Fig. 1.
Standardised mean differences and 95%
confidence intervals for proportion of patients who showed improvement
in overall premenstrual syndrome (progesterone versus placebo).
Negative values indicate reduction in symptoms, favouring active
treatment