Hysterosalpingo-foam sonography versus hysterosalpingography during fertility work-up: an economic evaluation alongside a randomized controlled trial.

Tubal pathology is one of the main underlying causes of infertility and affects between 11% and 30% of all women suffering infertility ( Farquhar et al. , 2019 ; Anyalechi et al. , 2021 ). During the fertility work-up, it is therefore important to assess the patency of the fallopian tubes. Tubal patency can be assessed by a variety of tests ( Saunders et al. , 2011 ). Laparoscopy with chromotubation is historically considered to be the reference test, as it can assess tubal patency and visualize pelvic anatomy at the same time, thus allowing simultaneous treatment ( NICE, 2013 ). However, laparoscopy is an invasive and expensive procedure, it has shown to be less cost-effective compared to hysterosalpingography (HSG) ( Verhoeve et al. , 2013 ), and therefore it is not the preferred test for women at low risk for tubal pathology ( NICE, 2013 ; ASRM, 2020 ). HSG and hysterosalpingo-foam sonography (HyFoSy) are widely used alternatives for tubal patency testing. Both tests have a high and comparable diagnostic accuracy for evaluating tubal patency, compared to laparoscopy with chromotubation with a sensitivity of respectively 0.95 and 0.94 and a specificity of 0.93 and 0.92 ( Broeze et al. , 2011 ; Maheux-Lacroix et al. , 2014 ; Alcázar et al. , 2020 ). During HSG, the uterus and fallopian tubes are made visible on X-rays by infusing iodinated contrast medium into the uterine cavity. HSG has been considered as first-choice tubal patency test for decades ( NICE, 2013 ; ASRM, 2020 ). Although HSG is less invasive than laparoscopy, it causes exposure to ionizing radiation and iodinated contrast medium, and women generally experience HSG as painful ( Dreyer et al. , 2014 ; Engels et al. , 2021 ; Serrano Gonzalez et al. , 2022 ). To avoid these disadvantages of HSG, sonographic tubal patency methods have been introduced, such as HyFoSy. During HyFoSy, an echogenic foam is infused into the uterine cavity to evaluate patency of the fallopian tubes ( Emanuel and Exalto, 2011 ). As HyFoSy is a more patient-friendly test than HSG, it has been argued that HSG should be replaced by sonographic tubal patency testing ( Hamed et al. , 2009 ; Luciano et al. , 2011 ; Lim et al. , 2015 ; Lo Monte et al. , 2015 ). For many years, there were no studies that compared the diagnostic effectiveness of HyFoSy and HSG during fertility work-up. We recently reported the results of the FOAM trial, which compared clinical management based on the results of HyFoSy and HSG during fertility work-up, with ongoing pregnancy leading to live birth as primary outcome. This study showed that management decisions based on the results of HyFoSy resulted in a comparable live birth rate at 12 months compared to HSG (46% versus 47%; difference −1.2%, 95% CI: −3.4% to 1.5%; P  = 0.27) ( van Welie et al. , 2022 ). Since healthcare costs are increasing, cost-effectiveness is becoming ever more relevant in healthcare ( OECD, 2023 ). Before adjusting fertility work-up guidelines and implementing HyFoSy as first-choice diagnostic tubal patency test, it may be relevant to compare the costs as well as the effects of HyFoSy and HSG. We here report a cost-effectiveness analysis of tubal patency testing by HyFoSy and HSG during fertility work-up, performed alongside the FOAM trial.

