CX-ToM: Counterfactual explanations with theory-of-mind for enhancing human trust in image recognition models.

We propose CX-ToM, short for counterfactual explanations with theory-of-mind, a new explainable AI (XAI) framework for explaining decisions made by a deep convolutional neural network (CNN). In contrast to the current methods in XAI that generate explanations as a single shot response, we pose explanation as an iterative communication process, i.e., dialogue between the machine and human user. More concretely, our CX-ToM framework generates a sequence of explanations in a dialogue by mediating the differences between the minds of the machine and human user. To do this, we use Theory of Mind (ToM) which helps us in explicitly modeling the human's intention, the machine's mind as inferred by the human, as well as human's mind as inferred by the machine. Moreover, most state-of-the-art XAI frameworks provide attention (or heat map) based explanations. In our work, we show that these attention-based explanations are not sufficient for increasing human trust in the underlying CNN model. In CX-ToM, we instead use counterfactual explanations called fault-lines which we define as follows: given an input image I for which a CNN classification model M predicts class c pred , a fault-line identifies the minimal semantic-level features (e.g., stripes on zebra), referred to as explainable concepts, that need to be added to or deleted from I to alter the classification category of I by M to another specified class c alt . Extensive experiments verify our hypotheses, demonstrating that our CX-ToM significantly outperforms the state-of-the-art XAI models.

Detecting clinically related content in online patient posts.

Patients with chronic health conditions use online health communities to seek support and information to help manage their condition. For clinically related topics, patients can benefit from getting opinions from clinical experts, and many are concerned about misinformation and biased information being spread online. However, a large volume of community posts makes it challenging for moderators and clinical experts, if there are any, to provide necessary information. Automatically identifying forum posts that need validated clinical resources can help online health communities efficiently manage content exchange. This automation can also assist patients in need of clinical expertise by getting proper help. We present our results on testing text classification models that efficiently and accurately identify community posts containing clinical topics. We annotated 1817 posts comprised of 4966 sentences of an existing online diabetes community. We found that our classifier performed the best (F-measure: 0.83, Precision: 0.79, Recall:0.86) when using Naïve Bayes algorithm, unigrams, bigrams, trigrams, and MetaMap Symantic Types. Training took 5 s. The classification process took a fraction of 1 s. We applied our classifier to another online diabetes community, and the results were: F-measure: 0.63, Precision: 0.57, Recall: 0.71. Our results show our model is feasible to scale to other forums on identifying posts containing clinical topic with common errors properly addressed.

Live birth and cumulative live birth rates in expected poor ovarian responders defined by the Bologna criteria following IVF/ICSI treatment.

OBJECTIVE: To determine the live birth and cumulative live birth rates of expected poor ovarian responders according to the Bologna criteria and to compare their outcomes with those of expected normal responders. DESIGN: Retrospective analysis. SETTING: University infertility clinic. PATIENTS: A total of 1,152 subfertile women undergoing their first in vitro fertilization (IVF) cycle. INTERVENTIONS: Women were classified into 4 groups according to the Bologna criteria for comparison. MAIN OUTCOME MEASURE(S): Live birth and cumulative live birth rates. RESULTS: Women with expected poor response (POR) had the lowest live birth rate than the other 3 groups (23.8%, p = 0.031). Cumulative live birth rates were significantly lower in those with expected POR than those with expected normal ovarian response (NOR) (35.8% vs 62.8%, p<0.0001). In the subgroup analysis, the cumulative live birth rates in expected PORs were significantly lower in those who had ≤3 oocytes retrieved (18.6% for ≤3 oocytes vs 44.0% for >3 oocytes, p = 0.006) whereas the live birth rates in fresh cycle did not differ (17.8% vs 30.9%, p = 0.108). CONCLUSION: Women who were expected POR according to the Bologna criteria had lower live birth and cumulative live birth than expected NOR but they still can achieve reasonable treatment outcomes and IVF treatment should not be precluded.

The effect of endometrial injury on ongoing pregnancy rate in unselected subfertile women undergoing in vitro fertilization: a randomized controlled trial.

