Every trial passes through the same stages, whatever its phase or size: concept, design, approvals, start-up, conduct, data lock, analysis, reporting and archive. This guide shows what each stage involves and the decisions that save time later.
Free sandbox · No credit card · 21 CFR Part 11 aligned
Month
Protocol, endpoints
Ethics, authority
Sites, build, UAT
Visits, monitoring
Cleaning, lock
Report, archive
The lifecycle at a glance
Overview
People often meet a clinical trial one stage at a time, as a protocol writer, a coordinator, a data manager or a statistician. Seeing the whole sequence explains why certain choices matter. A vague endpoint becomes an ambiguous form, then dozens of queries, then a delay at lock. A forgotten consent version becomes a deviation. An archive plan nobody wrote becomes a scramble years later. The lifecycle view is a way to make the downstream cost of upstream decisions visible.
This guide uses the same stages for any trial, from a Phase 1 first-in-human study to a Phase 3 pivotal trial. The scale changes: more sites, more oversight, more formal documents. The order does not. If you are after the data-specific sequence in more depth, read the clinical data management process. If you are after the cost side, see how much a clinical trial costs.
The stages
A question, a population and a rough design. Check whether participants exist in sufficient numbers and whether a trial is the right tool. The protocol synopsis generator help shape it.
Endpoints, eligibility, visit schedule, sample size and statistical plan. This document defines every later step. Write it so the schedule of assessments can be turned directly into forms.
Budget, agreements with sites and vendors, insurance and sponsor responsibilities. Unclear responsibilities here cause delay in every later stage.
Ethics committee or IRB review, competent authority authorisation where required, and trial registration. In the EU, applications for medicinal product trials run through CTIS; see the CTIS submission guide. In the US, see the IRB submission checklist.
Site selection and activation, training, essential documents, and the build of data capture, consent and diaries. The fastest way to launch a trial covers shortcuts that remain safe.
Screening, consent, randomization, visits, safety reporting and monitoring. Data flow from sites, queries open and close, and deviations are logged.
Edit checks, reviews, SDV or risk-based monitoring, safety reconciliation and amendments. Done continuously, this shortens the next stage. See risk-based monitoring.
A planned freeze after cleaning, coding and reconciliation are complete, followed by unblinding in blinded trials. Use the database lock checklist.
Statistical analysis on the locked dataset, clinical study report or publication, results posting where required and lay summaries.
Site close-out visits, final notifications, reconciliation of study supplies, archiving of essential records for the required period and decommissioning of systems. See archiving and retention.
Upload your protocol, review the AI-drafted visits and forms, and test everything in the free sandbox during approvals.
By phase
The stages are constant. Emphasis shifts. These are general patterns, not rules.
| Phase | Typical emphasis | Data-system note |
|---|---|---|
| Phase 1 | Safety and dosing, small cohorts, tight timing of procedures, rapid decision between cohorts | Frequent data review and fast query turnaround. See EDC for Phase 1. |
| Phase 2 | Signal of efficacy, dose finding, more sites, richer endpoints | Amendments are common, so versioning matters. See EDC for Phase 2. |
| Phase 3 | Confirmatory evidence, many participants and sites, long follow-up, formal oversight | Scale, consistency across sites and inspection readiness. See EDC for Phase 3. |
| Phase 4 and observational | Long-term safety, real-world use, registries | Large volumes and patient-reported data. See EDC for observational studies. |
Time
Delays cluster in a few predictable places. Approvals take longer when the dossier is incomplete, so a careful first submission is faster than a quick one. Start-up stalls when contracts and essential documents are collected late, which is why a tracker with owners and dates is worth the effort. Enrolment is the stage most likely to overrun, and a feasibility check that was too optimistic shows up here. At the other end, lock is delayed by cleaning that was left for the end.
Two habits help in every phase. First, treat the data system as a stage-five deliverable with its own test, not an afterthought: if you can build and test during approvals, you can go live the day sites are ready. Second, decide the end at the start: who owns the archive, which exports will be taken and where they will live. The cost of those decisions is small early and large late. For teams looking at the whole bundle, clinical trial platform choices for startup biotech outlines what to ask.
In Capture
Capture is a self-serve platform for the data side of stages five to ten: build the study or let AI draft it from the protocol, test free, go live with approved forms, collect EDC, ePRO and eConsent data under one audit trail, clean with edit checks and queries, and export with a data dictionary. Hosting is available in the EU (Frankfurt) and the USA (N. Virginia). It does not replace your regulatory submission or your operational tracker.
Stage 1
DraftBuild and edit freely
Stage 2
ApprovedLocked for live use
Stage 3
In the casebookOnly approved forms
SDV requirement changed on Systolic BP
Data manager · Reason: critical safety value
Form approved: Vital Signs (VS) v2
Study owner · Locked for live use
Answer changed: Weight 68.0 to 68.5 kg
Site coordinator · Reason: transcription error
Checklist
Can every endpoint be traced to a form field and an analysis?
Are approved versions of the protocol and consent filed and consistent with the build?
Have forms been tested with scripted scenarios, and do sites have roles and training?
Is someone reviewing data, queries and deviations on a schedule?
Is the checklist complete with approvals recorded?
Are exports, audit trail and records archived with a named owner?
Concept and feasibility, protocol design, funding and contracts, approvals and registration, start-up and system build, enrolment and conduct, ongoing cleaning and oversight, database lock, analysis and reporting, and close-out and archive.
It varies widely with phase, design, country and enrolment rate. Small early-phase studies can run in months, while confirmatory trials often run for years. Enrolment is the stage most likely to overrun.
Phases 1 to 4 describe the stage of development of a product. The lifecycle describes the stages every individual trial goes through, regardless of phase, from concept to close-out.
During approvals and start-up, so that testing is done before the first participant. Building from the draft protocol saves time, as long as you update it after any amendment.
Sites are closed, final notifications and reports are made, supplies are reconciled, records are archived for the required retention period and systems are decommissioned or exported. Retention rules differ by region.
The stages are similar but lighter: approvals and consent depend on design, there may be no investigational product, and data collection may rely on registries or patient-reported data.
Capture covers the data side from build and test through conduct, cleaning, lock and export. Regulatory submissions, contracts and operational tracking are handled outside it.
Keep exploring
Free sandbox with every feature. No credit card, and you pay only when you go live with real participants.