What is the most relevant standard of success in assisted reproduction? The singleton, term gestation, live birth rate per cycle initiated: the BESST endpoint for assisted reproduction

Jason K. Min1, Sue A. Breheny2, Vivien MacLachlan2 and David L. Healy3,4

1 Department of Obstetrics and Gynaecology, University of Alberta, Edmonton, Canada T5H 3V9, 2 Monash IVF, Epworth Hospital, Richmond 3121 and 3 Centre for Women’s Health Research, Monash University Department of Obstetrics and Gynaecology, Monash Medical Centre, Level 5, 246 Clayton Road, Clayton, Victoria 3168, Australia

4 To whom correspondence should be addressed. e-mail: david.healy{at}med.monash.edu.au


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Assisted reproduction programmes do not report success consistently. Rates vary with the definition used. Success must reflect delivery of healthy babies, and the burden of treatment to couples. We report the singleton, term gestation, live birth rate of a baby per assisted reproductive technology cycle initiated for a large IVF programme. We defined assisted reproductive technology cycles as those initiated with the intention of oocyte collection. We examined cycles conducted through Monash IVF in 2001. All women with positive pregnancy tests had first trimester ultrasonography. Obstetric outcomes were recorded. All babies had neonatal examinations conducted by paediatricians. A total of 644 positive pregnancy tests were recorded in 2600 cycles; 509 showed fetal heart motion. Of 448 deliveries, 328 were singleton and 120 were multiple. There were 290 singleton deliveries at term gestation. In 2001, a couple had an 11.1% chance of delivering a singleton, term gestation, live baby per assisted reproductive technology cycle begun. We suggest that delivery of a single, term gestation, live baby per cycle initiated is the most relevant standard of success. This statistic was 11.1% at Monash IVF. We encourage programmes to report this BESST (Birth Emphasizing a Successful Singleton at Term) outcome.

Key words: assisted reproduction/live birth rate per cycle/singleton pregnancies/success rates/term gestation


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Since the birth of the first IVF baby in 1978, the treatment of subfertility has significantly advanced. We no longer practice in an era where assisted reproductive technology is experimental. Over the past 25 years, these techniques have become well-established. The practice of assisted reproduction is now based on a mature science, and with maturity has come a redefinition of therapeutic goals. Pregnancy, without consideration of obstetric and neonatal outcomes, is no longer the objective. Increasingly, successful practice is reflected by the delivery of a healthy baby.

Practitioners readily acknowledge the significant contribution of multiple pregnancy to the risks and complications of assisted reproductive technology. Despite universal agreement on the need for a reduction of this iatrogenic complication (Hock et al., 2002Go), trends in multiple pregnancies and deliveries have not declined (Nygren and Andersen, 2002Go; SART/ASRM, 2002Go; Jones, 2003Go). Subsequently, the Chairman of the ESHRE has recently suggested that the most appropriate outcome variable of all assisted reproductive technology procedures is the singleton live birth rate per cycle initiated (Evers, 2002Go). Implicit in this suggestion is the need for all modern assisted reproductive technology clinics to adopt a new endpoint for success. We present this, and related outcomes, for a large assisted reproduction programme.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
We retrospectively examined all assisted reproductive technology cycles initiated by the Monash IVF programme, Monash University, Melbourne, Victoria, Australia in 2001. We defined assisted reproductive technology as any treatment involving oocyte retrieval, for subsequent manipulation in vitro, and their replacement into the female reproductive tract either as oocytes or embryos. This definition would include: IVF, with or without ICSI, gamete intra-Fallopian transfer (GIFT), zygote intra-Fallopian transfer (ZIFT), and microinjection intra-Fallopian transfer (MIFT). ‘Cycle’ described any assisted reproductive technology treatment regimen started with the intended goal of oocyte retrieval, regardless of whether the retrieval was actually performed.

