Article HMG versus rfsh for ovulation induction in developing countries: a cost effectiveness analysis based on the results of a recent meta-analysis

Transcription

1 RBMOnline - Vol 12. No Reproductive BioMedicine Online; on web 19 December 2005 Article HMG versus rfsh for ovulation induction in developing countries: a cost effectiveness analysis based on the results of a recent meta-analysis Dr Hesham Al-Inany obtained his MD in 1998 from Cairo University, Egypt. He is presently Assistant Professor at The Department of Obstetrics and Gynecology, Cairo University and IVF consultant in the Egyptian IVF-ET centre, Maadi, Cairo. Dr Al-Inany has been an Editor in the Cochrane Menstrual Disorders SubFertility Cochrane group since 2002 and has been a member of the Editorial Board of the British Journal of Obstetrics and Gynecology since He has been author or co-author of more than 50 research papers in both national and international journals. Dr Hesham Al-Inany Hesham G Al-Inany 1,2,3, Ahmed M Abou-Setta 1, Mohamed A Aboulghar 1,2, Ragaa T Mansour 1, Gamal I Serour 1 1 The Egyptian IVF-ET Centre, Maadi; 2 Cairo University, Department of Obstetrics and Gynecology, Cairo, Egypt 3 Correspondence: 8 Moustapha Hassanin Street, Manial, Cairo, 11451, Egypt. kaainih@link.net Abstract Both cost and effectiveness should be considered conjointly to aid judgments about drug choice. Therefore, based on the results of a recent published meta-analysis, a Markov model was developed to conduct a cost effectiveness analysis for estimation of the cost of an ongoing pregnancy in IVF/intracytoplasmic sperm injection (ICSI) cycles. In addition, Monte Carlo micro-simulation was used to examine the potential impact of assumptions and other uncertainties represented in the model. The results of the study reveal that the estimated average cost of an ongoing pregnancy is 13,946 Egyptian pounds (EGP), and 18,721 EGP for a human menopausal gonadotrophin (HMG) and rfsh cycle respectively. On performing a sensitivity analysis on cycle costs, it was demonstrated that the rfsh price should be 0.61 EGP/IU to be as cost-effective as HMG at the price of 0.64 EGP/IU (i.e. around 60% reduction in its current price). The difference in cost between HMG and rfsh in over 100,000 cycles would result in an additional 4565 ongoing pregnancies if HMG was used. Therefore, HMG was clearly more cost-effective than rfsh. The decision to adopt a more expensive, cost-ineffective treatment could result in a lower number of cycles of IVF/ICSI treatment undertaken, especially in the case of most developing countries. Keywords: cost effectiveness, HMG, infertility, Markov model, meta-analysis, recombinant FSH Introduction The number of couples requiring advanced fertility treatment is increasing, and hence economic evaluation of different therapeutic interventions is important as the demand for services increases (Daya, 2000). This is especially important if one considers that despite remarkable advances in infertility understanding and its management, success is limited and multiple treatment cycles may be needed to achievea successful pregnancy. Ovarian stimulation represents an excellent example. Multifollicular development via gonadotrophin administration is still an integral component for ovarian stimulation in IVF/ intracytoplasmic sperm injection (ICSI) cycles (Keck et al., 2005). The technological developments of pharmaceutical gonadotrophins over the last 40 years have shown improvements in specific activity, purity, degradation and impurities (Bassett and Driebergen, 2005). In the market, there are two major players: recombinant FSH (rfsh) and human menopausal gonadotrophin (HMG). Both effectiveness and costs should be considered together to aid judgment about whether one drug should be preferred over another (Wandwalo et al., 2005). Recombinant FSH has been available for almost 10 years (Out, 2005), but despite its proven efficacy, the relatively high cost 163

