What is a Cost Effective Approach?

by Bradley J Van Voorhis, MD

Abstract

Few cost-effectiveness studies about infertility treatment have been published; however, the available studies share some common findings. This article reviews the published literature and discusses the cost-effectiveness of various approaches to infertility treatment. In the absence of tubal blockage and severe male factor infertility, use of intrauterine inseminations and human menopausal gonadotropin (hMG) injections are more cost-effective than in vitro fertilization (IVF). IVF is at least as cost-effective as tubal surgery. Although IVF costs are high, they fall well within the range of those for other accepted medical treatments and below the upper limit of what the general public is willing to pay for these treatments. Cost-effectiveness analysis is an important means of improving quality of care while controlling costs. Further work regarding cost-effectiveness of treatments among different diagnostic groups of infertile patients is needed. [Drug Benefit Trends 10(7):20, 22-29, 1998. © 1998 SCP Communications, Inc.]

Introduction

Infertility is defined as failure to conceive after 1 year of unprotected intercourse. It is a common disorder, affecting 10% of couples in the US. Low sperm count or motility, blocked fallopian tubes from previous infection or surgery, and lack of ovulation are among the causes. In many couples, no cause for infertility can be found, even with intensive evaluation. Despite the prevalence of infertility, insurance coverage for its treatment is uncommon in the US. Currently only 12 states have laws regarding insurance coverage of infertility treatment. Although several states require comprehensive coverage, including in vitro fertilization (IVF), many other states have restrictions on the procedures that are covered. In addition, many insurance plans are exempt from having to provide coverage, or they set limits on the amount that will be reimbursed.

The costs of infertility evaluation and treatment are frequently passed directly to the patient because of limited insurance coverage. Often, the physician and patient must make infertility treatment decisions based on economic considerations rather than strictly on medical criteria. This underscores the importance of understanding the economics and cost-effectiveness of treatments offered to infertile couples.

Determination of Costs for Treatment

In performing a cost-effectiveness analysis for infertility treatment, a number of different costs may be included in the analysis. Costs to be considered include direct medical costs, direct nonmedical costs, indirect morbidity and mortality costs, and intangible costs[1-3] (B.J. Van Voorhis, MD, D.W. Stovall, MD, B.D. Allen, MD, et al, unpublished data). The types of direct medical costs usually taken into account include those for hospitalization, drugs, physician fees, laboratory tests, and radiologic procedures. Calculating the direct medical costs of infertility care is more complicated than calculating the costs for other medical conditions because it is necessary to consider both the costs of the infertility treatment and those incurred by the infants resulting from the treatment; a substantial consideration in infertility treatment is the cost of neonatal care for prematurely born multiple gestations resulting directly from these treatments.

In addition to direct medical costs, cost-effectiveness analyses often include direct nonmedical costs, indirect costs, and intangible costs. Direct nonmedical expenses include expenditures for food, lodging, and transportation connected with seeking such medical treatment. These expenses can be substantial, particularly when services are not readily accessible, as is often the case for highly specialized procedures (eg, IVF, which is offered at a limited number of centers). Indirect costs are those that occur because of a loss of life or livelihood and may result from morbidity or mortality arising from infertility treatment. Indirect costs include lost wages or decreased earning potential, which may occur while seeking treatment or because of disability. Finally, intangible costs include the pain, suffering, and grief that may occur as the result of disease or medical care. Intangible costs are very difficult to measure and have not been included in most cost-effectiveness analyses of infertility treatments.

Cost-effectiveness Analysis of IVF

IVF is becoming an increasingly common treatment for all causes of infertility. IVF involves stimulating multiple ovarian follicles to develop by injecting the woman with gonadotropin medications. The eggs within these follicles are retrieved using a needle under ultrasound guidance. The eggs are fertilized in the laboratory, and the resulting embryos are then transferred to the uterus through the cervix. Typically, 3 of 4 embryos are placed into the uterus in order to improve the chances for pregnancy. As might be expected, this practice also increases the risk of multiple births. Embryos not used for transfer at the time of retrieval are often cryopreserved for later procedures. Some programs use variations of IVF such as gamete intrafallopian tube transfer (GIFT), which involves retrieved eggs and sperm, or zygote intrafallopian tube transfer (ZIFT), which involves embryos. Both procedures require laparoscopy for placing the material in the fallopian tube. IVF, GIFT, and ZIFT are sometimes collectively referred to as assisted reproductive techniques (ART).

