Outcome of IVF pregnancies following severe OHSS.

Because severe ovarian hyperstimulation syndrome (OHSS) is a potentially life-threatening iatrogenic complication, much effort is made to prevent it and the anticipated pregnancy naturally becomes of secondary importance. There are many publications on OHSS, but very few on pregnancy outcomes. This work is to review the effect of OHSS on pregnancy outcome along the pregnancy course. Hospitalized patients with severe OHSS are exposed to several insults that could affect pregnancy outcome in its early stages: the ovarian hyperstimulation for IVF itself, haemodynamic instability that involve haemoconcentration, hypoxia, liver and renal dysfunction, and exposure to high endogenous oestrogens, cytokines, renin, angiotensin and prostaglandins. There is a paucity of data on the relation of OHSS and pregnancy complications. The incidence of multiple pregnancies, gestational diabetes mellitus, placental abruption prematurity and low birthweight is higher in cases of pregnancy complicated by severe OHSS. Therefore, these pregnancies should be considered as high-risk pregnancies, and followed/treated as such. As prevention is the best 'treatment' for OHSS, this may imply the need for more patient-friendly or mild stimulation protocols.

Can pregnancy rate be improved in gonadotropin-releasing hormone (GnRH) antagonist cycles by administering GnRH agonist before oocyte retrieval? A prospective, randomized study.

OBJECTIVE: To examine whether the addition of one dose of preovulatory GnRH agonist could improve implantation and pregnancy rates in GnRH antagonist IVF cycles. DESIGN: Prospective, randomized clinical trial. SETTING: University-affiliated IVF and infertility unit. PATIENT(S): Two hundred twenty-one patients intended for GnRH antagonist protocol IVF. INTERVENTION(S): Patients were prospectively randomized to two groups. The control group received hCG (5,000 U) 34 hours before oocyte aspiration, and the study group received triptorelin (0.2 mg SC) in addition to hCG. All other treatment parameters were identical. MAIN OUTCOME MEASURE(S): Oocyte pick-up day serum levels of E(2), P, LH, and FSH and implantation and pregnancy rates per started cycle and per completed cycle. RESULT(S): A total of 200 ET cycles were carried out: 97 in the study group and 103 in the control group. None of the cycle parameters of the study or control groups differed, excepting mean oocyte pick-up day FSH (11.26 IU/L [95% confidence interval (CI) 9.88-12.52] vs. 6.27 IU/L [95% CI 5.76-8.77]) and LH levels (5.19 IU/L [95% CI 4.47-5.9] vs. 3.28 IU/L [95% CI 2.22-4.18]). The implantation rate was 19.9% (52 of 261) for the study group and 13.9% (35 of 251) for the control group. The pregnancy rate in completed cycles and the ongoing pregnancy rate per ET were significantly higher in the study group than in the control group: 29.1% (30 of 103) in the control group and 44.3% (43 of 97) in the study group, and 22.3% (23 of 103) and 36.1% (35 of 97), respectively. However, the improvement in pregnancy rate per started cycle did not reach statistical significance (40.9% vs. 28.3%). CONCLUSIONS: The administration of triptorelin (0.2 mg) at the time of hCG administration in GnRH antagonist IVF cycles significantly improved overall and ongoing pregnancy rates in completed cycles but not in all started cycles. It is possible that this was achieved owing to an endometrial GnRH receptor effect, which should also be examined by direct endometrial studies.

A randomized clinical trial comparing recombinant hyaluronan/recombinant albumin versus human tubal fluid for cleavage stage embryo transfer in patients with multiple IVF-embryo transfer failure.

BACKGROUND: We aimed to examine the efficacy of using an embryo transfer medium enriched with hyaluronan (HA) to improve implantation in a selected group of patients aged <43 years with repeated (>4) implantation failures after IVF-embryo transfer. METHODS: About 101 patients, meeting our selection criteria, were randomly allocated to undergo embryo transfer either using our routine embryo transfer medium without HA (control group) or a HA enriched commercial embryo transfer medium (study group). The primary outcome was clinical pregnancy rate. RESULTS: After a similar treatment protocol, the ovarian hormonal response, the mean number of ova retrieved and injected per patient, fertilization and cleavage rates and mean embryo quality were comparable between the study and control groups. Although a similar number of embryos was transferred in both groups (3.1 +/- 0.7 versus 2.9 +/- 0.6, mean +/- SD), a significantly higher implantation rate (16.3% versus 4.8%, P = 0.002) and clinical pregnancy rate (35.2% versus 10.0%, P = 0.004) and delivered or ongoing pregnancy rate (31.3% versus 4.0%, P = 0.0005) were observed in the study group. When mean implantation rate per patient was calculated, the difference between the study (0.148 +/- 0.23) and control (0.04 +/- 0.13) group was significant (P = 0.003). CONCLUSIONS: In this selected group of patients after multiple IVF-embryo transfer failures, the use of HA enriched embryo transfer medium is beneficial.

