Hansen, Michele MPH; Kurinczuk, Jennifer J. MD; de Klerk, Nicholas PhD; Burton, Peter PhD; Bower, Carol MBBS, PhD
Children born after assisted reproductive technology treatment currently represent 3.3% of Australian births, and there has been an important shift in this country toward single embryo transfer leading to major improvements in perinatal outcomes after assisted reproductive technology.1,2 The percentage of babies conceived through assisted reproductive technology in other developed countries varies from just more than 1% in the United States to almost 5% in Denmark, with the largest European countries (Germany, France, United Kingdom, and Italy) ranging from 1.2% to 1.8%.3,4
Our previous record linkage study5 examined the birth prevalence of birth defects in children born as a result of assisted reproductive technology treatment in Western Australia from 1994 to 1998. The results of that study indicated a two-fold increased risk of a major birth defect being diagnosed by 1 year of age compared with a random sample of 4,000 nonassisted reproductive technology births. For the present study, we were able to increase our original sample size by including all births in Western Australia from 1994 to 2002, as well as birth defects diagnosed at up to 6 years of age and information about all terminations of pregnancy for fetal anomaly occurring over the study period. Our aim was to estimate birth-defect prevalence overall and for a range of different assisted reproductive technology techniques (in vitro fertilization [IVF], intracytoplasmic sperm injection, fresh embryo transfer, frozen-thawed embryo transfer). In the absence of early ultrasound data, we compared birth-defect prevalence in assisted reproductive technology singletons and twins according to the number of embryos transferred. We also examined whether birth-defect prevalence had changed over time comparing data from the time period of our earlier study (1994–September 1998) with the additional data available in this study (October 1998–December 2002).
MATERIALS AND METHODS
Western Australian fertility clinics have reported the details of all assisted reproductive technology procedures performed in Western Australia since April 1993 to the statutory Reproductive Technology Register.6 These name-identified data were linked to the Midwives' Notification of Births System, which collects demographic and clinical information about all births from 20 weeks of gestation (or birthweight of 400 g or more), including stillbirths, regardless of the place of delivery and accoucheur within the state of Western Australia. This allowed us to identify all assisted reproductive technology and, by subtraction, all nonassisted reproductive technology births in Western Australia from January 1994 to December 2002. These data were then linked to the Western Australian Register of Developmental Anomalies (formerly called the Western Australian Birth Defects Registry) to identify all live births and stillbirths in which a birth defect was diagnosed. Reproductive technology register data also were linked directly to the Western Australian Register of Developmental Anomalies to identify all terminations of pregnancy for fetal anomaly after assisted reproductive technology and, by subtraction, all nonassisted reproductive technology terminations of pregnancy for fetal anomaly.
The Western Australian Register of Developmental Anomalies collects information on birth defects occurring in live births and stillbirths delivered in Western Australia, and on pregnancies terminated because of fetal anomalies (regardless of the length of gestation).7 Birth defects are defined as structural or functional abnormalities that are present at conception or occur before the end of pregnancy and are diagnosed by 6 years of age.8 Each individual defect (up to a maximum of 10 defects per case) is coded according to British Pediatric Association International Classification of Diseases, 9th revision (ICD-9).7 All defects are classified as major or minor according to a method devised by the Centers for Disease Control and Prevention. Most minor defects are excluded unless they are disfiguring or require treatment. A list of exclusions can be found in the Western Australian Register of Developmental Anomalies Annual Report.7 Cases are reported to the Register by multiple statutory and voluntary sources with a high level of ascertainment and accuracy.9
Records were linked between the three registers by the Western Australian Health Department Data-Linkage Branch using probabilistic matching, and de-identified data files were provided to the researchers. Linked data from the Western Australian Data-Linkage Branch have been validated previously and have been used extensively for health research.10 Birth records were available for all assisted reproductive technology and nonassisted reproductive technology births in Western Australia from January 1994 to December 2002, and records of birth defects were available for those for whom a link was found within the Western Australian Register of Developmental Anomalies (including terminations of pregnancy for fetal anomaly). All individuals in our dataset had the potential for 6 years or more of follow-up, which is the upper age limit of recording of birth defects on the Western Australian Register of Developmental Anomalies. Approximately 3% of women conceiving by assisted reproductive technology in Western Australia were lost to follow-up because of out-of state migration over the course of the study period.11 This figure closely resembles the outward migration rate for the whole of the Western Australian population (approximately 2.8% per annum) and is low by international standards.12
Data relating to Aboriginal children (who account for 6% of all births in Western Australia) were excluded throughout, because Aboriginal women are much less likely to receive assisted reproductive technology treatment than other women in the population and their children have a lower prevalence of birth defects on the Western Australian Register of Developmental Anomalies before 2005 (thought to be attributable to under-ascertainment of birth defects in Aboriginal children in previous years).5,7 We also excluded births after gamete intrafallopian transfer as well as triplets and higher-order multiples because they were too few in number to allow meaningful comparisons in birth-defect prevalence.
