Randomised controlled trials that evaluated non-surgical interventions for non-pregnant women with irregular or abnormal cyclic uterine bleeding for three months or more were eligible for inclusion. Studies of women with polycystic ovarian syndrome were included with studies of irregular bleeding as long as the patient baseline and outcome data included information regarding cycle regularity. Studies of women with abnormal uterine bleeding caused by coagulation defects, systemic disease, structural abnormalities, cancer, medication side-effects or infertility (with a primary treatment goal of conception) were excluded. Outcomes of interest included bleeding profile, quality of life, pain, sexual function and patient satisfaction with treatment outcomes. For evidence of harms, observational studies with at least 1,600 participants/records were included.
Interventions evaluated by the included trials included levonorgestrel-releasing intrauterine system (LNG-IUS) combined oral contraceptive pills, nonsteroidal anti-inflammatory drugs (NSAID), progestogens, lifestyle/behavioural changes, acupuncture and decision aids. Most studies were head-to-head comparisons; a smaller proportion were placebo-controlled. Studies were performed in North America, Europe (including one UK study for irregular uterine bleeding and two UK studies for abnormal cyclic uterine bleeding), South America, Africa and Asia. The duration of most studies was six months or less.
Two reviewers independently screened studies for inclusion at the full text stage; any discrepancies were resolved by discussion with senior investigators.