Economic evaluation of highly purified menotropin compared with recombinant follicle-stimulating hormone in assisted reproduction

Transcription

1 FERTILITY AND STERILITY VOL. 80, NO. 5, NOVEMBER 2003 Copyright 2003 American Society for Reproductive Medicine Published by Elsevier Inc. Printed on acid-free paper in U.S.A. Economic evaluation of highly purified menotropin compared with recombinant follicle-stimulating hormone in assisted reproduction Adam Lloyd, M.Phil., a Richard Kennedy, M.B.Ch.B., D.A., F.R.C.O.G., b Julia Hutchinson, M.Sc., a and Will Sawyer, Ph.D. c Fourth Hurdle Consulting Ltd. and London School of Economics and Political Science, London, and University Hospitals Coventry and Warwickshire NHS Trust, Coventry, United Kingdom Received October 10, 2002; revised and accepted May 10, Funded by a grant from Ferring Pharmaceuticals. Reprint requests: Adam Lloyd, M.Phil., Fourth Hurdle Consulting Ltd, 2 Fisher Street, Holborn, London WC1R 4QA, United Kingdom (FAX: ; adamlloyd@ fourthhurdle.com). a Fourth Hurdle Consulting Ltd. b University Hospitals Coventry and Warwickshire NHS Trust. c London School of Economics and Political Science /03/$30.00 doi: /s (03) X Objective: To determine the cost of achieving pregnancy with different gonadotropin preparations. Design: Cost-minimization analysis of a prospective randomized clinical trial. Setting: Twenty-two centers in six countries. Patient(s): Women 18 to 36 years of age with infertility for more than 1 year who were undergoing IVF or ICSI. Intervention(s): or recombinant FSH. Main Outcome Measure(s): Mean cost of achieving an ongoing pregnancy. Result(s): The mean cost per patient treatment cycle was estimated to be 2,423 with highly purified (95% CI, 2,356 to 2,495) and 2,745 with recombinant FSH (95% CI, 2,658 to 2,830). The ongoing pregnancy rate was 22% with highly purified and 19% with recombinant FSH. The cost per ongoing pregnancy was 10,781 with highly purified (95% CI, 9,056 to 12,919) and 14,284 with recombinant FSH (95% CI, 11,883 to 17,891). Conclusion(s): and recombinant FSH are equally effective, but highly purified is less expensive per cycle. Using highly purified instead of recombinant FSH would translate into a 13% increase in the number of cycles that could be offered. (Fertil Steril 2003;80: by American Society for Reproductive Medicine.) Key Words: Assisted reproductive technology, menotropin, recombinant FSH, economics, cost effectiveness, cost, economic evaluation Infertility affects up to one in seven couples in the United Kingdom at some point in their lives (1). According to the Royal College of Obstetricians and Gynaecologists guideline published in 2000, a typical district health authority in the United Kingdom may see around 230 new consultant referrals each year (1). A proportion of these couples may be able to ultimately conceive (i.e., they are subfertile rather than infertile), but for the majority conception is unlikely without some form of medical intervention. In vitro fertilization and, more recently, ICSI are now commonly used treatments for infertility attributable to tubal factors, significant endometriosis, and male factor. In vitro fertilization is also used to treat persistent unexplained infertility. Although the effectiveness of IVF and ICSI is established, the techniques can be expensive (2, 3) and failure rates remain high: in the United Kingdom in 2000 to 2001, 21.8% of cycles initiated resulted in live birth (4). Effectiveness can be enhanced by freezing embryos. The live birth rate resulting from thawed cycles, while lower than for fresh embryos (13.8% [4]), is sufficient to make this an appealing alternative to repeated stimulation cycles for couples from whom spare healthy embryos can be obtained during IVF. In the United Kingdom National Health Service (NHS), assisted reproduction operates 1108

