Multiple-birth risk associated with IVF and extended embryo culture: USA, 2001

D.M. Kissin1,3, L.A. Schieve2 and M.A. Reynolds2

1 Epidemic Intelligence Service, Epidemiology Program Office, Division of Reproductive Health and 2 Assisted Reproductive Technology Epidemiology Team, Women's Health and Fertility Branch, Division of Reproductive Health, Centers for Disease Control and Prevention, Atlanta, GA 30341-3724, USA

3 To whom correspondence should be addressed at: Epidemic Intelligence Service, Epidemiology Program Office, Division of Reproductive Health, Centers for Disease Control and Prevention, 4770 Buford Highway NE, MS K-34, Atlanta, Georgia 30341-3724, USA. Email: dkissin{at}cdc.gov


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
BACKGROUND: Multiple births are associated with serious adverse infant and maternal outcomes. The objective of this study was to assess the multiple-birth risk (MBR) associated with IVF and determine whether the risk is impacted by stage of embryo development at transfer. METHODS: A population-based sample of 50 819 IVF transfers utilizing day 3 or day 5 embryos performed in the USA in 2001 on women aged 20–40 years was used to assess MBR and live-birth rate (LBR), stratified by patient age, supernumerary embryo availability, and number of embryos transferred. RESULTS: Although significantly more day 5 than day 3 transfers used ≤2 embryos (69.2 versus 27.7%), the former were not associated with decreased MBR. MBR was high when >1 embryo was transferred, irrespective of embryo development stage. LBR were generally maximized with 2 embryos transferred, and for some (day 5 transfers, patients aged 35–37 years) with one embryo. Electing to transfer a single day 5 embryo appeared efficacious for some patients: women aged 20–37 years with supernumerary embryos cryopreserved had LBR of 31.6–39.5%. CONCLUSIONS: MBR is high when ≥2 embryos are transferred. Single embryo transfer is the only way to prevent many multiple births and associated adverse health outcomes.

Key words: blastocyst/extended culture/IVF/multiple pregnancy


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Multiple pregnancy and births are the most common and most serious complications of assisted reproductive technology. Multiple births are associated with higher rates of preterm delivery, low birthweight, congenital anomalies, infant death, and disability among survivors (Martin and Park, 1999Go; ESHRE, 2000Go; Martin et al., 2003Go; Mathews et al., 2003Go). Carrying a multiple gestation pregnancy is associated with adverse maternal outcomes, such as haemorrhage, pregnancy-induced hypertension and anaemia, as well as maternal mortality (Senat et al., 1998Go; Nowak et al., 2003Go). The rate of multiple births in the USA has increased markedly during the past 2 decades, largely due to the increased use of fertility treatments, including assisted reproductive technology (Martin and Park, 1999Go; Reynolds et al., 2003Go). Transferring multiple embryos in order to increase the chance of pregnancy and live birth is a common practice among assisted reproductive technology providers, and it results in a high proportion of multiples among assisted reproduction-conceived infants. In 2001, 54% of infants born as a result of assisted reproductive treatment were multiples, compared with only 3% in the general population (Wright et al., 2004Go).

The desire to decrease multiple-birth risk (MBR) while maintaining a high live-birth rate has prompted researchers to investigate techniques that will improve the ability to identify high-quality embryos with greater potential for implantation, thus allowing acceptable rates of live birth to be maintained while transferring fewer embryos. One technique currently proposed as a method that can reduce MBR by achieving high pregnancy rates while transferring fewer embryos is the use of extended embryo culture (Gardner et al., 1998Go; Vidaeff et al., 2000Go).

