Background: Varicocele repair is a widely used treatment for subfertility. Our aim was to identify and combine the results from randomised controlled trials published to ascertain whether the pregnancy rates after varicocele repair are higher than those with no treatment.
Methods: We did a systematic review of seven studies identified by searching Medline and a register of controlled trials. We also searched the contents of specialist journals and the annual meeting programmes of relevant societies by hand. Inclusion criteria were treatment of varicocele in subfertile couples, random allocation to treatment and control groups, and pregnancy or livebirth rates as an outcome measure. We pooled data by use of fixed and random effects models.
Findings: None of seven eligible studies published between 1979 and 2002 described a strategy for concealment of the allocation sequence. There were 61 pregnancies among 281 treated couples and 50 pregnancies among 259 controls. The overall relative benefit of treatment was 1.01 (95% CI 0.73-1.40) by the fixed effects model and 1.04 (0.62-1.75) by the random effects model. The overall risk difference was 0.2% (-7 to 7) and 3% (-7 to 14), respectively. In subgroup analyses, varicocele treatment was not effective in trials restricted to male subfertility with clinical varicocele, or in those that included men with subclinical varicocele or normal semen analysis. However, this systematic review, done with a meta-analytical method, might have had insufficient power to detect small effects because of the small number of patients in some subgroups.
Interpretation: Varicocele repair does not seem to be an effective treatment for male or unexplained subfertility.