STUDY QUESTION: Does endometrial injury in the cycle preceding ovarian stimulation for in vitro fertilization (IVF) improve the ongoing pregnancy rate in unselected subfertile women? SUMMARY ANSWER: Endometrial injury induced by endometrial aspiration in the preceding cycle does not improve the ongoing pregnancy rate in unselected subfertile women undergoing IVF. WHAT IS KNOWN ALREADY: Implantation failure remains one of the major limiting factors for IVF success. Mechanical endometrial injury in the cycle preceding ovarian stimulation of IVF treatment has been shown to improve implantation and pregnancy rates in women with repeated implantation failures. There is limited data on unselected subfertile women, especially those undergoing their first IVF treatment. STUDY DESIGN, SIZE, DURATION: This randomized controlled trial recruited 300 unselected subfertile women scheduled for IVF/ICSI treatment between March 2011 and August 2013. Subjects were randomized into endometrial aspiration (EA) (n = 150) and non-EA (n = 150) groups according to a computer-generated randomization list. PARTICIPANTS/MATERIALS, SETTING, METHODS: Subjects were recruited and randomized in the assisted reproductive unit at the University of Hong Kong. In the preceding cycle, women in the EA group underwent endometrial aspiration using a Pipelle catheter in mid-luteal phase. All women were treated with a cycle of IVF/ICSI. Pregnancy outcomes were compared. MAIN RESULTS AND THE ROLE OF CHANCE: There were no significant differences in baseline or cycle characteristics between the groups. There were 209 subjects (69.7%) who were undergoing their first IVF cycle and 91 (30.3%) subjects who had repeated cycles. There was no significant difference in ongoing pregnancy rates [26.7% (40/150) versus 32.0% (48/150); RR 0.833 (95% CI 0.585-1.187), P = 0.375] in the EA and non-EA groups. The implantation rates [32.8% (67/204) versus 29.7% (68/229); RR 1.080 (95% CI 0.804-1.450), P = 0.120], clinical pregnancy rates [34.0% (51/150) versus 38.0 (57/150); RR 0.895 (95% CI 0.661-1.211), P = 0.548], miscarriage rates [30.3% (17/56) versus 18.6% (11/59), RR 1.628 (95% CI 0.838-3.164), P = 0.150] and multiple pregnancy rates [31.3% (16/51) versus 19.3% (11/57), RR 1.626 (95% CI 0.833-3.172), P = 0.154] were all comparable between the EA and non-EA groups. Subgroup analysis in women having first embryo transfer (n = 209) also demonstrated no significant difference in ongoing pregnancy rates, but for women undergoing repeated cycles (n = 91), the on-going pregnancy rate was significantly lower in the EA group than in the non-EA group. LIMITATIONS, REASONS FOR CAUTION: The study aimed at assessing an unselected population of subfertile women by recruiting consecutive women attending our fertility clinic. However, since the majority of the recruited women (69.7%) were having their first IVF treatments, the results may not be generalizable to all women undergoing IVF. WIDER IMPLICATIONS OF THE FINDINGS: Previous RCTs and meta-analyses have suggested improved pregnancy rates after pretreatment endometrial injury in women with repeated implantation failure. A recent RCT also showed increased pregnancy rates in unselected subfertile women after endometrial injury, although that study was terminated early and thus underpowered. Our study showed with adequate power that no significant improvement in pregnancy rates was observed after endometrial injury in unselected women undergoing IVF treatment. STUDY FUNDING/COMPETING INTERESTS: The study was supported by the Small Project Funding 201309176012 of the Committee on Research and Conference Grants, University of Hong Kong. The authors have nothing to disclose. TRIAL REGISTRATION NUMBER: HKCTR-1646 and NCT 01977976.

A highly-sensitive anti-Müllerian hormone assay improves analysis of ovarian function following chemotherapy for early breast cancer.