Between May 2015 and January 2019, 1160 women were included; a total number of 1026 women underwent both tubal patency tests and had data available (the flowchart of the FOAM trial is presented in Supplementary Fig. S1 ). The baseline characteristics for the entire study group and the two arms in those randomized after discordant results were comparable ( Supplementary Table S1 ). Seven hundred and forty-seven (73%) women had concordant test results, 143 (14%) women had discordant test results, and 136 (13%) women had at least one inconclusive test result, meaning either HyFoSy or HSG had failed. Of the 143 women with discordant test results, 105 gave consent and were randomly assigned to clinical management guided by either the results of HyFoSy or HSG. The live birth rate in women with concordant test results was 48% (361/747) and 40% (55/136) in women with inconclusive test results. For women with discordant test results allocated to clinical management based on the results of HyFoSy, live birth rate was 41% (22/54) and 49% (25/51) for women allocated to management based on the results of HSG. This resulted in a difference in live birth of −8% (95% CI: −27% to 10%). When these results were extrapolated to the entire study group, management based on the results of either HyFoSy or HSG was estimated to lead to a live birth rate of respectively 46% (474/1026) and 47% (486/1026) (difference −1.2%; 95% CI: −3.4% to 1.5%; P  = 0.27). Miscarriage, ectopic pregnancy, and multiple pregnancy rates were low and not significantly different for both groups. Additional information on the subsequent fertility treatment after tubal patency testing is presented in Supplementary Table S2 . In the entire study group, considering the costs in those with concordant, inconclusive, and discordant results, the average total costs for tubal patency testing, fertility treatments, pregnancy, and childbirth were €3307 (95% CI: €3104–€3523) for management based on the results of HyFoSy and €3427 (95% CI: €3223–€3642) for management based on the results of HSG, giving a mean difference of €−119 (95% CI: €−125 to €−114). So, while HyFoSy-based management led to lower costs per couple, live birth rates were also slightly lower, although this was not statistically significant. This yielded an ICER of €10 042, meaning that by using HyFoSy instead of HSG we would save €10 042 per each additional live birth lost. Vice versa, additional costs of €10 042 on HSG are incurred per additional live birth. In Fig. 1 , we report the bootstrap samples in the cost-effectiveness plane. Of all samples 81.1% appeared in the south-west (HyFoSy-based management is less expensive but also less effective) and 18.9% of the samples appeared in the south-east (HyFoSy-based management is less expensive and more effective). Because of crossing quadrants, simple 95% CIs of the ICER cannot be interpreted. Cost-effectiveness plane showing differences in costs and proportions of live birth in all bootstrap samples. HyFoSy, hysterosalpingo-foam sonography (HyFoSy-based management); HSG, hysterosalpingography (HSG-based management). The CEAC in Fig. 2 provides more information on the uncertainty surrounding the cost-effectiveness, as the ICER is difficult to interpret when crossing quadrants. We found that at willingness to pay of €5000 per additional live birth, HyFoSy-based management was cost-effective in 81.6% of bootstrap samples compared to HSG-based management, which decreased further to 32.4% for a willingness to pay threshold of €20 000 and 24.8% for a willingness to pay threshold of €40 000. Cost-effectiveness acceptability curve showing the proportion of bootstrap samples ( y -axis) that were found cost-effective when compared to a range of threshold monetary values ( x -axis). HyFoSy, hysterosalpingo-foam sonography (HyFoSy-based management); HSG, hysterosalpingography (HSG-based management). The cost-effectiveness planes and CEACs of the sensitivity analyses are shown in Supplementary Figs S2 , S3 , S4 , S5 , S6 , S7 , S8 , S9 , S10 , and S11 . First, in the subgroup with discordant test results (analysis 1), which is the subgroup which contributed to a difference between the two strategies in terms of management costs, the average total costs were €2823 (95% CI: €2335–€3300) for HyFoSy-based management and €3544 (95% CI: €2599–€4670) for HSG-based management, giving a mean (bootstrap) difference of €−722 (95% CI: €−1895 to €348). As before, the live birth rates were 40.7% after HyFoSy and 49.4% after HSG, giving a mean difference of −8.6% (95% CI: −27.6 to 10.2). HyFoSy was on average less expensive and less effective than HSG resulting in an ICER of €8349. We show the cost-effectiveness plane in Supplementary Fig. S2 and the cost-effectiveness acceptability curve in Supplementary Fig. S3 . Second, in the sensitivity analysis excluding the costs associated with pregnancy and childbirth, results were similar to the primary analysis (analysis 2). Third, in the sensitivity analysis using the societal perspective (analysis 3), we found that the results were similar to the ones in the primary analysis. Fourth and final, in the sensitivity analysis including costs if all HSG procedures were conducted with either water-based (analysis 4A) or oil-based contrast medium (analysis 4B), we found lower (€1062) and higher (€19 525) ICERs, respectively, compared to the primary analysis. There was still a considerable amount of uncertainty surrounding ICERs as shown in the CEACs such that overall conclusions were similar to the primary analysis.