2 within funding constraints, and centers operate within annual fixed budgets. It is important to understand the economic effect of treatments to ensure that the limited available budget is used efficiently. A significant component of the cost of IVF/ICSI is the drugs used for ovarian stimulation before oocyte retrieval. Several different drugs are available for use in assisted reproduction technology (ART). The most recent United Kingdom guideline on the management of infertility in tertiary care, produced by the Royal College of Obstetricians and Gynaecologists (1), recommends that recombinant FSH or highly purified be used for ovarian stimulation. The guideline states: There is a small but significant increase in pregnancy rates following the use of urinary derived high purity gonadotropin preparations when compared to human menopausal gonadotropin preparations. FSH produces significantly more oocytes in an IVF cycle compared to urinary derived high purity gonadotropin preparations, however pregnancy rates in fresh embryo replacement cycles are similar. Because of the small differences in outcome, additional factors should be considered when choosing a gonadotropin regimen, including patient acceptability, costs and drug availability. A recent large phase III clinical trial compared highly purified with a genetically engineered recombinant FSH (Gonal-F; Serono Laboratories, Norwell, MA) for ovarian stimulation in patients undergoing IVF or ICSI. The efficacy and safety results are reported in detail elsewhere (5). In the intention-to-treat analysis, rates of ongoing pregnancy at 10 weeks were 22% with highly purified and 19% with recombinant FSH (P not significant). In the per protocol analysis, patients who did not violate the protocol had ongoing pregnancy rates of 25% and 21%, respectively. The results of this trial allow us to use robust efficacy data to conduct an economic evaluation by using economic modeling. MATERIALS AND METHODS Overview A cost-minimization analysis was performed from the perspective of our hospital by using simple economic modeling techniques. Microsoft Excel software was used (Microsoft Corp., Redmond, WA). The analysis was done on the basis of efficacy results from the open-label, randomized, parallel-group, international multicenter clinical trial that compared highly purified and recombinant FSH (5). The analysis determined the average cost of achieving an ongoing pregnancy at 10 or more weeks with each treatment. The outcome measure, ongoing pregnancy at 10 weeks or more, was the primary measure of effectiveness. The total per patient cost per cycle of each of the treatment strategies included acquisition costs of the treatments and any other healthcare resource use associated with the assisted reproduction procedure, in addition to the costs of managing adverse events. The cost per ongoing pregnancy was calculated by dividing per patient cost by the rate of ongoing pregnancy. Clinical Study Details The study on which we based our analysis was an openlabel, randomized, parallel group, multicenter, multinational, comparative efficacy and safety trial (5). The protocol was approved by local independent ethics committees and institutional review boards before patients were enrolled. Seven hundred eighty-one patients at 22 centers in six countries (Belgium, Germany, Israel, the Netherlands, Switzerland, and the United Kingdom) were randomly assigned to receive highly purified or recombinant FSH for ovarian stimulation as part of a single cycle of IVF or ICSI after pituitary desensitization with LHRH analogues started in midluteal phase of the preceding cycle. Fertilization (through ART) was performed to the standard practice of the participating centers, as were the constituents of all the procedures. Women 18 to 36 years of age with regular ovulatory menstrual cycles of 24 to 36 days were eligible if they were considered candidates for IVF/ICSI and had had proven infertility for more than 1 year (except for proven bilateral occlusion or male factor). Patients were excluded if they had a body mass index of less than 18.0 or greater than 29.0; had undergone more than three previous unsuccessful IVF/ICSI cycles; were considered a poor responder to gonadotropinstimulated procedures (defined as a development of fewer than four follicles or more than 20 days of gonadotropin stimulation until hcg criteria were met); had a history of severe ovarian hyperstimulation syndrome (type III); or were known to use illegal drugs, drink excessive alcohol ( 30 units per week), or smoke more than 10 cigarettes per day. Pituitary down-regulation was established with injectable GnRH agonist formulation (depot or daily formulation) and was administered until ovarian suppression was achieved (according to local practice). Gonadotropins were then administered by the patient s.c. for 5 days at a fixed daily dose of 225 IU. The dosage was then individually titrated (to a maximum daily dose of 450 IU) until the patient had at least three follicles larger than 16 mm in diameter or estradiol levels exceeded 1,000 pmol/l. Final oocyte maturation was achieved with hcg, and the luteal phase was supported with progestagens, according to local practice. Ongoing pregnancy was assessed by a serum or urine pregnancy test and confirmed by ultrasonography 10 or more weeks after oocyte retrieval. Figure 1 shows the study visit schedule. Data Collection Data on efficacy, safety, and adverse events were collected in a case record form. During gonadotropin adminis- FERTILITY & STERILITY 1109