2 164 of rfsh as compared with HMG has an impact on consumption (Sykes et al., 2001), especially in developing countries (Dyer, 2002), where its price (150 Egyptian pounds for 75 IU rfsh) is almost three times that of HMG (50 Egyptian pounds for 75 IU). In developing countries, although the cost of one trial of IVF is much lower than that in Europe, the majority of people cannot afford this treatment because of the low per capita income (Aboulghar, 2005). Even if they can afford one IVF cycle, it is documented that for women who do not pursue a second IVF cycle after the first fails, the major reason was financial (Gol et al., 1997). In a recently published meta-analysis, it was demonstrated that there was no significant difference between HMG and rfsh regarding different IVF outcomes (Al-Inany et al., 2005) with special concern to the most important outcome: live birth/ongoing pregnancy rate (OR: 1.21, 95% CI ). Based on results of this meta-analysis, it was decided to conduct a cost effectiveness analysis for estimation of the cost of an ongoing pregnancy in an IVF/ICSI cycle from a perspective of a developing country: Egypt. Furthermore, this model is both representative of other developing countries in which the medical insurance and/ or government support for infertility treatment is markedly limited. Materials and methods The recurring nature of IVF/ICSI cycles dictated the building of a state transition model, also called a Markov model (Figure 1) (Tree Age Pro 2005 Software Inc., Williamstown, MA, USA) to simulate the IVF treatment cycle with its key steps to examine the costs and effectiveness of rfsh versus HMG. Monte Carlo micro-simulation was also used to examine the potential impact of assumptions and other uncertainties represented in the model. The transition probabilities for different health states used in the present analysis (Figure 1) were based on pooled outcome data that were collected from recently published meta-analysis (Al- Inany et al., 2005) (Table 1). The meta-analysis included eight truly randomized controlled trials (Jansen et al., 1998; Gordon et al., 2001; Ng et al., 2001; Strehler r et al., 2001; Westergaard et al., 2001; European and Israeli Study Group, 2002; Balasch et al., 2003; Kilani et al., 2003) showing no significant difference between recombinant FSH and HMG regarding different outcomes (ongoing pregnancy/live birth rate (OR 1.18; 95% CI ), clinical pregnancy rate (OR 1.2; 95% CI ), miscarriage rate (OR 1.2; 95% CI ), multiple pregnancy rate (OR 1.35; 95% CI ) and incidence of moderate/severe ovarian hyperstimulation (OHSS, OR 1.79; 95% CI ). There was a significant reduction in the amount of gonadotrophins utilized in favour of HMG over recombinant FSH (OR 317.8; 95% CI to 289.0) (Al- Inany et al., 2005). Drug costs for rfsh and HMG were based on the most recent wholesale acquisition costs for the two agents in Egypt obtained from the Egyptian Ministry of Health (i.e. retail cost). Although these are the official prices, it should be noted that there is a variation in these prices between centres (i.e. actual price). All other costs corresponding to the relevant health states were taken (Table 2) and included costs incurred for ovarian stimulation, monitoring, oocyte retrieval, laboratory procedures, and luteal phase support and pregnancy determination. The costs considered were for the year 2005, and were obtained from the Egyptian IVF ET Centre, Maadi, Cairo, Egypt. These were the charges made to the patients. Figure 1 represents the first complete cycle with all probabilities commencing with ovarian stimulation to ovum retrieval (i.e. yes or no); oocytes recovered (i.e. yes or no); IVF or ICSI; the subsequent health states are identical for each of IVF and ICSI); fertilization (i.e. yes or no); embryo transfer (i.e. yes or no); and ending with either clinical pregnancy or negative βhcg. If the patient became pregnant, then either ongoing pregnancy or miscarriage was obtained. If no pregnancy occurred, then either the patient restarted the cycle or discontinued for non-medical reasons; the probability of which varied by cycle number. Probability of discontinuation at the end of the cycle (failed