AIM: Anti-Müllerian hormone (AMH) shows promise as a biomarker of the ovarian reserve but current assays are insufficiently sensitive to allow assessment of this post-chemotherapy in most women. We have assessed a new highly sensitive AMH assay (Ansh picoAMH) in the evaluation of ovarian activity in women with very low ovarian reserve after chemotherapy. METHODS: A prospective cohort and an independent validation cohort of premenopausal women with early breast cancer (eBC) were recruited at the time of diagnosis (combined n=98), and ovarian reserve markers 2-5 years later following chemotherapy were assessed in relation to menstrual activity. RESULTS: The picoAMH assay had a limit of detection of 7.5 pg/ml. AMH clearly distinguished women with ongoing menses from those with amenorrhoea at 2 years after diagnosis (mean 522 ± 169 versus 8.9 ± 1.3 pg/ml, P<0.0001) with high predictive value for continuing menses or amenorrhoea for the subsequent 3 years. AMH was detectable in more women than using a previous assay (P=0.004). Other markers of the ovarian reserve (follicle-stimulating hormone (FSH), inhibin B) were also of discriminatory value but to lesser extents. This finding was validated in a second, independent cohort of women treated for eBC. CONCLUSION: The 10-fold increased assay sensitivity showed very clear distinction between groups based on ovarian activity with excellent prediction of future menses or amenorrhoea. This will improve assessment of post-chemotherapy ovarian function in women and may aid treatment decisions.

A randomized, controlled, pilot trial on the effect of dehydroepiandrosterone on ovarian response markers, ovarian response, and in vitro fertilization outcomes in poor responders.

OBJECTIVE: To evaluate whether pretreatment dehydroepiandrosterone (DHEA) supplementation improves ovarian response markers, ovarian response to standard low-dose gonadotropin stimulation, and in vitro fertilization (IVF) outcomes in poor responders. DESIGN: Randomized, double-blind, placebo-controlled pilot study. SETTING: Tertiary reproductive medicine unit. PATIENT(S): Thirty-two women with anticipated poor ovarian response. INTERVENTION(S): Randomization into DHEA group (n=16) receiving GNC (25 mg three times a day) or placebo (n=16) starting from at least 12 weeks before the scheduled IVF treatment according to a computer-generated randomization list. MAIN OUTCOME MEASURE(S): Measurement of monthly ovarian response markers, including antral follicle count (AFC), serum antimüllerian hormone (AMH), and follicle-stimulating hormone (FSH) levels; comparison of ovarian response to a standard dose of gonadotropin stimulation at week 8 and IVF outcomes; and AFC after 12 weeks (primary outcome). RESULT(S): The DHEA supplementation resulted in statistically significantly higher serum DHEA-S, free androgen index, and follicular DHEA-S levels. No statistically significant differences in the ovarian response markers (AFC, AMH, or FSH), the ovarian response to standard-dose gonadotropin stimulation, or IVF outcomes were found between the two groups. CONCLUSION(S): No statistically significant improvement in ovarian response markers, ovarian response to standard dose gonadotropin stimulation, or IVF outcomes was found in poor responders receiving pretreatment DHEA. CLINICAL TRIAL REGISTRATION NUMBER: HKCTR-1149 (www.hkclinicaltrials.com) and NCT01915186 (www.ClinicalTrials.org).

Effect of preovulatory progesterone elevation and duration of progesterone elevation on the pregnancy rate of frozen-thawed embryo transfer in natural cycles.

OBJECTIVE: To assess the incidence of P elevation (PE) in natural cycles and evaluate its effect on frozen-thawed embryo transfer cycles performed in natural cycles (FET-NC). STUDY DESIGN: Retrospective analysis. SETTING: A tertiary assisted reproductive unit. PATIENT(S): Subfertile woman who did not conceive in their stimulated IVF cycle and underwent the first FET-NC cycle. INTERVENTION(S): Achieved serum samples were assayed for P concentrations from the day of LH surge up to 3 days before the surge. The cutoff level of PE was defined as 5 nmol/L. MAIN OUTCOME MEASURE(S): Clinical and ongoing pregnancy rates. RESULT(S): The incidence of PE in natural cycles was 173 of 610 (28.4%). There were no significant differences in both clinical and ongoing pregnancy rates (39.0% vs. 37.3% and 32.5% vs. 31.7%) between those with vs. without PE on the day of LH surge. If PE lasted for 2 days or more, there was a significant reduction in the clinical pregnancy rate (39.4% vs. 20.7%). Using multivariate logistic regression, women's age, PE for 2 days or more, and the number of top-quality embryos were the significant factors for clinical pregnancy rates in FET-NC. CONCLUSION(S): The incidence of PE in FET-NC was similar to that in stimulated cycles. Progesterone elevation for 2 days or more before the LH surge impaired the clinical pregnancy rate of FET-NC, whereas PE on the day of LH surge only did not have such an adverse effect.