3 FIGURE 1 Sequence of visits associated with assisted reproduction in the study. tration, patients also recorded injection site reactions and adverse events on a diary card. Efficacy and safety results from this study are reported elsewhere (5). We accessed the full study data set from data held on file at Ferring Pharmaceuticals. Sample for Economic Evaluation The economic evaluation was conducted on all patients analyzed on an intention-to-treat basis. Medical Resource Use Data Medical resource use data was collected for each patient. These data included ampoules of study medication used; use of other medication as part of the IVF/ICSI procedure; number of visits to the clinic; number of clinic visits requiring injections; ultrasound examinations; blood tests and other diagnostic procedures; oocyte retrieval, fertilization, and implantation procedures; and medications, physician visits, and hospital stay for the management of adverse events. Resource use required to manage adverse events was estimated on the basis of descriptions of the events written by study investigators. Application of Unit Costs to Resource Use Resource use data for each patient cycle were multiplied by unit costs to calculate per patient costs. Unit costs of clinic visits, tests, and investigations and procedures were obtained from the published literature (cost of drugs) and the University Hospitals Coventry and Warwickshire (UHCW) finance department (costs of procedures, visits, ultrasonography, blood tests, and injections) or were estimated from published information if the former sources were unavailable TABLE 1 Unit costs for medical resource use. Resource type Source Unit cost ( ) Clinic visit UHCW Phone consultation UHCW Injection a UHCW Ultrasound UHCW Blood test UHCW 8.16 Other diagnostic tests UHFW 8.16 Oocyte retrieval UHFW Embryo implantation UHCW Fertilization technique IVF UHCW ICSI UHCW Hospitalizations Emergency department (for PSSRU overnight observation) General surgical ward PSSRU Gynecology ward PSSRU High-dependency unit NHS Outpatient procedure NHS Cost of study medication Gonal-F (per 75-U ampoule) BNF Menapur (HP-) BNF (per 75/75 ampoule) Cost of other drugs BNF Various Note: UHCW University Hospitals Coventry and Warwickshire, finance department; PSSRU Personal Social Services Research Unit, University of Kent; NHS National Health Service Reference costs; BNF British National Formulary, British Medical Association, London. a Staff cost attributable to giving injection. (as for some drugs and medical consumables). Details are provided in Table 1. For each medication used as part of the study protocol, we calculated the cost per day of the most economical available prescription that matched drug, brand, dose, and route of administration recorded in the case record form. The cost of concomitant medications was calculated in a similar manner. For each medication except depot injections, we calculated the cost per day and multiplied this value by the number of days for which the medication was taken. Depot injections were treated as a single day s treatment by using the full cost of the depot formulation. Effectiveness Measure for the Economic Evaluation The effectiveness measure for our analysis was the primary effectiveness measure from the study: rate of ongoing pregnancy at 10 weeks or longer. Cost Perspective for Analysis Costs were calculated from the perspective of the United Kingdom NHS Lloyd et al. Economics of assisted reproduction Vol. 80, No. 5, November 2003

