- Email: [email protected]
Impact of single embryo transfer policy on perinatal outcomes in fresh and frozen cycles—analysis of the Japanese Assisted Reproduction Technology registry between 2007 and 2012
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
Impact of single embryo transfer policy on perinatal outcomes in fresh and frozen cycles—analysis of the Japanese Assisted Reproduction Technology registry between 2007 and 2012 Kazumi Takeshima, M.D.,a Seung Chik Jwa, M.D., Ph.D., M.P.H.,a,b Hidekazu Saito, M.D., Ph.D.,a Aritoshi Nakaza, B.S.,b Akira Kuwahara, M.D., Ph.D.,c Osamu Ishihara, M.D., Ph.D.,d Minoru Irahara, M.D., Ph.D.,c Fumiki Hirahara, M.D., Ph.D.,e Yasunori Yoshimura, M.D., Ph.D.,f and Tetsuro Sakumoto, M.D., Ph.D.b a Division of Reproductive Medicine, Center of Maternal-Fetal, Neonatal and Reproductive Medicine, National Center for Child Health and Development, Tokyo; b SORA no MORI clinic, Okinawa; c Department of Obstetrics and Gynecology, School of Medicine, University of Tokushima, Tokushima; d Department of Obstetrics and Gynecology, Saitama Medical University, Saitama; e Department of Obstetrics and Gynecology, Yokohama City University Graduate School of Medicine, Hukuura, Kanazawa-ku, Yokohama City, Kanagawa; and f Yoshimura Bioteic Literacy Institute, Tokyo, Japan
Objective: To investigate whether the introduction of single embryo transfer (SET) policy in Japan has improved perinatal outcomes. Design: A retrospective cohort study. Setting: Not applicable. Patient(s): A total of 140,718 live births and 510 stillbirths (after 22 weeks of gestation) conceived by assisted reproductive technology in Japan between 2007 and 2012 were reviewed. Intervention(s): None. Main Outcome Measure(s): Preterm birth (PTB), low birth weight (LBW), very low birth weight (VLBW), small for gestational age (SGA), large for gestational age (LGA), perinatal mortality, and other pregnancy complications. Result(s): The rate of SET increased significantly from 52.2% in 2007 to 82.6% in 2012, while the rate of multiple pregnancy decreased significantly from 10.7% to 4.1% over the same period. The rates of PTB, LBW, and SGA decreased significantly, while that of LGA increased. Perinatal mortality decreased from 0.70% to 0.40% in fresh cycles, while that of frozen cycles did not change. Double ET or more was associated with a significantly increased risk for multiple pregnancy, placenta accreta, preterm premature rupture of membrane, cesarean section (CS), PTB, LBW, SGA, and early neonatal death compared with SET. Compared with before the SET policy was launched, the risks of multiple pregnancy, CS, early PTB before 32 weeks, LBW, VLBW, and SGA were significantly decreased after the policy was launched, with significant interactions of fresh/frozen status. Conclusion(s): The results suggest that the SET policy improved perinatal outcomes in Japan. The impact of SET policy was different in fresh and frozen cycles for several perinatal outcomes. Use your smartphone (Fertil SterilÒ 2015;-:-–-. Ó2015 by American Society for Reproductive Medicine.) to scan this QR code Key Words: Assisted reproductive technology, perinatal outcome, preterm delivery, low birth and connect to the weight, single embryo transfer Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/takeshimak-set-policy-perinatal-outcomes/
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
Received May 27, 2015; revised September 13, 2015; accepted October 5, 2015. K.T. has nothing to disclose. S.C.J. has nothing to disclose. H.S. has nothing to disclose. A.N. has nothing to disclose. A.K. has nothing to disclose. O.I. has nothing to disclose. M.I. has nothing to disclose. F.H. has nothing to disclose. Y.Y. has nothing to disclose. T.S. has nothing to disclose. K.T. and S.C.J. should be considered similar in author order. This study was supported by the grant of National Center for Child Health and Development (24-6) and a grant of the Ministry of Health, Labour and Welfare (no. H25-001). Reprint requests: Kazumi Takeshima, M.D., Division of Reproductive Medicine, Center of Maternal-Fetal, Neonatal and Reproductive Medicine, National Center for Child Health and Development, 2-10-1, Okura, Setagaya-ku, Tokyo, 157-8535, Japan (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.10.002 VOL. - NO. - / - 2015