The recent commercial availability of sequential culture media has greatly increased assisted reproductive technology providers' ability to extend the duration of embryo culture to 5 days, corresponding with the blastocyst stage of development. The technique allows selection of embryos that have improved viability and proven developmental capacity. It also allows for increased synchronization between the embryo and endometrium at the time of embryo transfer. Although the clinical advantages of extended embryo culture continue to be debated, several clinical studies report higher success rates with blastocyst transfers than with cleavage stage transfers (Milki et al., 2000Go; Karaki et al., 2002Go; Van der Auwera et al., 2002Go; Frattarelli et al., 2003Go). Day 5 transfer has become a routine procedure in many clinics, necessitating research on the association between blastocyst transfers and MBR (Wright et al., 2004Go). Recent studies have suggested that transferring fewer blastocysts does not lead to a reduction in multiple births, as expected (Coskun et al., 2000Go; Levron et al., 2002Go; Emiliani et al., 2003Go). Those studies, however, were limited by small sample sizes, which precluded detailed analyses of patient and other treatment factors.

Although we previously published an analysis of MBR based on the population of patients receiving assisted reproductive technology in the USA in 1996 (Schieve et al., 1999Go), at that time, nearly all embryo transfers were limited to cleavage stage embryos. The current popularity of transferring blastocysts necessitates a re-examination of MBR with particular focus on the effects of the stage of embryo development at transfer. We sought to assess MBR and factors associated with it among the population of patients undergoing IVF in the USA in 2001. We also assessed pregnancy and live-birth rates associated with these procedures. This study was approved by the Centers for Disease Control and Prevention (CDC) institutional review board.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The Fertility Clinic Success Rate and Certification Act of 1992 (US Department of Health and Human Services, 1992Go) mandates that every US medical centre performing assisted reproductive technology report its pregnancy success rate data annually to the CDC. The Society for Assisted Reproductive Technology (SART) annually creates a database of assisted reproductive technology cycles performed in US clinics (both clinics that are members of SART and clinics that are not) and shares the data with CDC under contract.

Our analyses include assisted reproductive technology cycles that were initiated in 2001. We used data from 384 clinics, an estimated 91% of the clinics that were in operation in 2001 (Centers for Disease Control and Prevention, 2003Go). Although we do not have complete information on the reasons for non-reporting on the part of some clinics, anecdotal reports suggest that many non-reporting clinics were smaller in size than average practices. Thus, we estimate that our data represent >91% of assisted reproductive technology cycles performed in the USA in 2001.

From the 107 587 assisted reproductive technology cycles initiated in 2001, we selected those that progressed to the transfer of ≥1 embryo (89 162 cycles). Of those, we selected the most common assisted reproductive technology cycle: IVF with freshly fertilized embryos, derived from patient oocytes (64 289 cycles). IVF cycles, both with and without ICSI, were included. Due to sample size limits imposed by key stratification factors we included only cycles among patients aged 20–40 years at the time of cycle initiation (56 908 cycles). Cycles performed for patients aged <20 (10 cycles) and patients aged >40 (7371 cycles) were excluded. We further excluded cycles in which embryos were transferred on days 2, 4 or 6 (11% of cycles). After the exclusions, a total of 50 819 IVF cycles remained for analysis: 43 265 cycles in which embryos were transferred on day 3 of culture and 7554 cycles in which embryos were transferred on day 5.

We examined total pregnancy rate, live-birth rate, multiple gestation risk (MGR) and MBR. Pregnancy was defined as a clinical intrauterine gestation detected on ultrasound, generally during the first trimester, and live-birth delivery was defined as the delivery of ≥1 liveborn infant. A live-birth delivery was classified as a multiple-birth delivery if ≥2 infants were delivered and ≥1 of them was liveborn. We examined the percentage of IVF cycles resulting in pregnancies and live-birth deliveries and the percentage of live-birth deliveries that were multiple births. Additionally, because procedures resulting in pregnancies with >2 fetuses are more likely to involve spontaneous or medical reductions in the number of fetuses, we examined the percentage of pregnancies that had been multiple gestations as the total potential for multiple births. A multiple gestation was defined as a pregnancy in which >1 fetal heart was detected on early ultrasound or that resulted in the birth of >1 infant. Because all multiple births are associated with substantially higher rates of infant morbidity and mortality than singletons, we did not separate either MGR or MBR by the number of fetuses or infants. All indices were assessed after stratifying the data according to patient age, number of embryos transferred, and cryopreservation of supernumerary embryos [all of which are factors known to be important predictors of MBR (Templeton and Morris, 1998Go; Schieve et al., 1999Go)], as well as duration of embryo culture.

