Department of Obstetrics and Gynecology, McMaster University, Hamilton and Department of Obstetrics and Gynaecology, Dalhousie University, Halifax, Canada
1 To whom correspondence should be addressed at: 400 Maders Cove Road, RR No.1 Mahone Bay, Nova Scotia, Canada, B0J 2E0. E-mail: collinsj{at}auracom.com
![]() |
Abstract |
---|
Key words: cost-effectiveness/gonadotrophins/intrauterine insemination/IVF
![]() |
Introduction |
---|
![]() |
What is the clinical setting of the question? |
---|
Faced with the need for empiric treatment, clinicians have devised treatment plans involving clomiphene citrate (CC), IUI or both (CC/IUI). Three relevant Level I studies (randomized controlled trials; RCTs) in which 1471 women were enrolled, evaluated CC treatment (Harrison and OMoore, 1983; Fisch et al., 1989
; Glazener et al., 1990
). One other trial involved a control group with normal fertility (Fujii et al., 1997
). The relative likelihood of pregnancy was 1.9-fold higher with clomiphene treatment than placebo, but this significant difference has no more than marginal clinical importance. A significant relative difference might represent either an extremely small or an extremely large actual difference: a relative risk of 1.9 when the baseline risk is ten per million would have a very different clinical meaning if the baseline risk were 10%. Thus, for clinical purposes it is better to make use of the simple difference in rates between alternative treatments (risk difference or RD) and the inverse of RD, which is the number needed to treat (NNT). The NNT estimates how many times the intervention would be used before there was a single additional event with treatment compared with control. These absolute measures (RD and NNT) directly reflect the experience of patients. From the clomiphene trials, on average one additional pregnancy would occur in 40 cycles (95% CI 22201) of CC compared with placebo treatment (The ASRM Practice Committee, 2000
). IUI treatment is not much better: from the combined data of two RCTs of IUI treatment involving 1691 cycles, NNT was 37 cycles (95% CI 23101) (Kirby et al., 1991
; Guzick et al., 1999
).
Data on the combination of CC/IUI compared with no treatment are available from one crossover RCT involving 51 women (Deaton et al., 1990). The pregnancy rates in the first phase of four cycles were 35% in the CC/IUI group and 15% in the control group. Considering all 298 cycles, the NNT for CC/IUI treatment was 16 cycles (95% CI 9165). Thus the evidence about the effectiveness of early treatment of unexplained infertility is based on a limited number of small RCTs; at the present time that evidence suggests that sensible treatment plans should begin with three to six cycles of CC/IUI therapy because treatment with CC or IUI alone appears to be inefficient. No high quality medical care research could be found that evaluates the effectiveness of extended treatment programmes lasting more than six cycles.
Therefore, low-tech treatment of unexplained infertility is effective in only a minority of couples and many of the unsuccessful couples would then wish to consider advanced therapy involving gonadotrophin stimulation and IUI (FSH/IUI) or IVF or both. Both treatments are complex and costly, so knowing the best approach is an important objective for clinical research.
![]() |
What is the economical setting of the question? |
---|
It is not surprising, then, to find Level II reports (cohort studies) indicating that FSH/IUI has a lower cost per pregnancy or live birth and that FSH/IUI is more cost-effective than IVF. In a 1997 Midwestern United States cohort study involving all diagnoses, costs per delivery were $10 282 and $37 028, respectively, for gonadotrophin/IUI cycles and ART cycles including IVF, gamete intra-Fallopian transfer (GIFT) and zygote intra-Fallopian transfer (ZIFT) (van Voorhis et al., 1998). In a UK modelling study published in 2000, involving four cycles of treatment, costs per pregnancy in couples with unexplained infertility were £3681 and £8770 respectively, for stimulated IUI and IVF (Philips et al., 2000
). If there were no further input, a policy of FSH/IUI cycles prior to IVF would appear to be prudent for couples with unexplained infertility who have not conceived after a trial of low-tech treatment.
![]() |
What evidence is relevant to the question? |
---|
First, it is important to choose the highest level of evidence about the effectiveness of FSH/IUI treatment. Several RCTs have compared gonadotrophin/IUI with various alternatives for the treatment of unexplained infertility; some of these Level I studies included a small proportion of cases with mild male infertility or treated endometriosis. Although not strictly free from diagnostic findings, the infertility in mild male and treated endometriosis cases also remains unexplained, and such couples usually are given the same treatment options as couples with no diagnostic findings (those with pure unexplained infertility). There is no good reason to believe that the effects of empiric treatments such as FSH/IUI or IVF differ because of the additional minimal diagnostic findings, considering the number and variety of unknown, undetected and unrecognized fertility defects that contribute to supposedly pure unexplained infertility.