Data analysis was performed using SPSS statistical software 19.0. Multivariable logistic regression was used to estimate odds ratios (ORs) and their 95% confidence intervals (CIs) to compare the relative odds of birth defects and other birth outcomes in assisted reproductive technology and nonassisted reproductive technology children. Singleton and twin births were analyzed separately. Birth-defect prevalence also was compared for children born as a result of IVF and intracytoplasmic sperm injection treatment and fresh compared with frozen-thawed embryo transfer.
The effects of the following covariates on the OR estimates were examined: maternal age (19 or younger, 20–24, 25–29, 30–34, 35–39, 40–44, 45 or older), year the child was born (as a categorical variable), and parity (0, 1–2, 3 or more). For analyses of rare defects, we were sometimes forced to reduce the number of subcategories within covariates. We estimated the prevalence of multiple major defects defined as two or more major birth defects affecting different systems as well as birth defects within a number of ICD-9 chapter headings. We tested for a difference in birth defect risk according to time period by adding time period (0=1994–September 1998, 1=October 1998–December 2002) as a covariate in the models including data for the whole study period. We tested for a trend in birth defect risk according to the number of embryos transferred per cycle by including numbers of embryos transferred (1, 2, or 3 or more) in a model restricted to assisted reproductive technology births.
We searched for any case of the following imprinting-related disorders in our data: Angelman syndrome; Beckwith-Wiedemann syndrome; Russell-Silver syndrome; Prader-Willi syndrome; pseudohypoparathyroidism type 1b; maternal and paternal uniparental disomy of chromosome 14; maternal hypomethylation syndrome; and transient neonatal diabetes (only the first five conditions were found).13
Termination of pregnancy data from the Western Australian Register of Developmental Anomalies were not linked to birth data from the Midwives' Notification of Birth System, so we were unable to tell which women delivering live children or stillborn fetuses in our dataset also may have had terminations of pregnancy for fetal anomaly. This meant that when we sought to allow for the potential correlation in results between births from the same mother using generalized estimating equations, we were unable to include terminations of pregnancy. Terminations of pregnancy for fetal anomaly data also were missing information about parity. Therefore, we have shown OR estimates including all births and cases of terminations of pregnancy for fetal anomaly adjusted for year of birth and maternal age, as well as OR estimates for births only when we were also able to adjust for parity and correlations within sibships. The first OR estimate, incorporating terminations of pregnancy for fetal anomaly data, represents our best estimate of the true difference in birth defect risk between assisted reproductive technology and nonassisted reproductive technology births. The second estimate, restricted to births only, may be more readily comparable with much of the data in the literature in which terminations of pregnancy for fetal anomaly data are often not available. An “exchangeable” correlation structure was specified for the generalized estimating equations analysis.
Comparisons for twins are shown for assisted reproductive technology twins compared with all nonassisted reproductive technology twin births regardless of zygosity and for assisted reproductive technology twins compared with nonassisted reproductive technology twins of unlike sex to allow for the differing proportions of monochorionic placentation in the two groups.
Given ethical restrictions on reporting data when the cell size is fewer than five, we were unable to show subgroup analyses for certain groups of conditions and for twins. The study had approval from the Human Research Ethics Committees of the Department of Health Western Australia and the University of Western Australia and the Western Australian Reproductive Technology Council.
