{"paper_id":"b1704119-309e-4e65-9f54-096a8d22af65","body_text":"Controlled ovarian hyperstimulation (COH) in\nconjunction with intrauterine inseminations (IUI)\nare commonly used to treat infertile couples ( 1 ).\nThe most important indications for IUI are male\nsubfertility, unexplained infertility, ovulatory dysfunction\nand cervical factor infertility ( 2 ). Several\nprognostic factors that determine IUI treatment\noutcome have been identified and include the\nwoman’s age, duration of infertility, follicle number,\nendometrial thickness, numbers of sperm inseminated,\nsperm morphology, progressive motile\nsperm count, and cause of infertility ( 3 - 5 ). Tomlinson\net al. ( 6 ) found no differences in age, duration\nof infertility, number of follicles, body mass index\n(BMI) and sperm quality in the pregnancy rates of patients undergoing IUI. Although IUI with or\nwithout ovarian stimulation is widely used, its success\nrate is lower than that of the assisted reproductive\ntechnologies (ART) ( 7 ). Nevertheless, in\ncomparison to ART controlled ovarian hyperstimulation\ncombined with intrauterine (COH+IUI)\nrequires less frequent clinic visits, and is simple,\nrelatively less invasive and comparatively inexpensive\n( 8 ). Regardless of the method of assisted\nconception utilized, the couples always desire to\nknow their chances of success. Thus, identifying\nthe factors which are influential in the success rate\nis highly crucial. The purpose of this study was to\nidentify the parameters which were most influential\nin the success rate using COH+IUI treatment\nmodality. Using logistic regression, we were able\nto devise a mathematical model to predict the success\nrate in COH+IUI. The data presented, will\nenable the healthcare providers to counsel their\npatients about their chances of getting pregnant by\nusing COH+IUI.\n\nIn this cross-sectional study, we included 994\nIUI cycles in 803 couples who referred to the infertility\nInstitute between 2010-2012. This study\nwas approved by the Institutional Review Board\nof the Royan Institute Research Center and the\nRoyan Ethics Committee according to the Helsinki\nDeclaration, signed informed written consent\nwas obtained from all participants. All couples had\nattempted to conceive for at least one year prior\nto undergoing COH+IUIs. The women completed\nthe self-administered questionnaire which was\nused to collect data about demographic, menstrual\nand obstetrical characteristics. A menstrual interval\nshorter than 21 days and longer than 35 days\nis defined as menstrual irregularities. Amount of\nbleeding is varied ( 9 ).\nThe study population comprised of all couples\nwho were candidates for COH+IUI and had diagnoses\nof subfertile male infertility, polycystic ovary\nsyndrome (PCOS), mild or minimal endometriosis\nor unexplained infertility and various ovulatory\ndisorders. Ovulatory disorders included diminished\novarian reserve, PCOS and hypothalamic amenorrhea.\nSubfertile male infertility was defined as per\ncriteria outlined by Molinaro et al. ( 10 ).\nThe following evaluations were performed prior\nto the initiation of COH+IUI. The women underwent\ncycle day 3 hormone evaluation [folliclestimulating\nhormone (FSH), luteinizing hormone\n(LH), estradiol (E 2 ), thyroid-stimulating hormone\n(TSH) and prolactin] and assessment of tubal patency\nby using hysterosalpingogram (HSG) and/\nor laparoscopy. Tubal patency of at least one tube\nwas mandatory. In case of either a tubal abnormality\nin HSG or dysmenorrhea and dyspareunia, a\nlaparoscopy was performed.\nInclusion criteria were: male factor, combined\ncauses, Ovulatory disorder (Pco, diminished ovarian\nreserve, and hypothalamic amenorrhea), unexplained\ndisorder, and all patients with normal TSH\nand prolactin levels. The couples with testicular\natrophy, hydrosalpinx, anatomical abnormalities,\ninfection, uterine fibroids, and systemic disease\nwere excluded from participation.