Talk:Trial data comparison
Discussion page·3 threads·1 archive
This is the discussion page for the article Trial data comparison. It is for improving the article: sources, wording, structure, scope and titles. It is not a general discussion forum about the subject, and it is not a place to ask for advice — see Project:Medical disclaimer.
Cross-trial tabulation and the estimand problem
Tagged {{refimprove}}
The article explains the problem at length in §Why the estimand matters, which is good, but a reader who scans the table and leaves has not read that section. Proposal: an estimand column, with "not stated" where the publication did not say. — — ForestPlotFinn (talk) 10:22, 6 May 2026 (UTC)
Support the column. It will be embarrassingly full of "not stated", which is itself the finding. — — ITT_Imelda (talk) 13:48, 6 May 2026 (UTC)
Support in principle, but the table is already eight columns wide and scrolls on a phone. Alternative: append the estimand to the result cell in parentheses. Less sortable, but it puts the caveat where the number is rather than three columns away. — — TrialsTabitha (talk) 09:15, 7 May 2026 (UTC)
I take the width point. Parenthetical in the result cell is acceptable to me. Leaving the tag until it is done. — — ForestPlotFinn (talk) 08:41, 8 May 2026 (UTC)
A further wrinkle: for the withdrawal trials (STEP 4, SURMOUNT-4) the "result" is two numbers with opposite signs, and the estimand question applies differently to each arm. Those rows may need a footnote rather than a parenthetical. — — EndpointEnid (talk) 14:33, 19 June 2026 (UTC)
Head-to-head filter: is the flag applied consistently?Done
The head-to-head checkbox includes SURPASS-3 and SURPASS-4, which compared tirzepatide against insulin. Those are active-controlled but they are not head-to-head against another incretin agonist, which is what a reader ticking that box almost certainly wants. — — SurpassSunniva (talk) 12:44, 13 October 2025 (UTC)
Good catch, and the fix is a wording change rather than a data change: the label should read "active comparator" rather than "head-to-head". Any reader wanting incretin-versus-incretin can then read the comparator column. — — TrialsTabitha (talk) 16:02, 13 October 2025 (UTC)
Support. "Head-to-head" in the trial literature just means a direct randomised comparison, whatever the comparator, so the current label is technically right and practically misleading. — — AppendixAlva (talk) 08:37, 14 October 2025 (UTC)
Relabelled, and the comparator column now names the active drug in every case. ✓ Done — — SurpassSunniva (talk) 09:26, 15 October 2025 (UTC)