1
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
S
ince the first IVF baby was born in 1978 (1), assisted reproductive technology (ART) has been widely performed globally. However, along with ART treatment, the rate of multiple pregnancy and poor perinatal outcome is on the rise. To that end, single embryo transfer (SET) has been introduced (2, 3) and advocated as a preventive measure in several countries (4). In Japan, multiple pregnancy caused by ART has also been a concern. In 1996, the Japan Society of Obstetrics and Gynecology (JSOG) recommended that the number of transferred embryos should be confined to three or less. Although the rate of high-order multiple pregnancy (triplet or more) has decreased since then, the real number of twin pregnancies has increased (5, 6). Subsequently, in 2008, the JSOG recommended that SET should normally be performed to prevent multiple pregnancy, while double embryo transfer (DET) is allowed in women who are over 35 years old and/or have experienced failure of implantation for more than two consecutive attempts. As a result, the rate of SET being performed between 2007 and 2010 surged from 49.9% to 73.0%, while the rate of multiple pregnancy decreased from 11.5 % to 4.8% (7). In view of these facts, it is worthwhile to investigate whether perinatal outcomes, including preterm delivery and low birth weight (LBW) rates, are affected by the introduction of the SET policy from 2008 onward. In the present study, we analyzed the data of the national ART registry from 2007 to 2012 to evaluate whether the SET policy improved perinatal outcomes including perinatal mortality, multiple pregnancy, preterm birth (PTB), LBW, small for gestational age (SGA), large for gestational age (LGA), and other pregnancy complications. Since it was suggested that frozen-thawed embryo transfer (frozen ET) has better perinatal outcome than fresh embryo transfer (fresh ET) (8), we also investigated whether the impact of SET policy on perinatal outcomes is different in fresh and frozen cycles.
MATERIALS AND METHODS Data Source
All data analyzed were obtained from the ART online registry provided by the JSOG. Since 2007, the JSOG has required all institutes that perform ART to register cycle-specific information including patient's age, number of ETs, and perinatal outcomes with an online system. Over 99% of registered facilities record their treatment data in the registry every year. The data were provided by the JSOG upon approval by the registration and research subcommittee of the JSOG ethics committee. Further, the Institutional Review Board at the National Center for Child Health and Development approved this study.
Patients The number of registered institutions in the database each year between 2007 and 2012 was 606, 609, 625, 591, 586, and 589, respectively. Pregnancy rate stratified by age from the data announced officially by JSOG are shown as Supplemental Table 1 (available online) (9). The pregnancy
rate decreased as age rose, and the trend was particularly prominent in patients over 40 years old. There was no trend in pregnancy rate from 2007 to 2012 in the groups under 40 years old. However, the pregnancy rate slightly increased by approximately 1% in these 6 years in the group over 40 years old. Those with live birth at 22–41 of gestational weeks and stillbirth after 22 weeks of gestation were analyzed. There were 144,808 eligible cycles for live births with known gestational age at delivery and birth weight. Among them, treatment cycle of gamete intrafallopian transfer (n ¼ 36), unknown fertilization method (n ¼ 465), cancellation cycles for ET (n ¼ 252), unknown number of ETs (n ¼ 1), cycles using frozen oocytes (n ¼ 57), and cases with missing or incomplete records (n ¼ 3,279) were excluded. There were 510 eligible stillbirth cases after 22 weeks of gestation. Consequently, the total number of analyzed subjects was 141,228. Since donor gametes or embryos are not allowed to be used in ART in Japan, all of the embryos transferred were autologous.