The study included 2,911 assisted reproductive technology births and terminations of pregnancy for fetal anomaly: 1,972 singletons (1,328 IVF, 633 intracytoplasmic sperm injection, and 11 partial zona dissection or subzonal insemination births) and 939 twins (643 IVF, 292 intracytoplasmic sperm injection, and four partial zona dissection-subzonal insemination). There were 210,977 nonassisted reproductive technology births and terminations for fetal anomaly in Western Australia over the same time period: 205,641 singletons and 5,356 twins (including 1,619 twins of unlike sex).
Assisted reproductive technology mothers were older, less likely to have had a previous child, and less likely to smoke than no-assisted reproductive technology mothers (Table 1). They were more likely to be married or cohabiting and were five times more likely to have private health insurance than nonassisted reproductive technology mothers. Assisted reproductive technology neonates were more likely to be delivered by caesarean (both elective and emergency), to have low birthweight, and to be born before term. Differences between assisted reproductive technology and nonassisted reproductive technology births were greater for singletons than for twins; however, assisted reproductive technology twins were still more likely than nonassisted reproductive technology twins of unlike sex to be born preterm and to have low birthweight (Table 1).
Table 2 shows the prevalence of major birth defects diagnosed by 6 years of age in assisted reproductive technology singleton and twin children over the whole period (1994–2002) and for the two periods 1994–September 1998 and October 1998–2002. The earlier time period corresponds to the timeframe of our previous study, although that study included only major birth defects diagnosed by 1 year rather than 6 years of age.5 A major birth defect was diagnosed by 6 years of age in 172 assisted reproductive technology singletons (8.7%), including births and terminations of pregnancy. There were 11,078 nonassisted reproductive technology (5.4%) singletons with a major birth defect diagnosed. Overall assisted reproductive technology singletons were 1.68 times more likely than nonassisted reproductive technology singletons to have a major birth defect diagnosed by 6 years of age (95% CI 1.43–1.96). The OR decreased slightly when adjusted for maternal age and year of birth to 1.53 (95% CI 1.30–1.79), but it barely changed when further restricted to births (live and stillbirths) only and also adjusted for parity and correlation within sibships.
When we examined the prevalence of birth defects in our current dataset for the time period corresponding to our previous publication (1994—September 1998), we found that 10.9% of assisted reproductive technology singletons had a major birth defect diagnosed by 6 years of age compared with 5.6% of nonassisted reproductive technology singletons. The prevalence of birth defects in assisted reproductive technology singletons decreased markedly in the subsequent time period (from October 1998 to 2002) to 7.5%, whereas the prevalence of birth defects in nonassisted reproductive technology singletons decreased only marginally and nonsignificantly from 5.6% to 5.2%. The odds of an assisted reproductive technology singleton having a major birth defect diagnosed by 6 years of age in the later time period was 1.32 (95% CI 1.06–1.63) compared with 1.87 (95% CI 1.47–2.37) in the earlier period.
The prevalence of birth defects in assisted reproductive technology twins (7.1%) was lower than in assisted reproductive technology singletons (8.7%). The reverse was true for nonassisted reproductive technology twins. If we consider data from the whole study period, then assisted reproductive technology twins were no more likely to have a major defect diagnosed than nonassisted reproductive technology twins of unlike sex (OR 1.08, 95% CI 0.77–1.51) after adjustment for maternal age and year of birth. Nevertheless, the pattern of birth defect risk over time was similar to that seen in singletons with a higher birth-defect prevalence in the earlier time period (assisted reproductive technology twins 9.2%, nonassisted reproductive technology twins of unlike sex 6.5%) and a marked decrease in birth-defect prevalence in assisted reproductive technology twins for the subsequent time period (5.7%).
