Before a trial can estimate treatment effects, it must identify an eligible population. This companion report examines that recruitment problem rather than tirzepatide's efficacy. Primary report
Study snapshot
The investigators described screening 1,583 people, with 651 proceeding to biopsy and 190 randomized. During recruitment, inclusion criteria were amended to incorporate a FAST-score threshold and a changed AST criterion. Overall qualification yield changed little, while patterns differed among investigator specialties. Primary report
Our appraisal: identify the denominator being improved
Qualification per biopsy and qualification per person screened are different operational measures. A screening strategy can change one without producing an equivalent change in the other. Costs, invasive procedures, and missed eligible participants also matter when evaluating recruitment efficiency.
The before-and-after amendment comparison is not itself randomized. Calendar time, site experience, referral patterns, and the mix of candidates may differ across periods. These possibilities should be considered before attributing every change to the amended criteria.
Recruitment yield is also not a complete diagnostic-validation study. Without a suitable reference-standard assessment across the relevant candidate population, one cannot infer sensitivity, specificity, or the number of eligible patients excluded by the screen from qualification rates alone.
This paper belongs beside the efficacy report as design context, not as another confirmation of treatment benefit. Full review should inspect site-level counts, protocol versions, screening reasons, and reference-standard availability. Its distinct contribution is explaining how a biopsy-defined trial population was assembled and why that population may differ from routine clinical referrals.
Frequently asked questions
Does this report test tirzepatide efficacy?
Its focus is trial design and screening yield.
Were the screening criteria unchanged throughout recruitment?
No. An amendment was introduced.
Was the amendment comparison randomized?
No.
Are qualification per biopsy and per screened person identical?
No.
Does recruitment yield alone establish diagnostic sensitivity?
No.
Is this independent efficacy replication?
No. It is a companion design analysis.
Why is it useful?
It clarifies population selection, recruitment efficiency, and limits on generalizing trial results.