Prospective, randomized trial of metformin and vitamins for the reduction of plasma homocysteine in insulin-resistant polycystic ovary syndrome.

One hundred and two women with insulin-resistant polycystic ovary syndrome were randomized to treatment with a vitamin B preparation, metformin, or both, in conjunction with standard infertility treatment. Plasma homocysteine levels were significantly reduced by both B vitamins and metformin, but to a greater degree by B vitamins, and higher pregnancy rates were associated with vitamin B treatment.

Favorable influence of local injury to the endometrium in intracytoplasmic sperm injection patients with high-order implantation failure.

Local injury to the endometrium prior to controlled ovarian stimulation may considerably improve implantation rates and pregnancy outcomes in intracytoplasmic sperm injection patients with high-order implantation failure (> or =4 IVF trials and > or =12 transferred embryos).

Ultrasonographic-guided percutaneous transabdominal puncture for oocyte retrieval in a rare patient with Rokitansky syndrome in an in vitro fertilization surrogacy program.

OBJECTIVE: To describe the option of ultrasonographic-guided percutaneous transabdominal puncture for ovum pickup in a rare case of Rokitansky syndrome with subcostal located ovaries and a right single kidney. DESIGN: Technique and instrumentation. SETTING: A university-based hospital IVF and Reproductive Medicine Unit. PATIENT(S): A woman with congenital agenesis of the uterus in a surrogate gestation program. INTERVENTION(S): Ultrasonographic-guided percutaneous transabdominal puncture for oocyte retrieval. MAIN OUTCOME MEASURE(S): Number of oocytes retrieved and fertilized. RESULT(S): Oocyte retrieval was feasible in four treatment cycles with subsequent embryo transfer to the surrogate mother. CONCLUSION(S): Ultrasonographic-guided percutaneous transabdominal oocyte collection is a feasible and efficacious procedure for IVF in a case of Rokitansky syndrome with elevated subcostal ovaries and a single kidney.

In vivo maturation of oocytes by extending the interval between human chorionic gonadotropin administration and oocyte retrieval.

OBJECTIVE: To find whether the percentage of mature oocytes can be increased in the following ICSI cycle, in a selected group of patients with >or=47% immature oocytes, with the prolongation of hCG-to-oocyte retrieval interval. DESIGN: Randomized study. SETTING: In vitro fertilization unit in a medical center in Israel. PATIENT(S): From January 2003 to June 2005, in 72 of 2,650 intracytoplasmic sperm injection cycles (2.8%), >or=47% of the total aspirated oocytes were immature (group A). In the following treatment cycle, with similar controlled ovarian hyperstimulation, hCG was planned to be injected in the same interval (group B) as well as 3-4 hours earlier (group C). In fact, the hCG-to-oocyte retrieval interval in group C was statistically significantly longer (38.6 +/- 1.2 hours) compared with that in group B (35.3 +/- 0.7 hours). INTERVENTION(S): The clinical characteristics of the patients, ovarian stimulation, hormonal profile, and the intracytoplasmic sperm injection outcome of the various groups were assessed. MAIN OUTCOME MEASURE(S): The number and maturity of the aspirated oocytes. RESULT(S): The mean number of ampules needed for stimulation, treatment duration, serum E(2), and P level on the hCG day were similar in the various groups. Escaped ovulation did not occur in any of the patients. A larger number of aspirated oocytes, a statistically significantly larger percentage of mature oocytes (72%), and a statistically significantly lower percentage of metaphase I (12%), germinal vesicle (18%), and degenerated oocytes (8%) were found in group C compared with group A (50%; 16%, 21%, and 13%, respectively). Similar morphological distribution of the aspirated oocytes was observed among groups B and C. Clinical implantation and pregnancy rates were higher in group C compared with group B. CONCLUSION(S): In a selected group of patients with >or=47% immature oocytes per retrieval, a larger cohort of mature oocytes can be recruited by extending the hCG- to-oocyte retrieval interval in a subsequent cycle. This improved outcome directly affects the quality of embryos available for transfer and thus increases pregnancy rate.