Included in our analysis were all gonadotrophin-stimulated cycles involving oocyte retrieval and insemination with either partner or donor sperm. Excluded from the analysis were: unstimulated or natural cycles, clomiphene citrate-stimulated cycles, cycles utilizing donor oocytes or embryos, and frozen oocytes. Frozen–thawed embryo transfer cycles were excluded since those cycles were not stimulated. Controlled ovarian stimulation cycles, which do not employ oocyte retrieval, were excluded by definition. Exclusions were not made based on female partner age, duration of infertility, aetiology of subfertility, or results of previous treatment cycles.

All women undergoing replacement of gametes or embryos had a maximum of three embryos transferred. All patients had quantitative determination of the {beta} subunit of {beta}hCG 16 days following oocyte retrieval. Pregnancy was defined as two rising serum {beta}hCG concentrations >5 IU/l. Those with positive pregnancy tests had transvaginal ultrasonography 6 weeks following oocyte retrieval to document the number of gestational sacs and the presence of fetal heart motion. A viable pregnancy was defined by the presence of at least one intrauterine gestational sac with fetal heart motion. Obstetric outcomes were monitored to determine pregnancy loss, preterm and term vaginal and Caesarean birth rates.

Deliveries were classified as preterm if occurring between 20 and 36 6/7 completed weeks of gestation (between 18 and 34 6/7 completed weeks from the day of gamete or oocyte collection), and term if occurring after 37 completed weeks of gestation (after 35 completed weeks from the day of collection). Records of live birth or stillbirth were kept for each gestation.

The singleton, term gestation, live birth rate of a baby per cycle initiated was the primary outcome measured. This statistic was defined as the BESST (Birth Emphasizing a Successful Singleton at Term) statistic. Rates of preclinical and viable pregnancies were also determined. Deliveries were examined by gestational age and number of babies.

The primary outcome was also stratified by maternal age at cycle initiation and by cycle rank. For this subgroup analysis, patients were excluded if their entire assisted reproductive technology history was not known in detail. Cycles were divided into the following categories by rank: first, second, third, and fourth or greater. Cycles were assigned consecutive ranks following application of all exclusion criteria. Maternal age was partitioned according to reporting norms for our programme into three categories: <35, 35–39 and >39 years. {chi}2-Analysis was used to compare the singleton, term gestation, live birth rate of a baby per cycle initiated among subgroups of each variable. Statistical analyses were conducted using Analyse-It for Microsoft Excel (2003, Leeds, UK). P < 0.05 was considered significant.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Of all assisted reproductive technology cycles initiated through Monash IVF programmes in 2001, 2600 cycles were identified in 1860 women for analysis by inclusion and exclusion criteria. Of the 2600 cycles initiated, 2214 oocyte retrievals (84%) and 2041 transfers (78%) of gametes or embryos were performed. Table I lists the distribution of transfers among the various assisted reproductive technology techniques employed. IVF with or without ICSI accounted for 96.5% of all assisted reproductive technology cycles conducted.


View this table:
[in this window]
[in a new window]
 
Table I. Distribution of gamete/embryo transfers by assisted reproductive technology
 
Of 644 preclinical pregnancies, 509 were viable on first trimester ultrasound. A total of 448 pregnancies continued past 20 weeks gestation. Of these, 328 were singleton, 116 were twin, and four were triplet pregnancies. Quadruplet or higher-order gestations did not occur.

Six pregnancies, one twin and five singleton, ended in stillbirth. All stillborn deliveries were preterm. The remaining 442 pregnancies resulted in live-born neonates. In all, 119 (26.9%) were multiple gestations. Of all live deliveries, 24.2% were preterm. Thirty-three singleton, 65 twin and all four triplet gestations delivered preterm. Multiples accounted for 67.6% of all preterm deliveries and 82.0% (150 of 183) of premature babies born. At term gestation, there were 50 twin and 290 singleton deliveries. Of pregnancies carrying to term, 85.3% were singleton.

Table II shows that success rate varied widely depending upon the chosen endpoint. The rates for positive pregnancy test, fetal heart motion present on first trimester ultrasound, and live delivery were 24.8, 19.6 and 17.0% respectively.


View this table:
[in this window]
[in a new window]
 
Table II. Obstetric outcomes in 1860 women receiving 2600 cycles of assisted reproductive technology
 
The singleton, term gestation, live birth rate of a baby was 11.1% per cycle initiated. In all, 164 (58.2%) of these babies delivered vaginally and 118 (41.8%) delivered by Caesarean section. The distribution into elective and emergency Caesarean deliveries was not available.

