1 Reproductive Medicine, Department of Obstetrics and Gynecology, Sahlgrenska University Hospital, Göteborg, Sweden, 2 Fertility Unit, Department of Gynecology and Obstetrics, Roskilde Hospital, Copenhagen University, Roskilde, Denmark and 3 Swedish Council on Technology Assessment in Health Care, Stockholm, Sweden
4 To whom correspondence should be addressed. E-mail: annika.strandell{at}medfak.gu.se
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: cost-effectiveness/hydrosalpinx/incremental cost-effectiveness ratio/in vitro fertilization/salpingectomy
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The suggested treatments have focused on the disposal of the hydrosalpingeal fluid. Various surgical methods, such as salpingectomy, salpingostomy, proximal tubal ligation and transvaginal aspiration of fluid, have been studied (Stadtmauer et al., 2000; Surrey and Schoolcraft 2001
; Sowter et al., 1997
; Van Voorhis et al., 1998
) as well as the use of antibiotics alone (Hurst et al., 2001
). However, salpingectomy is the only method that has been evaluated in prospective randomized controlled trials (Déchaud et al., 1998
; Goldstein et al., 1998
; Strandell et al., 1999
), while all the other methods are presented in different retrospective designs.
The largest randomized trial including the first IVF cycle demonstrated a significant benefit in terms of a doubled birth rate after salpingectomy compared with no intervention prior to IVF, in patients with hydrosalpinx that were fluid-filled to the extent that they could be detected at transvaginal sonography (Strandell et al., 1999). A second analysis included all subsequent IVF cycles, before which patients without prior salpingectomy had the option to undergo the procedure after a failed first or second cycle (Strandell et al., 2001
). Also, the cumulative results demonstrated a doubled hazard ratio after salpingectomy. From a crude rate perspective, including spontaneous pregnancies and allowing for salpingectomy in the no-intervention group, the cumulative birth rates in the intention-to-treat analysis were equal and reached 55%.
There is clear evidence for the benefit of salpingectomy, but there are other aspects, such as a psychological dilemma about whether to remove the tubes, and also economic considerations, which have not been described previously. The present study compares the cost-effectiveness of two strategies among hydrosalpinx patients: performing salpingectomy prior to the first IVF cycle, and optional salpingectomy after failed cycles. The analysis is based on the cumulative result of the Scandinavian multicentre study (Strandell et al., 2001).
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
A total of 204 patients with hydrosalpinges, primarily diagnosed by either hysterosalpingography or laparoscopy, were randomized in a ratio 3:2 at each centre, to undergo a laparoscopic salpingectomy or to have no intervention before their first cycle. The power calculation, randomization procedure, IVF performance and group allocation are described in detail in the original publication (Strandell et al., 1999). A later publication includes all subsequent cycles, routinely not more than three transfer cycles per couple and allowing for salpingectomy in the control group (Strandell et al., 2001). These original publications have shown that the benefit of salpingectomy is particularly evident in patients with large hydrosalpinges that are visible on ultrasound, a subgroup recognized before the start of the study. It was concluded that patients with ultrasound-visible hydrosalpinges had twice as high a chance of achieving a live birth if they had a salpingectomy prior to IVF, and this is the group which is recommended salpingectomy. Thus, in the present report, the focus is on this group.
In the present analysis, group allocation is according to the on-treatment-based analysis, in which the intervention strategy is defined as salpingectomy preceding the first embryo transfer and the control strategy allowing salpingectomy after one transfer. Thus, six patients from each group are excluded since they never started IVF treatment and six patients are reallocated (Figure 1). Three patients randomized to salpingectomy never underwent surgery and two patients randomized to salpingectomy underwent a complete IVF cycle before the surgery was performed. One patient randomized to no intervention had her first and only embryo transfer after a salpingectomy, which was performed after freezing of all embryos. Two patients randomized to salpingectomy conceived spontaneously after completed IVF treatment but their pregnancies are not included in the present analysis.
|
The various surgical interventions and their potential complications in the intervention group have been described previously (Strandell et al., 1999). Data from the control group show that 28 of the 86 patients had a salpingectomy performed after a failed first or second cycle. A maximum of three complete treatment cycles (including ovarian stimulation, one fresh and additional frozen embryo transfers) per patient were included. For each patient, the outcomes of treatment have been recorded as number of complete cycles of standard IVF and ICSI, cycles ending after oocyte retrieval, started cycles cancelled before oocyte retrieval, frozen/thawed transfer cycles, and complications during any procedure.
