CTRI registration is a gate, not a formality
Study start-up is a set of parallel activities that mostly tolerate slipping. Site contracts move, investigator meetings are rescheduled, import licences arrive late and the study absorbs it.
Registration on the Clinical Trials Registry of India is the exception. It has to be complete before the first participant is enrolled, and there is no version of the schedule where that is recovered afterwards. Every other start-up task can be late. This one can only be wrong.
What prospective actually means here
Registration must precede enrolment of the first participant. Not the first site initiation, not the first screening log, not the database going live. The first participant.
This trips teams up because enrolment is a site-level event and registration is a sponsor-level task, and the two are usually tracked by different people in different systems. A site that is ready, has its approvals, has the drug in hand and has a willing participant will enrol. If the registration entry is still in query at that moment, nobody at the site is likely to know.
The consequence is not a fine. It is that the trial is retrospectively registered, permanently, in a public record. Journals following ICMJE policy will decline to publish the results. Partners doing diligence will see it. Regulators reviewing a later submission will see it. It follows the study for the rest of its life, for a reason that was administrative.
The dependency chain that decides the date
Registration is not the first step, and treating it as an independent task is how it ends up late. It sits downstream of two things.
The regulatory permission has to exist, because the entry records it. Under the New Drugs and Clinical Trials Rules, 2019, the licensing authority works to defined timelines, with a shorter clock for applications relating to drugs discovered in India or developed and manufactured in India than for everything else. Those timelines are the planning basis, but a query resets the practical schedule, and queries are normal rather than exceptional.
Ethics Committee approval has to exist for at least the first sites, because the entry lists them. That in turn depends on the committee being registered, on its meeting calendar, and on the submission being complete enough to be taken up rather than deferred. A committee that meets monthly and defers once has cost six weeks.
Both of those have to land, then the registry entry is prepared, then it is reviewed and queried, then it is published. Work backwards from the enrolment date through all of it and the honest start date is usually earlier than the plan assumes.
Put first participant in on the plan as a gate rather than a milestone, and put registration published to its left with real float. It is the one date in start-up that costs something permanent if it is missed.
The registry entry is a public description of your trial
The second common mistake is treating the entry as a form to be completed rather than a document to be drafted. It is visible to anyone, including investigators at competing sites, patient groups, journalists and the regulator.
Two fields deserve real attention.
The primary outcome, with its time point. This is the field a journal editor compares against the manuscript, and a mismatch between what was registered and what is reported is one of the most damaging findings a paper can attract. If the protocol says the primary endpoint is measured at twenty-four weeks, the registry has to say twenty-four weeks. Vagueness here is not safety, it is a problem deferred.
The eligibility criteria, which have to match the protocol. Summarising them loosely for the registry creates a public document that disagrees with the approved protocol, and reconciling the two later is harder than getting it right once.
Amendments are part of the record
The entry is not finished when it is published. Changes to outcomes, sample size, sites or the trial's status are expected to be reflected, and the registry keeps the history.
That history is the point. A trial whose primary outcome changed after enrolment began, with the change recorded and dated, is a trial with an explanation available. A trial whose primary outcome differs between the registry and the publication with no recorded amendment has no explanation available, and the absence reads badly regardless of the reason.
The practical habit is to update the registry in the same step as the protocol amendment, not as a follow-up task. Follow-up tasks in start-up survive; follow-up tasks during conduct do not.
What we ask clients before the plan is fixed
Three questions, all of which are cheap to answer early and expensive to answer late.
Which Ethics Committees are involved, when do they meet, and are they all currently registered. A committee whose registration is due for renewal during your start-up window is a risk that is invisible until it is not.
Who owns the registry entry, and do they have access to the protocol as approved rather than as drafted. Entries prepared from an earlier protocol version are the most common source of registry and protocol disagreement.
What is the first site likely to enrol, and what is the earliest date it could. Not the planned date. The earliest. That number, not the plan, is the one the registration date has to beat.