Clinical Trial Workflow: The Whole Trial as One Flow Chart

At a glance

• A clinical trial workflow is best understood as one end-to-end flow chart: four stages drawn as sequential blocks, with the operational handoffs and decision gates between them.
• The four stages are planning/protocol, study startup, conduct, and closeout. Each has inputs, outputs, and an owner, and each hands off to the next at a gate.
• “Stages” (the operational flow) are not the same as clinical-trial “phases” (Phase I to IV, the regulatory classification of a study’s purpose). This page is about stages.
• The gates between stages are where trials stall: approval before activation, activation before enrollment, clean data before lock. Drawing them makes the bottlenecks visible.
• The deliverable is a reusable diagram you can rebuild, not another prose checklist that buries the flow in paragraphs.

Most “clinical trial workflow” pages give you a prose checklist that buries the actual flow in paragraphs. What a reader usually wants is the shape: a single diagram of how a trial moves from idea to archive, with the handoffs and decision points drawn in. This guide is built around that diagram.

It renders the whole trial as one labeled flow chart of four stages, describes the chart in enough structural detail to rebuild it, clarifies the stages-versus-phases confusion, and gives each stage its inputs, outputs, and owner. It is diagram-led and deliberately stays at the flow level; the deep how-to for startup, the timeline, and oversight live in their own guides.

The at-a-glance flow chart

Picture four blocks left to right, each feeding the next through a labeled gate:

[1. Planning / Protocol] --(protocol final + approvals)--> [2. Study Startup]
        |                                                        |
   inputs: concept,                                       (sites activated)
   feasibility                                                   v
        |                                                 [3. Conduct]
        v                                                        |
   output: protocol,                                      (LPLV + clean data)
   plans                                                         v
                                                          [4. Closeout]
                                                                 |
                                                          output: CSR, archive

Read it as a pipeline: each stage produces an output that becomes the next stage’s entry condition, and the arrows are gates, the next stage cannot truly begin until the gate condition is met. The whole flow runs from a study concept to a final report and archive. ICH E6(R3) describes trial conduct as spanning exactly this arc: the term covers processes from planning to reporting, including planning, initiating, performing, recording, oversight, evaluation, analysis, and reporting (Introduction). The chart is that sentence drawn out.

Stages vs. phases (clear this up first)

A frequent confusion: “stages” and “phases” are different things.

• Stages are the operational workflow of running any trial: planning, startup, conduct, closeout. This page is about stages.
• Phases (Phase I, II, III, IV) classify a study by its purpose and where it sits in a drug’s development (first-in-human safety, efficacy, confirmatory, post-market). A Phase II trial and a Phase III trial both run through the same four operational stages.

So a trial has one phase (its type) and moves through four stages (its workflow). Keep them separate and the rest is clear.

Why the distinction matters in practice: phase tells you the scientific stakes and roughly the scale (a first-in-human Phase I and a large Phase III differ enormously in size and risk), but it does not change the operational shape. Both still need a protocol, still need approvals and activated sites, still enroll and collect and monitor, still lock a database and report. So when you are planning the operations of a trial, you reason about stages and gates; when you are reasoning about the science and risk, you reason about phase. Confusing the two leads to timelines built around “Phase II” as if that were a workflow step, which it is not.

Stage 1: Planning / Protocol

• Inputs: the study concept, scientific rationale, and feasibility work.
• Activity: designing the trial and finalizing the protocol, plus the supporting plans (project plan, data management plan, monitoring plan). This is where quality is designed in, ICH E8(R1) and ICH E6(R3) call for building quality into the design by identifying the factors critical to trial quality up front (ICH E8 § 3.1; ICH E6 § 6.2).
• Outputs: a final protocol and the operational plans.
• Owner: typically the sponsor’s clinical and medical leads, with the project manager assembling the plans.
• Gate to next stage: protocol finalized and ready for regulatory/ethics submission.

Stage 2: Study Startup

• Inputs: the final protocol and plans.
• Activity: regulatory and IRB/EC submissions, site selection, contracts and budgets, essential-document collection, and site initiation. This is the most gate-dense stage and where trials most often stall.
• Outputs: approvals, executed agreements, and activated sites. GCP gates this stage explicitly: no trial activity requiring approval proceeds without IRB/EC approval (§ 1.4.7), and agreements with sites and service providers are documented before activities begin (§ 3.6.1).
• Owner: the clinical operations / project management function, coordinating regulatory, legal, and sites.
• Gate to next stage: sites activated (the green light to enroll).