The outcomes of clinical pregnancies were recorded as spontaneous abortions (gestational week 612 or >12 weeks), managed conservatively or surgically by curettage, ectopic pregnancies treated by laparoscopic salpingectomy, termination of pregnancy after >12 gestational weeks, stillbirths and live births. Routes of delivery were recorded as vaginal delivery with/without complications and Caesarean section with/without complications. The number of children per birth (singleton or twin) was also registered. A twin live birth was counted as only one birth. Several pregnancies, but only one birth per patient could be included in the analysis.
Calculation of costs
The costs of treatments and interventions were calculated from standardized hospital charges at the main Swedish centre, Sahlgrenska University Hospital. These charges, which are partly based on diagnosis-related groups, are routinely used for invoicing between departments within the hospital and for referred patients.
Only direct medical costs, i.e. treatment costs and costs that are related to the outcome, such as costs for spontaneous abortion or delivery, were included. Direct non-medical costs, such as cost for transportation to and from the hospital, and indirect costs for production loss due to sick leave related to treatment, were not included (Eisenberg, 1989; Drummond et al., 1997
).
The costs for laparoscopic salpingectomy were based on the charges for using the anaesthetic and the surgical facilities and the charges for the post-operative gynaecological ward. Patients were generally discharged from the hospital on the first post-operative day. Hence, the hospital charge for 1 day was used to calculate the cost associated with post-operative care. Post-operative complications requiring additional hospitalization were recorded and costs for additional days in hospital were calculated. No follow-up visits were scheduled.
The medication costs were calculated as the average cost of drugs given to the patients undergoing IVF, including pituitary down-regulation with a GnRH analogue and gonadotrophins. The cost of GnRH analogues was based on the average cost of the total dose of three different preparations used in the study (buserelin acetate administered nasally or subcutaneously, or nasal nafarelin acetate). The majority of cycles used recombinant or highly purified FSH and the costs for gonadotrophins are based on the cycles using recombinant FSH. Costs for luteal support are based on the use of vaginal progesterone, which was the most common preparation used. The different types of medication were evenly distributed between groups. All medication costs are based on the retail prices of the Swedish pharmacy chain Apoteket, 2004. The vast majority of frozen/thawed cycles (88%) were performed in natural cycles and no drug costs for stimulation were thus included.
The charge for IVF treatment set by the hospital was levelled according to the completeness of each treatment cycle, considering cancellations before and after oocyte retrieval. All follow-up visits were included. Complications such as ovarian hyperstimulation syndrome (OHSS) and infections requiring hospitalization were registered as days in hospital and the corresponding costs were calculated. Complications such as adnexal torsion due to enlarged hyperstimulated ovaries were considered related to the IVF treatment and the corresponding costs for surgery were thus referred to IVF-related complications.
Pregnancy-related costs included routine antenatal controls, spontaneous abortions, ectopic pregnancies, termination of pregnancy and delivery. The calculation of delivery-related costs was based on number of children, route of delivery and complications such as intrauterine death and antenatal care due to risk of premature delivery. For instrumental deliveries by vacuum extraction or forceps, as well as for twin deliveries, the hospital charge for delivery-with-complication was used. Neonatal care was accounted for by using the hospital charges, which were set according to birth weight classes.
The effectiveness of the two strategies (salpingectomy prior to IVF versus option to undergo salpingectomy in case of failed cycle) was expressed in terms of live birth rate per patient.
The average cost per patient, was calculated from individual data, with the 95% confidence interval (CI) for the difference between groups, and the cost-effectiveness ratios were calculated as the average cost (excluding neonatal costs) per patient divided by the live birth rate. The incremental cost-effectiveness ratio (ICER), the extra cost per unit of outcome obtained, i.e. the extra cost to achieve one more live birth, was calculated as:
![]() |
We used bootstrapping to calculate 95% confidence intervals for the ICER and for the differences in cost-effectiveness ratios between groups. The bootstrap method is used to obtain a robust non-parametric estimate of the confidence interval. A bootstrap sample, of the same sample size as the original study, is generated with replacement from the obtained empirical distribution, i.e. the original study sample. The cost-effectiveness ratio is computed by using the bootstrap sample in place of the original sample. These procedures are repeated 1000 times and the 2.5th and 97.5th percentiles of the calculated cost-effectiveness ratio in the 1000 bootstrap samples form the 95% confidence interval of the cost-effectiveness ratio (Chernick, 1999).
