Counts and proportions can favor different groups without contradicting each other. This donor-cycle study illustrates why both should be retained. Primary report
What was compared
Fifty oocyte donors participated in a prospective observational comparison of recombinant FSH plus LH versus highly purified hMG plus urinary FSH. The recombinant group yielded more total oocytes, while the urinary combination had a higher mature-oocyte proportion. The report found no significant differences in the cited follicular steroid measurements and reported similar ongoing-pregnancy rates. Primary report
Our appraisal: do not remove a component from the intervention
Describing the urinary group as “hMG alone” would misstate the comparison. When both groups contain combinations, a result cannot automatically be attributed to one ingredient. A component-specific claim requires a design that isolates that component.
The count-versus-proportion distinction is equally important. A larger total with a smaller mature fraction can still produce a different absolute number of mature oocytes. The relevant clinical question determines which quantity matters; selecting whichever statistic looks favorable creates a biased summary.
Because allocation was observational, baseline differences and treatment selection remain possible explanations for group contrasts. Prospective data collection improves temporal clarity, but it does not by itself supply the protection of random assignment.
Donor and recipient outcomes also occupy different levels. The hormonal environment belongs to the donor cycle, while pregnancy depends on later processes as well. Full review should trace that linkage and account for clustered observations. This small study is useful for endocrine characterization, not proof of equivalence or a universal fertility advantage.
Frequently asked questions
Was this a randomized trial?
No; it was prospective and observational.
Did the urinary arm receive only hMG?
No; urinary FSH was also used.
Did total yield and mature proportion favor the same group?
No.
Is a mature percentage the same as a mature count?
No.
Do nonsignificant hormone differences prove equivalence?
No.
Can a donor hormone measurement alone explain pregnancy?
No.
What should the comparison retain?
Both complete regimens, the allocation method, and each endpoint's denominator.