Patient age at cycle initiation was categorized as 20–29, 30–34, 35–37 and 38–40 years, and the number of embryos transferred was categorized as 1, 2, 3, 4 and ≥5. Because clinical information on embryo quality is not collected as part of the national Assisted Reproductive Technology Registry, embryo availability was used as a proxy measure. Embryo availability indicates that supernumerary embryos from the same treatment cycle were cryopreserved for future use; in such cases, the clinician would have chosen the embryos with the seemingly highest quality for transfer in the IVF cycle included in the study. Day of embryo transfer was used as a proxy measure for stage of embryo development, because specific data were not available for laboratory assessment of embryo developmental stage or for type of culture medium. Day of transfer was calculated by subtracting the oocyte retrieval date from the embryo transfer date. Typically, embryos transferred on day 3 correspond to the cleavage stage of development, whereas embryos transferred on day 5 correspond to the blastocyst stage.

To assess whether stratification by key variables was sufficient to control for any differences among patient and treatment groups, we repeated our analyses restricting to even more select patient groups, defined as the ‘best’ and the ‘worst’ prognosis groups, based on available variables consistently shown to be predictors of pregnancy and live-birth rates in our population-based registry. The ‘best’ patient prognosis group consisted of cycles meeting all of the following criteria: (i) ≥10 oocytes retrieved, (ii) no assisted hatching used, (iii) no ICSI performed in the absence of male factor infertility diagnosis, (iv) documented day 3 FSH level that was not elevated, (v) no diagnosis of diminished ovarian reserve, and (vi) either first assisted reproductive technology attempt or a history of live birth(s). In total, 6839 transfer cycles met the criteria for the best patient prognosis group. This number is likely smaller than the true number of best prognosis procedures because FSH data were missing for 35% of all transfers; thus none of these procedures could be classified as best prognosis. The ‘worst’ patient prognosis group consisted of cycles in which at least one of the following was true: (i) <5 oocytes retrieved, (ii) diagnosis of diminished ovarian reserve, or (iii) elevated FSH level. A total of 8316 transfer cycles was included in the worst patient prognosis group.

Differences in patient and treatment characteristics, pregnancy and live-birth rates, MGR and MBR were evaluated using {chi}2-tests.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In 43 265 (85%) of the 50 819 IVF cycles included in our final study population day 3 embryos were transferred, while in 7554 (15%) of the cycles day 5 embryos were used. The proportion of IVF cycles in which supernumerary embryos were cryopreserved was 30% among day 3 transfers, compared with 49% among day 5 transfers (P<0.0001). Initial analyses revealed substantial variation in apparent patient prognosis by both duration of embryo culture and cryopreservation of supernumerary embryos. Thus, we combined these two surrogate measures of embryo quality to create four separate presumed ‘embryo quality’ groups: group I consisted of day 3 embryo transfers in which no extra embryos were cryopreserved for future use (59.8%); group II consisted of day 3 embryo transfers in which supernumerary embryos were available and cryopreserved (25.3%); group III consisted of day 5 embryo transfers in which no extra embryos were cryopreserved (7.5%); group IV consisted of day 5 embryo transfers in which supernumerary embryos were cryopreserved (7.4%). Results are presented for these four presumed embryo quality groups.