There was little difference in overall birth defect risk when intracytoplasmic sperm injection and IVF singletons were compared (OR 1.09, 95% CI 0.77–1.54) or when intracytoplasmic sperm injection twins were compared with IVF twins (OR 0.93, 95% CI 0.52–1.67). Fresh embryo transfer appeared to slightly, although not significantly, increase the risk of birth defects in assisted reproductive technology singletons compared with frozen-thawed embryo transfer (adjusted OR 1.21, 95% CI 0.88–1.68). This difference was not present for fresh and frozen-thawed assisted reproductive technology twins (OR 1.01, 95% CI 0.58–1.75). When we examined birth-defect prevalence in fresh compared with frozen-thawed embryo transfer within assisted reproductive technology treatment types, we found that those born after fresh IVF had a greater prevalence of birth defects than those born after frozen-thawed IVF for both singletons (9.7% compared with 7.3%) and twins (7.9% compared with 6.9%). For intracytoplasmic sperm injection, there was little difference between fresh and frozen-thawed embryo transfer (8.7% compared with 9.0% for singletons and 6.1% compared with 6.2% for twins).
Table 3 shows the prevalence of major birth defects in assisted reproductive technology singleton and twin children according to the number of embryos that were transferred in the treatment cycle that gave rise to their birth. The results are shown for the whole study and then grouped according to time period. Data for the whole study suggest a pattern of increasing birth defect risk with increasing number of embryos transferred, although the test for trend was nonsignificant (P=.131). When we looked for this trend in the two different time periods, it was no longer apparent and the main pattern we saw was a dramatic decrease in birth-defect prevalence between the early and late time periods even within births after double embryo transfer or transfers of three or more embryos. This pattern of decreasing birth-defect prevalence in the second time period was evident for both singleton and twin children across all three clinics contributing data to both time periods. Two new clinics came into operation in the second time period but contributed only a small number of births to the study.
When we restricted our analyses to the three larger clinics contributing data to both time periods, we saw a decrease in birth-defect prevalence over time for both IVF and intracytoplasmic sperm injection births, but it was much more pronounced for IVF singletons (11.4–6.2%, P=.001) than intracytoplasmic sperm injection singletons (10.3–8.5%, P=.467). The decrease in birth-defect prevalence also occurred for twins but was not statistically significant because of the smaller sample size. Table 4 shows the prevalence of birth defects in singleton assisted reproductive technology births in the different time periods grouped according to assisted reproductive technology treatment type: fresh IVF; frozen-thawed IVF; fresh intracytoplasmic sperm injection; or frozen-thawed intracytoplasmic sperm injection. A big decrease in birth-defect prevalence was seen from the first time period to the second for IVF singletons born after transferring both fresh and frozen-thawed embryos, and both of these were statistically significant (P<.05). For intracytoplasmic sperm injection singletons resulting from fresh embryo transfer, there was no change in birth-defect prevalence over time; however, there was a large (although statistically nonsignificant) decrease in birth-defect prevalence for intracytoplasmic sperm injection singletons born after frozen-thawed embryo transfer (P=.159).
The change in birth-defect prevalence over time for IVF twins mirrored that seen for IVF singletons, although comparisons from one time period to the other did not reach statistical significance because of smaller sample size (data not shown). The number of intracytoplasmic sperm injection twins was small in both time periods, particularly those conceived as a result of frozen-thawed embryo transfer. The prevalence of birth defects for intracytoplasmic sperm injection twins born after fresh transfer decreased markedly from the first time period (9.7%) to the second (4.2%), but this change did not reach statistical significance because of small sample size.
Table 5 shows the prevalence of major birth defects in assisted reproductive technology and nonassisted reproductive technology children according to a range of ICD-9 chapter headings and individual defect diagnoses. The OR estimates for assisted reproductive technology compared with nonassisted reproductive technology children also are shown. Results for a number of assisted reproductive technology subtypes, ie, IVF, intracytoplasmic sperm injection and fresh and frozen-thawed embryo transfer, are shown in Tables 1 and 2 in the Appendix (available online at http://links.lww.com/AOG/A314). Assisted reproductive technology singletons had a significantly greater risk of having multiple major defects (OR 1.97, 95% CI 1.38–2.80) as well as cardiovascular, musculoskeletal, gastrointestinal, and genital defects diagnosed compared with nonassisted reproductive technology singletons. Assisted reproductive technology singleton boys had an increased risk of having hypospadias and cryptorchidism diagnosed, even when we restricted our sample to boys born at term (37 weeks of gestation or more) (OR hypospadias 2.61, 95% CI 1.52–4.48; OR cryptorchidism 1.74, 95% CI 1.05–2.87).