Live birth rate, multiple pregnancy rate, and obstetric outcomes of elective single and double embryo transfers: Hong Kong experience.

OBJECTIVE: To compare the live birth rate, multiple pregnancy rate, and obstetric outcomes of elective single and double embryo transfers. DESIGN: Case series with internal comparisons. SETTING: University affiliated hospital, Hong Kong. PARTICIPANTS: Between October 2009 and December 2011, 206 women underwent their first in-vitro fertilisation cycle. Elective single embryo transfer was offered to women who were aged 35 years or below, and had endometrial thickness of 8 mm or more and at least two embryos of good quality. MAIN OUTCOME MEASURES: Live birth rate, multiple birth rate, and obstetric outcomes. RESULTS: Among the 206 eligible women, 74 underwent an elective single embryo transfer and 132 a double embryo transfer. The live birth rate was comparable in the two groups, being 39.2% in the elective single embryo transfer group and 43.2% in the double embryo transfer group, while the multiple pregnancy rate was significantly lower in the elective single embryo transfer group than the double embryo transfer group (6.9% vs 40.4%; P<0.001). Gestational ages and birth weights were comparable in the two groups. There was no significant difference between the two groups with respect to the rate of preterm delivery and antenatal complications (27.6% vs 43.9%, respectively; P>0.05). CONCLUSION: In this selected population, an elective single embryo transfer policy decreases the multiple pregnancy rate without compromising the live birth rate. The non-significant difference in antenatal complications may be related to the small sample size.

Live birth rates following in vitro fertilization in women with thyroid autoimmunity and/or subclinical hypothyroidism.

OBJECTIVE: To investigate whether the live birth rate following in vitro fertilization (IVF) is affected by thyroid autoimmunity (TAI) and/or subclinical hypothyroidism in subfertile women. DESIGN AND SETTING: Retrospective study in a university infertility clinic. PATIENTS: A total of 627 women without past or current history of thyroid disorder undergoing their first IVF cycle. INTERVENTION: Pre-IVF archived blood serum samples were tested for TAI and thyroid function tests. MAIN OUTCOME MEASURE: Live birth rate. RESULTS: The clinical pregnancy rate, live birth rate and miscarriage rate were similar among women with or without TAI and/or subclinical hypothyroidism using a TSH threshold 4·5 mIU/l. Thyroid autoantibody level did not affect these IVF outcomes. CONCLUSION: The live birth rate and miscarriage rate of women with TAI and/or subclinical hypothyroidism following IVF were not impaired.

A pilot study on the combined use of letrozole, mifepristone and misoprostol in termination of first trimester pregnancy up to 9 weeks' gestation.

OBJECTIVE: To assess the feasibility of adding letrozole to the standard regimen of mifepristone and misoprostol for termination of pregnancy up to 63 days. STUDY DESIGN: We recruited 50 subjects who had requested legal termination of pregnancy up to 63 days. Medical abortion was performed with a singe dose of 200 mg mifepristone and 10 mg of letrozole daily for 3 days followed by 800 mcg vaginal misoprostol. RESULTS: The complete abortion rate was 98% (95% CI: 94-100%). The median induction-to-abortion interval of the regimen was 5.1 h (range 1.2-56 h). No serious adverse effects were reported. CONCLUSIONS: The results of this pilot study suggest that a regimen of mifepristone, letrozole and misoprostol is associated with a high complete abortion rate without major adverse events.

Identification of CD147 (basigin) as a mediator of trophoblast functions.

