Serum estradiol level and oocyte number in predicting severe ovarian hyperstimulation syndrome

Ovarian hyperstimulation syndrome (OHSS) is a relatively common and potentially life-threatening complication of ovarian stimulation, the pathogenesis of which remains unclear. To clarify the predictive values of serum estradiol levels and oocyte number in severe OHSS, and to investigate the impact of high serum estradiol levels on pregnancy outcome, we retrospectively analyzed clinical data from 431 cycles of ovarian stimulation for assisted reproduction performed from 1993 through 1995. Receiver operating characteristic plots were used to estimate the predictive power of the measured variables. The overall frequency of severe OHSS was 5.5%. Using a serum estradiol level of 3,600 pg/mL as the minimum cut-off value, the sensitivity was 58%, with a specificity of 92%, a positive predictive value of 29%, and a negative predictive value of 97%. The predictive power was similar when a cut-off point of 20 oocytes retrieved was used. The two criteria together gave a sensitivity of 33%, a specificity of 92%, a positive predictive value of 40%, and a negative predictive value of 98%. One of seven oocyte donors developed severe OHSS. The pregnancy rate was higher in patients with severe OHSS than in patients who did not develop this syndrome (73.9% vs 32.5%) but the pregnancy outcomes were not significantly different. We conclude that elevated estradiol concentrations and oocyte number appear to be helpful in predicting severe OHSS, but neither parameter by itself is predictive. This syndrome is rare in the absence of luteal hCG support, either exogenous or pregnancy-derived; when it occurs, there are usually extremely high preovulatory estradiol concentrations and numerous oocytes retrieved. High serum estradiol levels are unlikely to have adverse effects on pregnancy outcome in patients with severe OHSS.