Luteal phase characteristics following GnRH antagonist or agonist treatment - a comparative study.

Due to inherent differences between gonadotrophin-releasing hormone (GnRH) antagonists and agonists, their late effect on ovarian steroidal production during the luteal phase of IVF cycles may differ. The aim of this study was to characterize and compare the luteal phase hormonal profile after the use of GnRH antagonists or agonists in ovarian stimulation protocols for IVF, in non-conception cycles, to avoid the effect of human chorionic gonadotrophin (HCG) during the luteal phase in conception cycles. Seventy-eight normo-ovulatory patients <35 years old, undergoing IVF due to male or tubal infertility were randomly allocated either to a GnRH antagonist (study group) or GnRH agonist treatment (control group). Similar standard luteal support was given to all patients, using vaginal micronized progesterone. In non-conception cycles, no statistically significant differences were found comparing luteal phase. oestradiol or progesterone levels in the study and control groups. No statistically significant differences were found comparing the hormonal profile dynamics, the mid-luteal (HCG day +8) oestradiol/progesterone ratio and the percentage of mid-luteal oestradiol decline between the study and control groups. In conclusion, similar characteristics and dynamics of luteal phase oestradiol and progesterone were demonstrated comparing ovarian stimulation for IVF using GnRH agonist or antagonists, under similar luteal support.

Oocyte maturity and preimplantation development in relation to follicle diameter in gonadotropin-releasing hormone agonist or antagonist treatments.

OBJECTIVE: To analyze the distribution of oocytes nuclear maturational stages and intracytoplasmic sperm injection (ICSI) outcome in relation to follicular size, after controlled ovarian hyperstimulation (COH), using gonadotropin-releasing hormone (GnRH) antagonist or agonist protocols. DESIGN: Prospective comparative study. SETTING: Fertility unit in an academic hospital. PATIENT(S): Normo-ovulatory patients with unexplained, male and/or tubal factor infertility undergoing ICSI treatment. INTERVENTION(S): Oocytes were retrieved from small (7-12 mm) and large follicles (>12 mm). MAIN OUTCOME MEASURE(S): Oocyte nuclear maturation stage, fertilization and pre-implantation embryonic development of the mature ova evaluated. RESULT(S): Distribution of oocyte maturation stages from small follicles did not differ between the two COH protocols. From large follicles, a statistically significantly higher percentage of immature oocytes was retrieved after antagonist compared with agonist treatment (16 vs 11%, respectively). No statistically significant differences existed regarding ICSI outcome of mature ova within the same follicle diameter range between COH protocols. Inseminated ova from small follicles presented a significantly higher number of embryos with delayed development on day 3 after ICSI. CONCLUSION(S): A greater heterogeneity of maturity was found in the population of oocyte stages that were present after the antagonist protocol. Matured oocytes retrieved from small follicles generated embryos of lower developmental potential than oocytes derived from larger follicles.

Eight years' experience with an IVF surrogate gestational pregnancy programme.

The aim of this study was to retrospectively audit eight years' experience of an IVF surrogate gestational programme and to compare the outcome of surrogacy due to absence of the uterus with surrogacy indicated for repeated IVF failure and recurrent abortions. A total of 60 cycles of IVF surrogate pregnancy were initiated in 19 treated couples. Absence of the uterus was the indication for surrogacy in 10 cases: Rokitansky syndrome (eight cases) and post-hysterectomy (two cases) designated as group A. The indications in the remaining nine patients (group B) were: IVF implantation failure (three cases), habitual abortions (four cases) and deteriorating maternal diseases (two cases). IVF performance and subsequent pregnancy outcome of groups A and B were compared. There was no difference in ovarian stimulation parameters and in IVF performance between the groups A and B. The overall pregnancy rate per transfer was 10/60 (17%). The pregnancy rates per patient and per transfer were 7/10 (70%) and 7/35 (20%) in group A compared with 3/9 (33%) and 3/25 (12%) in group B. A median number of three treatment cycles were needed to achieve pregnancy. In conclusion, the existence or absence of the uterus in the commissioning mothers is irrelevant for their IVF performance and conception rates. In patients who conceived after more than three IVF cycles, an additional 'oocyte factor' might be present.

Diagnosis of conjoined twins before 16 weeks' gestation: the 4-year experience of one medical center.