4 TABLE 2 Demographic, characteristic, and baseline data. TABLE 3 Ongoing pregnancy rate. Recombinant FSH Recombinant FSH Age (y) Body weight (kg) Height (m) Ethnicity (%) White Black Asian Mixed No. of previous ART cycles Missing 1 0 Assisted reproductive treatment IVF (%) 139 (35) 143 (37) ICSI (%) 257 (65) 242 (63) Note: Data with the plus/minus sign are means ( SD). Cost-Minimization Analysis After assigning unit costs to the medical resource use recorded in the study, we determined the average cost per patient and the cost of achieving an ongoing pregnancy at 10 weeks or longer (i.e., the trial effectiveness measure) for each of the two treatments. The latter value was calculated by dividing the per patient cost by rate of ongoing pregnancy in the clinical trial. Confidence intervals were generated by using nonparametric bootstrapping. Role of the Funding Source The study sponsor allowed access to the clinical trial database and provided funds for analysis of the data to be undertaken. A report of the findings and this article were written independently and were not influenced or restricted by the sponsor in any way. The decision to seek publication for the article and the choice of journal was made by the authors. Ongoing pregnancy Total from the clinical trial Mean proportion of patients (95% CI) 0.22 ( ) 0.19 ( ) RESULTS Demographic Characteristics Table 2 shows demographic characteristics of the patients. Three hundred ninety-six patients received highly purified and 385 received recombinant FSH. Patients were well matched for all criteria. Effectiveness and Safety Data The treatments did not differ significantly in terms of the rate of ongoing pregnancies of greater than 10 weeks duration (Table 3). Both treatments were well tolerated, with few severe adverse events. The full clinical results are reported elsewhere (5). Medical Resource Use Table 4 shows the mean ( SD) values for use of medical resources. The number of clinic visits, ultrasonograms, and procedures per patient were all slightly higher in the highly purified group than the recombinant FSH group. Ninety-two percent of the highly purified group compared with 88% of the recombinant FSH group remained in the study long enough to undergo oocyte retrieval. These differ- TABLE 4 Medical resource use data for patients participating in clinical study. Recombinant FSH Ampoules of study medication GnRH depot injections GnRH daily injections Mean number of injections Daily GnRH 27 (8) 27 (7) hcg Clinic visits 7.7 (1.9) 7.4 (2.1) Clinic visits involving injections 3.8 (0.8) 3.7 (0.9) Number of ultrasonograms 4.4 (1.2) 4.3 (1.4) Blood tests 4.1 (1.0) 4.0 (1.2) Other diagnostic tests 1.7 (1.3) 1.5 (1.3) Proportion of patients undergoing oocyte retrieval Proportion of patients who had IVF Proportion of patients who had ICSI Proportion of patients who had implantation procedures s related to adverse events Hospitalization days Days of concomitant medication use Note: Data with the plus/minus sign are means ( SD). FERTILITY & STERILITY 1111

5 TABLE 5 Cost-minimization analysis. Recombinant FSH Total per patient cost (95% CI) ( ) 2,423 (2,356 2,495) 2,745 (2,658 2,830) Total cost per ongoing pregnancy (95% CI) ( ) 10,781 (9,056 12,919) 14,284 (11,883 17,891) ences were not statistically significant. Cost-Minimization Results Table 5 shows the mean cost per patient and the cost per ongoing pregnancy ( 10 weeks). The mean cost per patient in the recombinant FSH group was 2,745 (95% CI, 2,658 to 2,830), which was significantly higher than that for patients receiving highly purified ( 2,423 [95% CI, 2,356 to 2,495]). The mean per patient cost was significantly lower with highly purified than with recombinant FSH ( 322 [95% CI, 208 to 435]; P.001). This difference is mostly due to the lower acquisition cost of highly purified compared to recombinant FSH, which was more than enough to compensate for the slightly higher costs of procedures among highly purified recipients. The cost per ongoing pregnancy was 10,781 (95% CI, 9,056 to 12,919) with highly purified and 14,284 ( 11,883 to 17,891) with recombinant FSH. DISCUSSION Although highly purified and recombinant FSH were found to be equally effective per cycle in terms of the number of ongoing pregnancies arising from their use, we found that highly purified was significantly less expensive. Because our center has a fixed budget, use of highly purified instead of recombinant FSH would translate into a 13% increase in the number of cycles that could be offered. Unlike many other economic evaluations in assisted reproduction, our findings are based on analysis of a large prospective randomized clinical trial. Previous studies have evaluated both costs and effects on the basis of economic models (6 10), were cost studies that did