All costs were expressed in terms of the price level of 2004 and converted from Swedish kronor (SEK) into euros () at the exchange rate of 1
= 9.0 SEK.
Significance analyses were performed using the two-sample t-test for continuous variables and Fishers exact test for proportions. An -level of 0.05 was considered significant.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
Complications of IVF occurred in 12 patients (11%) in the intervention strategy; 10 patients developed OHSS (requiring a total of 20 days in hospital) and two patients with infections (fever and abdominal pain) required hospitalization for a total of 5 days. Among controls, 11 patients (13%) suffered from complications; seven patients developed OHSS (17 days in hospital), three patients were treated for pelvic infections (10 days in hospital) and one patient underwent laparotomy due to adnexal torsion.
In all, 95 patients were diagnosed with ultrasound-visible hydrosalpinges before the start of treatment. Among the 51 patients in the intervention group, 23.5% did not complete a full set of three stimulated cycles. The mean number of started cycles was 2.02 (SD 0.99) and mean number of transfers was 2.27 (SD 1.25). Among the 44 controls, 31.8% did not fulfil three stimulated cycles. The mean number of started cycles was 2.16 (SD 1.01) and the mean number of transfers was 2.25 (SD 1.24). Seventeen patients (39%) underwent salpingectomy after a failed cycle, of which 16 were after the first cycle. Six patients completed a full set of three stimulated cycles without undergoing salpingectomy. The live birth rates were 60.8% in the intervention group and 40.9% among controls, the difference being 19.9% (95% CI 0.4, 40.1).
The basic data for calculation of costs associated with treatment, pregnancy and neonatal care are described in detail in Tables IIV.
|
|
|
The average treatment costs per patient, including surgery and IVF, were 10 957 in the intervention strategy and
9996 among controls, i.e. a difference of
961 (95% CI 863, 2 787). The average cost per patient, including treatment and pregnancy-related costs, in the intervention group was
13 943 compared with
12 091 among controls, a difference of
1852 (95% CI 57, 3646).
The cost-effectiveness, expressed as cost per live birth, excluding neonatal costs, was 22 823 in the intervention group and
29 517 among controls, a difference of
6 693 (95% CI 22 575, 2 991). The ICER, was calculated as: (13 943 12 091)/(0.608 0.409) = 1852/0.199 =
9306 (95% CI 8653, 60 867). That is, the intervention strategy (among patients with ultrasound visible hydrosalpinges) results in an extra live birth at a cost of
9306. A summary of all costs is displayed in Table V.
|
A cost-effectiveness analysis, including also patients with hydrosalpinges that were not visible on ultrasound, was also conducted, and the summary results are shown in Table VI.
|
The ICER was 12 925 (95% CI 92 685, 131 492). Thus, despite the doubled sample size, the 95% CI is even wider than for the targeted group of ultrasound-visible hydrosalpinges, due to the small difference in effect size.
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The ICER of 9306 in the main analysis, concerning patients with ultrasound-visible hydrosalpinges, seems quite reasonable. Even the upper limit of the 95% CI seems to be a reasonable cost for saving another child. The wide confidence interval is due to the rather small sample in this group. From the perspective of the payers, of course, it is relevant to take into account that the average cost per patient is higher if salpingectomy is performed before the first cycle, even if the cost per live birth is lower. From the patients perspective, the reimbursement system for IVF and surgery may influence the choice of treatment strategy.