Perinatal Outcomes Our primary outcomes are gestational age and birth weight. Gestational age was divided into three groups: delivery at over 37 gestational weeks (term delivery), delivery before 37 weeks (PTB), and delivery before 32 weeks (early preterm birth [EPTB]). In terms of birth weight, we classified the newborns into three groups: newborns of over 2,500 g (normal birth weight), newborns of 1,500–2,499 g (LBW), and newborns of less than 1,500 g (very low birth weight [VLBW]). Further, according to the new standard of average size and weight of newborns of the Japan Pediatric Society, we also classified birth weight into SGA defined as below the 10th percentile of standard and LGA defined as above the 90th percentile (10). As secondary outcomes, stillbirth, early neonatal death within 7 days after birth, multiple pregnancy, cesarean section (CS), and other pregnancy complications including pregnancy-induced hypertension (PIH), gestational diabetes mellitus (GDM), placenta previa, placenta accreta, placenta abruption, and preterm premature rupture of membrane (pPROM) were included for the analysis. Perinatal mortality was calculated as incidence of stillbirth or early neonatal death among known cycles for each year. Multiple pregnancy was defined based on the number of live births.
Statistical Analyses We evaluated whether there is a linear trend for each variable over the years 2007–2012 by calculating linear regression using each variable as a dependent variable and the calendar year as the independent and ordinal variable. If a variable was continuous or dichotomous, it was put into the regression model directly as a dependent variable. If a variable consisted of more than two categories, indicator variables were created and those indicator variables were separately put into the regression model. For perinatal outcomes, whether there is a linear trend from year 2007 to 2012 was separately evaluated according to fresh/frozen status. Further, we evaluated whether these trends were changed after restricting SET VOL. - NO. - / - 2015
2
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Fertility and Sterility® cycles. Crude and adjusted odds ratios (ORs) of DET and multiple embryo transfer (MET, i.e., three or more ET) compared with SET were calculated by logistic generalized estimating equations, with robust variance estimation to adjust the effect of clustering of births among the mothers stratified by fresh and frozen cycles. Priori covariates for adjusted analysis were maternal age (categorized into <30, 30–34, 35–39, R40 years), years of cycles reported, and embryo stage at transfer (early cleavage/blastocyst stage). Since fertilization method (i.e., IVF or intracytoplasmic sperm injection [ICSI]) was only available in fresh cycles, it was adjusted in the analysis of fresh cycles. Trends across the number of embryos transferred (i.e., SET, DET, and MET) for perinatal outcomes were tested by treating SET, DET, and MET as ordinal variables in the logistic regression model. We also investigated interactions between embryo stage at transfer, maternal age (before/after 35 years), and number of ETs for perinatal outcomes, and significant interactions were observed for the outcomes of multiple birth, PIH, placenta previa, placenta accreta, CS, PTB, LBW, SGA, and LGA between embryo stage at transfer and number of ETs and for the outcomes of multiple birth, PIH, CS, PTB, LBW, LGA, and early neonatal death between maternal age and number of ETs. Thus, we considered embryo stage at transfer and age to serve as effect modifiers for those outcomes and additionally stratified by embryo stage at transfer and age. Finally, since the SET policy was announced by JSOG in April 2008, we divided samples into before/after the SET policy was announced based on the date of ET and evaluated whether risks for those perinatal outcomes were changed before/after the SET policy was
launched. We also investigated whether the effect of the SET policy on perinatal outcomes differed by fresh/frozen status by adding an interaction term between number of ETs and fresh/frozen status in adjusted model. All analyses were performed with the STATA/SE statistical package, version 12.1 (Stata Corp.). Two-tailed P values < .05 were considered as statistically significant.