Intracytoplasmic sperm injection singletons had a more than two-fold increased risk of having chromosomal (OR 2.32, 95% CI 1.26–4.26) and cardiovascular (OR 2.04, 95% CI 1.09–3.83) defects diagnosed. However, when we restricted cardiovascular defects to those that were not occurring in conjunction with a chromosomal defect, the prevalence decreased after intracytoplasmic sperm injection and was similar to that reported after IVF (Tables 1 and 2 in the Appendix).
Although their overall risk of having any major birth defect diagnosed was not increased, assisted reproductive technology twins were 2.36 times more likely to have multiple major defects diagnosed compared with nonassisted reproductive technology twins of unlike sex (95% CI 1.09–5.11). Assisted reproductive technology twins had increased ORs for cardiovascular, musculoskeletal and urinary defects, although urinary defects were the only subgroup with a statistically significant excess risk (OR 3.00, 95% CI 1.14–7.90). Assisted reproductive technology twin boys were at increased risk for cryptorchidism (OR 2.32, 95% CI 0.88–6.11) but not hypospadias.
Imprinting-related disorders (Angelman syndrome, Beckwith-Wiedemann syndrome, Prader-Willi syndrome) were diagnosed in three intracytoplasmic sperm injection singletons, in one IVF singleton (0.20% assisted reproductive technology), and in 30 nonassisted reproductive technology singletons (0.01%). No assisted reproductive technology twins were identified with an imprinting disorder. The odds of an assisted reproductive technology singleton having any imprinting-related disorder diagnosed after adjustment for maternal age and year of birth was 8.21 (95% CI 2.81–24.00).
There were 1,184 terminations of pregnancy for fetal anomaly in nonassisted reproductive technology singletons (5.8 per 1,000 births) and 19 terminations of pregnancy for fetal anomaly in assisted reproductive technology singletons (9.7 per 1,000 births). The rate of terminations of pregnancy for fetal anomaly in nonassisted reproductive technology twins was the same as that for nonassisted reproductive technology singletons (5.8 per 1,000 births; 31 terminations of pregnancy for fetal anomaly); however, there were only two terminations in assisted reproductive technology twin pregnancies (2.1 per 1,000 births).
The proportion of major birth defects diagnosed antenatally in singletons and twins did not differ according to assisted reproductive technology status (19.8% for both assisted reproductive technology and nonassisted reproductive technology singletons, 19.4% for assisted reproductive technology twins, and 19.9% for nonassisted reproductive technology twins). When we examined the proportion of fetuses with an antenatal diagnosis of a major birth defect in which the pregnancy then was terminated, there was again little difference between assisted reproductive technology and nonassisted reproductive technology singletons (55.9% compared with 53.5%). Although based on small numbers, and not statistically significant (P=.067), it appears that couples using assisted reproductive technology were less likely to terminate a twin pregnancy with an antenatal diagnosis of a major birth defect than couples with a nonassisted reproductive technology twin pregnancy (15.4% compared with 44.3%).
This study examined the prevalence of birth defects (including terminations of pregnancy for fetal anomaly) in all children born after assisted reproductive technology treatment in Western Australia compared with all nonassisted reproductive technology births over the course of a 9-year period. All children in the study had 6 years of follow-up, offsetting any potential bias from an initial higher level of scrutiny of assisted reproductive technology pregnancies and births. Singleton assisted reproductive technology children were approximately 50% more likely to have a major birth defect diagnosed by 6 years of age compared with nonassisted reproductive technology children in Western Australia. In contrast, assisted reproductive technology twins were not at increased risk for birth defects compared with nonassisted reproductive technology twins of unlike-sex when terminations of pregnancy for fetal anomaly were included as well as birth defects in live and stillbirths.