\nAll IUI cycles were performed with ovarian\nstimulation and included either clomiphene citrate\n(Iran Hormone Pharmaceutical Company,\nIran), gonadotropin only, or letrozol (Femara,\nNovartis Pharma AG, Switzerland) or the combination\nof either clomiphene citrate or letrozol\nwith gonadotropin. On days 11-12 of the menstrual\ncycle, we assessed follicular development\nand endometrial thickness by transvaginal\nultrasound. If the endometrial thickness was <7\nmm, 4 mg/per day E 2  was administered (2 mg,\nAburaihan Co., Iran) and continued during the\nluteal phase. Once a leading follicle of ≥18 mm\nwas identified, human chorionic gonadotropin\n(5,000 IU IM, Pregnyl ® , Darou Pakhsh Pharmaceutical,\nIran) was administored to induce\nthe final stage of oocyte maturation and a single\nIUI was planned 36-38 hours later. If more than\nfive follicles ≥18 mm in size developed, the cycle\nwas cancelled.\nSemen samples were obtained from patients\nwho attended the unit for infertility treatments.\nSemen samples (n=994) were collected following\n3-7 days of sexual abstinence. They were allowed\nto liquefy at (add room temperature or 37˚C) for\n15-30 minutes and each was subject to an analysis\naccording to the 2010 World health Organization\n(WHO) guidelines ( 11 ). After analysis, samples\nwere prepared for IUI using to discontinuous density\ngradient centrifugation (DGC). For this purpose,\nwe prepared a two layer gradient consisted of solutions of 100 and 50% Allgrade ®  (LifeGobal,\nBelgium). The 50% concentration was made by\ndiluting 100% Allgrade ®  with Ham’s F10 (Sigma,\nUSA) medium. The density gradients were performed\nby layering 2 mL of each concentration\ninto a conical tube (15 mL, Falcon, Becton Dickinson,\nNJ, USA). These tubes were pre-incubated for\nat least 2 hours in a 37˚C incubator. After semen\nliquefaction, 2 mL of ejaculation was layered on\nthe top of the Allgrade ®  gradient and centrifuged\nfor 30 minutes at 300 x g. After centrifugation, the\nsperm was collected at the bottom of the tube by\na clean Pasteur pipette and transferred to a 5 mL\nclean tube (Falcon), and washed twice with Ham’s\nF10 medium by using centrifugation at 300 x g for\n5 minutes. The pellet was resuspended in 1 mL of\nHam’s F10 medium, then the sperm concentration\nand motility were evaluated. In our study the sperm\nanalysis data haven’t been recorded after processing\nand sperm analysis data before processing was\navailable.\nIntrauterine inseminations was performed\nby a soft catheter (INDOVASIVE, Biorad, India)\nwith an insemination volume of 0.6 mL.\nThe IUI catheter was passed gently through the\ncervical canal until the tip passed the internal\nos. Then, the sperm suspension was deposited\nslowly through the uterine cavity. All patients\nwere provided with luteal support by using cyclogest\naccording to the treatment physician’s\npreference. Clinical pregnancy was defined as\na positive pregnancy test followed by the presence\nof a gestational sac visualized by transvaginal\nsonography 4weeks after IUI.\nIn order to build a prediction model, we used\nstepwise logistic regression analysis, in which a\nP value of 0.15 was used as an entry criterion,\nwhereas a P value of 0.10 was the threshold for\na variable to stay in the model. We check the\nperformance of the model by the area under the\nreceiver operating characteristic (ROC) curve\n(AUC). An AUC of 0.5 indicates no discriminative\nperformance, whereas an AUC of 1.0 indicates\nperfect discrimination.\nCalibration of the model was assessed by comparing\nthe predicted probability of pregnancy in a\ncategory of patients and the observed percentage\nof pregnant woman in that category. We first categorized\nthe predicted probabilities of pregnancy\nin 10 groups, then we compared the mean predicted\nprobability of pregnancy in that particular category\nwith the observed probability, i.e. pregnancy\nrate in that category.