It should be pointed out that the original study designated salpingectomy prior to the first cycle and the management of subsequent cycles varied between centres, implying that the discussion of salpingectomy after failed cycles was not always initiated. This design results in a heterogeneous control group, considering that salpingectomy could be performed after the first or second failed cycle, or not at all. Among controls, 39% had a salpingectomy performed and only six patients (14%) completed a full set of three stimulated cycles without having undergone a salpingectomy and not achieving a live birth, suggesting that the vast majority of patients were offered surgery. While the study was conducted and no result regarding the benefit of salpingectomy was yet available, patients were obviously offered, and accepted, salpingectomy. At present, several years after the positive result was published, the dissemination and implementation of salpingectomy as a recommendation vary between clinics and physicians. A survey of IVF clinics in the UK demonstrated that 12% of units did not recommend treatment of hydrosalpinx prior to IVF treatment, while 36, 33 and 19% recommended treatment weakly, strongly and very strongly respectively (Hammadieh et al., 2004). Considering this scenario, it is likely that the control group in the present study is a strong reflection of clinical practice.
The cost of salpingectomy depends very much on the length of hospitalization. In the present analysis, the cost of one overnight stay was included, since these surgeries were performed later in the afternoon, in the hospital we used as a model. The salpingectomy procedure could certainly be performed in a day-surgery setting, which would further reduce the cost of salpingectomy.
On the other hand, if indirect costs due to sick leave after surgery (general recommendation of 1 week) had been included, the cost of the salpingectomy strategy would have increased, since all patients in the active intervention group underwent surgery, compared with only 39% of controls. The cost of production loss is difficult to calculate since the actual number of days on sick leave was not recorded. On average, however, this means an extra cost of approximately 665 per patient (5 days x
133) in the intervention group and
260 per patient (0.39 x 5 days x
133) among controls. This difference of
405 implies an ICER of
11 342 among those with ultrasound-visible hydrosalpinges. Other indirect costs are supposed to be evenly distributed between the two strategies.
The rather high complication rate will increase the costs of IVF treatment, but only marginally. More important is that complications like OHSS and infections were evenly distributed between the groups and will not affect the difference in costs.
Inclusion of neonatal costs in the analysis would skew the results in favour of the intervention strategy, because of the fairly low numbers of births and the skewed distribution of multiple births. The higher twin rate among controls (22 versus 16%) was not statistically significant, and there is no rationale for patients in the control group to have a higher twin rate than patients having salpingectomy prior to their first IVF. In a larger sample, the need for neonatal care would probably have been more evenly distributed. Thus, it is reasonable to exclude the costs of neonatal care when calculating the cost-effectiveness ratios. Currently, single embryo transfers are more common and performed as a routine procedure in many Scandinavian centres and the problem of low birth weight in twin births has already decreased in these countries.
In conclusion, the incremental cost to achieve the higher birth rate by adhering to the intervention strategy, i.e. salpingectomy prior to the first IVF cycle, seems quite reasonable in patients with ultrasound-visible hydrosalpinges.
![]() |
Acknowledgements |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
![]() |
References |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Andersen AN, Lindhard A, Loft A, Ziebe S and Andersen CY (1996) The infertile patient with hydrosalpinges-IVF with or without salpingectomy. Hum Reprod 11,20812084.[ISI][Medline]
Barmat LI, Nasti K, Yang X, Spandorfer S, Kowalik A and El-Roeiy A (1999) Are cytokines and growth factors responsible for the detrimental effects of hydrosalpingeal fluid on pregnancy rates after in vitro fertilization-embryo transfer? Fertil Steril 72,11101112.[CrossRef][ISI][Medline]
Bedaiwy MA, Goldberg JM, Singh M, Sharma R, Wang X, Nelson D and Falcone T (2002) Relationship between oxidative stress and embryotoxicity of hydrosalpingeal fluid. Hum Reprod 17,601604.
Camus E, Poncelet C, Goffinet F Wainer B, Merlet F, Nisand I and Philippe HJ (1999) Pregnancy rates after IVF in cases of tubal infertility with and without hydrosalpinx: Meta-analysis of published comparative studies. Hum Reprod 14,12431249.
Chernick MR (1999) Bootstrap methods, a practitioners guide. John Wiley New York, USA.
Choe J and Check JH (1999) Salpingectomy for unilateral hydrosalpinx may improve in vivo fecundity. Gynecol Obstet Invest 48,285287.[CrossRef][ISI][Medline]
Drummond MF, OBrien B and Stoddart GL (1997) Methods for the economic evaluation of health care programmes. 2nd edition. Oxford: Oxford University Press.