When data on terminations of pregnancy for fetal anomaly were excluded, assisted reproductive technology twins did appear to be at some increased risk for birth defects compared with nonassisted reproductive technology twins of unlike sex (OR 1.31, 95% CI 0.87–1.96), reflecting the lower rate of pregnancy termination in assisted reproductive technology compared with nonassisted reproductive technology twins. Couples in general may be less likely to terminate a pregnancy when only one twin is affected, particularly if it is likely to put the unaffected twin at considerable risk. This may be even more likely for parents who have had difficulty conceiving and have used assisted reproductive technology. The possibility of lower pregnancy termination rates in assisted reproductive technology twin pregnancies has been discussed by Kallen et al14,15 in their analyses of Swedish assisted reproductive technology data.
In keeping with most other studies, the risk of birth defects was similar among children conceived by IVF and intracytoplasmic sperm injection and after using fresh compared with frozen-thawed embryo transfer.16–18 The lowest birth-defect prevalence in assisted reproductive technology singleton subgroups was seen in IVF children conceived after frozen-thawed embryo transfer (7.3%).
When we compared the prevalence of birth defects in our current dataset for the time period corresponding to our previous study (1994–September 1998)5 and the subsequent time period (October 1998–December 2002), we found that the prevalence of birth defects diagnosed by 6 years of age decreased markedly for assisted reproductive technology singletons and twins from the first time period to the second. During the first time period, assisted reproductive technology singletons had an 87% increased risk of having a major birth defect diagnosed compared with nonassisted reproductive technology singletons. However, when data from 1998 to 2002 were considered, the excess risk was 32%. Kallen et al14 also reported a decrease in crude risk ratio for “relatively severe” malformations over time from 1.46 (1986–March 2001) to 1.27 (April 2001–2006). They postulate that differences in risk for certain malformations may be attributable to different case-mix during the two periods.
Assisted reproductive technology is a rapidly changing field and there have been many changes in clinical practice over time. Unfortunately, for the time period covered by this study, the Reproductive Technology Register has no information about the number of fetal hearts on early ultrasound examination or about a number of changes to clinical and laboratory practices. Western Australia clinicians report the following changes to clinical practice during our study period that may have affected birth-defect prevalence: using manufactured culture medium with improved formulations rather than medium prepared inhouse; using human serum albumin rather than maternal serum as a protein supplement in culture medium; a move away from large-volume incubators to mini-incubators, which had improved temperature and CO2 regulation; and a move away from urine-derived gonadotropins to gonadotropins produced through recombinant technologies. Changes to culture media and culture conditions of this kind were reported in the literature at approximately this time to improve embryo quality and pregnancy rates, suggesting that better-quality embryos were being transferred.19–23 The use of recombinant gonadotropins may have led to a lower dosage requirement to achieve ovulation and a slightly shorter stimulation period.24,25
We can exclude certain other changes that would not have influenced our results. Blastocyst culture and vitrification were not in use in Western Australia clinics during the study period. All embryo freezing used the traditional slow-freezing method and the selection of embryos for transfer were largely based on the number of blastomeres and the degree of fragmentation. In terms of case-mix, Western Australia clinicians report an increasing proportion of overweight and obese clients and increasing maternal age over time; however, both these changes would be expected to increase rather than decrease birth defect risk.
The information that is available on the reproductive technology register does allow us to conclude that the change in birth-defect prevalence was evident in children conceived by IVF and by intracytoplasmic sperm injection, and it could not be explained by the decrease in number of embryos transferred over time.
A number of studies suggest that subfertile couples have an increased risk of birth defects in their offspring, even without the use of assisted reproductive technology.26,27 We are unable to examine the effect of underlying infertility on birth defect risk because we did not have access to information about cause of infertility for couples undertaking assisted reproductive technology treatment in Western Australia. Finally, the reproductive technology register does not record treatment with ovulation induction or intrauterine insemination, so some of these children will have been included in the nonassisted reproductive technology comparison group. Given that use of these techniques has been associated with a small increase in birth defect risk,27,28 their inclusion in the nonassisted reproductive technology comparison group may have biased our results toward the null.
