Clinical Trial Start-Up: The Gates, Not the Checklist

At a glance

• Start-up is a dependency chain, not a task list. The regulations are unusually specific about order, and the order is what makes a timeline realistic.
• Roles and activities are allocated before activities begin. Agreements are documented before the activities they cover.
• For a US IND trial, the sponsor obtains a signed Form FDA 1572 before permitting an investigator to begin participation.
• Essential records should generally be in place before the trial starts, and may be updated afterwards. Filing them later is not the same obligation.
• Most of start-up genuinely runs in parallel. A small number of things cannot, and those are the only ones worth sequencing a plan around.

Why a checklist is the wrong shape

Almost every start-up resource is a list: contracts, budgets, regulatory submissions, supplies, training, site initiation. The list is not wrong, but it is flat, and flatness is exactly what makes start-up timelines slip. A list tells you what to do. It does not tell you what you are not allowed to do yet.

The regulations are more useful than the lists, because several of their requirements are expressed as timing. “Prior to initiating.” “Before permitting.” “Before the trial starts.” Where a requirement names a moment, the moment is the obligation, and everything that depends on it inherits that constraint.

Read start-up as a set of gates and two things follow. The genuinely blocking items become obvious, and everything else can be run in parallel without anxiety.

Gate one: decide who does what, before anyone does it

The first gate is the least visible because it produces no deliverable a project plan tracks.

Prior to initiating clinical trial activities, the sponsor should determine the roles and allocate their trial-related activities accordingly (ICH E6(R3) §3.3).

This precedes contracting. It is the decision about which party owns which activity, and it has to be made rather than discovered. Trials that skip it do not fail immediately; they accumulate seams, and duties fall into the seams because each party assumed another held them.

The contracting gate follows directly. Agreements made by the sponsor with the investigator or institution, service providers and any other parties involved with the clinical trial should be documented prior to initiating the activities (ICH E6(R3) §3.6.1).

“Prior to initiating the activities” is the operative phrase and it is routinely violated by work beginning under a letter of intent while the master agreement is negotiated. The regulatory problem is not the commercial informality. It is that for the duration of that gap, the allocation of responsibilities is undocumented.

What the agreement has to secure is also specified. The sponsor should obtain the investigator’s, institution’s and where applicable service provider’s agreements to conduct the trial in accordance with the approved protocol and in compliance with GCP and applicable regulatory requirements, to comply with procedures for data recording and reporting, and to retain the essential records (ICH E6(R3) §3.6.3).

Note that those are commitments about conduct, not about scope and payment. A contract that allocates deliverables and fees without securing those agreements has satisfied the commercial need and not the GCP one.

Gate two: the site and the investigator have to be qualified, demonstrably

Selection is a sponsor obligation, and it carries an evidentiary standard.

The sponsor is responsible for selecting the investigators and institutions. Each investigator should be qualified by education, training and experience and should demonstrate they have adequate resources and facilities to properly conduct the trial (ICH E6(R3) §3.7.1).

Two limbs, and the second is the one that bites during start-up. Qualified but not resourced is the commonest failure, and it does not announce itself at selection. It shows up months later as slow enrolment, late data entry, and thin documentation, at which point the cause is expensive to fix.

The investigator side of the same requirement makes the evidence explicit: investigators should be qualified by education, training and experience to assume responsibility for the proper conduct of the trial and should provide evidence of such qualifications (ICH E6(R3) §2.1.1). The obligation is to provide evidence, not merely to be qualified, which is why CVs, licences and training records are start-up artifacts rather than administrative residue.

Gate three: approval, and the permission to begin

Ethics approval is the gate everyone knows about, and the one most often treated as a scheduling problem rather than a constraint. The review body’s written procedures cover conducting initial and continuing review of trials (ICH E6(R3) §1.4.3), and its purpose is to safeguard the rights, safety and well-being of all trial participants (ICH E6(R3) §1.2.1). An approval issued subject to modifications is not yet permission to proceed, and treating it as one is how sites end up enrolling against a protocol version the committee did not approve.

For a trial under a US IND there is a second, separate permission gate that is easy to under-read. Before permitting an investigator to begin participation in an investigation, the sponsor shall obtain a signed investigator statement, Form FDA-1572, containing the investigator’s details, the protocol, the research and laboratory facilities, and the IRB responsible for the study (21 CFR §312.53(c)(1)).

“Before permitting an investigator to begin participation” is a sponsor-side gate, not a site-side formality. The commitments that signature carries are a subject of their own and are covered separately. What matters for start-up sequencing is narrower: until it is signed and obtained, that investigator may not begin.