Eisenberg JM (1989) A guide to the economic analysis of clinical practices. J Am Med Assoc 262,28792886.[CrossRef][ISI][Medline]
Déchaud H, Daurès JP, Arnal F, Humeau C and Hedon B (1998) Does previous salpingectomy improve implantation and pregnancy rates in patients with severe tubal factor infertility who are undergoing in vitro fertilization? A pilot prospective randomized study. Fertil Steril 69,10201025.[CrossRef][ISI][Medline]
Goldstein DB, Sasaran LH, Stadtmauer L and Popa R (1998) Selective salpingostomy-salpingectomy (SSS) and medical treatment prior to IVF in patients with hydrosalpinx [abstract]. Fertil Steril 70(3, Suppl 1)S320.[CrossRef]
Hammadieh N, Afnan M, Evans J, Sharif K, Amso N, Olufowobi O (2004) A postal survey of hydrosalpinx management prior to IVF in the United Kingdom. Hum Reprod 19,10091012.
Hammarberg K, Astbury J and Baker H (2001) Womens experience of IVF: a follow-up study. Hum Reprod 16,374383.
Hurst BS, Tucker KE, Awoniyi CA and Schlaff WD (2001) Hydrosalpinx treated with extended doxycyclin does not compromise the success of in vitro fertilization. Fertil Steril 75,10171019.[CrossRef][ISI][Medline]
Meyer WR, Castelbaum AJ, Somkuti S, Sagoskin AW, Doyle M, Harris JE and Lessey BA (1997) Hydrosalpinges adversely affect markers of endometrial receptivity. Hum Reprod 12,13931398.[Abstract]
Mukherjee T, Copperman AB, McCaffrey C, Cook CA, Bustillo M and Obasaju MF (1996) Hydrosalpinx fluid has embryotoxic effects on murine embryogenesis: a case for prophylactic salpingectomy. Fertil Steril 66,851853.[ISI][Medline]
Sowter MC, Akande VA, Williams JA and Hull MG (1997) Is the outcome of in-vitro fertilization and embryo transfer treatment improved by spontaneous or surgical drainage of a hydrosalpinx? Hum Reprod 12,21472150.[Abstract]
Stadtmauer LA, Riehl RM, Toma SK and Talbert LM (2000) Cauterization of hydrosalpinges before in vitro fertilization is an effective surgical treatment associated with improved pregnancy rates. Am J Obstet Gynecol 183,367371.[CrossRef][ISI][Medline]
Strandell A, Waldenström U, Nilsson L and Hamberger L (1994) Hydrosalpinx reduces in-vitro fertilization/embryo transfer rates. Hum Reprod 9,861863.[Abstract]
Strandell A, Sjögren A, Bentin-Ley U, Thorburn J, Hamberger L and Brännström M (1998) Hydrosalpinx fluid does not adversely affect the normal development of human embryos and implantation in vitro. Hum Reprod 13,29212925.
Strandell A, Lindhard A, Waldenström U, Thorburn J, Janson PO and Hamberger L (1999) Hydrosalpinx and IVF outcome: a prospective, randomized multicentre trial in Scandinavia on salpingectomy prior to IVF. Hum Reprod 14,27622769.
Strandell A, Lindhard A, Waldenstrom U and Thorburn J (2001) Hydrosalpinx and IVF outcome: cumulative results after salpingectomy in a randomized controlled trial. Hum Reprod 16,24032410.
Strandell A, Thorburn J and Wallin A (2004) The presence of cytokines and growth factors in hydrosalpingeal fluid. J Assist Reprod Genet 21,241247.[CrossRef][ISI][Medline]
Surrey ES and Schoolcraft WB (2001) Laparoscopic management of hydrosalpinges before in vitro fertilization-embryo transfer: salpingectomy versus proximal tubal occlusion. Fertil Steril 75,612617.[CrossRef][ISI][Medline]
Van Voorhis BJ, Sparks AE, Syrop CH and Stowall DA (1998) Ultrasound-guided aspiration of hydrosalpinges is associated with improved pregnancy and implantation rates after in-vitro fertilization cycles. Hum Reprod 13,736739.[Abstract]
Zeyneloglu HB, Arici A and Olive DL (1998) Adverse effects of hydrosalpinx on pregnancy rates after in vitro fertilization-embryo transfer. Fertil Steril 70,492499.[CrossRef][ISI][Medline]
Submitted on March 2, 2005; resubmitted on May 28, 2005; accepted on June 6, 2005.
|