not enroll sufficient patients to address the relative efficacy of procedures (3, 11), or have been comparisons of IUI with IVF in patients with idiopathic or mild male factor infertility (12, 13). In a previous modeled analysis using effectiveness data from a systematic review, routine NHS data, and United Kingdom public and private costs, IVF was the most costeffective option in severe tubal disease and severe endometriosis, compared with other treatment protocols (6). In our analysis, effectiveness is based on a large prospective randomized study; we therefore believe that these findings are the best available source of robust evidence relating to the cost-effectiveness of these drugs in ovarian stimulation. The third guideline on assisted fertility treatment produced by the Royal College of Obstetricians and Gynaecologists (1) suggested that recombinant FSH might offer additional benefit over highly purified gonadotropins of natural origin, reflecting the best available evidence at the time. Hypothesized advantages of recombinant products include higher purity, lower batch variance, and reduction of reliance on products of human origin, and meta-analysis has found that the pregnancy rate is significantly higher with recombinant FSH than with urinary FSH (14). However, the study on which we based our analysis (5), which compared highly purified and recombinant FSH, found no significant difference between the agents in terms of ongoing pregnancy rate despite adequate power to do so. Given the high costs of conducting such studies, clear evidence of differential effectiveness is unlikely to emerge in the near future. Previous economic analyses have used economic models to extrapolate from available evidence and estimate consequences over three cycles of therapy (8 10). The analysis considered only a single cycle of IVF/ICSI. Given that the live birth rate declines for repeated cycles, extrapolating these results for subsequent cycles might be misleading. Because the study was inadequately powered to demonstrate significant differences in effectiveness depending on the number of previous ART cycles, subgroup analyses for this purpose would not be a robust use of the data. The inability of the data we have to address the issue of repeated cycles remains a limitation of the analysis. Patients in the study received a mixture of IVF/ICSI procedures and a mixture of depot and daily GnRH agonists. The sample was stratified for IVF/ICSI, and both treatment arms had a similar number of ICSI and IVF procedures; therefore, it is unlikely that procedure use had a differential effect on one treatment over another. Because the study was not designed to compare these treatments, further studies will be needed to evaluate the relative cost-effectiveness of the two procedures. The appropriateness of using pooled medical resource use from six European countries for a United Kingdom based analysis might be questioned. However, the use of a common protocol in all participating countries ensured consistency in approach to patient management. As part of our sensitivity analysis, an alternative determination was performed by using aggregate United Kingdom charges per cycle (i.e., per patient charges) instead of bottom-up costing for the individual procedures performed. The results were consistent with those of the bottom-up analysis, suggesting that the aggregate pattern of resource use activity 1112 Lloyd et al. Economics of assisted reproduction Vol. 80, No. 5, November 2003

6 FIGURE 2 Sensitivity analysis of the relationship between discount rates for study medication and cost per cycle. information on the cost as well as the potential outcomes of treatment is important. Future research could proceed in a number of directions. The economics of repeated cycles could be assessed if appropriate data from randomized trials become available. In conclusion, recombinant FSH and highly purified were equally effective, but treatment with highly purified was significantly less expensive for the United Kingdom NHS. In our center, this would translate into an increase of approximately 13% in the number of cycles that can be offered within NHS budgets. did not materially differ from that in normal practice in the United Kingdom. In the United Kingdom, hospitals negotiate directly with drug companies and receive discounts from the list prices for some drugs. For agents that have several competing suppliers