The cost-effectiveness of IVF in the US has been examined in a number of reports (Table 1). Neumann and colleagues[4] were the first to report on this topic; they obtained costs of performing IVF by polling 6 IVF centers in the eastern US.[4] Direct costs of the treatment were estimated to be $8000 per cycle for medical charges. They also estimated the costs of complications, including multiple births. Indirect costs of the cycle were included in the analysis by calculating lost wages at $11 per hour. Outcomes of the treatment cycles were obtained from published reports, and the investigators estimated from those data that pregnancy rates varied from 12% for the first cycle of treatment to 7% for the sixth cycle of treatment. By this analysis, the estimated cost per delivery for the first cycle was $66,667 and increased to $114,286 per delivery after the sixth treatment cycle.

These researchers also performed a sensitivity analysis by varying the probabilities by 2% each way; they estimated that the cost per delivery for IVF would fall somewhere between a low of $55,143 and a high of $211,940.[4] It is important to note that the conclusions of these researchers were based on estimates and not on actual costs and outcomes of IVF centers.

Subsequently published analyses suggest that the actual cost per delivery for IVF may be somewhat, although not markedly, lower than that estimated by Neumann and associates[4] (Table 1). There are several reasons for the initial high estimate. For instance, the investigators used a low estimate of the pregnancy rate obtained using IVF. Recent improvements in embryo culture techniques and increased experience with this relatively new treatment have led to significantly higher pregnancy rates in many centers. In addition, the report failed to account for the pregnancies that can result from cryopreserved embryos. Cryopreserved embryo transfer is much less expensive than transfer of "fresh" embryos, since the costs associated with ovarian stimulation and egg retrieval are avoided. Finally, many couples (particularly poor-risk couples) drop out of treatment, as opposed to continuing through 6 cycles of IVF, as assumed in some of the high-range estimates in the study.

In a cost-effectiveness study from Brigham and Women's Hospital,[5] 182 patients were followed until they had achieved a pregnancy or completed a maximum of 3 IVF cycles. An ongoing pregnancy rate of 27% was found after the first cycle of treatment. Largely as a result of this higher pregnancy rate, a lower cost per delivery for IVF was found compared with that reported by Neumann and coworkers.[4] Trad and coworkers[5] calculated the cost per delivery to be $29,120 after 1 cycle and $31,590 after a maximum of 3 cycles of IVF. Next, they calculated the different costs per delivery in patients with high, moderate, and low probability of pregnancy. These groupings were based on the woman's age and whether or not severe male factor infertility was present, since both of these are known prognostic factors for IVF pregnancies. The cost per delivery in this study was found to be $22,857, $34,000, and $42,666 for the high, moderate, and low probability groups, respectively. This study demonstrated that IVF treatments become significantly more cost-effective in "good prognosis" patients because of improved pregnancy rates.[5]

We determined the cost-effectiveness of a variety of assisted reproductive procedures during 1992 at the University of Iowa.[6] Rather than estimating costs, we tabulated all medical charges for all patients from the time of the initial consult until intrauterine fetal cardiac activity was confirmed by ultrasound examination. Costs of all immediate complications of treatment were included in the analysis. Obstetric charges and lost wages were not included, since these data were often not available to us due to deliveries occurring in other hospitals. In 1992, we had an average cost of $8071 per initiated cycle and an ongoing pregnancy rate of 18.3% per initiated cycle. Our calculated cost per delivery was $44,200 for IVF-again, a lower figure than earlier estimates, largely because of higher pregnancy rates achieved. We also found that embryo cryopreservation and subsequent transfer is a very cost-effective practice. Transfer of cryopreserved embryos resulted in an 18% pregnancy rate per initiated cycle, and the cost per delivery was only $10,953 for these cycles. Cost savings were accrued for cryopreserved embryo transfer because ovarian stimulation and oocyte retrieval are not required, and hospital and physician charges are reduced accordingly (Table 2). We also discovered that other assisted reproduction procedures, including GIFT and ZIFT, were more cost-effective than IVF because of the higher pregnancy rates achieved with these techniques in our program. When considering all the ART procedures offered in our program in 1992, the overall cost per delivery was $30,252, and the cost per infant was $22,991.

An important variable in the cost-effectiveness of IVF is the woman's age. We found the cost per delivery to be $31,597 for women younger than 38 years.[7] In contrast, for women 38 years of age or older, the cost per delivery was nearly triple, at $89,981. The use of donor oocytes is very cost-effective in older women, as the cost per delivery was $35,605 for those cycles despite the increased cost of paying for a donor. This reflects the well-documented efficacy of donor-oocyte cycles in older women.

