THE FLUID-DYNAMICS OF INJECTION - VARIABLES AS THEY RELATE TO TRANSVAGINAL GAMETE INTRAFALLOPIAN TRANSFER AND TUBAL EMBRYO-TRANSFER

Transvaginal gamete intra-Fallopian transfer (GIFT) and tubal embryo t ransfer have not, in most cases, produced pregnancy rates as high as t he equivalent transabdominal methods. We postulated that two parameter s, i.e. wide dispersion of the gametes and peritoneal spill of gametes or embryos, are of prime importance in explaining this difference. We designed this study to analyse the effect of varying the rate of inje ction (mu l/min) of radio-opaque contrast media and the distance the t ransfer catheter is placed into the tubal isthmus upon these two param eters. Selective salpingography was performed on 30 patients who were allocated to one of two experiments. In the first experiment (flow rat e), 20 patients were divided into four groups so that we could analyse four sets of 10 Fallopian tubes. Each group was subjected to injectio n of contrast at different flow rates: 100, 50, 25 or 12.5 mu l/min. P eritoneal spill occurred from 4, 2, 0 and 0 tubes and wide dispersion (> 4 cm) of contrast occurred in 6, 5, 1 and 0 tubes at the respective injection rates of 100, 50, 25 and 12.5 mu l/min. At 12.5 mu l/min co ntrast was localized to < 2 cm of tubal ampulla in all cases. In the s econd experiment (catheter placement), the remaining 10 patients were divided to allow analysis of two groups of 10 Fallopian tubes. The tra nsfer catheter was placed either 1.5 cm or 5 cm into the tube and the radio-opaque contrast was injected at 12.5 mu l/min. No difference in the degree of dispersion of contrast along the Fallopian tube was obse rved. No backflow of contrast into the uterine cavity was observed in any patient following removal of the catheter. No evidence of signific ant endometrial trauma was seen at subsequent hysteroscopy. These resu lts indicate that injection flow rate is likely to be a major factor a ffecting the prospect of pregnancy following transvaginal GIFT or tuba l embryo transfer, and there is a need to repeat studies of these tech niques at rates of injection much slower than previously thought appro priate.