STUDY QUESTION: Does CD147 regulate trophoblast functions in vitro? SUMMARY ANSWER: CD147 exists as a receptor complex on human trophoblast and regulates the implantation, invasion and differentiation of trophoblast. WHAT IS KNOWN ALREADY: CD147 is a membrane protein implicated in a variety of physiological and pathological conditions due to its regulation of cell-cell recognition, cell differentiation and tissue remodeling. Reduced placental CD147 expression is associated with pre-eclampsia, but the mechanism of actions remains unclear. STUDY DESIGN, SIZE, DURATION: A loss of function approach or functional blocking antibody was used to study the function of CD147 in primary human cytotrophoblasts isolated from first trimester termination of pregnancy and/or in the BeWo cell line, which possesses characteristics of human cytotrophoblasts. PARTICIPANTS/MATERIALS, SETTING METHODS: CD147 expression was analyzed by immunofluorescence staining and western blotting. CD147-associated protein complex on plasma membrane were separated by blue native gel electrophoresis and identified by reversed-phase liquid chromatography coupled with quadrupole time-of-flight hybrid mass spectrometer. Cell proliferation and invasion were determined by fluorometric cell proliferation assays and transwell invasion assays, respectively. Matrix metalloproteinases (MMPs) and urokinase plasminogen activator (uPA) activities were measured by gelatin gel zymography and uPA assay kits, respectively. Cell migration was determined by wound-healing assays. Cell fusion was analyzed by immunocytochemistry staining of E-cadherin and 4',6-diamidino-2-phenylindole. The transcripts of matrix proteinases and trophoblast lineage markers were measured by quantitative PCR. Extracellular signal-regulated kinase (ERK) activation was analyzed by western blot using antibodies against ERKs. MAIN RESULTS AND THE ROLE OF CHANCE: CD147 exists as protein complexes on the plasma membrane of primary human cytotrophoblasts and BeWo cells. Several known CD147-interacting partners, including integrin ß1 and monocarboxylate transporter-1, were identified. Suppression of CD147 by siRNA significantly (P < 0.05) reduced trophoblast-endometrial cell interaction, cell invasion, syncytialization, differentiation and ERK activation of BeWo cells. Consistently, anti-CD147 functional blocking antibody suppressed the invasiveness of primary human cytotrophoblasts. The reduced invasiveness was probably due to the restrained (P < 0.05) enzyme activities of MMP-2, MMP-9 and uPA. LIMITATIONS, REASONS FOR CAUTION: Most of the above findings are based on BeWo cell lines. These results need to be confirmed with human first trimester primary cytotrophoblast. WIDER IMPLICATIONS OF THE FINDINGS: This is the first study on the role of CD147 in trophoblast function. Further investigation on the function of CD147 and its associated protein complexes will enhance our understanding on human placentation. STUDY FUNDING/COMPETING INTEREST(S): This work was supported in part by the University of Hong Kong Grant 201011159200. The authors have no competing interests to declare.

Perfluorooctanoate suppresses spheroid attachment on endometrial epithelial cells through peroxisome proliferator-activated receptor alpha and down-regulation of Wnt signaling.

Exposure of animals to perfluorooctanoic acid (PFOA), a surfactant used in emulsion polymerization processes causes early pregnancy loss, delayed growth and development of fetuses. The mechanisms of action are largely unknown. We studied the effect of PFOA on implantation using an in vitro spheroid-endometrial cell co-culture model. PFOA (10-100µM) significantly reduced Jeg-3 spheroid attachment on RL95-2 endometrial cells. PFOA also suppressed ß-catenin expression in Jeg-3 cells. The Wnt agonist Wnt3a stimulated ß-catenin expression in Jeg-3 cells and reversed the PFOA suppression of the spheroid attachment. The putative PFOA receptors (PPARα, ß, γ) present in both cell lines were not affected by PFOA (0.01-100µM). The PPARα antagonist MK886 restored the ß-catenin and E-cadherin expression levels in Jeg-3 cells and reversed the suppression of the spheroid attachment caused by PFOA. Taken together, PFOA suppresses spheroid attachment through PPARα and Wnt signaling pathways via down-regulation of ß-catenin and E-cadherin expression.

A randomized clinical trial comparing the short-term side effects of sublingual and buccal routes of misoprostol administration for medical abortions up to 63 days' gestation.

