BACKGROUND: A possible therapeutic effect of diagnostic tubal patency testing has been debated in the literature for half a century. Further debate surrounds whether oil-soluble or water-soluble contrast media might have the bigger fertility-enhancing effect. Historically a variety of agents have been used to 'flush' the fallopian tubes, although tubal flushing does not currently form part of routine practice in the treatment of fertility delay. OBJECTIVES: To evaluate the effect of flushing a woman's fallopian tubes with oil- or water-soluble contrast media on subsequent pregnancy outcomes in couples with infertility. SEARCH STRATEGY: The search strategy of the Menstrual Disorders and Subfertility Group and a recent search of electronic databases completed in December 2001 was used for the identification of relevant randomised controlled trials. SELECTION CRITERIA: All randomised trials where tubal flushing with oil-soluble contrast media or tubal flushing with water-soluble media was compared with one another or with no treatment were considered for inclusion in the review. DATA COLLECTION AND ANALYSIS: Eight randomised controlled trials were identified and included in this review. A further one randomised controlled trial is ongoing. All trials were assessed for quality criteria. The studied outcomes were pregnancy, live birth (and ongoing pregnancy), miscarriage, ectopic pregnancy, treatment complications including pain, intravasation of contrast medium, infection and haemorrhage, and image quality. MAIN RESULTS: Tubal flushing with oil-soluble media versus no intervention was associated with a significant increase in the odds of pregnancy (OR 3.57, 95%CI 1.76-7.23). There were no data from RCTs to assess tubal flushing with water-soluble media versus no intervention. Tubal flushing with oil-soluble media was associated with a significant increase in the odds of live birth versus tubal flushing with water-soluble media (OR 1.49, 95%CI 1.05-2.11) but the odds of pregnancy showed no significant difference (OR 1.23, 95%CI 0.95-1.60) and there was evidence of statistical heterogeneity for these two outcomes. The addition of oil-soluble media to flushing with water-soluble media (water-soluble plus oil-soluble media versus water-soluble media alone) showed no significant difference in the odds of pregnancy (OR 1.16, 95%CI 0.78-1.70) or live birth (OR 1.06, 95%CI 0.64-1.77). REVIEWER'S CONCLUSIONS: There is some evidence of effectiveness of tubal flushing with oil-soluble contrast media in increasing the odds of pregnancy versus no intervention. The limited evidence of an increase in the odds of live birth from tubal flushing with oil-soluble contrast media versus water-soluble contrast media must be interpreted cautiously. Further robust randomised trials, comparing oil-soluble versus water-soluble media and comparing each versus no intervention, are required to provide convincing evidence as to whether the technique should be accepted into widespread clinical practice.