RESULTS Table 1 shows the characteristics of cycles reported in each year from 2007 to 2012. During that period, the reported number of cycles dramatically increased from 14,991 cycles in 2007 to 33,066 cycles in 2012, of which fresh cycles increased moderately from 7,797 cycles to 9,126 cycles, while frozen cycles increased more drastically from 7,194 cycles to 23,940 cycles, resulting in the proportion of fresh cycles significantly decreasing from 2007 to 2012. There was a significant upward trend in terms of the age of subjects both in fresh and frozen cycles (P for trends < .001). In particular, the rate of over 35 years old increased from nearly 50% in 2007 to 60% in 2012. In terms of the fertilization method of fresh ET, the rate of IVF significantly decreased from 49.2% in 2007 to 45.9% in 2012, while that of ICSI and split fertilization (IVFþICSI) increased slightly. After the introduction of SET policy by the JSOG in 2008, the rate of SET significantly increased during the analyzed period from 52.2% in 2007 to 82.6% in 2012. On the other hand, the rate of DET significantly decreased from 38.1% in 2007 to 17.0% in 2012. Similarly, the rate of three or more
TABLE 1 Characteristics of cycles transferred among ART patients stratified by year, 2007–2012 (n [ 141,228 cycles). Characteristic Maternal age <30 30–34 35–39 R40 Fresh or frozen Fresh Frozen Fertilization method in fresh cyclea IVF ICSI Split (IVFþICSI) No. of ETs One Two Three or more Embryo stage at transfer Early cleavage Blastocyst No. of single and multiple pregnancies Single Twin Triplet
2007 (n [ 14,991)
2008 (n [ 17,956)
2009 (n [ 22,366)
2010 (n [ 24,832)
2011 (n [ 28,017)
2012 (n [ 33,066)
34.4 3.8 1,519 (10.1) 6,008 (40.1) 6,118 (40.8) 1,346 (9.0)
34.6 3.8 1,696 (9.5) 6,793 (37.8) 7,678 (42.8) 1,789 (10.0)
34.8 3.8 2,010 (9.0) 8,049 (36.0) 9,811 (43.9) 2,496 (11.2)
35.0 3.8 2,063 (8.3) 8,609 (34.7) 11,136 (44.9) 3,024 (12.2)
35.2 3.9 2,266 (8.1) 9,166 (32.7) 12,825 (45.8) 3,760 (13.4)
35.4 3.9 2,493 (7.5) 10,501 (31.8) 15,208 (46.0) 4,864 (14.7)
7,797 (52.0) 7,194 (48.0)
7,634 (42.5) 10,322 (57.5)
8,603 (38.5) 13,763 (61.5)
8,557 (34.5) 16,275 (65.5)
8,728 (31.2) 19,289 (68.9)
9,126 (27.6) 23,940 (72.4)
3,834 (49.2) 3,349 (43.0) 614 (7.9)
3,800 (49.8) 3,233 (42.4) 601 (7.9)
4,207 (48.9) 3,661 (42.6) 735 (8.5)
4,011 (46.9) 3,743 (43.7) 803 (9.4)
3,961 (45.4) 3,885 (44.5) 882 (10.1)
4,190 (45.9) 4,026 (44.2) 910 (10.0)
7,825 (52.2) 5,718 (38.1) 1,448 (9.7)
12,347 (68.8) 5,199 (29.0) 410 (2.3)
16,930 (75.7) 5,246 (23.5) 190 (0.9)
19,567 (78.8) 5,114 (20.6) 151 (0.6)
22,642 (80.8) 5,232 (18.7) 143 (0.5)
27,316 (82.6) 5,620 (17.0) 130 (0.4)
7,400 (49.4) 7,591 (50.6)
7,132 (39.7) 10,824 (60.3)
8,471 (37.9) 13,895 (65.1)
8,413 (33.9) 16,419 (66.1)
9,085 (32.4) 18,932 (67.6)
10,464 (31.7) 22,602 (68.4)
13,382 (89.3) 1,580 (10.5) 29 (0.19)
16,824 (93.7) 1,122 (6.3) 10 (0.06)
21,193 (94.8) 1,160 (5.2) 13 (0.06)
23,634 (95.2) 1,182 (4.8) 16 (0.06)
26,840 (95.8) 1,156 (4.1) 21 (0.07)
31,722 (96.0) 1,327 (4.0) 17 (0.05)
Note: Data are shown mean SD for continuous variables and n (%) for dichotomous variables. a Denominators are numbers of fresh cycles in each year. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
3