The percentage distribution of patient and treatment characteristics among the four presumed embryo quality groups is presented in Table I. A comparison of the general patient profiles for each group suggests that patients in group I have the poorest prognosis in terms of factors associated with successful embryo implantation (i.e. higher proportion of patients: in the two oldest age categories; with a diagnosis of diminished ovarian reserve or both male and female infertility; with prior assisted reproductive technology cycles; and with elevated level of day 3 FSH). In contrast, the patient profile for group IV suggests the best prognosis [i.e. higher proportion than other groups: in the two youngest patient age categories; with diagnoses of tubal factor infertility and ovulatory dysfunction; with no prior assisted reproductive technology cycles (despite having similarly high rates of previous pregnancies and births as the other three groups); and with normal day 3 FSH level]. Additionally, among cycles included in group I, notably fewer oocytes were retrieved, a higher proportion of ICSI use without male factor infertility diagnosis took place and more frequent use of assisted hatching occurred than in other groups; again group IV had the largest contrast on each treatment factor. Groups II and III had similar values for many key patient and treatment factors, which tended to be intermediate between groups I and IV.


View this table:
[in this window]
[in a new window]
 
Table I. Percentage distribution of patient and treatment characteristics by embryo stage and embryo availability, IVF cycles, 2001a

 
With respect to number of embryos transferred, the presumed embryo quality groups with day 5 embryos (groups III and IV) were more similar to each other than to those with day 3 embryos (groups I and II). A sharp difference in the distribution across groups was found. For example, the proportion transferring ≤2 embryos was 26.9, 29.7, 58.3 and 80.3% for groups I through IV, in order. The distributions of patient and treatment characteristics presented in Table I were assessed separately for each age group and were found to be similar (data not shown). Thus, the differences observed between the presumed embryo quality groups were not merely attributable to the different age distributions. Because of the substantial differences in patient and treatment profiles, we continued to use the presumed embryo quality grouping as a key stratification variable throughout the remaining analyses.

Of 50 819 transfer cycles, 22 279 (43.8%) resulted in a clinical intrauterine pregnancy. Substantial variation in pregnancy rates was found across strata in terms of both patient age and embryo quality (Table II). In general, pregnancy rates were lowest for the group with the poorest presumed embryo quality (group I), intermediate for embryo quality groups II and III, and highest for the highest presumed embryo quality group (group IV). Moreover, an inverse association was found between patient age and pregnancy rate for the poorest presumed embryo quality group (group I): the pregnancy rate associated with 2 embryos transferred was nearly 44% for women aged 20–29 years, compared to 22.0% for women aged 38–40 years. As embryo quality increased, the age effect lessened, such that no clear trend in pregnancy rate was seen for the highest presumed embryo quality group (group IV). For group IV, the pregnancy rate associated with transferring 2 embryos was 64.1% for women aged 20–29 years, whereas the corresponding rates for women aged 30–34, 35–37, and 38–40 years were 65.5, 56.9 and 59.7% respectively.


View this table:
[in this window]
[in a new window]
 
Table II. Pregnancy rate (%) by number of embryos transferred, patient's age, stage of embryo development, and embryo availability, IVF cycles, 2001a

 
No significant increase in pregnancy rate occurred when 2 versus 3 embryos were transferred for women of any age in groups II, III or IV. In fact, among women aged 35–37 years in embryo quality groups III and IV, no significant difference was found between pregnancy rates achieved with double embryo and single embryo transfer (SET) (48.3 versus 45.8% for group III and 56.9 versus 54.2% for group IV). Among women aged ≥30 years in group I, pregnancy rates were increased by transferring >2 embryos.

The overall live-birth rate was 36.6%. Although lower than pregnancy rates in all but a few instances, live-birth rates had a pattern of results similar to pregnancy rate results (Table III). Generally, highest live-birth rates were achieved in the highest presumed embryo quality group (group IV), whereas the lowest rates were associated with the poorest presumed embryo quality group (group I). Additionally, increasing the number of transferred embryos from 2 to 3 did not increase the chance of a live birth for most patient age and presumed embryo quality groups.