BACKGROUND: Buccal misoprostol 800 mcg and sublingual misoprostol 800 mcg show high efficacy when used with 200 mg mifepristone for early pregnancy termination but have different side effect profiles. This is the first double-blind randomized trial comparing the side effect profiles of these two routes of administration of misoprostol when used with mifepristone for termination of pregnancies up to 63 days' gestation. STUDY DESIGN: Eligible women (n=90) who requested legal termination of pregnancy up to 63 days' gestation were randomized to two groups and given 200 mg of oral mifepristone followed 48 h later by 800 mcg of either sublingual (n=45) or buccal (n=45) misoprostol. RESULTS: Most of the side effects including fever were more common in the sublingual group, but only the incidence of chills was significantly higher in the sublingual group (55.6% vs 91.1%, p=.0001). Complete abortion occurred in 95.4% [95% confidence interval (CI): 84.9-99.5] of women in the buccal group and 97.8% (95% CI: 88.2-99.9) in the sublingual group. CONCLUSIONS: When combined with mifepristone for termination of pregnancy up to 63 days, both the buccal and sublingual routes are effective routes of administration. The sublingual route tended to be associated with more side effects.

2,3,7,8-Tetrachlorodibenzo-p-dioxin (TCDD) suppresses spheroids attachment on endometrial epithelial cells through the down-regulation of the Wnt-signaling pathway.

The environmental toxicant 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) affects embryo development, implantation and fertility in humans. The underlying molecular mechanism by which TCDD suppresses implantation remains largely unknown. We used the trophoblastic spheroids (embryo surrogate)-endometrial cells co-culture assay to study the attachment of trophoblastic spheroids (BeWo and Jeg-3) onto the endometrial epithelial (RL95-2 and Ishikawa) cells. TCDD dose-dependently induced cytochrome P450 1A1 (Cyp1A1) expression in trophoblastic and endometrial epithelial cells. Moreover, TCDD at 1 and 10 nM suppressed ß-catenin (a Wnt-signaling molecule) and E-cadherin expression, as well as spheroids attachment onto endometrial cells. Interestingly, activation of the canonical Wnt-signaling pathway via Wnt3a or lithium chloride reverted the suppressive effect of TCDD on ß-catenin and E-cadherin expressions in the BeWo and RL95-2 cells, and restored the spheroids attachment rate to be comparable to the untreated controls. Taken together, TCDD induces Cyp1A1 expression, modulates the Wnt-signaling pathway and suppresses spheroids attachment onto endometrial cells.

Ovarian stimulation modulates steroid receptor expression and spheroid attachment in peri-implantation endometria: studies on natural and stimulated cycles.

OBJECTIVE: To compare the effect of high serum E(2) levels on endometrial steroid receptors in gonadotropin-stimulated cycles (hCG + 7) and natural cycles (LH + 7), and to study its effect on spheroid attachment. DESIGN: Observational. SETTING: University hospital. PATIENT(S): Infertile patient with normal menstrual cycles undergoing IVF treatment. INTERVENTION(S): Gonadotropin stimulation and endometrial biopsy; trophoblast spheroid (embryo surrogate, Jeg-3)-endometrial cell (Ishikawa) coculture assay. MAIN OUTCOME MEASURE(S): Steroid receptor expression by quantitative polymerase chain reaction and immunohistochemistry; spheroid attachment rate. RESULT(S): Endometrial biopsies from natural (n = 12) and stimulated (n = 23) cycles were obtained. The expression of estrogen receptor α (ERα) but not ERß or progesterone receptor (PR) transcript was significantly reduced in stimulated cycles compared with natural cycles. Glucocorticoid receptor (GR) transcript was significantly increased in the excessive responders of the stimulated cycle. There was no difference in ERα immunoreactivity in endometrial stroma, but a higher immunoreactivity was seen in endometrial glands of stimulated cycles. The endometrium of stimulated cycles had a lower expression of PR protein in glands, but a higher expression in stroma. Although no GR protein was detected in glands, GR protein expression was significantly up-regulated in stroma of the stimulated cycles. Endometrial cells treated with high steroid concentrations had a reduced spheroid attachment rate compared with the controls. CONCLUSION(S): High serum E(2) level affects the expression of steroid receptors in the endometrial cells and suppresses spheroid attachment.

A pilot study of pre-operative misoprostol in reducing operative blood loss during hysterectomy.