In summary, the measured cost-effectiveness of IVF as a treatment for infertility is somewhat improved over the initial published estimate. The major reason for this is the improved pregnancy rates that have been seen over the years, particularly in large centers but also in the general experience of clinics throughout the US. It is likely that the cost-effectiveness of this treatment will continue to evolve as new discoveries lead to improvements in IVF success rates. It is important to remember that IVF is a very new technology, so continued advances in this field can be expected. Therefore, policy makers and insurance companies will need to continually monitor the cost-effectiveness of this procedure if they are to make valid conclusions and decisions regarding patient care and access to IVF treatment.

Other Infertility Treatments

There have been relatively few analyses of alternative treatments for infertile couples. The effectiveness of intrauterine inseminations (IUIs) with and without ovarian follicular stimulation with human menopausal gonadotropin injections (hMG-IUIs) has been demonstrated in multiple series for couples whose infertility is not associated with fallopian tube obstruction or severe male factor infertility.[25-28] IUI treatment is performed by isolating the motile sperm from an ejaculate using centrifugation procedures. The motile sperm are then placed into the uterus, bypassing the cervix, and thus allowing greater numbers of sperm to reach high in the female reproductive tract. This insemination is performed when the woman is ovulating during a natural cycle. Alternatively, the insemination can be performed after multiple egg development and ovulation have been stimulated by the oral medication clomiphene citrate or injectable gonadotropin treatment (hMG). Although the effectiveness per cycle is inferior to that of IVF, the costs associated with these treatments are also lower; therefore, they are commonly used in treating infertile couples. In 1997, we published the first study designed to compare the cost-effectiveness of a wide variety of infertility treatments including IUI alone, clomiphene citrate administration plus IUI (CC-IUI), and hMG-IUI.[7] Although the cost-effectiveness of these various treatments would ideally be determined after a prospective, randomized allocation of couples to treatments, such a study is not feasible and is not likely to be done. Therefore, we are left comparing cost-effectiveness of procedures as they are commonly performed in an infertility practice. We found that IUI, CC-IUI, and hMG-IUI are similar procedures in terms of cost per delivery (Table 3). Despite the increased costs of hMG cycles due to medication and cycle monitoring, higher pregnancy rates reduced the cost per delivery of hMG-IUI cycles to the range seen in IUI and CC-IUI cycles. IUI, CC-IUI, and hMG-IUI are all more cost-effective procedures than IVF due to the higher costs associated with IVF cycles. These results support the findings of an earlier study from Utah by Peterson and coworkers[26] demonstrating that the cost per delivery for hMG-IUI cycles was roughly one third that of IVF cycles at their academic institution. The researchers suggested that a cost-effective approach to treating infertility would be to perform 3 to 4 cycles of hMG-IUI in most couples before resorting to IVF.

Although general recommendations can be made regarding a cost-effective approach to treating infertility based on cost-effectiveness studies, there are factors present in individual patients that would change the efficacy, and therefore influence the cost-effectiveness, of certain treatments. Two factors that influenced the cost-effectiveness of IUI, CC-IUI, hMG-IUI, and IVF cycles in our study were age of the female partner and low post-wash sperm numbers.[7] Women who were at least 38 years old had a cost per delivery nearly 3 times that of younger women. Despite the increased costs associated with compensating an oocyte donor, the use of donor oocytes in recipients 38 years of age or older reduced the cost per delivery to levels seen in younger patients. The numbers of sperm available for insemination also affected the success rates and the cost-effectiveness of IUI, CC-IUI, and hMG-IUI procedures. If fewer than 10 million motile sperm were available for insemination, then substantial reductions in pregnancy rates were noted, leading to higher costs per delivery for IUI, CC-IUI, and hMG-IUI cycles. Our data suggest that a policy of proceeding to IVF rather than hMG-IUI in cases of low sperm counts would be cost-effective. Larger studies will be required to determine the cost-effectiveness of infertility procedures for different diagnostic groups. Just as age and the number of spermatozoa inseminated affect the success of the procedures studied, it is possible that certain diagnostic groups will have better success with specific treatments, resulting in altered cost-effectiveness of procedures for those groups.