View this table:
[in this window]
[in a new window]
 
Table III. Live-birth rate (%) by number of embryos transferred, patient's age, stage of embryo development, and embryo availability, IVF cycles, 2001a

 
The overall risk of multiple gestations for cycles resulting in pregnancy was 40.0%. When 2 embryos were transferred, an inverse association of MGR with patient age was found (e.g. for group I, MGR was 28.4% for the youngest age group versus 16.1% for the oldest age group). A positive association between MGR and presumed embryo quality group also was observed, such that group IV had the highest MGR within each age group (Table IV). For example, among women aged 20–29 years, group IV had an MGR of 53.7% compared with 28.4% for group I. Even among patients aged 38–40 years, the MGR was almost 30% for group IV but 16.1% for group I. With 3 embryos transferred, MGR was >30% in all but one group (group I patients aged 38–40 years had an MGR of 25.0%) and reached 56.6% for women in group IV aged 38–40 years.


View this table:
[in this window]
[in a new window]
 
Table IV. Multiple-gestation risk (%) by number of embryos transferred, patient age, stage of embryo development, and embryo availability, IVF cycles, 2001a

 
The total MBR was 36.8%. The pattern of results by age, presumed embryo quality group, and number of embryos transferred was similar to that observed with MGR (Table V). When 2 embryos were transferred, MBR declined with age from 23.2 to 13.8% for the youngest to oldest age category in group I, and from 52.1 to 25.0% for the youngest to oldest age category in group IV. In general, when 2 embryos were transferred, group IV had the highest MBR.


View this table:
[in this window]
[in a new window]
 
Table V. Multiple-birth risk (%) by number of embryos transferred, patient age, stage of embryo development, and embryo availability, IVF cycles, 2001aGo

 
Results from the analyses of pregnancy rates, live-birth rates, MGR, and MBR restricted to the subsample of best (n=6839) and worst (n=8316) patient prognosis groups were generally consistent with the pattern of results found for the original population. Although not all subgroups could be evaluated due to sample size limitations, rates for the best patient prognosis group were generally higher than for the worst patient prognosis group, as expected. For example, overall pregnancy rate was 51.8% in the best patient prognosis group but 31.4% in the worst patient prognosis group. The overall MGR was 43.6% in the best patient prognosis group and 31.4% in the worst. Selected data for pregnancy rate and MGR from these analyses are presented in Table VI. The pattern of results for outcome variables by patient age, presumed embryo quality group, and number of embryos transferred in both the best and the worst patient prognosis groups was generally similar to that observed in the original sample. Pregnancy rates for the best and worst patient prognosis groups varied substantially across strata based on both patient age and presumed embryo quality. In general, pregnancy rates were lowest for group I and highest for group IV. An inverse association between patient age and pregnancy rate was found for the poorest presumed embryo quality group only. In most cases, no significant increase in pregnancy rate occurred when 2 versus 3 embryos were transferred for embryo quality groups II, III or IV. MGR remained high, and group IV had generally the highest MGR in both the best and the worst patient prognosis groups. Additionally, an inverse association of MGR with patient age was evident in both patient prognosis groups; however, in the best patient prognosis group the association was only observed in group I, rather than being evident in all four presumed embryo quality groups, as was the case with the original sample.


View this table:
[in this window]
[in a new window]
 
Table VI. Pregnancy rate (%) and multiple-gestation risk by patient's age(%), stage of embryo development, and embryo availability, besta and worsta patient prognosis groups, 2 and 3 embryos transferred, IVF cycles, 2001b

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Previous studies have shown that number of embryos transferred, patient age and embryo availability are important predictors of multiple gestations and births (Templeton and Morris, 1998Go; Schieve et al., 1999Go). Our analyses revealed that the number of days in culture and, thus, stage of embryo development at transfer, is another important determinant of MGR and MBR. This factor needs to be considered when deciding on the number of embryos to transfer.

Day 5 transfer has become an increasingly viable assisted reproduction treatment option since 1999, and many providers have looked to this promising technique as a feasible opportunity to reduce the number of embryos transferred. Our population-based data show that cycles using day 5 embryos were, in fact, associated with fewer embryos transferred than day 3 transfer cycles. The reduction in number of embryos transferred, however, did not translate into decreased risk for multiple gestations or births. Although in most day 5 IVF cycles, only 2 embryos were transferred, the increased implantation rate associated with day 5 embryos led to an MGR nearly or exceeding 30% in all age groups and nearly or exceeding 40% in the two youngest age groups.

