The intended use statement is the most expensive line in your study.

Most of the IVD studies I am asked to review begin with a draft protocol, a target subject count, and a timeline. Sometimes there is a preliminary budget. Almost never is there a finalized intended use statement. The intended use is treated as something that will be tightened later, after the protocol is mostly drafted, after the validation results begin to come in. It will be cleaned up at submission. This is the most common error in IVD development I see, and it is responsible for more wasted study budget than any other single decision.

The intended use statement is not a document. It is a sentence — sometimes two — that names what the assay is for, who it is for, and under what conditions. It is the sentence the FDA will read first and will use to decide what evidence the validation has to provide. It is the sentence that determines whether your study needs 200 subjects or 20,000. And one word in it can change everything.

Consider the difference between an assay intended to aid in the diagnosis of a condition and an assay intended to diagnose the same condition. To the casual reader, these phrases sound nearly identical. To the FDA, they are different submissions, with different evidentiary bars, different validation requirements, different post-market obligations, and different liability profiles. "Aid in the diagnosis" can be supported by a clinical performance study of a few hundred subjects, with the assay positioned as one input alongside clinical judgment. "Diagnose" requires a clinical performance study of thousands, with the assay positioned as a sufficient determinant of the diagnostic decision. The same assay. The same scientific data. Two completely different validation paths, separated by three words.

The same dynamic applies across every dimension of the intended use. The population — "adult patients" versus "adult patients with suspected sepsis" versus "adult patients in emergency department settings with suspected sepsis" — determines the recruitable pool and the comparator structure. The condition — "rule out" versus "screen for" versus "monitor" — determines the analytical performance characteristics that have to be demonstrated. The setting — "professional use" versus "near-patient testing" versus "home use" — determines whether the validation has to include lay-user studies, whether the device has to be CLIA-waived, and whether the kit becomes a separately regulated artifact under the medical device statute. One adjective shifts the entire submission.

Sponsors rarely lead with the intended use statement because it is hard to write and feels premature. They want to design the collection, run the validation, see the data, and then describe the assay around what the data showed. The instinct is reasonable on its face and wrong in practice. The intended use determines what data you need to collect. If you collect the data first and write the intended use later, you have, in most cases, collected the wrong data — or collected the right data but in insufficient quantity, or under the wrong conditions, or with the wrong comparator. The validation is built backward from the intended use, not forward into it.

What I push every sponsor to do, before any protocol drafting begins, is to write a draft intended use statement and bring it to a pre-submission with FDA. The pre-submission discussion will not lock the statement in — the statement will continue to evolve through the validation — but the agency will tell you what evidence they would expect to see if that statement were submitted today. That evidence list is the protocol's actual job to produce. The protocol exists to produce the evidence the intended use requires; it does not exist as an independent document the intended use is later squeezed into.

The corollary is the diagnostic question I ask sponsors who want me to review their protocol before they have a finalized intended use. What is the difference between an assay that is right and an assay that is clinically useful? The first is a technical achievement. The second is a regulatory achievement. The intended use statement is the bridge between them, and a sponsor who has not written the bridge has not yet decided whether they are building toward the first goal or the second. The protocol that follows from each goal is different.

This is unglamorous advice. The intended use statement does not produce a deliverable; it produces a constraint. The constraint is the most valuable thing in early IVD development, because it converts a vague desire to "validate the assay" into a specific question the validation has to answer. Specific questions get answered. Vague desires get spent against, expensively, until the budget runs out.

What I tell sponsors I work with: do not write the protocol before you write the intended use. Do not start collection before you have done a pre-submission against the intended use. Do not let the intended use drift during the validation — and if it has to drift, drift it deliberately, with a return-trip pre-submission, not by accident in the submission narrative. The intended use is the most important word in your study, and the rarity of its appearing in the first sponsor conversation is, in my view, the largest single source of wasted IVD development capital in the industry today. Lead with it. Everything else descends from it.