BACKGROUND: Descriptions of the rarely occurring condition of conjoined twins are sparse. METHODS: We report a case series of four conjoined twin pregnancies diagnosed up to 16 weeks of gestation. One was in a quadruplet in vitro fertilization intracytoplasmic sperm injection (IVF-ICSI) pregnancy of cryopreserved embryos and the others were spontaneous conceptions. Three were thoracopagus and had increased nuchal translucency (NT) or big nuchal cysts and one was omphalopagus. RESULTS: All the three affected twin pregnancies were terminated by cervical dilatation and uterine evacuation. In another case, selective fetal reduction, shortly followed by embryo reduction to singleton was performed. Histological analysis of one case revealed dissimilar gender. No further cytogenetic or molecular studies were conducted by parental request. The pathophysiology of increased NT, subcutaneous edema or nuchal cysts in thoracopagus conjoined twins, the possible embryonic events that may lead to dizygotic conjoined twins and the clinical management of conjoined twins in sets of multiple gestation are discussed. CONCLUSIONS: The current case series illustrates that there are many problems related to this rare, but intriguing, accident of nature that still need to be overcome.

A comparison of the incidence, presentation, and management of ovarian pregnancies between two periods of time.

STUDY OBJECTIVE: To compare the prevalence, presentation diagnostic modalities, and management of 37 ovarian pregnancies in one institution during two time periods. DESIGN: A retrospective study (Canadian Task Force classification II-2). SETTING: Department of Gynecology in a university-affiliated hospital. PATIENTS: Twenty patients with ovarian pregnancy between 1971-1989 compared with 17 patients with ovarian pregnancy between 1990-2000. MEASUREMENTS AND MAIN RESULTS: Seventeen ovarian pregnancies, diagnosed between 1990-2000, comprised 2.6% (17/634) of all ectopic pregnancies, 1:3500 of all live births leading to a mean ovarian pregnancy rate per year of 1.5, as opposed to 3.1% (20/647) of all ectopic pregnancies, 1:3600 of all live births, and a mean ovarian pregnancy rate per year of 1.1 in the previous period. Circulatory collapse was present in 4/17 (23%) patients in the 1990-2000 time period and in 6/20 (30%) in the previous period. Culdocentesis for diagnostic purposes was practically abandoned in the recent decade. Wedge resection by laparotomy was the treatment of choice in the past, and after 1997, it was performed exclusively by laparoscopy. When an ovarian pregnancy was diagnosed, an intrauterine device (IUD) was present in 65% of all patients and in 73% of women who had previously conceived in the 1999-2000 time period, compared with higher rates in the previous period (90% and 100%, respectively). CONCLUSION: The absolute number of ovarian pregnancies in the last 11 years increased when compared with the previous 19 years; however, the prevalence rate per delivery was stable. Patients still experienced circulatory collapse despite modern diagnostic modalities. Culdocentesis for diagnostic purposes has become an unnecessary procedure. Laparoscopic wedge resection was the treatment of choice. The relationship between IUD use and ovarian pregnancies was still strong.

Vasa previa: the impact of prenatal diagnosis on outcomes.

OBJECTIVE: To evaluate outcomes and predictors of neonatal survival in pregnancies complicated by vasa previa and to compare outcomes in prenatally diagnosed cases of vasa previa with those not diagnosed prenatally. METHODS: We performed a multicenter study of 155 pregnancies complicated by vasa previa. Cases were obtained from the Vasa Previa Foundation and 6 large hospitals. Comparisons were made between groups based on prenatal diagnosis status and neonatal survival. RESULTS: The overall perinatal mortality was 36% (55 of 155). In 61 cases (39%), vasa previa was diagnosed prenatally; 59 of 61 (97%) infants from these pregnancies survived compared with 41 of 94 (44%) in cases not diagnosed prenatally (P <.001). Median 1- and 5-minute Apgar scores in cases diagnosed prenatally were 8 and 9, respectively, compared with 1 and 4 among survivors in cases not diagnosed prenatally (P <.001). More than half (24 of 41) of surviving neonates born to women without prenatal diagnosis required blood transfusions compared with 2 of 59 diagnosed prenatally (P <.001). Multivariable logistic regression analysis showed that the only significant predictors of neonatal survival were prenatal diagnosis (P <.001) and gestational age at delivery (P =.01). CONCLUSIONS: Good outcomes with vasa previa depend primarily on prenatal diagnosis and cesarean delivery at 35 weeks of gestation or earlier should rupture of membranes, labor, or significant bleeding occur.