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION ETs decreased from 9.7% in 2007 to 2.3% in 2008 and remained below 1% after 2009. Accordingly, singleton pregnancies gradually increased from 89.3% in 2007 to 96.0% in 2012, and twin pregnancies decreased from 10.5% in 2007 to 4.0% in 2012 (P for trend < .001), while multiple pregnancies (triplet or more) decreased from 0.19% in 2007 to 0.06% in 2008 and became almost stable after 2008. Table 2 shows perinatal outcomes and pregnancy complications for fresh ET and frozen ET stratified by calendar year. For the gestational weeks, the rates of PTB and LBW decreased both in fresh and frozen cycles (P for trend < .001). The mean birth weight in frozen ET was heavier than that in fresh ET in all the analyzed years. The rate of SGA decreased (11.3% to 8.4% in fresh ET, 6.9% to 5.6% in frozen ET, from 2007 to 2012), while that of LGA increased (7.3% to 8.6% in fresh ET, 10.7% to 13.2% in frozen ET, from 2007 to 2012; P for trend < .001). In frozen ET, the rate of SGA was lower, while that of LGA was higher compared with those in fresh ET throughout the analyzed years. The rate of stillbirth did not change over the study period in either fresh or frozen cycles, while the rate of early neonatal death decreased from 0.21% in 2007 to 0.11% in 2012 in fresh cycles with significant trend (P for trend ¼ .004) but did not change in frozen cycles, resulting in a significant decreasing trend in perinatal mortality in fresh cycles. For pregnancy complications, PIH and GDM significantly increased from 2007 to 2012, while other pregnancy complications of placenta previa, placenta accreta, placenta abruption, and pPROM did not demonstrate significant trend over the study period. These trends for PIH and GDM remained significant after adjusting for maternal age and restricting SET cycles, except for PIH in fresh cycles. CS decreased significantly from 35.6% in 2007 to 32.8% in 2012 in fresh cycles but did not change in frozen cycles. When we restricted SET cycles, significant trends disappeared or were attenuated in the outcomes of PIH, mode of deliveries, gestational weeks at birth, birth weight, SGA, LGA, early neonatal death, and perinatal mortality in fresh cycles and vaginal deliveries, gestational weeks at birth, birth weight, and SGA in frozen cycles, suggesting that these significant trends in perinatal outcomes were related to the increased rate of SET over time. In Table 3, crude and adjusted ORs of DET and MET compared with SET for perinatal outcomes are presented. In fresh DET and MET, the adjusted OR of multiple pregnancy was 19.2 (95% confidence interval [CI], 17.2–21.5) and 26.4 (95% CI, 22.2–31.3) with significant trend for number of ET (P for trend < .001). Similar significant results were observed in frozen cycles. MET was associated with a significantly higher risk for early neonatal death in fresh cycles. On the other hand, DET was associated with a significantly higher risk for pPROM in both fresh and frozen cycles but a significantly lower risk for placenta previa in frozen cycles only. Interestingly, both DET and MET demonstrated a significantly decreased risk for PIH and GDM in fresh cycles. For delivery outcomes of PTB, EPTB, LBW, VLBW, SGA, and CS, DET and MET demonstrated a significantly increased risk compared with SET and a significantly decreased risk for LGA in both fresh and frozen cycles with significant trend.
