Repeated administrations can generate more observations than the number of people enrolled. They do not generate more independent participants. That distinction matters in this small Melanotan II experiment. Primary study
Study and findings
Ten men with erectile dysfunction and organic risk factors completed a placebo-controlled crossover study. Melanotan II produced more reported erections and longer measured rigidity than placebo; nausea was more frequent, including severe episodes. The publisher's report explains the unequal administration counts: one participant received the wrong vial sequence. Primary report
Our appraisal: preserve person, period, and event
The probability of a response per administration is not the same as the proportion of patients who respond at least once. Someone who responds once and someone who responds consistently may both count as responders in a person-level analysis. A summary should state which definition is being used.
Crossover analysis should account for repeated observations within a person, treatment order, and possible carryover. Treating every administration as an unrelated observation can overstate precision. The administration error also needs explicit handling in assignment-based and received-treatment analyses.
Laboratory rigidity and reported desire are different endpoints, neither of which automatically establishes sustained improvement in everyday sexual function. Short observation windows cannot settle long-term effectiveness, acceptability, or safety.
The severe nausea observations should not be hidden behind an overall statement that no participant withdrew. Completion does not mean absence of clinically important adverse effects. Conversely, event-level adverse-effect rates should not be presented as if they were patient-level risks.
This study provides early human pharmacological evidence under supervised conditions. It does not establish a general treatment recommendation, safe self-use, or equivalence between the investigated material and an unverified commercial preparation. Full review should retain the small sample and the crossover-analysis limitations.
Frequently asked questions
How many independent participants were enrolled?
Ten.
Are repeated injections independent patients?
No.
Were administration counts perfectly balanced?
No; an error was reported.
Was severe nausea observed?
Yes.
Does completing the study imply no important side effects?
No.
Does an acute rigidity measure establish durable benefit?
No.
What must a response rate identify?
Whether its denominator is people, administrations, or another unit.