The FOAM trial was a multicenter prospective, comparative study with a randomized design, performed in the Netherlands between May 2015 and January 2019 ( van Welie et al. , 2022 ). In this study, the effectiveness of clinical management based on the results of tubal patency testing by HyFoSy and HSG during fertility work up was compared. The study was approved by the Institutional Review Board of the Amsterdam UMC, location VU Medical Centre (No. 2014.454), the National Central Committee on Research involving Human Subjects (CCMO, The Netherlands; No. NL50484.029.14), and the local board of directors of the participating hospitals. The study is registered in the International Clinical Trials Registry Platform (No. NTR4746). The protocol of the FOAM trial, including a description of this economic evaluation, has been published previously ( van Rijswijk et al. , 2018b ). In summary, eligible and consenting women underwent both tubal patency tests in a randomly assigned order (HyFoSy-HSG or HSG-HyFoSy). The results of both tests were compared and assessed as concordant, inconclusive, or discordant. In case of discordant test results, women were randomly allocated to management based on the results of either HyFoSy or HSG. Infertile women were eligible if they were between 18 and 41 years of age and scheduled for tubal patency testing. The male partner or sperm donor should have normal or mildly impaired semen quality with a total post washed motile sperm count above 3 × 10 6 per milliliter. Women known with endometriosis, anovulatory cycles not responding to ovulation induction, and/or allergy to iodinated contrast medium were not invited. Eligible and consenting women underwent both tubal patency tests in the follicular phase of their cycle. During HyFoSy, 5–10 cc of echogenic foam (created by mixing 5 cc ExEm ® -gel with 5 cc sterile purified water (IQ Medical Ventures BV, Rotterdam, The Netherlands)) was infused through a special catheter with cervical cannula into the uterine cavity, while a transvaginal ultrasound was made. This ultrasound showed whether the foam passed through the fallopian tubes into the abdominal cavity, thereby demonstrating patency of the tubes or not. During HSG, contrast medium (oil- or water-based, according to local protocols) was infused into the uterine cavity through a special HSG-balloon catheter, cervical vacuum cup, or hysterophore. During infusion of the contrast medium, approximately four to six radiographs were taken to evaluate the patency of the fallopian tubes. The evaluation was done by a gynaecologist and/or radiologist. Due to the design of the study, neither of the clinicians nor participating couples were blinded, however HyFoSy and HSG were performed by different clinicians who were not informed about the results of the other test. Subsequent clinical fertility management was either based on the results of both tests in women with concordant or inconclusive test results or based on the results of the randomly assigned test in women with discordant test results. In women with at least one patent fallopian tube, expectant management for a minimum of 6 months or IUI was offered, depending on their prognosis for natural conception based on the Hunault model ( Hunault et al. , 2004 ). After six failed IUI cycles, women were offered IVF. In women with bilateral occlusion, IVF was offered or diagnostic laparoscopy with chromotubation was performed to confirm tubal occlusion and if confirmed, IVF was offered. Women with ovulation disorders and tubal patency continued using ovulation induction with or without IUI according to the local protocols. The primary outcome was ongoing pregnancy leading to live birth within 12 months after randomization. Ongoing pregnancy was defined as an intrauterine pregnancy with a fetal heartbeat on ultrasound examination between 10 and 12 weeks of gestation. Live birth was defined as a live birth after 24 weeks of gestation. Secondary outcomes considered in this economic evaluation were ongoing pregnancy rate, miscarriage rate, ectopic pregnancy rate, and multiple pregnancy rate. The economic evaluation was performed from a healthcare (direct medical costs) and societal (direct and indirect medical costs) perspective and expressed in European currency (Euro), in line with the International Society for Pharmacoeconomics and Outcomes Research (ISPOR) guidelines ( McGhan et al. , 2009 ) and following the Consolidated Health Economic Evaluation Reporting Standards statement ( Husereau et al. , 2022 ). From a healthcare perspective, we