Second, the cost-effectiveness of clinical interventions should be based on full economic evaluations involving a randomized evaluation of FSH/IUI treatment with a parallel assessment of costs. Such studies are not common but they are the best source of information on which to make judgements about relative cost.
Third, in order to address fairly the most relevant clinical outcome, treatment effects should be established relative to live birth rather than pregnancy or other surrogate outcomes. Furthermore, a singleton live birth is the outcome of greatest interest, and consequently the effects and costs should be calculated for this expression of the success rate. Effects and costs are often expressed per delivery, however, meaning that multiple birth rateswhich may differ between FSH/IUI and IVFhave to be subtracted from the overall live birth or delivery rates to estimate singleton birth rates.
Finally, in order to ensure that the observed differences are clinically as well as statistically significant, the study results should be expressed as NNT, a rendering of clinical study results which is currently as close as we are able to come to meeting the patients needs for digestible information (Cook and Sackett, 1995). Along the same lines, the analysis should be according to intention to treat: couples should be analysed within the group to which they were randomly allocated regardless of whether they made a different choice after randomization. The basis for using an intention to treat analysis is that pragmatic trials evaluate what happens in real clinical life, where discontinuation and changes in treatment plans are common events. The trials results will then be applicable to other clinical settings with all of their imperfections, and in particular will apply to couples at the time that they are considering whether to use a treatment, when their own future is as uncertain as that of the trial subjects at enrolment.
![]() |
Is FSH/IUI an effective treatment? |
---|
Guzick et al. (1999) compared intracervical insemination (ICI) alone, IUI alone, FSH/ICI and FSH/IUI in a randomized controlled trial with a factorial design. The authors described the study as a trial of treatment for unexplained infertility, (Guzick et al., 1998
) but some female partners had stage I or II endometriosis (if six months had elapsed after treatment) and 25% of the male partners had oligozoospermia. Also, the trial design assumes that ICI is equivalent to no treatment. In this trial, live birth rates per cycle and per couple were higher with FSH/IUI than with ICI. Table I presents results based on the intention to treat; all births during an assigned protocol of treatment, including conceptions during a cycle of rest, are included. Estimated singleton birth rates per cycle were 5.7 and 2.4% respectively, in FSH/IUI and ICI cycles.
|
By statistical standards, FSH/IUI is an effective treatment for persistent unexplained infertility, judged on the outcome of singleton live births. The effect is so small, however, that it becomes clinically appreciable only in association with a protocol of four cycles. Naturally this alters the cost-effectiveness, an issue that will be dealt with momentarily.
![]() |
Is FSH/IUI treatment better than CC/IUI? |
---|
Thus, while cycles of gonadotrophin treatment with IUI are marginally superior to no treatment, FSH/IUI has not yet been proven superior to CC/IUI in couples with unexplained infertility. A trial is needed to compare extended CC/IUI treatment (say three cycles) with a similar number of FSH/IUI treatment cycles in couples with unexplained infertility who have completed an initial three cycles of CC/IUI treatment for unexplained infertility without success.
![]() |
Is IVF treatment superior to FSH/IUI? |
---|
More couples withdrew from IVF cycles (42%) than FSH/IUI cycles (16%). The overall live birth rates per cycle and per couple were not significantly different with FSH/IUI or IVF. The calculated outcomes for couples with unexplained infertility are shown in Table II. Singleton live birth rates are based on the authors report that the multiple birth rates were 29 and 21% respectively, in FSH/IUI and IVF cycles. During an average of 3.5 years, the singleton live birth rates per cycle were 6 and 10% respectively, in the FSH/IUI and IVF groups. The singleton live birth rates per couple were 26 and 31% respectively, in the FSH/IUI and IVF groups. The clinical differences between FSH/IUI and IVF treatment of unexplained infertility were small; the NNTs were 25 cycles and 20 couples and they were not statistically significant. No trial to date appears to have been powered specifically for the comparison of IVF and FSH/IUI in unexplained infertility. Also, the reported trials involved an exceedingly small fraction of the patients worldwide undergoing treatment, so that generalizations have to be limited. Nevertheless, the best evidence available indicates that, at least in this diagnostic grouping, the difference in success rates between FSH/IUI and IVF may be quite small. Since the effects of FSH/IUI are known to be small, the true importance of the effect of IVF also may be small. In any case, there is a need for further study of the effectiveness of these advanced treatments.
|
![]() |
Is FSH/IUI treatment more cost-effective than IVF? |
---|
|
![]() |
Problems with the evidence |
---|
A related problem concerns the number of patients in these studies. If 20% of the couples undergoing IVF cycles have unexplained infertility, then the studies include a minute fraction of the volume of advanced treatment for unexplained infertility. The 184 and 86 IVF cycles in the Amsterdam and ESHRE trials, for example, represent only 0.2% of the total 140 600 IVF cycles in the United States and Europe in just one year, 1998 (Nygren and Nyboe Andersen, 2001; Society for Assisted Reproductive Technology & American Society for Reproductive Medicine, 2002
). Thus, so few of the patients in clinical practice are reported that one cannot determine whether the study subjects are representative.