At the time of cycle initiation, the median age of women undergoing treatment was 35 years (range 20–48 years). The mean was 35.3 years. Of the 2600 cycles originally identified, 197 were excluded because their precise rank could not be determined. This left 2403 cycles for subgroup analysis. Table III shows the singleton, term gestation, live birth rate of a baby stratified by cycle rank and maternal age at the time of cycle initiation.


View this table:
[in this window]
[in a new window]
 
Table III. The singleton, term gestation, live birth rate of a baby per cycle initiated; stratified by female age and cycle rank
 
Of 2403 ranked cycles, 43.9% were conducted in women <35 years of age. Delivery rates showed a significant decrease, from 13.9 to 3.3% with increasing maternal age (P < 0.001). This finding was observed within each cycle rank except for the third. Significance was maintained for first (P = 0.0001) and fourth or greater (P = 0.0008) ranks.

Of the 2403 ranked cycles, 43.7, 26.8, 15.8 and 31.3% were first, second, third and fourth or greater cycles respectively. For all ages, and within each age subgroup, delivery rates were stable across cycle ranks (P = non-significant).


    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Multiple pregnancy is the most frequent and most serious iatrogenic complication of assisted reproductive technology (Evers, 2002Go). This definitive statement by the Chairman of the ESHRE reflects the scope of the burden imposed by multiple gestation. It is universally recognized that multiple gestation, and its attendant prematurity, is associated with increased mortality and morbidity, both for mothers and fetuses (Fathalla, 2002Go; Schieve et al., 2002Go; Wennerholm, 2002Go). Indeed, complications are so common that some have classified multiple pregnancy as a major pathology (Ozturk and Templeton, 2002Go). This characterization is supported by the financial costs of delivering multiple pregnancies arising from assisted reproductive technology cycles. Per gestation, the associated costs through delivery were recently estimated at US$58 865 for twins, US$170 282 for triplets and US$281 698 for quadruplets (ESHRE, 2000Go).

The main strategy to combat multiple gestation associated with IVF and ICSI is to limit the maximum number of embryos transferred for any particular cycle (ASRM, 1999Go; ESHRE, 2000Go). In many centres, the transfer of no more than two embryos is considered standard (ESHRE, 2001Go; Ozturk and Templeton, 2002Go). Unfortunately, the most recent reports from the Society for Assisted Reproductive Technology Registry in the USA and the European IVF Monitoring Program Registry, and the Australian National Perinatal Statistics Unit have failed to demonstrate a reduction in multiple delivery rate from 1998 to 1999. In the USA, this rate was 36.5% (SART/ASRM, 2002Go); in Europe, it was 26.3% (Nygren and Andersen, 2002Go); and in Australia, it was 20.9%.

The maintenance of high multiple birth rates stems partially from the continued transfer of three or more embryos (Jones, 2003Go). However, even in IVF programmes with a two-embryo transfer policy, the twin pregnancy rate is still significant, estimated at 20–35% (ESHRE, 2001Go). In 1999, the USA, European and Australian assisted reproductive technology twin birth rates were 31.8% (SART/ASRM, 2002Go), 24.0% (Nygren and Andersen, 2002Go) and 19.6% respectively. Clearly, the majority of multiple gestations arising from assisted reproductive technology are twins, and their contribution to obstetric and perinatal complications are substantial (ESHRE, 2000, 2001). As high-risk pregnancies, twin gestations should be considered complications of assisted reproductive technology treatment, and not counted as successes (Olivennes, 2000Go; ESHRE, 2001Go; Ozturk and Templeton, 2002Go). Fortunately, elective single embryo transfer provides a potentially viable solution, minimizing twin gestations while maintaining pregnancy rates in selected patient groups (ESHRE, 2001Go; Martikainen et al., 2001Go; Tiitinen et al., 2001Go; De Sutter et al., 2002Go; Gerris et al., 2002Go; Ozturk and Templeton, 2002Go). Unfortunately, a policy of single embryo transfer in a few programmes will have only a small impact on pooled data unless it is globally adopted.

