Premium
Sperm count as a surrogate endpoint for male fertility control
Author(s) -
Benda Norbert,
Gerlinger Christoph
Publication year - 2007
Publication title -
statistics in medicine
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.996
H-Index - 183
eISSN - 1097-0258
pISSN - 0277-6715
DOI - 10.1002/sim.2967
Subject(s) - fertility , sperm , surrogate endpoint , pregnancy , azoospermia , clinical endpoint , medicine , gynecology , pregnancy rate , obstetrics , clinical trial , infertility , andrology , population , biology , environmental health , genetics
Abstract When assessing the effectiveness of a hormonal method of fertility control in men, the classical approach used for the assessment of hormonal contraceptives in women, by estimating the pregnancy rate or using a life‐table analysis for the time to pregnancy, is difficult to apply in a clinical development program. The main reasons are the dissociation of the treated unit, i.e. the man, and the observed unit, i.e. his female partner, the high variability in the frequency of male intercourse, the logistical cost and ethical concerns related to the monitoring of the trial. A reasonable surrogate endpoint of the definite endpoint time to pregnancy is sperm count. In addition to the avoidance of the mentioned problems, trials that compare different treatments are possible with reasonable sample sizes, and study duration can be shorter. However, current products do not suppress sperm production to 100 per cent in all men and the sperm count is only observed with measurement error. Complete azoospermia might not be necessary in order to achieve anacceptable failure rate compared with other forms of male fertility control. Therefore, the use of sperm count as a surrogate endpoint must rely on the results of a previous trial in which both the definitive‐ and surrogate‐endpoint results were assessed. The paper discusses different estimation functions of the mean pregnancy rate (corresponding to the cumulative hazard) that are based on the results of sperm count trial and a previous trial in which both sperm count and time to pregnancy were assessed, as well as the underlying assumptions. Sample size estimations are given for pregnancy rate estimation with a given precision. Copyright © 2007 John Wiley & Sons, Ltd.