clinical pregnancy) was obtained from another recent study, as it was not possible to obtain it from the meta-analysis (Schroder r et al., 2003). The probability of discontinuation after the 1st cycle is 0.489, after the second cycle would reach and after the 3rd cycle is (Schroder r et al., 2003). Patients ending with miscarriage were re-entered into the cycle. Probabilities were again drawn from the meta-analysis (Al-Inany et al., 2005). The OHSS complication is independent of IVF outcome. Hence, the cost incurred for OHSS management was placed early in the cycle to accumulate them regardless of where the cycle may end (i.e. ongoing pregnancy, re-start, or stopping IVF). Patients, who cancelled early in the cycle for failed stimulation were re-entered into the start cycle state, re-incurring the costs of ovarian stimulation. It was decided to run a Markov model for three cycles, as many patients pursue a live birth through infertility treatment over a long treatment course, sometimes up to 10 or more cycles of treatment. However, financial and other personal costs often limit most patients to only three cycles of treatment. Then, to perform the analysis, a virtual population of 100,000 patients (the`markov cohort ) was treated in the computer simulation of assisted reproduction treatment in Monte Carlo simulations. These large numbers of patients and simulations provided a high degree of statistical accuracy and allowed confidence limits around the outcome estimates to be generated with precision. Results Running the Markov model through three treatment cycles for rfsh resulted in the individual s probability of ending at restarting the cycle of 6.6%, an ongoing pregnancy of 35.9%, and in discontinuing IVF of 57.5% (Figure 2). For HMG, by the end of the 3rd cycle, the individual s probability of ending at re-starting the cycle was 6%, of ongoing pregnancy 40.8%, and of discontinuing IVF 53.2% (Figure 3). On running Monte Carlo Simulation (100,000 runs, random allocation, transition probabilities used as a guide), the HMG cycle cost on average was 13,946 ± 6512 EGP per ongoing pregnancy, versus an average of EGP 18,721 ± 8749 per ongoing pregnancy for rfsh (Table 3). On repeating the calculations assuming identical effectiveness for both treatments of 25% of ongoing life birth rate, and then re-examining their cost effectiveness, an ongoing pregnancy in HMG cycles would cost 9297 ± 1058 EGP, whilewith rfsh it would cost 13,118 ± 2541 EGP, still in favour of HMG (Table 3). Assuming discontinuation rate of 0%, cost of ongoing pregnancy in recfsh arm would be 43,287EP and that of hmg arm would reach 36,093EP.

3 On performing sensitivity analysis on cost of HMG versus rfsh, it was found that the rfsh price should be 0.61 EGP/IU to be as cost-effective as HMG at the price of 0.64 EGP/IU (i.e. around 60% reduction in its current price) (Figure 4). The total cost of 100,000 cycles in the HMG arm is 1,392,633, EGP. The total cost of 100,000 cycles in the rfsh arm is 1,870,117, EGP, with the difference estimated (i.e. rfsh HMG) equalling 477,483,412 EGP. The number of HMG cycles that can be performed using this difference (i.e. 477,483,412/13,946) is 34,238 cycles, with 4565 more ongoing pregnancies if HMG was used. Figure 1. The IVF cycle state transition (Markov) model. 165

4 Table 1. IVF transition probabilities. Name Description Value P_cancel_HMG a Probability of failed stimulation or fertilization in vitro HMG P_cancel_rFSH a Probability of failed stimulation or fertilization in vitro rfsh P_clin_preg_HMG a Probability of clinical pregnancy HMG P_clin_preg_rFSH a Probability of clinical pregnancy rfsh P_miscar_HMG a Probability of miscarriage HMG P_miscar_rFSH a Probability of miscarriage rfsh P_OHSS_HMG a Probability of OHSS HMG P_OHSS_rFSH a Probability of OHSS rfsh a Al-Inany et al., 2005; Sykes et al., HMG = human menopausal gonadotrophin, OHSS = ovarian hyperstimulation syndrome, rfsh = recombinant FSH. Table 2. IVF model costs and units. Name Description Value EGP Name Description Value EGP C_oocyte retrieval Oocyte