The establishment of a new definition of a successful outcome in assisted reproductive technology is appropriate. Although a seemingly simple act, the importance of setting a new BESST endpoint cannot be overstated. It will focus the philosophical change already underway—the recognition that delivery of a single healthy baby is the most appropriate endpoint. Singleton live birth rate per cycle started has been advocated, in response to the high incidence of multiples (ESHRE, 2000Go, 2001; Evers, 2002Go) and by the World Health Organization (2002Go). If the object is a healthy baby, the specification of ‘term gestation’ is also justified.

Term gestation is a term that is well defined, internationally agreed and able to be retrieved in all countries. Rates of term gestation may differ in different countries depending on race. In the USA the pre-term birth rate seems higher in black than in white women (Alexander et al., 1999Go). This may be pertinent when comparing BESST rates between countries.

Assisted reproductive technology programmes should express outcome as a proportion of all treatment cycles initiated. Oocyte retrieval is a significant component of assisted reproductive technology, accounting for much of the stress, financial burden, and almost all of the surgical risk (Kovacs et al., 2001Go). Moreover, the exclusion of cycles from which oocyte retrieval is not attempted is inappropriate. Although less daunting, the cost of follicular stimulation is not insignificant; nor is the emotional burden of a cycle that is terminated prior to oocyte retrieval. In the USA, statute requires the reporting of assisted reproductive technology outcomes as a proportion of cycles initiated (US Congress, 1992Go). Inclusion of all cycles initiated, regardless of outcome, is most appropriate because it best represents the burden of treatment endured by a couple.

The World Health Organization (2002Go) has recommended that assisted reproductive technology statistics emphasize the birth rates of healthy infants as well as rates of malformations, neonatal morbidity and mortality, and abnormalities of pregnancy. Registries reporting outcomes for various national and international bodies generally provide sufficient data to calculate the live singleton birth rate per cycle initiated (Nygren and Andersen, 2002Go; SART/ASRM, 2002Go). Further stratification of those births into term gestation would be simple enough. However, the determination of birth defects would prove more difficult. Registration in Sweden utilizes extensive coordination of national assisted reproductive technology, birth, and malformation registries to track such outcomes (Nygren, 2002Go). Other large registries, such as those in Australia, the USA and Europe lack similar infrastructure, necessitating more extensive follow-up by individual assisted reproductive technology programmes to capture requisite data (FSA, 2002Go; Nygren and Andersen, 2002Go; Schieve et al., 2002Go; US Congress, 1992Go). To our knowledge, the singleton, term gestation, live birth rate of a baby per assisted reproductive technology cycle begun has not been reported by these registries or individual assisted reproductive technology programmes. In spite of employing multiple search strategies, we were unable to identify any reference in the literature explicitly reporting this outcome. Yet this is precisely what a subfertile couple wishes to know.

At Monash IVF, our typical woman was 35 years old (median; range 20–48 years) in 2001. Her prospect of a singleton, term gestation, live birth of a baby per cycle begun was 11.1%. We must note that this rate excludes frozen embryos. Outcomes derived from frozen embryo transfers would have contributed to the numerator but not the denominator, since the embryos used would have been obtained from oocytes collected in the original stimulation cycle (Kovacs et al., 2001Go). Subsequently, the success rate we report represents a minimum value.

Overall, the singleton, term gestation, live birth rate of a baby per cycle initiated showed a significant decrease with increasing maternal age (Table III). This finding was consistent with known decreases in fecundity with advancing maternal age (Hull et al., 1996Go; Chuang et al., 2003Go).

Meldrum et al. (1998Go) found no significant decline in assisted reproductive technology success rates with three cycle attempts. Similarly, we found that the delivery rate was stable across cycle ranks. Our analysis also examined cycles from a single year, limiting the effect of improved assisted reproductive technology techniques on outcome. Our population was not biased towards those with superior prognoses as we took all-comers. Unlike previous studies that examined pregnancy or delivery rates alone (Tan et al., 1992Go; Yovel et al., 1994Go; Templeton et al., 1996Go; Meldrum et al., 1998Go; Jansen, 2003Go), we have reported a more relevant and valid endpoint.