For perinatal outcomes with significant interaction between embryo stage at transfer and number of ETs, adjusted ORs additionally stratified by embryo stage are shown in Supplemental Table 2 (available online). Significant ORs of DET and MET for multiple pregnancy were attenuated in blastocyst ET in fresh cycles but were still significant. A significantly decreased risk for PIH in fresh cycles was observed in early cleavage ET, whereas in blastocyst ET, an increased risk was observed, although it was not significant. A significantly decreased risk of DET in frozen cycles for placenta previa and a significantly increased risk for placenta accreta were observed only in blastocyst ET. Similarly, for perinatal outcomes with significant interaction between age (before/after 35 years) and number of ETs, adjusted ORs additionally stratified by age are shown in Supplemental Table 3 (available online). Significant ORs of DET and MET for multiple pregnancy were attenuated after the age of 35 years both in fresh and frozen cycles but were still significant. A significantly decreased risk of DET and MET for PIH in fresh cycles was observed only in women after the age of 35 years, while a significantly increased risk of DET for PIH in frozen cycles was observed only in women before 35 years of age. For the delivery outcomes of CS, LBW, and SGA, adjusted ORs of DET and MET were attenuated in women older than 35 years in both fresh and frozen cycles. Table 4 shows adjusted ORs of the effect of SET policy for selected perinatal outcomes. After the SET policy was introduced in April 2008, SET increased from 54.2% to 78.9%, while DET and MET decreased from 34.7% to 20.5% and 8.4% to 0.70%, respectively. After SET policy was announced, adjusted ORs for multiple pregnancy were significantly decreased in both fresh and frozen cycles (adjusted OR in fresh cycles, 0.42; 95% CI, 0.39–0.45; adjusted OR in frozen cycles, 0.50; 95% CI, 0.46–0.54, respectively). A similar significantly decreased risk was observed in stillbirth, CS, PTB, LBW, and SGA in both fresh and frozen cycles. For EPTB, VLBW, and early neonatal death, adjusted ORs were significantly decreased only in fresh cycles after SET policy was announced. P values for interaction between before/after SET policy and fresh/frozen status were significant for the outcomes of multiple birth, CS, EPTB, LBW, VLBW, and SGA, suggesting that the effect of the SET policy on those perinatal outcomes is significantly different by fresh/frozen status.
