15. Reading and Reporting a Trial: Flow, Registration, Integrity
After this lesson you can
- Interpret an estimate and interval for importance, not just significance
- Draw a participant flow diagram that adds up from assessed to analyzed
- Build a results table that names the analysis population in every row
- Recognize outcome switching and the four forms of spin
- Appraise a trial report with eight questions
Video coming soon
Lessons publish daily from November 17, 2026.
The closing lesson shows how to interpret and report a trial: the estimate and its interval read for importance, the participant flow diagram and the results table. It covers registration, outcome switching, the forms of spin, the integrity ladder and eight questions for appraising any trial report.
Questions and answers
Questions a learner would ask about this lesson. Think through your own answer before you open each one.
- What does a reported result need besides a p value?
Show answer
Three things: the number (how much the endpoint changed), the interval (how precisely it is known) and the registered endpoint (whether it was the pre-specified primary). In the fictional Oakhaven trial, the abstract should open with 20 versus 40 percent poor control and its interval, not with "dramatically cut poor control."
- What is the difference between importance and significance?
Show answer
Significance asks whether the interval excludes zero; importance asks whether the effect matters to a patient or a clinic. A tiny effect can be significant in a huge trial, and a large effect can be non-significant in a tiny one. Oakhaven reads its 20-point estimate against the smallest important difference of ten points: likely important, size uncertain.
- What does the Oakhaven participant flow diagram show?
Show answer
1,000 assessed for eligibility; 500 ineligible, 200 declined and 50 failed the run-in, so 750 left before randomization. 250 were randomized, 125 and 125; 10 per arm were lost by twelve months and 115 per arm were measured. All 125 and 125 were analyzed by intention to treat under the pre-specified rule for the missing.
- What does registration prevent?
Show answer
The registry entry is public, time-stamped and filed before the first participant enrolls, and it records the primary endpoint, the design and the planned sample size. It exposes outcome switching, where the reported primary differs from the registered one, and it makes unreported trials findable.
- What is spin, and how does it differ from fabrication?
Show answer
Spin is reporting that leads a reader to a stronger conclusion than the results support, even when every number is correct. The lesson names four forms: reporting a secondary as if primary, emphasizing a subgroup, the language of "trend toward" or "approached significance," and selective reporting or not reporting at all. Fabrication, inventing or altering data, sits at the top of the integrity ladder above honest error and questionable practice.
Practice quiz
- Question 1How many people does the Oakhaven flow diagram show as assessed for eligibility?
- Question 2Of the 1,000 assessed, 250 were randomized. How many left before randomization, and which three boxes account for them?
- Question 3The risk difference for the Oakhaven primary result is:
- Question 4Trial registration before enrollment mainly prevents:
- Question 5Reporting a secondary endpoint as the main finding when the registered primary was null is:
- Question 6An abstract says "the program works especially well in younger adults," based on risk ratios of 0.4 and 0.65 with an interaction p of about point three. The honest statement is:
- Question 7Why does the Oakhaven report show 125 analyzed per arm when only 115 were measured?
- Question 8Coordinator B recorded 51 run-in failures instead of 50; it was found, corrected and logged. On the integrity ladder this is:
- Question 9How should the baseline table by arm be presented?
- Question 10The first appraisal question to ask of a trial report is:
Saved in your browser only: no account, no server.
Design of Clinical Studies is education only. Every lesson follows the fictional Oakhaven Adherence Trial with invented numbers; the course uses no real data and cites no real studies.