Cost-effectiveness studies can be useful in establishing general guidelines of care for large groups of patients. However, these analyses are less helpful in counseling individual patients since decisions regarding treatments are influenced by considerations other than cost and effectiveness. Factors commonly considered by patients include the time away from work, the convenience, and the many ethical issues surrounding infertility treatments. In addition, cost-effectiveness comparisons are most comfortably made when the treatments being compared are similar in effectiveness.[2] Although we found IUI to be cost-effective compared with ART in our study, IUI was nearly 5-fold less effective than ART in terms of pregnancy rate per cycle. To some infertile couples, costs may be less relevant when considering the increased time, frustration, and grief that may accompany more failed cycles with the less successful procedure.

Discussion

From this review, it is apparent that the evaluation of cost-effective approaches to infertility treatment is in its infancy. Nevertheless, some important principles have emerged from the initial studies in this field. In general, it appears that for infertile couples without tubal disease or severe male factor infertility, the most cost-effective approach is to start with IUI or superovulation-IUI treatments before resorting to IVF procedures. Factors that have been identified as influencing the cost-effectiveness of a procedure include the woman's age and the number of sperm present for insemination. More work needs to be done to identify the influence that certain diagnoses have on the cost-effectiveness of infertility treatments. IVF is expensive, particularly if one considers the costs associated with multiple gestations and premature deliveries. However, the cost of IVF still falls within the range of that for other medical procedures considered to be cost-effective and certainly falls below the upper limit of what the public is willing to pay for these procedures. Indeed, for patients with severe tubal disease, IVF has been found to be more cost-effective than surgical repair. In addition, the cost-effectiveness of IVF will likely improve as success rates of this treatment show gradual improvement over time. The common policy of excluding infertility treatments from insurance plans is puzzling and accentuates the importance to the clinician of understanding the economics of infertility treatment, since costs are often passed directly to the patient. Current policies that have led to inequities in access to health care should not be tolerated.

Editorial Comment

DSM-An Option For Managing Infertility Infertility affects 10% to 15% of reproductive-aged couples in the US. In spite of the large numbers of individuals affected, infertility has remained a relatively silent disease. And disease it is! The emotional and physical impact of childlessness on a couple can be devastating. However, third-party payers have long argued that infertility is not a disease, but rather a "condition." While the semantics seem trivial to the casual observer, the legal impact of this wording is profound. Medical treatment of "conditions" is considered to be cosmetic in nature, and therefore deemed medically unnecessary (and uncovered). Medical procedures for "diseases" are considered warranted, necessary, and covered. Payers also perceive that medical therapies for infertility are largely unsuccessful. They are also rightly concerned about the impact of multiple pregnancy rates.

This reluctance to provide benefits for infertility treatment has two detrimental effects: (1) it discourages patients from seeking effective medical attention, and (2) it promotes "gaming" of the reimbursement system (coding for disorders that are present and covered, but ignoring the underlying primary concern-infertility). This leads to a highly inaccurate database, as it is impossible to manage conditions that are not identified.

Infertility actually lends itself very well to disease state management. The goal of disease state management is to provide the best possible outcome at the lowest possible cost. This method "manages" all aspects of the disease (awareness, education, prevention, and intervention). The endpoint for infertility is clearly defined: pregnancy resulting in childbirth. Life table analysis and the statistical probabilities of pregnancy-limited somewhat by the quality of the studies at the current time-are available in the literature for many of the disease processes causing infertility (eg, endometriosis, tubal factor, male factor). Confounding factors known to affect outcomes are recognized, including age, smoking, and body weight. And as is pointed out in the article by Dr. Van Voorhis, pregnancy rates resulting from in vitro fertilization have risen steadily over the last few years. Using this body of knowledge, it should be relatively simple to establish rational treatment plans for couples with infertility over a limited period of time.

The current article provides an excellent summary of what we know about cost-effective care for the infertile couple. There are 3 points that should be emphasized:

IVF success rates are increasing, making it more cost effective for more diagnoses; several programs in the US have pregnancy rates consistently above 50% per embryo transfer.

It is incumbent on providers of infertility services to decrease the rate of high-order (3 or more) multiple pregnancy. Most of the programs mentioned above transfer 3 or fewer embryos in women under the age of 40.

It is hoped that the cost of IVF will decrease, either because of competitive pricing of drugs and services or because emerging technologies will lessen the need for expensive, controlled, ovarian hyperstimulation. All these advancements will result in improvements of the cost-effectiveness of IVF, thereby lowering the cost per delivery for infants conceived by means of advanced technologies.


David A. Grainger, MD
Associate Professor & Director, Division of Reproductive Endocrinology
Department of Obstetrics and Gynecology
University of Kansas School of Medicine-Wichita
Wichita, Kan.