These analyses demonstrate that the chance of a successful assisted reproduction treatment cycle, as measured by pregnancy and live-birth rates, does not always increase linearly with each additional embryo transferred (see Tables II and III). For many cycles, transferring >2 embryos did not increase pregnancy or live-birth rates. Note that comparisons of various numbers of embryos transferred is perhaps most reasonable for the group of women known to have extra embryos available and, thus, known to have electively selected fewer embryos for transfer than were available. Among all such cycles included in this 2001 study population (both day 3 and day 5 transfers in every age group), rates of pregnancy and live birth did not improve when the number of embryos transferred increased from 2 to 3. In fact, the rates did not significantly improve from 1 to 2 embryos transferred for some groups (day 5 transfers, patients aged 35–37 years). As noted, however, limiting transfer to 2 embryos still posed a substantial MGR and MBR. Thus, success rates of single embryo transfer are important to consider. Although a few groups in this study did appear to benefit from transferring >2 embryos (i.e. those in the oldest age group with poorest presumed embryo quality), pregnancies in these groups still resulted in high MGR and MBR when high numbers of embryos were transferred.

This analysis is subject to certain limitations. The unit of analysis was an embryo transfer procedure rather than a patient who underwent assisted reproduction treatment. Therefore, it is possible that women undergoing >1 embryo transfer in a given year (who are also most likely to be those who failed at least one treatment) are included in the analysis more than once. This lack of independence could have resulted in underestimation of the true per-patient pregnancy rates and live-birth rates, for which the denominator is the number of IVF transfers. It is highly unlikely that a woman would achieve >1 live birth during a 1 year period, and therefore we do not think that analyses of MBR are affected.

An important limitation to the data is that women were not randomized for number of embryos transferred or for receipt of day 3 or day 5 embryos. The number of embryos transferred and stage of embryo development at transfer were based on patient and provider choice. If a large number of transfer cycles were from clinics with policies to exclusively transfer embryos at either cleavage or blastocyst stage, our results could be influenced by clinic practices that would be difficult to separate from the effect of the embryo stage. While we do not have data on clinic transfer polices, we found that of the 384 clinics included in these analyses, only a limited number exclusively transferred either day 3 (n=42 clinics) or day 5 embryos (n=5 clinics). Moreover, many of these 47 practices tended to be small: 87% had <100 transfers, and 60% had <50 transfers. Thus, this practice pattern might simply reflect patient and provider decision, rather than actual clinic policy. Regardless, these clinics (n=47) contributed only 5% of transfer cycles to our sample.

Nonetheless, it is likely that multiple personal and clinical factors played a role in the decision to transfer a certain number of embryos at a certain stage of development. Although analyses limiting to cycles with known extra embryos does address some of the heterogeneity present in the population regarding number of embryos transferred, unmeasured differences between various patient and treatment groups likely remained. As observed in our early descriptive analysis, both days in culture and cryopreservation of supernumerary embryos showed variation with the few potential predictors of prognosis available in the dataset. In our analyses we controlled for the two most predictive factors of both success and multiple gestation and birth risk available to us: patient age and presumed embryo quality group. Nevertheless, residual differences likely remained. Our analyses of the ‘best’ and ‘worst’ prognosis groups, however, did not show different patterns of MGR and MBR distribution by patient age, presumed embryo quality group, or number of embryos transferred from the original sample. Thus, we believe that our results are robust even though they are observational.

To date, published data do not provide conclusive evidence on whether transferring embryos at the blastocyst stage versus the cleavage stage is advantageous for achieving pregnancy and birth (Blake et al., 2002Go). Our study supports a positive association. Even though uncertainty remains, blastocyst transfers have become widely available during the past several years. Our population-based study suggests that many clinics may be limiting embryo transfer to 2 blastocysts to avoid high-order multiple gestations. Our study also shows, however, that transferring 2 embryos that have been cultured for 5 days is associated with high MGR and MBR. This finding necessitates consideration of single-blastocyst transfer.