\nAll statistical analyses were done by using\nSPSS software (version 20, USA). Chi-square\nand t tests were used for analyses. We performed\nunivariate logistic regression for each\nfactor and reported the odds ratio (OR) and 95%\nconfidence interval (CI). In order to predict the\nIUI result, we used multiple logistic regression\nanalyses. Data were expressed as mean ± standard\ndeviation (SD). A Pvalue of <0.05 was considered\nto be statistically significant.\n\nWe studied a total of 994 IUI cycles in 803 couples.\nEach couple underwent 1.23 ± 0.4 (mean ±\nSD) COH+IUI cycles (range: 1-3). Causes for infertility\nwere: unexplained disorder (290, 29.2%),\nmale factor (395, 39.7%), combined causes (108,\n10.9%), and ovulatory disorder (201, 20.2%). In\nour study population, ovulatory disorders included\ndiminished ovarian reserve, 0.5% (n=1), PCOS,\n93.5% (n=188) and hypothalamic amenorrhea, 6%\n(n=12).\nIn our study combined cause including; ovulatory\ndisorder and male factor 83.3% (n=90), tub\nperitoneal and male factor 6.5% (n=7), uterine factor\nand ovulatory disorder 3.7% (n=4), uterine factor\nand male factor 2.8% (n=3), male factor and\nrecurrent abortion 1.9% (n=2), ovulatory disorder\nand recurrent abortion 0.9% (n=1), uterine factor\n& recurrent abortion 0.9% (n=1).\nTable 1  compares the demographic characteristics\nbetween pregnant and nonpregnant women.\nThe pregnancy rate in younger women was significantly\nhigher than those of older women. In\naddition, an infertility duration of ≤4years was\nassociated with a significantly higher pregnancy\nrate (OR:1.5, CI:1.1-2.2, P=0.01). Infertility type\n(primary or secondary) did not significantly affect\nthe outcome. With regards to the diagnosis\nof infertility, the highest pregnancy rate (27.8%)\nwas achieved in couples with combined infertility,\nwhereas the lowest (13.4%) rate was observed in\ncouples who suffered from male factor infertility\n(P<0.001,  Table 1 ).\nCharacteristics of study patients who underwent IUI\nIUI; Intrauterine insemination, OR; Odds ratio, CI; Confidence interval, BMI; Body mass index,†; Values are mean ± SD, §; Reference category\nand *; P<0.05 was considered as statistically significant.\nThe pregnancy rates according to female characteristics\nand sperm parameters (according to Strict\nCriteria) are summarized in  table 2 . Pregnancy\nrate was not related to sperm count. There were no\nsignificant differences in total sperm concentration\namong the pregnant and nonpregnant study population.\nSperm parameters did not significantly affect\nthe outcome of COH+IUI treatment. Seminal\nvolume did not significantly affect the success of\nCOH+IUI. The total dose of gonadotropin in nonpregnant\nwomen was significantly lower than that\nof the pregnant women (P=0.03,  Table 2 ).\nNo significant difference was found between\nthe two groups in different types of gonadotropins\n(data not shown).\nPregnancy outcome, in our study included, there\nwere 3 (1.8%) ectopic pregnancies, 9 spontaneous\nmiscarriages of which 5 (3%) occurred during the\nfirst trimester and 4 (2.4%) during the second trimester;\n8 (4.8%) cases of blighted ovum, and 145\n(87.9%) live births. This corresponded to an ongoing\npregnancy rate of 14.9% (149/994) per IUI\ncycle. Of the 165 clinical pregnancies (ongoing\npregnancies and early pregnancy loss), 22 were\ntwin pregnancies (13.3%). There were 7 (4.2%)\ntriplet pregnancies, of which one ended with a late\nabortion and another terminated at 24 weeks from\nwhich no fetus survived. Of the remaining triplet\npregnancies, two mothers gave birth at 32 and 34\nweeks (two healthy sets of one girl and two boys)\nand 3 triplet pregnancies were reduced to twins.\nThe mean birth weight was 1488 ± 395 g and mean\ngestational age at delivery was estimated to be 32\nweeks for the triplet pregnancies that reduced to\ntwins. Of these, all neonates were well and healthy.\nThe mean birth weight of singletons was 3000 ±\n525.7 g, twins weighed 2081 ± 557.4 g and triplets\nweighed 1588.3 ± 549.1 g. All singletons ended at\n28-40 weeks and twins at 30-38 weeks.