retrieval 4000 c_rfsh a Cost of rfsh unit 1.93 C_Fertilization Lab procedure 500 u_rfsh b Units rfsh required 2192 ± C_Trans Embryo transfer 1000 c_hmg a Cost of HMG unit 0.64 C_OHSS OHSS management 4000 u_hmg b Units HMG required 2077 ± ( ,000) C_Miscar Miscarriage 1,000 ± 500 c_gnrha a GnRHa unit cost 35 C_βHCG One βhcg test 50 u_gnrha b Units GnRHa required 11 C_antenatal_care Antenatal care 500 c_lp_support a Leuteal phase support 1400 a Ministry of Health list price. b Al-Inany et al., 2005; Sykes et al., EGP = Egyptian pounds, GnRHa = gonadotrophin-releasing hormone analogues, βhcg = β human chorionic gonadotrophin, HMG = human menopausal gonadotrophin, OHSS = ovarian hyperstimulation syndrome, rfsh = recombinant FSH. 166 Figure 2. rfsh: By the end of the 3rd cycle, the individual s probability of ending at re-starting the cycle is 6.6%, an ongoing pregnancy is 35.9%, and in discontinuing IVF is 57.5%. Figure 3. Human menopausal gonadotrophin: by the end of the 3rd cycle, the individual s probability of ending at restarting the cycle is 6%, in ongoing pregnancy is 40.8%, and in discontinuing IVF is 53.2%.

5 Table 3. Monte Carlo simulation human menopausal gonadotrophin (HMG) versus recombinant FSH (rfsh) cycle cost statistics per ongoing pregnancy and assuming identical effectiveness. Parameter HMG rfsh Parameter HMG rfsh Using transitional probabilities Assuming identical effectiveness Mean 13,926 18,701 Mean ,118 SD ±6512 ±8749 SD ±1058 ±2541 Minimum ,566 Minimum ,566 Maximum 36,093 43,247 Maximum 12,593 21,297 25th percentile , percentile ,566 Median ,116 Median ,066 75th percentile 27,043 35, percentile 12,093 20, ,000 runs, random allocation. Prices in Egyptian pounds. Figure 4. Sensitivity analysis on cost of human menopausal gonadotrophin (HMG) versus recombinant FSH (rfsh). Discussion The construction of economic models in an IVF/ICSI setting involves multiple decision points and many necessary assumptions in building realistic models of a quite complex care process (Barlow, 2001). The uncertainty plays an important role in dealing with assumptions, and it is important in terms of validity and acceptability. For this reason, the best available evidence from practice should always support decision models. The current Markov model was based on results of the most recently published systematic review and meta-analysis, which is considered the most thorough analysis in evidence-based medicine (Al-Inany et al., 2005). In addition, the results of a meta-analysis usually reflect a wide range of clinical settings and patient characteristics; thus, the results can be broadly generalized among typical patients. Infertility management is not, and probably will not be covered at any time in the near future, by the health authorities in developing countries (Aboulghar, 2005). Even in developed countries, where the government partly or completely covers infertility treatment, cost effectiveness is of utmost importance (Gerli et al., 2005). The contribution of the cost of medication to the overall cost of assisted reproduction is significant. In the medical literature, there are a number of cost effectiveness analyses comparing rfsh versus HMG (Daya et al., 2001; Sykes et al., 2001; Lloyd et al., 2003; Silverberg et al., 2003; Hatoum et al., 2005), with a number of similarities: direct involvement of pharmaceutical companies, model gonadotrophin drug therapies over three cycles of IVF, using expert clinical panels to provide estimates of outcomes (except Hatoum et al., 2005), all compared recombinant versus urinary FSH (except Lloyd et al., 2003). Differences in their approach include the choice of evidence and information on which they based their modelling. The present analysis is distinguished from those mentioned above being based on a recently published metaanalysis (Al-Inany et al., 2005) (assumptions are best evidence and bias is minimized) with no involvement of pharmaceutical companies. As can be seen, the HMG arm exerts absolute dominance over the rfsh arm (Figures 2 and 3). Further analyses, such as incremental cost effectiveness and threshold analysis, were therefore not possible. 167