Our study suggests that transferring a single embryo, either cleavage or blastocyst, appears to be the only way to reduce MGR and MBR in all subgroups. Our analyses showed, however, that SET is an extremely rare procedure. Although we could not fully assess the efficacy of SET due to sample size limitations, our data are suggestive that SET may be efficacious for some groups of women. Other research supports the feasibility of SET, with pregnancy rates ranging from 29.7 to 60.9% in specific groups of women (Gerris et al., 1999Go; Vilska et al., 1999Go; Dhont, 2001Go; Martikainen et al., 2001Go; Kovacs et al., 2003Go; Gardner et al., 2004Go; Milki et al., 2004Go; Thurin et al., 2004Go). Further studies on SET efficacy that would address the issue of generalizability of the evidence are strongly encouraged.


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Blake D, Proctor M, Johnson N and Olive D (2002) Cleavage stage versus blastocyst stage embryo transfer in assisted conception. Cochrane Database Syst Rev 2, CD002118.[Medline]

Centers for Disease Control and Prevention, American Society for Reproductive Medicine, Society for Assisted Reproductive Technology, RESOLVE (2003) 2001 assisted reproductive technology success rates. Centers for Disease Control and Prevention, Atlanta, USA, pp. 1–505.

Coskun S, Hollanders J, Al-Hassan S, Al-Sufyan H, Al-Mayman H and Jaroudi K (2000) Day 5 versus day 3 embryo transfer: a controlled randomized trial. Hum Reprod 15, 1947–1952.[Abstract/Free Full Text]

Dhont M (2001) Single-embryo transfer. Semin Reprod Med 19, 251–258.[CrossRef][ISI][Medline]

Emiliani S, Delbaere A, Vannin AS, Biramane J, Verdoodt M, Englert Y and Devreker F (2003) Similar delivery rates in a selected group of patients, for day 2 and day 5 embryos both cultured in sequential medium: a randomized study. Hum Reprod 18, 2145–2150.[Abstract/Free Full Text]

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

Frattarelli JL, Leondires MP, McKeeby JL, Miller BT and Segars JH (2003) Blastocyst transfer decreases multiple pregnancy rates in in vitro fertilization cycles: a randomized controlled trial. Fertil Steril 79, 228–230.[CrossRef][ISI][Medline]

Gardner DK, Schoolcraft WB, Wagley L, Schlenker T, Stevens J and Hesla J (1998) A prospective randomized trial of blastocyst culture and transfer in in-vitro fertilization. Hum Reprod 13, 3434–3440.[Abstract]

Gardner DK, Surrey E, Minjarez D, Leitz A, Stevens J and Schoolcraft WB (2004) Single blastocyst transfer: a prospective randomized trial. Fertil Steril 81, 551–555.[CrossRef][ISI][Medline]

Gerris J, De Neubourg D, Mangelschots K, Van Royen E, Van de Meerssche M and Valkenburg M (1999) Prevention of twin pregnancy after in-vitro fertilization or intracytoplasmic sperm injection based on strict embryo criteria: a prospective randomized clinical trial. Hum Reprod 14, 2581–2587.[Abstract/Free Full Text]

Karaki RZ, Samarraie SS, Younis NA, Lahloub TM and Ibrahim MH (2002) Blastocyst culture and transfer: a step toward improved in vitro fertilization outcome. Fertil Steril 77, 114–118.[CrossRef][ISI][Medline]

Kovacs G, MacLachlan V, Rombauts L, Healy D and Howlett D (2003) Replacement of one selected embryo is just as successful as two embryo transfer, without the risk of twin pregnancy. Aust NZ J Obstet Gynaecol 43, 369–371.[CrossRef][ISI][Medline]