\nWe used linear-by-linear test for calculation, the\ncorrelation between age of women and clinical,\nongoing and multiple pregnancies rate. When one\nof the variables is ordinal and the other variable is\nordinal or nominal with 2 level, this trend test can\nbe used ( 12 ).\nThe proportion of clinical, ongoing and multiple\npregnancies, decreased with age (P=0.041,\nP=0.044 and P=0.046, respectively,  Table 3 ).\nCycle parameters of the patients who underwent IUI\nIUI; Intrauterine insemination, OR; Odds ratio, CI; Confidence interval, FSH; Follicle-stimulating hormone , LH; Luteinizing hormone , †; Values are mean ± SD, §; Reference category,*; Significant statistical differences between the two groups.\nClinical and ongoing pregnancy rates per couple and the frequency of multiple pregnancies for women according to age group\n*; Significant statistical differences between the groups.\nOf the 145 (87.8%) live births, 139 resulted in\nlive deliveries at term, 123 (88.5%) patients underwent\ncaesarean sections and 16 (11.5%) had\nnormal vaginal deliveries. There were no major\ncongenital anomalies reported. The live birth rate/\ncycle was 14.5% (145/994).\nThe clinical pregnancy rate per couple was\n20.5% (165/803) with an ongoing pregnancy rate\nper couple of 18.5% (149/803). Pregnancy rates\nper cycle were as follows: first (21%), second\n(19.4%) and third (15.3%).\nStepwise multiple linear regression analysis was\nperformed to compare the association between\ndependent (total dose of gonadotropin) and independent\n(age, BMI, menstrual irregularities, duration\nof infertility, type of infertility, endometrial\nthickness, number of dominant follicle, etiology\nof infertility) variables. Age (P<0.001), menstrual\nirregularities (P<0.001), and duration of infertility\n(P=0.01) were the main variables that significantly\ninfluenced the total dose of gonadotropin in couples\nundergoing IUI ( Table 4 ).\nVariables influencing total dose of gonadotropin in couples undergoing IUI\n*; P value multiple regression, IUI; Intrauterine insemination and\nSE; Standard error.\nAccording to logistic regression, female age, duration\nof infertility, menstrual irregularities, seminal\nvolume and total dose of gonadotropin were significantly\nassociated with pregnancy outcome. Higher\nfemale age, prolonged duration of infertility and regular\nmenstruation showed a negative association with\npregnancy outcome, while seminal volume and total\ndose of gonadotropin were positively associated with\npregnancy outcome ( Table 5 ).\nResult of logistic regression analysis\nOR; Odds ratio, CI; Confidence interval, §; Reference category, *;\nSignificant statistical differences and AUC; Area under curve.\nThe ROC curve was used to assess the discriminative\nperformance of the fitted logistic model ( Fig .1 ).\nAn AUC equal to 0.5 indicates no discriminative\npower whereas an AUC of 1.0 shows a perfect discrimination.\nIn our study, the AUC for the fitted logistic\nmodel was found to be 0.65 with the 95% CI of\n0.60 - 0.70, indicative of a reasonable prognostic potency\nfor predicting pregnancy following COH+IUI.\nWith the data obtained, we were able to construct\na formula for calculation of the probability\nof pregnancy following COH+IUI as carried out in\nour study (see below).\nIn this formula, the duration of infertility is\nthe number year each couple has been attempting\nto conceive without success. The menstrual\nhistory is=1 if history of menstrual is irregular\nand 0 if history of menstrual is regular. The performance\nof the prediction model for pregnancy\nfollowing COH+IUI was calibrated as shown in\n figure 2 . The predictive performance appears to\nbe acceptable because the 95% confidence intervals\nof the observed pregnancy rates overlap\nwith the predicted pregnancy rate.