DISCUSSION Our study demonstrated that the introduction of SET policy by the JSOG in 2008 resulted in significant improvements in many perinatal outcomes including multiple pregnancy, PTB, LBW, SGA, perinatal mortality, and other pregnancy complications over the study period. Our results suggest that introduction of SET policy may have a different impact on several perinatal outcomes in fresh and frozen cycles. To date, many studies have reported on the improvement of perinatal outcome by SET. A study from Turkey reported that mandatory SET resulted in a decrease in multiple pregnancy and admissions to the neonatal intensive care unit (11). According to Sazonova et al. (12), women with two VOL. - NO. - / - 2015
4
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
VOL. - NO. - / - 2015
TABLE 2 Perinatal outcomes stratified by calendar year and fresh/frozen cycles, 2007–2012. Outcome
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
.16 .012 < .001 .85 .55 .34 .81
.76 .39 < .001 .22 .81 .86 .27
< .001 < .001 < .001 < .001 < .001 < .001 .01
.011 .45 < .001 .81 .23 .37 .38
< .001 < .001 < .001 < .001
.07 .10 .25 .13
4,693 (49.6) 4,699 (49.7) 62 (0.7) 786 (8.4) 806 (8.6) 10 (0.11) 1,021 (10.8) 0.40
.017 .012 .84 < .001 .001 .004 < .001 .002
.09 .12 .09 .046 .41 .26 < .001 .21
n ¼ 19,289 62 (0.32) 657 (3.4) 238 (1.23) 122 (0.63) 83 (0.43) 36 (0.19) 71 (0.37) n ¼ 19,227
n ¼ 23,940 78 (0.33) 873 (3.7) 324 (1.35) 204 (0.85) 89 (0.37) 41 (0.17) 107 (0.45) n ¼ 23,862
.58 < .001 < .001 .25 .15 .94 .11
.99 < .001 < .001 .30 .014 .72 .021
11,097 (57.7) 7,754 (40.3) 376 (2.0)
13,628 (57.1) 9,776 (41.0) 458 (1.9)
< .001 .60 < .001
.32 < .001 < .001
n ¼ 7,797 38 (0.49) 139 (1.8) 29 (0.14) 72 (0.92) 3 (0.04) 17 (0.22) 35 (0.45) n ¼ 7,759
n ¼ 7,634 26 (0.34) 142 (1.9) 32 (0.42) 74 (0.97) 6 (0.08) 20 (0.26) 28 (0.37) n ¼ 7,608
n ¼ 8,603 37 (0.43) 164 (1.9) 40 (0.46) 78 (0.91) 9 (0.10) 20 (0.23) 25 (0.29) n ¼ 8,566
n ¼ 8,557 45 (0.53) 168 (2.0) 72 (0.84) 65 (0.76) 2 (0.02) 35 (0.41) 35 (0.41) n ¼ 8,512
n ¼ 8,728 34 (0.39) 200 (2.3) 108 (1.24) 78 (0.89) 8 (0.09) 22 (0.25) 35 (0.40) n ¼ 8,694
n ¼ 9,126 26 (0.28) 197 (2.2) 139 (1.52) 93 (1.02) 7 (0.08) 26 (0.28) 34 (0.37) n ¼ 9,100
4,576 (59.0) 2,760 (35.6) 423 (5.5) 38.2 2.2 6,709 (86.5) 895 (11.5) 155 (2.0) n ¼ 8,675 2,781 556 6,424 (74.1) 2,012 (23.2) 239 (2.8)
4,802 (63.1) 2,597 (34.1) 209 (2.8) 38.4 2.2 6,712 (88.2) 771 (10.1) 125 (1.6) n ¼ 8,161 2,839 537 6,457 (79.1) 1,519 (18.6) 185 (2.3)
5,534 (64.6) 2,761 (32.2) 271 (3.2) 38.4 2.1 7,645 (89.3) 785 (9.2) 136 (1.6) n ¼ 9,060 2,855 526 7,274 (80.3) 1,587 (17.5) 199 (2.2)
5,478 (64.4) 2,774 (32.6) 260 (3.1) 38.4 2.1 7,649 (89.9) 718 (8.4) 145 (1.7) n ¼ 8,960 2,868 517 7,303 (81.5) 1,470 (16.4) 187 (2.1)
5,648 (65.0) 2,901 (33.4) 145 (1.7) 38.5 2.1 7,880 (90.6) 697 (8.0) 117 (1.4) n ¼ 9,061 2,891 506 7,516 (83.0) 1,385 (15.3) 160 (1.8)
5,937 (65.2) 2,984 (32.8) 179 (2.0) 38.5 2.0 8,251 (90.7) 708 (7.8) 141 (1.6) n ¼ 9,454 2,891 506 7,866 (83.2) 1,399 (14.8) 189 (2.0)