Levron J, Shulman A, Bider D, Seidman D, Levin T and Dor J (2002) A prospective randomized study comparing day 3 with blastocyst-stage embryo transfer. Fertil Steril 77, 1300–1301.[CrossRef][ISI][Medline]

Martikainen H, Tiitinen A, Tomas C, Tapanainen J, Orava M, Tuomivaara L, Vilska S, Hyden-Granskog C and 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]

Martin JA and Park MM (1999) Trends in twin and triplet births: 1980–97. Natl Vital Stat Rep 47, 1–16.[Medline]

Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F and Munson ML (2003) Births: final data for 2002. Natl Vital Stat Rep 52, 1–113.[Medline]

Mathews TJ, Menacker F and MacDorman MF (2003) Infant mortality statistics from the 2001 period linked birth/infant death data set. Natl Vital Stat Rep 52, 1–28.[Medline]

Milki AA, Hinckley MD, Fisch JD, Dasig D and Behr B (2000) Comparison of blastocyst transfer with day 3 embryo transfer in similar patient populations. Fertil Steril 73, 126–129.[CrossRef][ISI][Medline]

Milki AA, Hinckley MD, Westphal LM and Behr B (2004) Elective single blastocyst transfer. Fertil Steril 81, 1697–1698.[CrossRef][ISI][Medline]

Nowak E, Blickstein I, Papiernik E and Keith L (2003) Iatrogenic multiple pregnancies. Do they complicate perinatal care? J Reprod Med 48, 601–609.[ISI][Medline]

Reynolds MA, Schieve LA, Martin JA, Jeng G and Macaluso M (2003) Trends in multiple births conceived using assisted reproductive technology, United States, 1997–2000. Pediatrics 111, 1159–1162.[Abstract/Free Full Text]

Schieve LA, Peterson HB, Meikle SF, Jeng G, Danel I, Burnett NM and Wilcox LS (1999) Live-birth rates and multiple-birth risk using in vitro fertilization. J Am Med Assoc 282, 1832–1838.[Abstract/Free Full Text]

Senat MV, Ancel PY, Bouvier-Colle MH and Breart G (1998) How does multiple pregnancy affect maternal mortality and morbidity? Clin Obstet Gynecol 41, 78–83.[Medline]

Templeton A and Morris JK (1998) Reducing the risk of multiple births by transfer of two embryos after in vitro fertilization. New Engl J Med 339, 573–577.[Abstract/Free Full Text]

Thurin A, Hausken J, Hillensjö T, Jablonowska B, Pinborg A, Strandell A and Bergh C (2004) Elective single-embryo transfer versus double-embryo transfer in in vitro fertilization. N Engl J Med 351, 2392–2402.[Abstract/Free Full Text]

US Department of Health and Human Services (1992) Fertility Clinic Success Rate and Certification Act of 1992 (FCRSCA). Section 2(a) of Pub L No. 102–493, October 24, 1992.

Van der Auwera I, Debrock S, Spiessens C, Afschrift H, Bakelants E, Meuleman C, Meeuwis L and D'Hooghe TM (2002) A prospective randomized study: day 2 versus day 5 embryo transfer. Hum Reprod 17, 1507–1512.[Abstract/Free Full Text]

Vidaeff AC, Racowsky C and Rayburn WF (2000) Blastocyst transfer in human in vitro fertilization. A solution to the multiple pregnancy epidemic. J Reprod Med 45, 529–539.[ISI][Medline]

Vilska S, Tiitinen A, Hyden-Granskog C and Hovatta O (1999) Elective transfer of one embryo results in an acceptable pregnancy rate and eliminates the risk of multiple birth. Hum Reprod 14, 2392–2395.[Abstract/Free Full Text]

Wright VC, Schieve LA, Reynolds MA, Jeng G and Kissin D (2004) Assisted reproductive technology surveillance—United States, 2001. MMWR Surveill Summ 53, 1–20.[Medline]

Submitted on January 10, 2005; resubmitted on February 25, 2005; accepted on March 7, 2005.