considered all costs associated with achieving live births in the primary analysis: tubal patency testing, fertility treatments, and costs associated with pregnancy and childbirth. Costs associated with pregnancy and childbirth included costs for miscarriage, ectopic pregnancy, and birth (singleton and multiple). The Dutch Health Care Authority has not assigned a standard price for HyFoSy. The unit price for HyFoSy was therefore estimated by including the costs to perform a transvaginal ultrasound and the costs of the ExEm ® -Foam and catheter. The costs for HSG were according to the Dutch Health Care Authority plus the mean costs for water- and oil-based contrast medium (retrieved from the Dutch Formulary for medication ( Hakkaart-Van Roijen, 2016 ) and Guerbet, respectively). The costs for fertility treatments and pregnancy losses (miscarriages and ectopic pregnancies) were obtained from the costs as retrieved by an expert panel on cost-effectiveness from the Dutch Consortium for Research in Women’s Health. The expert panel, consisting of gynaecologists, economists, and a methodologist, determined the actual per unit medical costs from resources that are being used in fertility studies within our consortium from two university hospitals and one general hospital. For the costs of a singleton and multiple pregnancy and birth, we used the study of Lukassen et al. (2004) . From a societal perspective, we added costs due to productivity loss to all other mentioned costs. Productivity loss due to absenteeism from work during fertility treatment were calculated using Dutch governmental guidelines ( Hakkaart-Van Roijen, 2016 ). All prices were standardized for the year 2018 using consumer price index data. The unit costs that were included in this economic evaluation are presented in Table 1 . Unit costs for the different direct medical cost categories: tubal patency tests, fertility treatment, pregnancy/birth and the indirect medical costs. CBS, Statistics Netherlands. Prices of the year of 2018. HyFoSy costs were estimated based on the costs for transvaginal ultrasound and the costs for the ExEm ® -foam and catheter (€62.10 and €70.00, respectively). HSG costs were calculated by averaging the costs of water- and oil-based contrast medium (€168.21 and €399.90, respectively). Ovulation induction costs were calculated by averaging the costs for ovulation induction by Clomiphene Citrate and gonadotrophines (€198.43 and €583.15, respectively). Costs for miscarriage were based on the mean costs for spontaneous miscarriage, medical treatment and curettage. Data on resource use were collected from the individual case record forms of the clinical trial. For each woman, the number of subsequent fertility treatments and pregnancy outcome within twelve months after inclusion was reported. For the analysis, we created two datasets for the 1026 participating women: one in which all would only have HyFoSy and HyFoSy based management and a second in which all would only have HSG and would follow HSG based management. We calculated the weighted average of all live births for women with concordant, inconclusive, and discordant test results, in which only the one with discordant results (n = 143) contributed to a difference between the two strategies in terms of management costs. For the women with at least one inconclusive test results, we included the costs for the initial (inconclusive) test and the alternative other test. The incremental cost-effectiveness ratio (ICER) was calculated by dividing the difference in mean total costs by the difference in chance of live birth. We bootstrapped the costs and effects for 5000 samples with replacement to obtain the 95% confidence intervals (95% CI) of skewed costs and effects. We reported the results from bootstrapping in the cost-effectiveness plane with the four quadrants that represent the four possible conclusions (HyFoSy-based management is more effective, but also more expensive than HSG based management; HyFoSy-based management is less effective and less expensive; HyFoSy-based management is less effective and more expensive; and HyFoSy-based management is more effective and less expensive). We used the bootstrap samples to construct a cost-effectiveness acceptability curve (CEAC) following Fenwick et al. (2004) . In CEACs, we vary the willingness-to-pay threshold from €0 to €100 000. The purpose of the CEAC is to show the probability that HyFoSy-based management would be considered cost-effective at these varying thresholds, defined as the proportion of bootstrap samples