Stage 3: Conduct

• Inputs: activated sites.
• Activity: enrollment, data collection, monitoring, safety reporting, and ongoing oversight. The trial is live; this is the longest stage. Oversight here is a sponsor obligation: appropriate oversight of delegated activities must be maintained throughout (§ 10.2, § 10.3).
• Outputs: collected data through last patient last visit (LPLV), and a maintained oversight record.
• Owner: clinical operations, with sites and the CRO doing the work under oversight.
• Gate to next stage: LPLV reached and data cleaning underway toward a clean database.

Stage 4: Closeout

• Inputs: LPLV and cleaned data.
• Activity: database lock, final analysis, the clinical study report, reconciliation, and archiving.
• Outputs: the locked database, the study report, and a complete archive. Record-keeping and retention obligations apply here (§ 3.16.3).
• Owner: data management and the sponsor, with the project manager driving closeout.
• Gate: the trial is complete and inspection-ready in the archive.

The hand-offs: who passes what to whom

The arrows in the chart are not just dates; they are hand-offs between functions, and each one is a place a trial can drop the baton. Reading the flow by who-hands-to-whom makes the coordination explicit:

• Planning to startup: the clinical and medical leads hand a final protocol and the operational plans to clinical operations. If the protocol is still moving, startup cannot stably begin, because every downstream document keys off it.
• Startup to conduct (the activation hand-off): clinical operations hands activated sites to the site teams and monitors. This is the trial’s most important single hand-off: the protocol becomes a live, enrolling study only once approvals, contracts, and site initiation are all complete. Everything upstream exists to reach this point, and everything downstream depends on it.
• Conduct to closeout: the site teams and CRO hand a complete, cleaned dataset (through LPLV) to data management. A messy hand-off here, queries still open, data not locked, stalls the entire closeout.

Naming the owner on each side of the arrow is what keeps a hand-off from becoming a gap. The diagram’s value is that it shows not just the work but the seams between teams.

Reading the flow: where it stalls

The value of drawing the workflow is that the gates become visible, and the gates are where trials stall:

• Planning to startup: a protocol that keeps changing delays submission.
• Startup to conduct: the big one, approvals, contracts, and activation gate enrollment, and startup delays push the entire downstream flow. Industry experience is consistent that startup is the usual long pole, so it is where management attention pays off most.
• Conduct to closeout: slow or messy data cleaning delays database lock.

Seeing the trial as a gated flow tells you where to put management attention: on the handoffs, not just the work inside each stage. The practical move is to assign each gate an owner and a clear entry condition, then review the trial by gate rather than by calendar, because a gate that has not been met is a more honest status than a date that has merely arrived.

A note for lean teams: because the workflow is a sequence of handoffs and owners, it benefits from a single place that tracks where the trial is and who owns the next gate. A clinical project management tool can hold that operational picture; TrialTrack is one option that maps tasks and milestones to the studies, sites, and vendors involved, so the flow’s current state is visible (it is a coordination tool and does not confer compliance). The diagram is the model; a tool keeps it current.

Frequently asked questions

What are the stages of a clinical trial workflow? Four: planning/protocol, study startup, conduct, and closeout, each producing an output that gates the next.

How are stages different from phases? Stages are the operational workflow every trial runs through. Phases (I-IV) classify a study by purpose and development position. A trial has one phase and moves through all four stages.

Which stage causes the most delay? Study startup. Its gates (approval, contracts, activation) sit before everything else, so a startup delay pushes the whole timeline.

What gates each stage transition? Protocol finalization gates startup; site activation gates conduct; LPLV and clean data gate closeout. Each gate is a condition the next stage needs.

Who owns each stage? Roughly: sponsor clinical/medical leads own planning, clinical operations owns startup and conduct, and data management with the sponsor owns closeout, with the project manager coordinating across all four.

The bottom line

A clinical trial workflow is one end-to-end flow of four stages, planning, startup, conduct, closeout, connected by gates where each stage’s output becomes the next stage’s entry condition. Draw it as a diagram, keep stages distinct from phases, and the bottlenecks (especially the startup gate) become obvious. The workflow is the map; managing the handoffs between stages is how you keep a trial moving.

Sources

• ICH E6(R3) Good Clinical Practice
• ICH E8(R1) General Considerations for Clinical Studies