Assisted reproductive technology treatment costs of a live birth: an age-stratified cost–outcome study of treatment in Australia

Assisted reproductive technologies (ART), such as in-vitro fertilisation (IVF) in which a woman’s egg is retrieved and fertilised and the resultant fresh (or thawed frozen embryo) is implanted into her uterus, are now widely accepted as effective and acceptable treatments for infertility. However, the question of how ART should be funded and who should have access to treatment is generating considerable debate in Australia, Europe and the United States.1-5 Before the May 2005 federal Australian budget, there was much speculation that the number of publicly funded ART treatment cycles would be capped, based on a woman’s age and the number of treatment cycles offered. Hindering this debate, and ultimately causing the issue to be referred to an independent review committee,6 is the complete lack of economic studies evaluating the costs and outcomes of ART treatment in Australia.

During 2002, 32 958 ART treatment cycles (including donor insemination) were initiated in Australia, resulting in 5953 live-born babies.7 This equates to a utilisation rate of about 1600 cycles per million population, which is similar to rates found in Europe.8-10 Despite this, a review of the literature over the past decade found no economic studies that used Australian data or that could be usefully applied to the unique Australian health funding environment.

Currently, all treatment cycles undertaken in Australia attract a rebate through the Medicare Benefits Scheme (MBS). More than half of the estimated $6000–9000 direct treatment costs are covered by the MBS or the Pharmaceutical Benefit Scheme (PBS). The most significant change to funding of ART in the past decade has been the introduction of the Medicare Plus Safety Net in January 2004 — this pays 80% of the out-of-pocket expenses for medical services provided outside of a hospital once an individual’s or family’s threshold is reached. This policy has effectively reduced patient expenses for ART services by up to half, and resulted in a 57% increase in MBS benefits paid for ART services in 2004 (to $79 million).11

It is not just the absolute amount spent on ART services by government, insurers and patients that is important; the return on that expenditure in terms of ART outcomes must also be considered. This measure of cost-effectiveness is most commonly expressed in economic evaluations of ART by the cost per birth of at least one live-born baby (a live-birth event). The aims of this cost–outcome study were to calculate the average cost of an ART treatment cycle, and the average and age-specific cost per live birth for all ART treatment cycles conducted in Australia in 2002. Direct costs borne by government, private insurers and patients were included to provide a societal perspective of total health care resources consumed by ART services.

A total of 29 608 ART cycles were initiated in Australia in 2002: 18 870 fresh cycles and 10 738 frozen embryo cycles (Box 2).

The total direct health care cost of ART cycles undertaken in 2002 was $151.7 million ($130.9 million for fresh cycles; $20.8 million for frozen embryo cycles). The average cost of a fresh cycle was $6940 for all initiated cycles. The average cost of a frozen embryo cycle was $1937. Average treatment costs were up to 6.1% more expensive in the younger age groups because they included a larger proportion of completed cycles, and were more likely to include cryopreservation.

The average cost per live birth for all non-donor ART cycles was $32 903 (Box 3). The cost per live birth varied from $12 491 for frozen embryo cycles in women aged 30–34 years to $133 705 for fresh embryo cycles in women aged 40 years and older. The average cost of the 5474 live-born babies conceived following non-donor ART treatment was $27 722.

The most health care resources were consumed by the 35–39 years age group (34%) and the 30–34 years age group (32%). The ≥ 40 years age group consumed 21% of resources, despite having the highest cost per live birth (Box 4).

A subanalysis of cycles undertaken by women aged ≥ 42 years is presented in Box 5. This older age grouping resulted in a cost per live delivery of $182 794 and consumed 9% of ART health care resources.

This is the most comprehensive population-based costing study of resources consumed during ART treatment. The strength of this study lies in the use of ANZARD, which is a national registry of all ART cycles initiated in Australia. Use of detailed data from the 25 Australian fertility clinics eliminated the need to use a single estimate for the proportion of cycles abandoned before embryo transfer. Furthermore, this study reflected total health care costs by including government, private insurance and patient out-of-pocket expenses, rather than relying solely on fertility clinic charges. The costs were also nationally adjusted to account for differences in fertility clinic fees, which vary by up to 20% between states. However, the costs were not adjusted for differential age-related medication consumption and early cycle cancellation, or for complications arising from ART treatment that required hospitalisation (0.7% of cycles7). The ART success rates were based on the most recently available national data, which are at least 3 years old. Given the continuing advances in ART treatment and success rates, our findings need to be considered in this context.

