In the realm of clinical research, elapsed time is a strategic performance measure. Every day between protocol approval and final data lock can affect sponsor confidence, site engagement, resource utilisation, and the speed at which evidence becomes available.
Yet clinical trials rarely slow down because one person is unable to complete one task. They slow down because work waits between functions: regulatory submissions sit for review, contracts move through sequential approvals, sites re-enter information into multiple systems, and data issues accumulate until the final cleaning cycle.
Value Stream Mapping (VSM) provides a disciplined way to see the complete flow of work, information, decisions, and data. Instead of optimising one department in isolation, the team examines the value stream from end to end and designs a future state with fewer queues, clearer ownership, and earlier quality checks.
The fundamental purpose is not to remove necessary regulatory controls. It is to remove avoidable delay while protecting participant safety, data integrity, and inspection readiness.
A healthcare VSM review found that published applications commonly focus on lead time, waiting, process efficiency, and value-added activity. The same principle applies to clinical trials: map the flow, quantify the delay, identify the constraint, and improve the system.
1. Define the Value Stream and Select the Scope
A clinical-trial value stream includes the activities required to convert an approved research concept into reliable, analysis-ready evidence.
For this guide, define the scope as:
- Start: Protocol approval by the sponsor governance body
- End: Final database lock
- Flow unit: One clinical trial protocol moving through a defined site network
- Primary customer: Sponsor, regulatory stakeholders, investigators, and ultimately patients who depend on timely evidence
- Supporting functions: Clinical operations, regulatory affairs, contracts, finance, site start-up, data management, biostatistics, quality, and technology
Do not begin with “the entire clinical development process.” Select one protocol family, therapeutic area, or study type. A focused scope produces usable data and prevents the mapping workshop from becoming a general discussion about organisational complexity.
Before mapping, capture the:
- Voice of the Customer (VOC): rapid site activation, complete data, predictable milestones
- Voice of the Business (VOB): controlled cost, compliant execution, portfolio throughput
- Voice of the Process (VOP): actual cycle times, queue times, rework, and handoffs
These voices establish the CTQs: critical-to-quality requirements: that the future state must protect.
2. Build the Current-State Map
A VSM is more than a process flowchart. It combines the work steps with information flow, queue sizes, lead times, process times, decision points, systems, and performance data.
Conduct a cross-functional gemba review with people who perform the work. Include study managers, regulatory specialists, contract negotiators, site staff, data managers, and quality representatives. Document the process as it operates today, not as the procedure says it should operate.
A typical current-state sequence is:
- Protocol approved and distributed
- Country and site feasibility completed
- Site selected
- Regulatory and ethics packages prepared
- Budget and contract negotiated
- Essential documents collected
- Site initiation visit completed
- Site activated
- Participants enrolled and visits conducted
- eCRF data entered and queries raised
- Queries resolved and data reconciled
- Last patient last visit completed
- Final cleaning and review performed
- Database locked
At each step, record:
- Process time: Hands-on time required to complete the activity
- Waiting time: Time the work remains idle before the next action
- Lead time: Total elapsed time from start to completion
- First-pass yield: Percentage completed correctly without rework
- Work in process (WIP): Sites, documents, queries, or approvals currently open
- Systems used: CTMS, eTMF, EDC, e-signature, email, spreadsheets, and other platforms
- Approval count: Formal checkpoints and average queue time at each checkpoint
This approach aligns with Lean Six Sigma process mapping practice: observe the actual workflow, validate it with stakeholders, and use baseline data before designing improvements.
3. Worked Example: Quantifying the Site-Activation Bottleneck
Consider a hypothetical Phase II study with 24 planned sites. The current-state data below represents a four-month sample.
| Process segment | Process time per site | Average waiting time | First-pass yield |
|---|---|---|---|
| Feasibility and selection | 6 hours | 8 days | 92% |
| Regulatory and ethics package | 12 hours | 18 days | 78% |
| Budget and contract negotiation | 10 hours | 31 days | 65% |
| Essential document collection | 5 hours | 14 days | 72% |
| Site initiation and activation approval | 8 hours | 22 days | 83% |
| Total before activation | 41 hours | 93 days | : |
The map shows an important pattern: the team performs approximately 41 hours of work, but the site waits an average of 93 days to become active. Process cycle efficiency is therefore:
[
\text{Process Cycle Efficiency} =
\frac{41\text{ hours}}{41\text{ hours} + (93 \times 24\text{ hours})}
]
[
= 1.8%
]
That does not mean every waiting hour is removable. Regulatory review and investigator availability are legitimate constraints. However, the result signals a system dominated by queues rather than value-adding work.
The bottleneck is budget and contract negotiation. It has the longest wait, the lowest first-pass yield, and multiple handoffs between the sponsor, site, finance, and legal teams.
A published clinical research improvement project provides a useful real-world comparison: its protocol intake process recorded an average 121.5 days for budget negotiation, with budget build and approval adding approximately 20.89 days in one segment. The proposed countermeasures included standardised templates, electronic document routing, checklists, reminders, and dedicated ownership.
4. Identify the Eight Wastes
Use the eight DOWNTIME wastes to examine the current state without blaming individuals.