Ovarian pregnancy-a 12-year experience of 19 cases in one institution.

OBJECTIVES: To report the prevalence, presentation, diagnostic modalities, and treatment of ovarian pregnancy in one institution. STUDY DESIGN: Retrospective case control study of 19 cases of ovarian pregnancy treated between 1990 and 2001 at Assaf Harofeh Medical Center, Zerifin, Israel. MAIN OUTCOME MEASURES: Prevalence, presentation, diagnostic modalities, surgical treatment, and relation to intrauterine device (IUD) use. RESULTS: Nineteen ovarian pregnancies, diagnosed between 1990 and 2001, comprised (19/694) 2.7% of all ectopic pregnancies, 1:3000 of all live births leading to a mean ovarian pregnancy per year of 1.6. Presenting symptoms were similar to those of tubal pregnancies including circulatory collapse which was present in 4/19 (21%) of patients. Culdocentesis for diagnostic purposes, has become an unnecessary procedure. Wedge resection by laparotomy was the treatment of choice in the past, and from 1994, it was performed exclusively by laparoscopy. When an ovarian pregnancy was diagnosed, intrauterine device was present in 68% of the patients and in 76% of the fertile women. CONCLUSIONS: The absolute number of ovarian pregnancies between 1900 and 2001 increased but the prevalence rate per delivery was stable. Despite modern diagnostic modalities patients still present in circulatory collapse-conservative approach may underestimate the potential risk of bleeding. Culdocentesis has no clinical diagnostic benefits. Laparoscopy is invaluable, as diagnosis and treatment can be carried out as a single treatment. Laparoscopic wedge resection is the treatment of choice. The relation between IUD use and ovarian pregnancies is still strong.

Birth of healthy twins resulting from donated oocytes and posthumous use of frozen-thawed spermatozoa obtained prior to chemotherapy.

The improvement in life expectancy, following better cancer therapy combined with new options in treating male infertility have increased the use of sperm freezing. We describe a rare case of twin pregnancy of a childless widow, using donated oocytes after ICSI with her deceased husband's banked frozen spermatozoa. Sperm was frozen before chemotherapy treatment and a written consent for future thawing and injection to donor oocytes was given. Use of thawed spermatozoa after death was not mentioned in the consent form. During the husband's illness years, ICSI of donated oocytes fertilized with thawed spermatozoa yielded three embryos, but no pregnancy was achieved after ET. While waiting for a subsequent oocyte donation, the male partner died of his malignancy. The woman expressed her wish to be treated with the frozen spermatozoa of her deceased husband. Since the original informed consent did not cover postmortem use of the husband's sperm, a special application for its use was made and subsequently approved by the Israeli Ministry of Health legal advisor. Another six donated oocytes were fertilized with the deceased's thawed spermatozoa, and the transfer of four embryos to the uterine cavity resulted in a twin pregnancy and later delivery of two healthy babies. This case emphasizes the importance of written informed consent signed by patients with a life-threatening disease at the time of banking spermatozoa, and discussing the possibility of its posthumous use. Several legal and ethical concerns are discussed.

Insulin resistance in patients with polycystic ovary syndrome is associated with elevated plasma homocysteine.

BACKGROUND: Elevated levels of plasma homocysteine have recently been implicated as a significant risk factor for cardiovascular disease, pre-eclampsia, and recurrent pregnancy loss, and have been found to be associated with insulin resistance in a number of clinical situations. We examined the relationship between plasma homocysteine and insulin resistance in patients with polycystic ovary syndrome (PCOS). METHODS: A total of 155 infertile patients with PCOS as defined by clinical, biochemical and ultrasound criteria were screened for insulin resistance utilizing single-sample fasting insulin and glucose measurement, calculated by glucose:insulin ratio or homeostasis model assessment (HOMA) index. Total plasma homocysteine was measured by fluorescence polarization immunoassay. One hundred normo-ovulatory women with normal ovaries being treated for other infertility diagnoses served as a control group. RESULTS: Insulin resistance was found in the majority of PCOS patients: -53.5% (83/155), 60.6% (94/155) and 65.8% (102/155), when defined by fasting insulin, glucose:insulin ratio, or logHOMA respectively. Mean plasma homocysteine in the PCOS group was significantly higher than in the normal ovary group (11.5 +/- 7.4 versus 7.4 +/- 2.1 micromol/l, P < 0.001). Insulin-resistant PCOS patients had significantly higher plasma homocysteine (12.4 +/- 8.4 micromol/l) than non-insulin-resistant PCOS patients (9.6 +/- 4.4 micromol/l) regardless of body mass index (P = 0.003 by groups, P = 0.005 by correlation of single samples). Thirty-four per cent (53/155) of the PCO patients had homocysteine values >95th percentile of the controls (11.0 micromol/l, P < 0.0001). Statistically significant correlations were found between all insulin resistance indices and homocysteine levels. Multiple logistic regression defined insulin resistance as the major factor examined that influenced homocysteine levels. CONCLUSIONS: Insulin resistance and hyperinsulinaemia in patients with PCOS is associated with elevated plasma homocysteine, regardless of body weight. This finding may have important implications in the short term regarding reproductive performance, and in the long term regarding cardiovascular complications associated with insulin-resistant PCOS.