Assisted reproductive technology techniques are now older than some of our patients. Clinical assisted reproductive technology is now established worldwide in both developed and developing nations. The objective of treatment and reporting of endpoints must parallel this mature science. We propose the singleton, term gestation, live birth rate of a baby per cycle begun as the BESST measure of assisted reproductive technology success.


    Acknowledgements
 
We are indebted to the staff at Monash IVF, and to our patients.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Alexander GR, Kogan MD, Himes JH, Mor JM and Goldenberg R (1999) Racial differences in birth weight for gestional age and infant mortality in extremely low-risk US populations. Paediatr Perinat Epidemiol 13,205–217.[CrossRef][ISI][Medline]

ASRM (American Society for Reproductive Medicine) (1999) Guidelines on Number of Embryos Transferred. A Practice Committee Report: A Committee Opinion. Birmingham, USA.

Chuang CC, Chen CD, Chao KH, Chen SU, Ho HN and Yang YS (2003) Age is a better predictor of pregnancy potential than basal follicle-stimulating hormone levels in women undergoing in vitro fertilization. Fertil Steril 79,63–68.[CrossRef][ISI][Medline]

DeSutter P, Gerris J and Dhont M (2002) A health-economic decision-analytic model comparing double with single embryo transfer in IVF/ICSI. Hum Reprod 17,2891–2896.[Abstract/Free Full Text]

ESHRE Capri Workshop Group (2000) Multiple gestation pregnancy. Hum Reprod 15,1856–1864.[Abstract/Free Full Text]

ESHRE Campus Course Report (2001) Prevention of twin pregnancy after IVF/ICSI by single embryo transfer. Hum Reprod 16,790–800.[Abstract/Free Full Text]

Evers JLH (2002) Female subfertility. Lancet 360,151–159.[CrossRef][ISI][Medline]

Fathalla M (2002) Current challenges in assisted reproduction. In Vayena E, Rowe PJ and Griffin PD (eds) Medical, Ethical and Social Aspects of Assisted Reproduction (2001, Geneva), Current Practices and Controversies in Assisted Reproduction: Report of a WHO Meeting. WHO Publications, Geneva, pp 3–12.

FSA (Fertility Society of Australia) (2002) Reproductive Technology Accreditation Committee code of practice for centres using assisted reproductive technology. Accessed February 17, 2003, at http://www.fsa.au.com/pdfs/RTAC-guidelines-2002.pdf.

Gerris J, De Neuborg D, Mangelschots K, Van Royen E, Vercruyssen M, Barudy-Vasquez J, Valkenburg M and Ryckaert G (2002) Elective single day 3 embryo transfer halves the twinning rate without decrease in the ongoing pregnancy rate of an IVF/ICSI programme. Hum Reprod 17,2626–2631.[Abstract/Free Full Text]

Hock DL, Seifer BD, Kontopoulos E and Ananth CV (2002) Practice patterns among board-certified reproductive endocrinologists regarding high-order multiple gestation: a United States national survey. Obstet Gynecol 99,763–770.[Abstract/Free Full Text]

Hull MG, Fleming CF, Hughes AO and McDermott A (1996) The age-related decline in female fecundity: a quantitative controlled study of implanting capacity and survival of individual embryos after in vitro fertilization. Fertil Steril 65,783–790.[ISI][Medline]

Jansen RPS (2003) The effect of female age on the likelihood of a live birth from one in-vitro fertilisation treatment. Med J Aust 178,258–261.[ISI][Medline]

Jones HW (2003) Multiple births: how are we doing? Fertil Steril 79,17–21.[CrossRef][ISI][Medline]

Kovacs GT, MacLachlan V and Breheny SA (2001) What is the probability of conception for couples entering an IVF program? Aust NZ J Obstet Gynecol 41,207–209.[ISI][Medline]

Martikainen H, Tiitinen A, Tomas C, Tapanainen J, Orava M, Tuomivaara L, Vilska S, Hyden-Granskog C, Hovatta O and the Finnish ET Study Group (2001) One versus two embryo transfer after IVF and ICSI: a randomized study. Hum Reprod 16,1900–1903.[Abstract/Free Full Text]