6 168 One may argue that if there is no significant difference between the drugs and one is cheaper than the other, then it would be logical that HMG is more cost-effective, and there is no need for such a complex analysis. However, IVF/ICSI cycles involve numerous steps. Each has its outcome probabilities and associated uncertainties. Therefore, it cannot simply be judged that the least costly treatment is the most effective, or vice versa. Uncertainty plays a major role here. Moreover, it is important to know precisely how large the difference between the two drugs is. Some may also argue that the meta-analysis from which the transition probabilities were drawn found no statistically significant difference between HMG versus rfsh. It was decided to repeat the calculations, assuming identical effectiveness of both treatments of 25% of ongoing life birth rate, and reexamine their cost effectiveness (Table 3). The results were in the same direction, although the difference was less. From the calculation shown in Figure 4, price of recombinant FSH should be reduced by more than 50% to be equivalent to HMG in its cost effectiveness. The current cost effectiveness analysis did not consider cryocycles in the calculations. It was clearly stated that the Markov model was built on assumptions from meta-analysis (Al- Inany et al., 2005). The meta-analysis did not consider cryocycles because most trials included did not report cryo-cycles. Accordingly, it was not possible to reach evidence-based assumptions regarding this specific issue. In Egypt, only around 50% of Egyptian infertile couples could afford to pay for IVF (Serour et al., 1991). It is unlikely that Egypt, as a developing country with some private health care structures in place to offer tertiary level infertility care, stands alone in this matter. It was shown that in many developing countries around 10% of all women s visits to doctors (all categories of medical doctors) are related to problems of childlessness; Inhorn et al., 1991). Whereas the real cost for each live IVF birth in the US is estimated at about $50,000, in a developing country setting, that cost could reach as high as $100,000 (Okonofua et al., 1991). If IVF services are provided entirely by the private sector, then new reproductive technologies will benefit only a small proportion of infertile women, primarily an elite group, who can afford the costs associated with this technology (Inhorn et al., 1991). Even among those who can afford treatment, if couples feel that they can only afford one cycle they are more likely to press for multiple embryo transfer, with the resultant additional neonatal costs of multiple pregnancy. In conclusion, from an economic evaluation point of view, HMG was more cost-effective than rfsh. The decision to adopt a more expensive treatment could result in a lower number of cycles of IVF/ICSI treatment especially if patients are paying for it. References Aboulghar MA 2005 The importance of fertility treatment in the developing world. British Journal of Obstetrics and Gynaecology 112, Al-Inany H, Aboulghar MA, Mansour RT, Serour GI 2005 Ovulation induction in the new millennium: recombinant follicle-stimulating hormone versus human menopausal gonadotropin. Gynecological Endocrinology 20, Balasch J, Penarrubia J, Fabregues F et al Ovarian responses to recombinant FSH or HMG in normogonadotrophic women following pituitary desensitization by a depot GnRH agonist for assisted reproduction. Reproductive BioMedicine Online 7, Barlow DH 2001 Cost-effectiveness modelling. Human Reproduction 16, Bassett RM, Driebergen R 2005 Continued improvements in the quality and consistency of follitropin alfa, recombinant human FSH. Reproductive BioMedicine Online 10, Daya S 2000 Cost-effective, evidence-based infertility care. Current Opinion in Obstetrics and Gynecology 12, Daya S, Ledger W, Auray JP et al Cost-effectiveness modelling of recombinant FSH versus urinary FSH in assisted reproduction techniques in the UK. Human Reproduction 