Increased frequency of female partner chromosomal abnormalities in patients with high-order implantation failure after in vitro fertilization.

OBJECTIVE: To find the type and frequency of chromosomal abnormalities in a selected group of high-order implantation failure (> or =6 IVF trials and > or =15 transferred embryos) and to evaluate its impact on pregnancy outcome. DESIGN: A retrospective study. SETTING: In vitro fertilization (IVF) unit in a university affiliated hospital. PATIENT(S): Sixty-five couples with high-order implantation failure in IVF and embryo transfer. INTERVENTION(S): In vitro fertilization/embryo transfer (ET), work-up for implantation failure, cytogenetic analysis of the couple. MAIN OUTCOME MEASURE(S): We studied the type and frequency of chromosomal changes, quality of embryos, cumulative pregnancy rates, and pregnancy outcome. RESULT(S): The mean number of treatment cycles per patient, before karyotyping was 7.8 +/- 2.4 (range: 6 to 16 cycles). The mean cumulative number of all transferred embryos per patient was 25.7 +/- 10.3 (range: 9 to 65 embryos). Chromosomal abnormalities were found in 10 of 65 (15.4%) cases: translocations in six, mosaicism in two, and inversion or deletion in another two. The morphologic characteristics of the transferred embryos and the cumulative pregnancy rates were similar in patients with implantation failure with and without chromosomal changes. Three of the 16 patients with abnormal karyotype delivered and three miscarried within a follow-up period of 1 year. CONCLUSION(S): A high frequency of chromosomal aberrations was found in a selected group of high-order implantation failures, a similar frequency to recurrent miscarriages. Karyotyping is recommended as part of the work-up for repeated implantation failure in assisted reproduction. Treatment options include further IVF trials, preimplantation genetic diagnosis, or oocyte donation, tailored according to the type of chromosomal change. An international registry should be considered to assist in counseling these patients.

Increased early pregnancy loss in IVF patients with severe ovarian hyperstimulation syndrome.

BACKGROUND: Since severe ovarian hyperstimulation syndrome (OHSS) is a potentially life-threatening complication of assisted reproduction, the focus of attention in such cases is placed firmly upon the health of the patient, with the endeavour to achieve a pregnancy being considered of secondary importance. The aim of this study was to focus on the pregnancy rate and pregnancy outcome in IVF patients hospitalized for severe or critical OHSS, in one centre, during a period of 6 years. METHODS: We compared the characteristics of patients with severe OHSS: those who conceived with the ones who did not conceive, and among pregnant IVF patients, those with ongoing pregnancies with those that miscarried. RESULTS: Pregnancy was achieved in 60 of 104 (58%) patients with severe OHSS. Pregnancy continued until delivery in 37 of these 60 patients (62%), whereas the remaining 23 (38%) aborted. The pregnancy and abortion rates in patients with severe OHSS were significantly higher than those of IVF patients without OHSS, during the same time period [23% (1138/4922) and 15% (169/1138) respectively, P < 0.001]. The mean duration of hospitalization for OHSS was significantly shorter in those who delivered compared with those who aborted (5.9 +/- 3.2 versus 10.5 +/- 9.6 days, P < 0.01) and in the non-pregnant patients compared with the pregnant patients (5.2 +/- 3.2 versus 7.6 +/- 6.6 days, P < 0.02). CONCLUSIONS: The clinical pregnancy rate of IVF patients with severe OHSS was significantly higher than that of patients without the syndrome. A longer stay in hospital-reflecting a more severe form of OHSS-was correlated with a higher frequency of abortions. OHSS, necessitating hospitalization, is a detrimental clinical situation not only for the mother but also for the developing pregnancy.