Our data on specific subgroups of birth defects by treatment type should be interpreted with caution because no adjustment was made for multiple comparisons. In agreement with many other studies, we found increased risks of cardiovascular, musculoskeletal, genital, and gastrointestinal defects in assisted reproductive technology compared with nonassisted reproductive technology singletons.14,15,27,29–32 We also found an increased risk of hypospadias and surgically treated cryptorchidism in assisted reproductive technology singleton boys. Similar to other studies,33 we found evidence of three conditions within our assisted reproductive technology cohort that are sometimes caused by imprinting errors (Beckwith-Wiedemann syndrome, Angelman syndrome, Prader-Willi syndrome); however, whether this was the case in these children is not known.
Based on a large record linkage study of all assisted reproductive technology and nonassisted reproductive technology births and birth defect registrations over the course of a 9-year period, our data suggest an encouraging decline in major birth-defect prevalence in children born as a result of assisted reproductive technology in Western Australia. Although we cannot explain this decrease with the data available, changes to clinical practice may be largely responsible with improved culture media and more optimal culture conditions leading to the transfer of “healthier” embryos.
1. Chambers GM, Illingworth PJ, Sullivan EA. Assisted reproductive technology: public funding and the voluntary shift to single embryo transfer in Australia. Med J Aust 2011;195:594–8.
2. Wang YA, Macaldowie A, Hayward I, Chambers GM, Sullivan EA. Assisted reproductive technology in Australia and New Zealand 2009. Assisted reproduction technology series number 15, category number PER 51. Canberra (Australia): AIHW; 2011.
3. Centers for Disease Control and Prevention, American Society for Reproductive Medicine, Society for Assisted Reproductive Technology. 2009 Assisted reproductive technology success rates: National summary and fertility clinic reports. Atlanta (GA): Centers for Disease Control and Prevention; 2011.
4. de Mouzon J, Goossens V, Bhattacharya S, Castilla JA, Ferraretti AP, Korsak V, et al.. Assisted reproductive technology in Europe, 2007: results generated from European registers by ESHRE. Hum Reprod 2012;27:954–66.
5. Hansen M, Kurinczuk JJ, Bower C, Webb S. The risk of major birth defects after intracytoplasmic sperm injection and in vitro fertilization. N Engl J Med 2002;346:725–30.
6. The Human Reproductive Technology Act 1991—Directions. Perth (Australia): Western Australian Reproductive Technology Council; 1997.
7. Bower C, Rudy E, Callaghan A, Quick J, Cosgrove P, Watson L. Report of the Western Australian Register of Developmental Anomalies 1980–2010. Perth (Australia): King Edward Memorial Hospital Women and Newborn Health Service; 2011.
9. Bower C, Ryan A, Rudy E. Ascertainment of pregnancies terminated because of birth defects: effect on completeness of adding a new source of data. Teratol 2001;63:23–5.
10. Holman CD, Bass AJ, Rouse IL, Hobbs MS. Population-based linkage of health records in Western Australia: development of a health services research linked database. Aust NZ J Public Health 1999;23:453–9.
11. The Western Australian Reproductive Technology Council. Annual reports spanning Reproductive Technology Register reporting years 1994–2002. Perth (Australia): Health Department of Western Australia; 1996–2004.
12. Clark A, Preen DB, Ng JQ, Semmens JB, Holman CD. Is Western Australia representative of other Australian States and Territories in terms of key socio-demographic and health economic indicators? Aust Health Rev 2010;34:210–5.
13. Amor DJ, Halliday J. A review of known imprinting syndromes and their association with assisted reproduction technologies. Hum Reprod 2008;23:2826–34.
14. Kallen B, Finnstrom O, Lindam A, Nilsson E, Nygren KG, Otterblad PO. Congenital malformations in infants born after in vitro fertilization in Sweden. Birth Defects Res A 2010;88:137–43.
15. Kallen B, Finnstrom O, Nygren KG, Olausson PO. In vitro fertilization (IVF) in Sweden: risk for congenital malformations after different IVF methods. Birth Defects Res A 2005;73:162–9.