The average cost of a fresh cycle in Australia was $6940, and for a frozen embryo transfer cycle, $1937. It is difficult to compare these estimates with results from other studies because of differences in study design, health care setting, and definitions of cost. However, our figures are substantially less than US estimates for ART cycle costs, and similar to cost estimates from the United Kingdom and other European countries. To aid comparison with other countries, the following estimates have been inflated to March 2005 using the Australian Health Services Consumer Price Index13 and converted to Australian dollars using September 2005 foreign exchange rates. In a 2002 survey, the estimated average cost of an IVF cycle in the US was US$9547 ($15 017), and the average cost from 25 other countries, excluding Australia, was US$3518 ($5534).8 The UK National Institute for Clinical Excellence (NICE) estimated the average cost in England for an IVF cycle as £2771 ($6961), for an ICSI cycle as £2936 ($7376), and for a frozen embryo transfer cycle as £1000 ($2512).9 A recent study estimated the cost of a complete IVF cycle undertaken in Finland as €3291 ($5582).14

Funding of ART services differs between countries, and often within countries. Australia is unique in its unrestricted approach to public funding of ART services, and cannot be easily compared with funding arrangements internationally.2 For example, in the UK there is a wide variation in the public provision of ART services based on the area health authority. The NICE guidelines developed in 2004 recommend that couples be offered up to three cycles of IVF on the National Health Service (NHS) if the couple meet strict selection criteria.9 From April 2005, all women with appropriate clinical need should have at least one cycle of treatment paid by the NHS.15 Fertility treatment in other European countries is covered by a mix of restricted public and private sector financing.2,16 In the US, financing of ART is primarily through third-party payer arrangements or directly by patients, and only 14 states are required to provide some level of insurance for infertility treatment.17

Although ART services are costly from an individual’s perspective, they only account for a small proportion of national health care expenditure. ART services in Australia accounted for less than 1% ($79 million) of the $8.6 billion MBS benefits paid in 2004,11 and, as we found, 0.2% ($151.7 million) of the $72.2 billion public and private expenditure on health care in 2002–03.18 It will be difficult for the government to make significant savings by limiting funding in this area. Furthermore, because the cost of a live birth is a function of ART pregnancy rates, as the overall success of ART improves, so should its cost-effectiveness. Indicative of this is the increase in live births per embryo transfer cycle from 13% in 1993 to 21% in 2002.7

The age-specific cost per live birth has been investigated in a small number of studies, and although the absolute costs vary between studies, the relative change between age groups is similar, and is directly associated with the age-related decline in female fecundity.9,19,20 For cycles recorded in ANZARD, women aged 40–44 years undergoing fresh non-donor ART in 2002 had a 6% chance per initiated cycle of achieving a live birth, compared with an 18% chance in women aged 35–39 years and a 25% chance in women younger than 35 years.7 Success rates for women older than 40 years vary considerably with each successive year, and the use of donor eggs improves success rates with ART, particularly in older women.12

Our aim in this study was to look at the cost of ART treatment in terms of live births achieved; a second study will consider the downstream costs associated with ART treatment. Multiple births, which occur in 18%–22% of all deliveries following ART treatment in Australia, despite fewer embryos being transferred over the past decade,7 remain the most significant challenge facing assisted reproduction. The practice of selective single embryo transfer in women at risk of twinning and the augmenting effect of cryopreservation of embryos to reduce the incidence of multiple births, requires further economic evaluation.21,22 The goal of assisted reproduction must be the birth of a healthy baby, and any economic framework must consider the long-term economic and non-economic costs of multiple births following ART procedures.

We found ART to be three to four times less cost-effective in women aged ≥ 40 years than in younger women, but age is not the sole determinant of potential success with ART treatment. Factors such as ovarian reserve, successful child-bearing, duration of infertility, number of previous unsuccessful attempts at treatment, and the use of donor eggs all influence success rates.23,24 Furthermore, broader arguments relating to community values, ethical and responsible practice and equity of access, should be judged alongside economic considerations to inform public policy on ART provision.

4 Total non-donor ART treatment costs* and cost per live birth, Australia 2002

---

* Cost is expressed in 2005 Australian dollars.