where the ICER was below the threshold value. This reflects the uncertainty around cost-effectiveness even when bootstrap samples cross quadrants. We first repeated the previously mentioned economic evaluation in the discordant patient group (instead of the overall 1026 women), as although this group can in practice not be found without conducting both tests on a patient, this subgroup is the driving force behind the difference in the overall group. Second, to evaluate the influence of costs associated with pregnancy and childbirth on the results, we repeated the analysis but now excluded these costs. As the primary goal of fertility treatment is to achieve live births, these costs as consequences of those live births could be considered as secondary, less important, or even warranted. Third, to evaluate the influence of the societal costs on cost-effectiveness, we conducted the same analysis in which we added costs for productivity loss due to hospital visits during their fertility treatment according to Dutch guidelines as well ( Hakkaart-Van Roijen, 2016 ). Fourth, as HSG can be performed with either water-based or oil-based contrast medium and this leads to considerable differences in price, sensitivity analyses were undertaken for the price of HSG with exclusively water-based contrast (€160) or oil-based contrast (€400).

In summary, this study showed that in infertile women with indication for tubal patency testing during their fertility work-up, clinical management based on the test results of HyFoSy leads to slightly lower, though not statistically significant, live birth rates, at lower costs, compared to management based on HSG results. This comparison resulted in an ICER of €10 042, meaning that by using HyFoSy instead of HSG we would save €10 042 per additional live birth lost. Although there was considerable uncertainty around the ICER, the more valuable we consider a live birth to be, the less likely it is that the HyFoSy strategy would be cost-effective compared to the HSG strategy. Importantly, the direct fertility enhancing effect of both tubal patency tests is not incorporated in this conclusion as women underwent both tubal patency tests in this study.

We here report results of an economic evaluation, performed alongside the FOAM trial. In women scheduled for tubal patency testing during their fertility work-up, clinical management based on the results of HyFoSy leads to slightly lower live birth rates (though not statistically significant) at slightly lower costs compared to management based on the results of HSG. This yielded an ICER of €10 042, meaning that by using HyFoSy instead of HSG, we would save €10 042 per additional live birth lost. There was considerable uncertainty around the ICER. The primary outcome of this study was ongoing pregnancy leading to live birth, the most relevant outcome for couples during their fertility work-up. We performed two sensitivity analyses to also evaluate the influence of the costs for contrast media used on the cost-effectiveness, as the costs for HSG are mainly driven by the type of contrast medium (oil-based contrast medium is approximately 12 times more expensive than water-based contrast medium). As a consequence, this had a considerable influence on our results given that the costs of the tests was the primary difference between strategies. The FOAM trial included women with numerous causes of infertility, therefore the results of this study may be widely applicable within fertility care. Analyses were performed from two perspectives: a healthcare perspective including direct medical costs only and a societal perspective including direct and indirect medical costs. A limitation of this study is the number of inconclusive HyFoSy test results, which was higher than anticipated. This situation could partly be explained by the fact that HyFoSy was a relatively new method for assessing tubal patency and that clinicians may not have mastered performing HyFoSy, despite the mandatory training before start of the study. Another explanation can be that HyFoSy is an operator dependent test, which makes it more difficult to assess the sonographic images once the test is performed. In the situation of inconclusive test results, an alternative tubal patency test should be performed which involves additional costs. We expect that once HyFoSy is more established in daily practice, clinicians’ experience will grow, resulting in a reduced number of inconclusive tests. We observed a significant uncertainty surrounding ICER comparing both tubal patency