A further problem is the lack of knowledge about prognostic factors beyond the usual surrogates for severity: duration and female partners age. No specific prognostic factors are known because the causes are both masked and multifactorial, so the severity of the disease cannot be graded. Variability among studies often is due to differences in the mix of severity among the subjects from study to study, but in this case there is no means of determining how that mix might affect the comparison of the effectiveness of two treatments.
Finally, an unidentifiable subset of couples with unexplained infertility may have conditions that cannot possibly respond to the effect of pre-implantation interventions. For the present, the treatment of unexplained infertility is limited to procedures that do no more than increase the probability of successful insemination. Numerous post-fertilization defects are imaginable that could contribute to this diagnostic grouping. Thus many patients and subjects of trials cannot conceive despite the use of empiric treatment. The presence of such unresponsive conditions reduces the power of RCTs to discriminate between effective and ineffective treatments.
Thus, much of the existing medical care evidence about the advanced treatment of unexplained infertility is valid, but it is not robust. Under such conditions, it is difficult for societies, associations and other authorities to formulate lasting guidelines because of the present uncertainty and the future potential for change as new data become available.
![]() |
Conclusions |
---|
Obviously, there is a need for more research: the important clinical questions call for a large international factorial study involving a no treatment arm as well as CC/IUI, FSH/IUI and IVF treatments. The study should have the power to address each of the six possible comparisons. While such a study might cost millions, if it is not done many more millions could be wasted on treatments that are not as effective as they now seem.
Inevitably, the results of such large definitive studies leave many unanswered important clinical questions. Observational studies and trials are needed to help sort out issues such as the optimal timing of treatment and the preferable order of treatments. Most important, however, is the need for scientific studies to elucidate some of the causes of unexplained infertility and discover specific treatments that are far more likely to be effective than the present choice of empiric remedies.
![]() |
References |
---|
Collins, J. (2002) An international survey of the health economics of IVF and ICSI. Hum. Reprod. Update, 8, 265277.
Collins, J.A., Burrows, E.A. and Willan, A.R. (1995) The prognosis for live birth among untreated infertile couples. Fertil. Steril., 64, 2228.[ISI][Medline]
Cook, R.J. and Sackett, D.L. (1995) The number needed to treat: a clinically useful measure of treatment effect. Br. Med. J., 310, 452454.
Crosignani, P.G., Walters, D.E. and Soliani, A. (1992) Addendum to the ESHRE multicentre trial: a summary of the abortion and birth statistics. Hum. Reprod., 7, 286287.[ISI][Medline]
Deaton, J.L., Gibson, M., Blackmer, K.M., Nakajima, S.T., Badger, G.J. and Brumsted, J.R. (1990) A randomized, controlled trial of clomiphene citrate and intrauterine insemination in couples with unexplained infertility or surgically corrected endometriosis. Fertil. Steril., 54, 10831088.[ISI][Medline]
Ecochard, R., Mathieu, C., Royere, D., Blache, G., Rabilloud, M. and Czyba, J.C. (2000) A randomized prospective study comparing pregnancy rates after clomiphene citrate and human menopausal gonadotropin before intrauterine insemination. Fertil. Steril., 73, 9093.[CrossRef][ISI][Medline]
Fisch, P., Casper, R.F., Brown, S.E., Wrixon, W., Collins, J.A., Reid, R.L. and Simpson, C. (1989) Unexplained infertility: evaluation of treatment with clomiphene citrate and human chorionic gonadotropin. Fertil. Steril., 51, 828833.[ISI][Medline]
Fujii, S., Fukui, A., Fukushi, Y., Kagiya, A., Sato, S. and Saito, Y. (1997) The effects of clomiphene citrate on normally ovulatory women. Fertil. Steril., 68, 997999.[CrossRef][ISI][Medline]
Glazener, C.M.A., Coulson, C., Lambert, P.A., Watt, E.M., Hinton, R.A., Kelly, N.G. and Hull, M.G.R. (1990) Clomiphene treatment for women with unexplained infertility: placebo-controlled study of hormonal responses and conception rates. Gynecol. Endocrinol., 4, 7583.[ISI][Medline]
Goeree, R., Labelle, R. and Jarrell, J.F. (1993) Cost-effectiveness of an in-vitro fertilization program and the costs of associated hospitalizations and other infertility treatments. New Reproductive Technologies and the Health Care System: The Case for Evidence-Based Medicine (ed. by Royal Commission on New Reproductive Technologies), pp. 569599. Ministry of Supply and Services, Ottawa.