16. Halliday JL, Ukoumunne OC, Baker HW, Breheny S, Jaques AM, Garrett C, et al.. Increased risk of blastogenesis birth defects, arising in the first 4 weeks of pregnancy, after assisted reproductive technologies. Hum Reprod 2010;25:59–65.
17. Lie RT, Lyngstadaas A, Orstavik KH, Bakketeig LS, Jacobsen G, Tanbo T. Birth defects in children conceived by ICSI compared with children conceived by other IVF-methods; a meta-analysis. Int J Epidemiol 2005;34:696–701.
18. Wennerholm UB, Soderstrom-Anttila V, Bergh C, Aittomaki K, Hazekamp J, Nygren KG, et al.. Children born after cryopreservation of embryos or oocytes: a systematic review of outcome data. Hum Reprod 2009;24:2158–72.
19. Andersen CY, Hay-Schmidt A, Byskov AG, Hornnes P, Starup J. [Human serum albumin as the protein source in culture of human oocytes, spermatozoa and pre-embryos]. Ugeskr Laeger 1991;153:3393–5.
20. Cooke S, Tyler JP, Driscoll G. Objective assessments of temperature maintenance using in vitro culture techniques. J Assist Reprod Genet 2002;19:368–75.
21. Fujiwara M, Takahashi K, Izuno M, Duan YR, Kazono M, Kimura F, et al.. Effect of micro-environment maintenance on embryo culture after in-vitro fertilization: comparison of top-load mini incubator and conventional front-load incubator. J Assist Reprod Genet 2007;24:5–9.
22. Quinn P. The development and impact of culture media for assisted reproductive technologies. Fertil Steril 2004;81:27–9.
23. Veeck Gosden L, Berrios R, Bodine R, Clarke RN, Zaninovic N. The human embryo: slow freezing. In: Gardner DK, Howles CM, Shoham Z, Weissman A, editors. Textbook of assisted reproductive technologies. 3rd ed. London (UK): Informa Healthcare; 2009.
24. Daya S. Follicle-stimulating hormone in clinical practice: an update. Treat Endocrinol 2004;3:161–71.
25. Goa KL, Wagstaff AJ. Follitropin alpha in infertility: a review. BioDrugs 1998;9:235–60.
26. Jaques AM, Amor DJ, Baker HW, Healy DL, Ukoumunne OC, Breheny S, et al.. Adverse obstetric and perinatal outcomes in subfertile women conceiving without assisted reproductive technologies. Fertil Steril 2010;94:2674–9.
27. Zhu JL, Basso O, Obel C, Bille C, Olsen J. Infertility, infertility treatment, and congenital malformations: Danish national birth cohort. BMJ 2006;333:679–83.
28. Reefhuis J, Honein MA, Schieve LA, Rasmussen SA, National Birth Defects Prevention Study. Use of clomiphene citrate and birth defects, National Birth Defects Prevention Study, 1997–2005. Hum Reprod 2011;26:451–7.
29. El-Chaar D, Yang Q, Gao J, Bottomley J, Leader A, Wen SW, et al.. Risk of birth defects increased in pregnancies conceived by assisted human reproduction. Fertil Steril 2009;92:1557–61.
30. Olson CK, Keppler-Noreuil KM, Romitti PA, Budelier WT, Ryan G, Sparks AE, et al.. In vitro fertilization is associated with an increase in major birth defects. Fertil Steril 2005;84:1308–15.
31. Pinborg A, Loft A, Aaris Henningsen AK, Rasmussen S, Andersen AN. Infant outcome of 957 singletons born after frozen embryo replacement: the Danish National Cohort Study 1995–2006. Fertil Steril 2010;94:1320–7.
32. Tararbit K, Houyel L, Bonnet D, De Vigan C, Lelong N, Goffinet F, et al.. Risk of congenital heart defects associated with assisted reproductive technologies: a population-based evaluation. Eur Heart J 2011;32:500–8.
33. Manipalviratn S, DeCherney A, Segars J. Imprinting disorders and assisted reproductive technology. Fertil Steril 2009;91:305–15.