Meldrum DR, Silverberg KM, Bustillo M and Stokes L (1998) Success rate with repeated cycles of in vitro fertilization–embryo transfer. Fertil Steril 69,1005–1009.[CrossRef][ISI][Medline]

Nygren K (2002) The Swedish experience of assisted reproductive technologies surveillance. In Vayena E, Rowe PJ and Griffin PD (eds) Medical, Ethical and Social Aspects of Assisted Reproduction (2001, Geneva), Current Practices and Controversies in Assisted Reproduction: Report of a WHO Meeting. WHO Publications, Geneva, pp 351–354.

Nygren KC and Andersen AN (2002) Assisted reproductive technology in Europe, 1999. Results generated from European registers by ESHRE. Hum Reprod 17,3260–3274.[Abstract/Free Full Text]

Olivennes F (2000) Avoiding multiple pregnancies in ART. Double trouble: yes a twin pregnancy is an adverse outcome. Hum Reprod 15,1663–1665.[Free Full Text]

Ozturk O and Templeton A (2002) Multiple pregnancy in assisted reproduction techniques. In Vayena E, Rowe PJ and Griffin PD (eds) Medical, Ethical and Social Aspects of Assisted Reproduction (2001, Geneva), Current Practices and Controversies in Assisted Reproduction: Report of a WHO Meeting. WHO Publications, Geneva, Switzerland, pp 220–234.

SART/ASRM (Society for Assisted Reproductive Technology and the American Society for Reproductive Medicine) (2002) Assisted reproductive technology in the United States: 1999 results generated from the American Society for Reproductive Medicine/Society for Assisted Reproductive Technology Registry. Fertil Steril 78 918–931.[CrossRef][ISI][Medline]

Schieve LA, Wilcox LS, Zeitz J, Jeng G, Hoffman D, Brzyski R, Toner J, Grainger D, Tatham L and Younger B (2002) Assessment of outcomes for assisted reproductive technology: overview of issues and the US experience in establishing a surveillance system. In Vayena E, Rowe PJ and Griffin PD (eds) Medical, Ethical and Social Aspects of Assisted Reproduction (2001, Geneva), Current Practices and Controversies in Assisted Reproduction: Report of a WHO Meeting. WHO Publications, Geneva, pp 363–376.

Tan SL, Royston P, Campbell S, Jacobs HS, Betts J, Mason B and Edwards RG (1992) Cumulative conception and live birth rates after in-vitro fertilization. Lancet 339,1390–1394.[CrossRef][ISI][Medline]

Templeton A, Morris JK and Parslow W (1996) Factors that affect outcome of in vitro fertilization treatment. Lancet 348,1402–1406.[CrossRef][ISI][Medline]

Tiitinen A, Halttunen M, Harkki P, Vuoristo P and Hyden-Granskog C (2001) Elective single embryo transfer: the value of cryopreservation. Hum Reprod 16,1140–1144.[Abstract/Free Full Text]

US Congress (1992) Fertility Clinic Success Rate and Certification Act (Publ. No. 102-493 October 24, 1992). Washington, DC.

Wennerholm U-B (2002) Obstetric and neonatal complications in multiple gestation. In Healy DL, Kovacs, GT, McLachlan R and Rodriguez-Armas O (eds) Reproductive Medicine in the Twenty-First Century. Proceedings of the 17th World Congress on Fertility and Sterility. Parthenon, London, pp 182–193.

World Health Organization (2002) In Vayena E, Rowe PJ and Griffin PD (eds) Medical, Ethical and Social Aspects of Assisted Reproduction (2001, Geneva), Current Practices and Controversies in Assisted Reproduction: Report of a WHO Meeting. WHO Publications, Geneva, Switzerland, pp 381–396.

Yovel I, Geva E, Lessing JB, Yaron Y, Botchan A and Amit A (1994) Analysis of the fourth to eighth in-vitro fertilization treatments after three previously failed attempts. Hum Reprod 9,738–741.[Abstract]