\nProbability of pregnancy after IUI = e ( -2.154 - 0.147 × Durationofinfertility + 0.865 × Menstrualhistory + 0.105 × Volume + 0.022 × ( Total dose of gonadotropin ) ) 1 + e ( -2.154 - 0.147 × Durationofinfertility + 0.865 × Menstrualhistory + 0.105 × Volume + 0.022 × ( Total dose of gonadotropin ) )\nROC curve for assessment of logistic regression discrimimative\npreformance.\nROC; Receiver operating characteristic.\nCalibration plot, showing the relationship between predicted\nand observed rate of pregnancy after intrauterine insemination\n(IUI).\n\nAmong the various parameters that were studied,\nfemale age, duration of infertility, menstrual\nirregularities, semen volume, cause of infertility\nand the dose of gonadotropin significantly affected\ntreatment success.\nIn the current study, we have shown a statistically\nsignificant association between reduced\nCOH+IUI success rate and increased age. Several\nstudies have illustrated the decline in pregnancy\nwith advancing age, ( 13 - 15 ); however, Erdem et\nal. ( 15 ) did not find female age to be a prognostic\nfactor in the prediction of a live birth in ovarian\nstimulation and IUI cycles. The current study, IUI\nwas offered for women over the age of 40 years.\nAccording to studies, woman over the age of 40\nare not good candidates for IUI ( 16 ,  17 ).\nWe observed a significant decrease in pregnancy\nrate with increased duration of infertility\n(OR: 0.8, CI: 0.8-0.9, P<0.001). This result was\nalso supported by the results observed in a study\nby Kamath et al. ( 18 ). In another study ( 17 ), a significantly\nhigher pregnancy rate (14.2%) was observed\nin couples with the duration of infertility of\nless than 6 years compared to 6.1% rate for those\nwith the duration of more than 6 years. Merviel et\nal. ( 14 ) did not observe this difference. However,\nour findings indicate that the duration of infertility\nmust be considered when counselted patients on\ntheir chances of a successful pregnancy.\nInfertility type (primary or secondary) did not\nsignificantly affect the outcome of COH+IUI the\nresult of which has been shown in some studies\n( 14 ,  15 ). Our study found a significant effect of the\ntotal gonadotropin dose on outcome in pregnancies\nconceived by COH+IUI (P=0.03). The data were\nalso evaluated to determine the variables which\nmay influence the total dose of gonadotropins. It\nappears that women with higher age, those with\nmenstrual irregularities and low number of dominant\nfollicles as well as those with shorter duration\nof infertility should not be given high doses of\ngonadotropins. I argue with such a strong statement\nbased on the findings presented in  table 4 . The data\npresented are simple correlation data. The only\ntime you can make such a statement is when for\nexample you give the same doses of gonadotropins\nto women of younger and older age and assess the\noutcome.\nA total of 17.5% of the recorded clinical pregnancies\nafter COH+IUI at our center were multiple\npregnancies; no case of hyperstimulation was\ndocumented during the study period. Other studies\nhave reported an incidence of twins (20%) and\nhigher-order (39%) multiple pregnancies that were\nthe result of ovulation induction ( 19 ,  20 ). Thus,\ncenters should choose appropriate stimulation protocols and attempt to achieve a balance between\nthe search for advanced success rate and suitable\nmultiple pregnancy rates. The present study, no\nsignificant difference was found in endometrial\nthickness between pregnant women and those who\ndid not become pregnant. This finding is similar to\nthe result of Kamath et al. ( 18 ).\nThe information available at present study indicates\nthat COH+IUI can be considered prior to\nmore expensive IVF in patients that have combined\n(27.8% per cycle) and ovulatory disorder\n(20.9% per cycle) infertilities. The success rate\nwas higher for ovulatory cases and for those who\nsuffered from more than one etiological factor. The\npatients of this group have been diagnosed as the\ncombination of mild male infertility and PCOS.