Gate four: the people and the product are ready

Readiness is usually tracked as training completion. The requirement is slightly different and more specific about the investigational product.

The investigator should be familiar with the appropriate use of the investigational products as described in the protocol, in the current Investigator’s Brochure, in the product information and other information sources provided by the sponsor (ICH E6(R3) §2.1.2).

Three sources, not one. Protocol familiarity alone does not satisfy it, and Investigator’s Brochure familiarity is the half most often assumed rather than evidenced. It also has a downstream consequence worth knowing during start-up: expectedness assessments for safety reporting are judged against the reference safety information, so an investigator who has not absorbed the brochure cannot make those assessments reliably later.

The supply gate, which is narrower than teams expect

Drug supply is usually tracked as a logistics milestone. There is one regulatory constraint sitting inside it, and it is about destination rather than timing.

A sponsor shall ship investigational new drugs only to investigators participating in the investigation (21 CFR §312.53(b)).

Read alongside the 1572 gate, the sequence resolves: an investigator becomes one who is participating when the sponsor has obtained their signed statement and permitted them to begin. Shipping ahead of that is not an early start, it is shipping to someone who is not yet a participating investigator.

The sponsor-side preparation that precedes it is separate and can run in parallel. The sponsor should ensure that an Investigator’s Brochure is developed and updated as significant new information on the investigational product becomes available (ICH E6(R3) §3.15.1), and that the product is characterised as appropriate to the stage of development (ICH E6(R3) §3.15.2). Neither waits on site activation, and both feed the investigator-familiarity requirement above.

Gate five: the records that should already exist

Some essential records should generally be in place prior to the start of the trial and may be subsequently updated during the trial (ICH E6(R3), Annex 1).

The sentence does two things. It sets a timing expectation, and it explicitly permits later updating, which is the part teams miss in both directions. Records generated after the fact are not compliant simply because the file is complete by database lock. And a record that changes during the trial has not failed; that is anticipated.

The practical test during start-up is not “is the file complete” but “would the absence of this document have prevented the activity that has already happened.” Where the answer is yes, the gap is a sequencing failure rather than a filing one.

The one thing start-up decides that cannot be added later

Everything above is about permission. One start-up activity is different, because it determines how the whole trial will be run rather than whether it may begin.

ICH E8(R1) places it explicitly at this stage: the guidance covers the consideration of quality in the design and conduct of clinical studies, including the identification, during study planning, of factors that are critical to the quality of the study (ICH E8(R1), objectives). Not during conduct, and not at the first protocol deviation. During planning.

That identification is what the sponsor’s quality system is then built around, since the sponsor should implement an appropriate system to manage quality throughout all stages of the trial process (ICH E6(R3) §3.10). A monitoring plan written without it is monitoring everything equally, which is the same as monitoring nothing in particular.

The practical consequence for a start-up timeline is that critical-to-quality identification belongs early and cannot be parallelised with the things that depend on it: the protocol, the monitoring plan, and the data that will be collected all inherit its output. It is the one start-up task where doing it late does not delay the trial, it degrades it, which is why it rarely shows up as a slipped milestone and often shows up as findings eighteen months later.

What actually runs in parallel

The honest answer is: most of it. Start-up feels serial because the gates are invisible, and once they are named the parallelism becomes obvious.

Genuinely blocking, in order:

• Allocation of activities, before contracting has anything to allocate
• Agreements, before the activities they cover begin
• Ethics approval, before participants are approached
• The signed 1572, before that investigator begins participation on an IND trial

Genuinely parallel: budget negotiation, supply forecasting and shipment planning, system build and user provisioning, training material preparation, site-facing document drafting, and essential-record collection. None of these depends on another finishing, and sequencing them serially is the most common self-inflicted delay in start-up.

The one that traps teams is contracting, because it sits between allocation and almost everything operational. If contracting is going to be slow, the response is to start it earlier rather than to begin the work it gates and document it afterwards.

The site initiation visit sits at the end of this chain, confirming that the gates above are actually closed before enrolment opens. It has its own requirements and its own checklist, covered separately; treating it as the moment start-up begins rather than the moment it is verified is a reliable way to arrive at it unprepared.

The useful reframing

Ask of every start-up item: what does this block? An item that blocks nothing can run whenever capacity allows. An item that blocks enrolment, or that blocks an investigator from beginning at all, belongs on the critical path regardless of how administrative it looks.

Start-up plans built that way tend to be shorter, more parallel, and considerably harder to slip, because the things that genuinely cannot move are the only things being protected.

Sources

• ICH E6(R3) Good Clinical Practice, version R3
• 21 CFR Part 312, Investigational New Drug Application (current as of 4/14/2026)