an impact assessment and a record. The responsible parties should establish which records and time period may be affected, whether source records remain available, and whether data correction is needed. The assessment should consider both comparisons that rely on the control data. The provider’s repair alone does not demonstrate that the affected trial data are complete or suitable for use.
Conclusion. The provider may perform the correction, but the responsible sponsor arrangements must ensure that the defect is investigated, documented, communicated to affected sponsors, and addressed under the trial’s applicable procedures. The example does not assume that one universal sponsor structure applies; it shows why the actual structure and escalation route must be specified.
Worked Example: An Intervention-Specific Safety Signal
Worked Example: An Intervention-Specific Safety Signal
Hypothetical situation. A platform has one sponsor responsible for the platform-wide trial and a separate sponsor responsible for Intervention B. A site reports a serious medical event in a participant receiving Intervention B. The event is initially reported through a shared safety intake team that serves all arms.
Step 1: Separate intake from accountability. The shared team’s receipt of the report is an operational function. It does not, by itself, establish which sponsor is responsible for evaluating or acting on the information. The documented arrangement must identify who receives the report, who assesses it under the applicable requirements, and who makes any required notifications.
Step 2: Route information according to its implications. The Intervention B sponsor needs the information relevant to its intervention-specific obligations. The platform sponsor also needs information when the event may affect shared trial procedures, participants in other arms, or the integrity of the platform. The route should be timely and defined in advance rather than invented after a report arrives.
Step 3: Preserve an auditable record. The record should show when the event was received, who reviewed it, what actions were taken, and which sponsors were informed. If the event does not affect platform-wide conduct, the rationale for that assessment should still be traceable.
Conclusion. Sharing a safety intake function can make reporting more consistent, but it does not collapse arm-specific and platform-wide accountability into one undifferentiated role. The allocation must address both the intervention sponsor’s obligations and any platform-level consequences.
Worked Example: A Change to a Common Procedure
Worked Example: A Change to a Common Procedure
Hypothetical situation. Sites use one common assessment procedure for control participants and participants in two intervention arms. A sponsor proposes a change to the assessment schedule because a measurement device is no longer available. The change could affect data collected in every arm.
Step 1: Classify the scope. Although the change was raised by one sponsor, it concerns common infrastructure and may influence all comparisons that use the assessment. It should not be treated as an arm-only operational adjustment without assessing the wider impact.
Step 2: Identify who must evaluate and document the change. The responsible parties should be clear about who assesses the consequences for participant safety, conduct, data comparability, and each affected sponsor’s obligations. The appropriate approvals and document updates depend on the applicable requirements and the trial’s established arrangements.
Step 3: Confirm implementation and evidence. Before relying on the revised procedure, the responsible parties should ensure that sites receive clear instructions, that implementation can be verified, and that any deviations are documented and communicated. The change record should identify the effective date and the affected sites or participants.
Conclusion. A common procedure has shared consequences even when one sponsor initiates a change. Explicit accountability makes it possible to assess the change across affected arms and confirm that sites implement the approved approach.
Diagnostics: How to Find Accountability Gaps
A responsibility map is only useful if it can be tested against realistic events. Review it using scenarios such as a data-system outage, a serious safety report, a missed control assessment, or a service provider’s failure. For each scenario, ask whether staff can identify the accountable sponsor, the performer, the evidence to review, the recipients of an escalation, and the person responsible for ensuring follow-up.
State which shared control, procedure, system, or service the affected comparison relies on.
Identify the sponsor or sponsors responsible under the actual trial arrangement for the relevant obligation.
Record who performs the task, what evidence shows it was done properly, and how performance is monitored.
Confirm who receives a failure report, who assesses effects on each relevant arm, and how corrective action is tracked to completion.
Warning signs include contracts that name a vendor but not the sponsor accountable for oversight; a shared control process with no owner for data completeness; arm-specific information that cannot reach the platform sponsor when it has broader implications; and change records that do not identify which sponsors were consulted or informed. A responsibility map should also be checked against contracts, protocols, operating procedures, and actual reporting channels. Conflicting documents are themselves a governance risk.
Comparison and Limitations
In a separately sponsored, single-intervention trial, sponsor responsibilities can often be described within one trial’s organizational boundary. A platform with shared controls has additional dependencies: more than one intervention comparison may rely on the same participants, procedures, data, or systems. A single coordinating sponsor can simplify common operations, but it does not automatically make that sponsor responsible for every intervention-specific obligation. Conversely, separate arm sponsors can preserve clear product-specific accountability, but require reliable coordination where their obligations depend on shared trial processes.
The practical comparison is not “centralized is always safer” versus “distributed is always safer.” Centralization can reduce inconsistent procedures while concentrating operational dependency; distribution can clarify arm-specific responsibility while increasing the need for dependable information exchange. The suitable arrangement depends on applicable requirements, agreements, the platform structure, and the duties at issue.
This framework does not determine legal sponsor status, settle contractual disputes, or replace jurisdiction-specific regulatory advice. Nor does it decide which body should approve a strategic or scientific choice. It helps make responsibility for ensuring performance, monitoring delegation, and responding to shared consequences explicit. A responsibility map should be reviewed when the sponsor arrangement, shared services, or dependencies materially change.
Common Mistakes
- “The contractor runs the system, so the contractor is accountable.” A contractor may perform and be contractually responsible for specified tasks, but that fact alone does not establish that sponsor accountability has transferred. State who oversees the work and how failures are escalated.
- “The control is shared, so every sponsor owns every control duty.” Shared use does not imply identical legal duties for all parties. Specify the accountable party or parties for each obligation and how others relying on the control are kept informed.
- “The issue belongs to one arm because it was first noticed there.” A problem originating in one arm may affect common systems, control data, or other comparisons. Assess its scope before limiting its communication or response.
- “A responsibility table is enough.” A table that names people but omits monitoring evidence, escalation, or follow-up may describe an intention without making it operational. Test the allocation using realistic failure scenarios.
Check Your Understanding
Answer each question, then compare your reasoning with the solutions.
- A provider operates a shared data system. What additional information is needed beyond the provider’s name to make accountability traceable?
- Why can a control-data problem affect more than one intervention sponsor?
- A safety report enters through a platform-wide intake team. Does that fact alone identify who is accountable for assessment and action? Explain.
- What should be checked when a common assessment procedure changes?
Solutions. (1) Identify the accountable sponsor or sponsors, the provider’s defined tasks, the evidence used to monitor performance, and the escalation and follow-up route. (2) Multiple intervention comparisons may use the same control participants or records, so a defect can affect the data and interpretation of each comparison. (3) No. Intake is an operational function; the arrangement must separately specify who assesses the report, who acts under applicable requirements, and how relevant sponsors are informed. (4) Assess effects across dependent arms and comparisons, identify the responsible parties and required approvals, document the change and effective date, communicate instructions to sites, and verify implementation.