Goverde, A.J., McDonnell, J., Vermeiden, J.P.W., Schats, R., Rutten, F.F.H. and Schoemaker, J. (2000) Intrauterine insemination or in-vitro fertilisation in idiopathic subfertility and male subfertility. Lancet, 355, 1318.[CrossRef][ISI][Medline]
Guzick, D.S., Sullivan, M.W., Adamson, G.D., Cedars, M.I., Falk, R.J., Peterson, E.P. and Steinkampf, M.P. (1998) Efficacy of treatment for unexplained infertility. Fertil. Steril., 70, 207213.[CrossRef][ISI][Medline]
Guzick, D.S., Carson, S.A., Coutifaris, C., Overstreet, J.W., Factor-Litvak, P., Steinkampf, M.P., Hill, J.A., Mastroianni, L., Buster, J.E., Nakajima, S.T., Vogel, D.L. and Canfield, R. (1999) Efficacy of superovulation and intrauterine insemination in the treatment of infertility. N. Engl. J. Med., 340, 177183.
Harrison, R.F. and OMoore, R.R. (1983) The use of clomiphene citrate with and without human chorionic gonadotropin. Irish Med. J., 76, 273274.[ISI]
Karande, V.C., Rao, R., Pratt, D.E., Balin, M., Levrant, S.G., Morris, R., Dudkiewicz, A. and Gleicher, N. (1996) A randomized prospective comparison between intrauterine insemination and fallopian sperm perfusion for the treatment of infertility. Fertil. Steril., 64, 638640.[ISI]
Karande, V.C., Korn, A., Morris, R., Rao, R., Balin, M., Rinehart, J., Dohn, K., and Gleicher, N. (1999) Prospective randomized trial comparing the outcome and cost of in vitro fertilization with that of a traditional treatment algorithm as first-line therapy for couples with infertility. Fertil. Steril., 71, 468475.[CrossRef][ISI][Medline]
Karlstrom, P.O., Bergh, T. and Lundkvist, O. (1993) A prospective randomized trial of artificial insemination versus intercourse in cycles stimulated with human menopausal gonadotropin or clomiphene citrate. Fertil. Steril., 59, 554559.[ISI][Medline]
Kirby, C.A., Flaherty, S.P., Godfrey, B.M., Warnes, G.M. and Matthews, C.D. (1991) A prospective trial of intrauterine insemination of motile spermatozoa versus timed intercourse. Fertil. Steril., 56, 102107.[ISI][Medline]
Meldrum, D.R., Silverberg, K.M., Bustillo, M. and Stokes, L. (1998) Success rate with repeated cycles of in vitro fertilization-embryo transfer. Fertil. Steril., 69, 10051009.[CrossRef][ISI][Medline]
Nygren, K.-G. and Nyboe Andersen, A. (2001) Assisted reproductive technology in Europe, 1998. Results generated from European registers by ESHRE. Hum. Reprod., 16, 24592471.
Philips, Z., Barraza-Llorens, M. and Posnett, J. (2000) Evaluation of the relative cost-effectiveness of treatments for infertility in the UK. Hum. Reprod., 15, 95106.
Society for Assisted Reproductive Technology & American Society for Reproductive Medicine (2002) Assisted reproductive technology in the United States: 1998 results generated from the American Society for Reproductive Medicine/Society for Assisted Reproductive Technology registry. Fertil. Steril., 77, 1831.[CrossRef][ISI][Medline]
Templeton, A.A. and Penney, G.C. (1982) The incidence, characteristics, and prognosis of patients whose infertility is unexplained. Fertil. Steril., 37, 175182.[ISI][Medline]
The ASRM Practice Committee (2000) Educational Bulletin: Treatment of Unexplained Infertility. 14. 2000. The American Society of Reproductive Medicine, Birmingham AB.
van Voorhis, B.J., Stovall, D.W., Allen, B.D. and Syrop, C.H. (1998) Cost-effective treatment of the infertile couple. Fertil. Steril., 70, 9951005.[CrossRef][ISI][Medline]