- Defects: Incomplete regulatory packets, incorrect budget assumptions, and inconsistent document versions
- Overproduction: Preparing duplicate reports or collecting documents before the requirements are confirmed
- Waiting: Sites waiting for contracts, signatures, ethics responses, or activation decisions
- Non-utilised talent: Experienced site-start-up specialists spending time chasing signatures manually
- Transportation: Moving information between disconnected systems or email threads
- Inventory: Large queues of inactive sites, unresolved queries, or unsigned agreements
- Motion: Repeated navigation across systems, shared drives, and spreadsheets
- Extra-processing: Duplicate data entry, repeated reviews, and approvals that do not change the decision
The most visible waste is often waiting, but the root cause may be variation in submission quality, unclear approval criteria, or an overloaded constraint.
During the Analyse Phase of DMAIC, use Pareto charts, cause-and-effect diagrams, stratification, and a detailed approval log to test the causes. For example, if 60% of contract delays come from missing budget assumptions, the solution should begin with standard input requirements: not with additional escalation meetings.
5. Design the Future-State Map
The future state should create a smoother flow while retaining required governance.
Key design principles include:
- Front-load protocol and budget quality. Use a structured pre-start-up review to identify unusual procedures, visit burdens, pass-through costs, and country-specific requirements before site negotiation begins.
- Run safe activities in parallel. Begin feasibility, regulatory preparation, data-management planning, and standard contract preparation concurrently where dependencies allow.
- Create a single digital intake. Use one controlled checklist with visible status, owner, due date, and escalation rule.
- Standardise approval criteria. Approval is essential for governance, but undefined approval thresholds create bottlenecks. Set decision rights and service-level targets.
- Build quality at the source. A submission that is complete on the first pass reduces rework and protects downstream capacity.
- Clean data continuously. Begin edit-check review, query monitoring, and reconciliation during the trial rather than waiting until last patient last visit.
- Use Agile coordination. Short cross-functional iterations, visible work queues, and weekly review cycles complement DMAIC by accelerating feedback without weakening control.
The future state should also define a practical control system. A dashboard might display active-site count, median activation lead time, contract first-pass yield, days in approval, open query ageing, and projected database-lock date.
6. Current Versus Future Performance
The following target state is hypothetical and should be validated through a pilot.
| Measure | Current state | Future-state target | Improvement |
|---|---|---|---|
| Site activation lead time | 93 days | 48 days | 48% reduction |
| Hands-on process time | 41 hours | 35 hours | 15% reduction |
| Contract first-pass yield | 65% | 90% | +25 percentage points |
| Regulatory package first-pass yield | 78% | 94% | +16 percentage points |
| Approval queue time | 22 days | 7 days | 68% reduction |
| Queries open at last patient last visit | 420 | 180 | 57% reduction |
| LPLV-to-database lock | 150 days | 90 days | 40% reduction |
| Process cycle efficiency before activation | 1.8% | 3.0% | 67% relative increase |
The target is not simply “work faster.” It is to reduce avoidable waiting, improve yield, and move data-quality work earlier. Throughput: the number of sites activated or clean records released per period: should rise without increasing compliance risk.
7. Sequence the Kaizen Work
Do not launch every improvement simultaneously. Sequence kaizen activities according to the constraint.
Kaizen 1: Stabilise the intake
- Define the minimum complete submission
- Create a standard budget and regulatory checklist
- Assign one accountable owner
- Establish baseline measures for lead time and rework
Kaizen 2: Improve the bottleneck
- Introduce standard contract clauses and negotiation ranges
- Set approval service-level agreements
- Add electronic signatures and automated reminders
- Establish an escalation path for ageing work
Kaizen 3: Connect the flow
- Integrate CTMS, eTMF, and contract status where possible
- Remove duplicate data entry
- Display WIP and ageing by site
- Review the constraint daily during the pilot
Kaizen 4: Move quality upstream
- Use early data review and risk-based monitoring
- Track query ageing throughout the study
- Complete reconciliation before the final lock cycle
- Confirm that control limits and response rules are understood
Kaizen 5: Control and sustain
- Re-map the process after 8–12 weeks
- Compare actual performance with the future-state targets
- Audit standard work adoption
- Escalate recurring variation through the governance forum
Build Capability to Improve the Whole Value Stream
A clinical-trial VSM is most effective when the organisation develops shared improvement capability. Yellow Belts can support data collection and local kaizen. Green Belts can lead focused DMAIC projects involving activation, contracts, or data cleaning. Black Belts can manage complex cross-functional programmes, mentor Green Belts, and connect project outcomes to enterprise strategy.
The relationship can be expressed as Y = f(x): database-lock performance is the output, while inputs such as protocol complexity, approval time, submission quality, site responsiveness, and query ageing influence the result.
If your organisation is ready to move from isolated process fixes to disciplined, data-driven transformation, Lean Six Sigma training provides the methods, tools, and practical structure to lead that change.
Pursue Lean Six Sigma certification with Lean 6 Sigma Hub to develop the capability to map value, remove bottlenecks, and deliver measurable improvement across complex clinical-trial processes. Explore the Lean Six Sigma Green Belt Online Training or advance to the Lean Six Sigma Black Belt Online Training, both designed for practical, self-paced application and accredited by CSSC.
For additional guidance, see the process mapping guide, project scope boundary calculator, and CTQ tree alignment calculator.
Kaizen. Kai-Care. Kai-Done. ( Lean Six Sigma)