\nCompared result has been reported regarding the\nhighest success rate in an ovulatory patient ( 13 ,  15 ,\n 17 ). The clinical pregnancy rate was significantly\nhigher in patients who had irregular menstruation.\nAll of these patients were diagnosed with PCOS,\naccording to the cause of infertility as discussed,\nthe success rate was higher in an ovulatory patient.\nWhen the effect of the infertility etiology was assessed,\nthere was a significantly lower pregnancy\nrate observed in endometriosis patients compared\nwith women who had unexplained infertility ( 16 ).\nPeterson et al. ( 21 ) have found the average pregnancy\nrate for unexplained infertility to be 18%.\nOur result showed a 13.8% average pregnancy rate\nfor unexplained infertility; however, Basirat and\nEsmaeilzadeh ( 22 ) determined that the etiology of\ninfertility was not significantly different between\npregnant and nonpregnant women (P=0.63).\nPredictive sperm parameters for successful IUI\nhave been controversial ( 23 ,  24 ). Total motile\ncount (TMC) is a potential predictive factor for\na successful COH+IUI ( 15 ). The pregnancy rates\naccording to Kamath et al. ( 18 ) were as follows:\na significantly higher pregnancy rate (18.2%) was\nobserved when TMC was in the range of 10-20\nmillion. TMC at a range of 5-10 million resulted\nin a 5.6% pregnancy rate, whereas in cases where\nTMC was <5 million, the rate was 2.7%. A TMC\nof <1 million was associated with poor pregnancy\nrates. When the TMC was <5 million, sperm\nmorphology appeared to play an important role. A\npregnancy rate of 18.4% was observed with a normal\nmorphology compared to a rate of <5.4% with\n<30% morphology ( 18 ). In our study, although the\nsperm analysis data was not recorded after processing,\nthe data prior to sperm processing was\navailable. Sperm parameters did not significantly\naffect the success of COH+IUI. These results also\nconfirm the findings achieved by other researchers\n( 16 ,  25 ). In contrast, other studies have described\nseveral semen parameters that correlated with IUI\noutcome, such as the number of motile sperm ( 24 ,\n 26 ) and normal morphology ( 24 ). In the current\nstudy there was a lower pregnancy rate when the\nTMC was in the range of 10-20 million (9.5%).\nThe pregnancy rate was 17% when the TMC was\n>20 million. The pregnancy rate increased when\nthere was higher normal morphology (OR=1.02,\n95% CI:0.7-1.4,  Table 2 ). Generally, published\nstudies have been inconsistent related to the association\nbetween the morphology readings and the\nsuccess in IUI ( 21 ,  22 ).\nWe observed four parameters that significantly\naffected success: duration of infertility, menstrual\nirregularities, seminal volume and total dose\nof gonadotropin. In a previous study there were\nfour prognostic factors: etiology and duration of\ninfertility, number of treatment cycles, and number\nof pre ovulatory follicles ( 16 ). Kamath et al.\n( 18 ). found a significant effect of the duration of\ninfertility and TMC on outcome of pregnancies\nconceived by IUI.\nOur study was a retrospective, no documents\nwere available about the number of follicles and\nthere was a limitation to our study. On the other\nhand, we did not have information about semen\nquality after processing and it was another ristriction.\n\nWith the data obtained, we were able to construct\na formula for to calculate the probability\nof pregnancy following COH+IUI as carried out\nin the present study. These results suggested that\nfemale age, duration of infertility, cause of infertility,\nmenstrual irregularities, ejaculatory volume\nand total dose of gonadotropin be the most important\nprognostic factors in predicting successful\noutcome of IUI. A larger study population might\nassist with the formulation of a better predictive\nmodel for IUI success. Such information could be\nused by couples and clinicians during counseling\nparticipants to arrive at a decision with regards to\ntheir treatment options.","source_license":"CC-BY-4.0","license_restricted":false}