16, Dyer SJ 2002 The conflict between effective and affordable health care a perspective from the developing world. Human Reproduction 17, European and Israeli Study Group on Highly Purified Menotropin versus Recombinant Follicle-Stimulating Hormone 2002 Efficacy and safety of highly purified menotropin versus recombinant follicle-stimulating hormone in in vitro fertilization/ intracytoplasmic sperm injection cycles: a randomized, comparative trial. Fertility and Sterility 78, Gerli S, Casini ML, Unfer V et al Ovulation induction with urinary FSH or recombinant FSH in polycystic ovary syndrome patients: a prospective randomized analysis of cost-effectiveness. Reproductive BioMedicine Online 9, Gol J, Austin C, Lisbona H et al Factors influencing patients decision not to repeat IVF. Journal of Assisted Reproduction and Genetics 14, Gordon UD, Harrison RF, Fawzy M et al A randomized prospective assessor-blind evaluation of luteinizing hormone dosage and in vitro fertilization outcome. Fertility and Sterility 75, Hatoum HT, Keye WR Jr, Marrs RP et al A Markov model of the cost-effectiveness of human-derived follicle-stimulating hormone (FSH) versus recombinant FSH using comparative clinical trial data. Fertility and Sterility 83, Inhorn MC 2003 Global infertility and the globalization of new reproductive technologies: illustrations from Egypt. Social Science and Medicine 56, Jansen CA, van Os HC, Out HJ, Coelingh Bennink HJ 1998 A prospective randomized clinical trial comparing recombinant follicle stimulating hormone (Puregon) and human menopausal gonadotrophins (Humegon) in non-down-regulated in-vitro fertilization patients. Human Reproduction 13, Keck C, Bassett RM, Ludwig M 2005 Endocrinology in ovarian stimulation. Factors influencing response to ovarian stimulation. Reproductive BioMedicine Online 11, Kilani Z, Dakkak A, Ghunaim S et al A prospective, randomized, controlled trial comparing highly purified HMG with recombinant FSH in women undergoing ICSI: ovarian response and clinical outcomes. Human Reproduction 18, Lloyd A, Kennedy R, Hutchinson J, Sawyer W 2003 Economic evaluation of highly purified menotropin compared with recombinant follicle-stimulating hormone in assisted reproduction. Fertility and Sterility 80, Ng EH, Lau EY, Yeung WS, Ho PC 2001 HMG is as good as recombinant human FSH in terms of oocyte and embryo quality: a prospective randomized trial. Human Reproduction 16, Okonofua F 1996 The case against new reproductive technologies in developing countries. British Journal of Obstetrics and Gynaecology 103, Out H 2005 Recombinant follicle-stimulating hormone: gold standard or not? Reproductive BioMedicine Online 11, Schroder AK, Katalinic A, Diedrich K, Ludwig M 2004 Cumulative pregnancy rates and drop-out rates in a German IVF programme: 4102 cycles in 2130 patients. Reproductive BioMedicine Online 8,

7 Serour GI, El Ghar M, Mansour RT 1991 Infertility: a health problem in the Muslim world. Population Sciences 10, Silverberg K, Daya S, Auray JP 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. Fertility and Sterility 77, Strehler E, Abt M, El-Danasouri I et al Impact of recombinant follicle-stimulating hormone and human menopausal gonadotropins on in vitro fertilization outcome. Fertility and Sterility 75, Sykes D, Out HJ, Palmer SJ, van Loon J 2001 The cost-effectiveness of IVF in the UK: a comparison of three gonadotrophin treatments. Human Reproduction 16, Wandwalo E, Robberstad B, Morkve O 2005 Cost and costeffectiveness of community based and health facility based directly observed treatment of tuberculosis in Dar es Salaam, Tanzania. Cost Effectiveness and Resource Allocation 3, 6. Westergaard LG, Erb, Laursen SB et al Human menopausal gonadotropin versus recombinant follicle-stimulating hormone in normogonadotropic women down-regulated with a gonadotropinreleasing hormone agonist who were undergoing in vitro fertilization and intracytoplasmic sperm injection: a prospective randomized study. Fertility and Sterility 76, Received 30 September 2005; refereed 21 October 2005; accepted 15 November