strategies. This uncertainty predominantly concerned the uncertainty surrounding the difference in live birth rate for both tubal patency strategies. Due to the design of the study, women underwent both tubal patency tests, which made it impossible to compare the potential fertility enhancing effect of either test. The fertility enhancing effect of HSG with oil-based contrast has widely been studied and confirmed ( Fang et al. , 2018 ; Wang et al. , 2020 ). For HyFoSy, only small and observational studies reported on pregnancy outcomes and such effect has not been confirmed yet ( Emanuel et al. , 2012 ; Exacoustos et al. , 2015 ; Tanaka et al. , 2018 ). If either of the tests appears to be more effective, this can potentially lead to a reduction in expensive fertility treatments. Finally, in this analysis, costs were used based on Dutch standards, given that the study was conducted in the Netherlands. It is important to note that the costs related to fertility care can differ across countries, especially the costs for oil-based contrast medium used during HSG, thus limiting the generalizability of the study findings to other countries. To our knowledge, this is the first economic evaluation that compared clinical management based on the results of tubal patency testing by HyFoSy and HSG during fertility work-up. Van Rijswijk et al. performed a cost-effectiveness analysis on the comparison of tubal patency testing by HSG with oil-based versus water-based contrast medium. They showed that HSG with oil-based contrast medium is cost-effective compared to water-based contrast medium if the society is willing to pay $8198 for an additional live birth ( van Rijswijk et al. , 2018a ). Subsequently, they performed a long-term cost-effectiveness analysis over a 5-year follow-up period showing that HSG with oil-based contrast medium resulted in more live births at comparable costs compared to HSG with water-based contrast medium, and therefore using oil-based contrast medium during HSG was the dominant strategy ( van Welie et al. , 2021 ). These studies showed that even though HSG with oil-based contrast medium incurred higher costs than water-based contrast medium, oil-based contrast medium was eventually less expensive thanks to the cost savings on fertility treatments. Furthermore, these results underline the importance of comparing the direct fertility enhancing effect of tubal patency tests. van Kessel et al. (2022) performed a cost-effectiveness analysis comparing tubal patency testing by HSG (with either water- or oil-based contrast medium) and transvaginal hydrolaparoscopy (THL) during fertility work-up. They showed that THL resulted in a higher live birth rate, at lower costs. The main driver for the lower costs for THL were the lower number of additional laparoscopies and fertility treatments. In the FOAM trial, HSG was performed with either water- or oil-based contrast medium, and no distinction between both contrast media was made, therefore the effect of either of the contrast media on live birth chance remains unknown. To determine the most effective treatment approach, it is essential to evaluate the cost-effectiveness of diagnostic tests in relation to the societal value placed on achieving an additional live birth (referred to as the willingness to pay). The ESHRE Capri Workgroup illuminated the challenges of performing cost-effectiveness analyses in infertility care. It is difficult to capture the value of infertility care in quality adjusted life years (QALYs), the standard metric used when comparing clinical interventions. QALYs represent an additional year in perfect health and not an additional healthy live birth, which is the primary aim of infertility care (Group ECW, 2015 ). Besides cost-effectiveness, we have to consider other aspects of the tests as well. HSG represents a higher burden for both patient and clinician than HyFoSy: it is experienced as significantly more painful (based on Visual Analogue Scale; 5.4 cm versus 3.1 cm; P  < 0.001 ( van Welie et al. , 2022 )) and causes exposure to ionizing radiation and iodinated contrast medium. HyFoSy can be easily performed by a gynaecologist or fertility doctor allowing a one-stop fertility work-up with no need for the radiology department and staff. However, the direct fertility enhancing effects of the two tests have never been compared. To determine the preferred first choice tubal patency test during fertility work-up a head-to-head comparison of tubal patency testing by HyFoSy and HSG with oil-based contrast medium would be required.