OBJECTIVE: To investigate the effectiveness of a single pre-operative dose of sublingual misoprostol on reducing blood loss in abdominal hysterectomies performed for symptomatic uterine leiomyomas. STUDY DESIGN: A cohort of 64 women undergoing total abdominal hysterectomy for symptomatic uterine leiomyomas were randomly assigned to receive a single dose of sublingual 400 mcg misoprostol (n=32) or placebo containing 20mg vitamin B(6) (n=32) 30 min before the operation. The primary outcome was the operative blood loss. The secondary outcomes were requirement for blood transfusion, change in haemoglobin level after operation, and the incidence of side effects. RESULTS: Women who had misoprostol were found to have similar operative blood loss to those who had placebo (570.9 ± 361.3 ml versus 521.4 ± 297.4 ml, for misoprostol and placebo group respectively; P=0.803). This study with a sample size of 64 was sufficient to have 80% power at the 5% level of significance to detect a reduction of blood loss greater than or equal to 30%. There were no observed differences in the need for post-operative blood transfusion (25% versus 15.6%, for misoprostol and placebo group respectively; P=0.536), the change in haemoglobin level after the operation, and the side effects profiles between the two groups. CONCLUSION: A single pre-operative dose of sublingual misoprostol is not effective in reducing intra-operative blood loss and need for post-operative blood transfusion after total abdominal hysterectomies for symptomatic uterine leiomyomas.

A prospective randomized trial to compare immediate and 24-hour delayed catheter removal following total abdominal hysterectomy.

OBJECTIVE: To assess whether early or immediate removal of a 12F in-dwelling Foley catheter after total abdominal hysterectomy affects the level of subjective pain assessment postoperatively. DESIGN: Randomized controlled trial. SETTING: University Hospital. POPULATION: Seventy women underwent total abdominal hysterectomies for various benign gynecological diseases. METHODS: Women were randomized to have the urinary catheter removed in the operating room after the surgical procedure or to have it removed on postoperative day 1. MAIN OUTCOME MEASURES: The primary outcome was patients' pain assessment and the secondary outcomes were rate of re-catheterization and symptomatic urinary tract infection. RESULTS: There was no difference in the pain assessment between the two groups. A significantly higher number of urinary retention episodes requiring re-catheterization were found in the immediate removal group compared with the delayed removal group (20 vs. 0%; p= 0.011). The incidence of symptomatic urinary tract infection did not differ between the two groups. CONCLUSIONS: There are pros and cons regarding the policy of one-day in-dwelling catheterization compared to immediate catheter removal.

Excessive ovarian stimulation up-regulates the Wnt-signaling molecule DKK1 in human endometrium and may affect implantation: an in vitro co-culture study.

BACKGROUND: High serum estradiol (E2) levels following ovarian stimulation lead to reduced implantation and pregnancy rates, yet the underlying mechanisms remain unknown. We investigated if aberrant expression of genes in the Wnt-signaling pathway may be involved. METHODS: Microarray and real-time PCR analysis were performed to analyze gene expression profiles of endometrial samples taken at day hCG + 7 in stimulated cycles, and days LH + 7 and LH + 10 in natural cycles. Expression of several Wnt-signaling transcripts, including Dickkopf homolog 1 (DKK1), DKK2 and secreted frizzled-related protein 4 (sFRP4), was analyzed throughout the menstrual cycle. JAr spheroid/Ishikawa endometrial cell co-culture experiments were established to study effects of DKK1 on spheroid attachment in vitro. RESULTS: We identified 351 differentially expressed genes. Endometrial samples taken at hCG + 7 had similar expression profiles to those at LH + 10. DKK1 transcripts were up-regulated and DKK2 and sFRP4 were down-regulated in the stimulated compared with LH + 7 group (all P < 0.05). DKK1 transcripts were low in proliferative phase (PS) and increased in late-secretory phase (LS, P < 0.05), although DKK2 peaked in mid-secretory phase (P < 0.05). sFRP4 transcripts were high in PS. Treatment of spheroid with recombinant human DKK-1 protein dose-dependently suppressed (P < 0.05 versus control) spheroids attachment onto endometrial cells (associated with decreased beta-catenin protein): this suppression was nullified by anti-DKK1 antibody. CONCLUSION: Gene expression patterns in stimulated cycles resembled those of LS in natural cycles, when the implantation window is about to close, suggesting high serum E2 and/or progesterone concentrations may advance endometrial development, altering the implantation window and possibly decreasing pregnancy rate. Aberrant expression of DKK1 might impair embryo attachment and implantation in vivo.