and at larger centers that have substantial buying power, these discounts can be substantial. However, the degree of discount is not, in general, freely available public information. We addressed this issue by conducting a sensitivity test that examined the cost per ongoing pregnancy at different discount rates. Figure 2 shows the economic cost per ongoing pregnancy at various rates of discount from list price. The cost of a single cycle is slightly higher than that reported by Phillips et al. (6). Those investigators estimated the cost per cycle to be 1,717 (corrected to 2001 costs), whereas our estimate was 2,423 for highly purified and 2,745 for recombinant FSH. This difference may be explained in part by the allowance for research and overheads made in charges from our Trust, or by different treatment of study medications in the two trials. However, the applicability of the results should be considered in the light of potentially different cost structures at other centers. The perspective of this analysis was the United Kingdom NHS, and we considered charges to health authorities. However, in the United Kingdom, patients commonly finance assisted reproduction privately, often at considerable cost. Although some providers may offer NHS purchasers discounted prices subsidized by private patient income, the results presented should be broadly applicable to privately funded patients, particularly those managed on a not for profit basis. We believe that offering these patients clear Acknowledgements: The authors thank the investigators and patients from the clinical study, who generated the data on which our evaluation is based, and Jens-Pierre Quartarolo, M.D., Ferring; Tony Waters, Ph.D., Ferring; Loretto Lacey, M.A., Lacey Consulting Ltd.; Alex Sexton, M.A., Fourth Hurdle; and Pippa Anderson, M.Sc., Fourth Hurdle, for support. References 1. Royal College of Obstetricians and Gynaecologists. National evidencebased clinical guidelines. The management of infertility in tertiary care. January Available at PageID 108&GuidelineID Collins J. Cost-effectiveness of in vitro fertilization. Semin Reprod Med 2001;19: Granberg M, Wikland M, Hamberger L. Financing of IVF/ET in the Nordic countries. Acta Obstet Gynecol Scand 1998;77: Human Fertilisation and Embryology Authority (HFEA) Annual report and accounts. Available at Annual_Report/HFEA_Annual%20Report2002.pdf. 5. The European and Israeli Study Group on Highly Purified Menotropin versus Recombinant Follicle-Stimulating Hormone. Efficacy and safety of highly purified menotropin versus recombinant follicle-stimulating hormone in in vitro fertilization/intracytoplasmic sperm injection cycles: a randomized comparative trial. Fertil Steril 2002;78: Phillips Z, Barraza-Llorens M, Posnett J. Evaluation of the relative cost-effectiveness of treatments for infertility in the UK. Hum Reprod 2000;15: Mantovani LG, Belisari A, Szucs TD. Pharmaco-economic aspects of in vitro fertilization in Italy. Hum Reprod 1999;14: Daya S, Ledger W, Auray JP, Duru G, Silverberg K, Wikland M, et al. Cost-effectiveness modelling of recombinant FSH versus urinary FSH in assisted reproduction techniques in the UK. Hum Reprod 2001;16: Silverberg K, Daya S, Auray JP, Duru G, Ledger W, Wikland M, et al. Analysis of the cost effectiveness of recombinant versus urinary follicle-stimulating hormone in in vitro fertilization/intracytoplasmic sperm injection programs in the United States. Fertil Steril 2002;77: Sykes D, Out HJ, Palmer SJ, van Loon J. The cost-effectiveness of IVF in the UK: a comparison of three gonadotropin treatments. Hum Reprod 2001;16: Poulain P, Le Lannou D, Le Martelot MT, Vialard J, Chevalier R, Corvez A. Socioeconomic approach to the practice of in vitro fertilization in the Brittany region in Contracept Fertil Sex 1995;23: Zayed F, Lenton EA, Cooke ID. Comparison between stimulated in vitro fertilization and stimulated intrauterine insemination for the treatment of unexplained and mild male factor infertility. Hum Reprod 1997;12: Goverde AJ, McDonnell J, Vermeiden JP, Schats R, Rutten FF, Schoemaker J. Intrauterine insemination or in vitro fertilisation in idiopathic subfertility and male subfertility: a randomised trial and cost-effectiveness analysis. Lancet 2000;355: Daya S, Gunby J. Recombinant versus urinary follicle stimulating hormone for ovarian stimulation in assisted reproduction. Hum Reprod 1999;14: FERTILITY & STERILITY 1113