4,485 (51.7) 4,120 (47.5) 70 (0.8) 972 (11.3) 628 (7.3) 18 (0.21) 1,307 (15.1) 0.70
4,136 (50.7) 3,973 (48.7) 52 (0.6) 766 (9.4) 669 (8.3) 17 (0.21) 1,762 (21.6) 0.53
4,624 (51.0) 4,392 (48.5) 44 (0.5) 853 (9.5) 732 (8.1) 16 (0.18) 2,017 (22.3) 0.59
4,530 (50.6) 4,364 (48.7) 66 (0.7) 788 (8.9) 763 (8.6) 11 (0.12) 2,036 (22.7) 0.64
4,605 (50.8) 4,403 (48.6) 53 (0.6) 735 (8.2) 791 (8.8) 8 (0.09) 983 (10.9) 0.46
n ¼ 7,194 23 (0.32) 213 (3.0) 32 (0.44) 51 (0.71) 15 (0.21) 7 (0.10) 32 (0.44) n ¼ 7,171
n ¼ 10,322 46 (0.45) 289 (2.8) 59 (0.57) 83 (0.80) 46 (0.45) 26 (0.25) 32 (0.31) n ¼ 10,276
n ¼ 13,763 35 (0.25) 416 (3.0) 73 (0.53) 83 (0.60) 43 (0.31) 26 (0.19) 39 (0.28) n ¼ 13,728
n ¼ 16,275 60 (0.37) 532 (3.3) 155 (0.95) 113 (0.69) 45 (0.28) 32 (0.20) 40 (0.25) n ¼ 16,215
3,784 (52.8) 2,931 (40.9) 456 (6.4)
5,912 (57.5) 4,102 (39.9) 262 (2.6)
7,933 (57.8) 5,432 (39.6) 363 (2.6)
9,413 (58.1) 6,338 (39.1) 464 (2.9)
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
2010
P value for trend restricting SET cyclesa
2008
2011
2012
5
Fertility and Sterility®
Fresh Pregnancy complications Still birth PIH GDM Previa Placental accreta Placenta abruption pPROM Delivery outcomes Mode of delivery Vaginal CS Unknown Gestational weeks at birth (wk)b R37 3236 <32 Neonates Birth weight, g R2,500 1,500–2,499 <1,500 Sex Male Female Unknown SGAc LGAd Early neonatal deathe Unknown Perinatal mortality, %e Frozen Pregnancy complications Still birth PIH GDM Previa Placental accreta Placenta abruption pPROM Delivery outcomes Mode of delivery Vaginal CS Unknown
2009
P value for trenda
2007
Continued. Outcome
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Gestational weeks at birthb R37 3236 <32 Neonates Birth weight, g R2,500 1,500–2,499 <1,500 Sex Male Female Unknown SGAc LGAd Early neonatal deathe Unknown Perinatal mortality, %e
2007
2008
2009
2010
2011
2012
38.4 2.2 6,259 (87.3) 805 (11.2) 107 (1.5) n ¼ 7,887 2,901 538 6,343 (80.4) 1,417 (18.0) 127 (1.6)
38.5 2.1 9,239 (90.0) 898 (8.7) 139 (1.3) n ¼ 10,863 2,961 519 9,186 (84.6) 1,511 (13.9) 166 (1.5)
38.6 2.1 12,402 (90.3) 1,138 (8.3) 188 (1.4) n ¼ 14,418 2,977 517 12,350 (85.7) 1,851 (12.8) 217 (1.5)
38.5 2.2 14,606 (90.1) 1,350 (8.3) 259 (1.6) n ¼ 16,977 2,967 528 14,472 (85.2) 2,201 (13.0) 304 (1.8)
38.5 2.1 17,342 (90.2) 1,594 (8.3) 291 (1.5) n ¼ 20,055 2,975 516 17,295 (86.2) 2,424 (12.1) 336 (1.7)
38.5 2.1 21,565 (90.4) 1,942 (8.1) 355 (1.5) n ¼ 24,864 2,980 519 21,443 (86.2) 3,002 (12.1) 419 (1.7)
4,105 (52.1) 3,738 (47.4) 44 (0.6) 544 (6.9) 836 (10.7) 6 (0.08) 1,685 (21.4) 0.38
5,710 (52.6) 5,089 (46.9) 64 (0.6) 632 (5.9) 1,293 (12.0) 17 (0.16) 3,191 (29.4) 0.58
7,540 (52.3) 6,795 (47.1) 84 (0.6) 834 (5.8) 1,785 (12.5) 10 (0.07) 3,922 (27.2) 0.31
8,889 (52.4) 7,989 (47.1) 99 (0.6) 1,045 (6.2) 2,115 (12.5) 15 (0.09) 4,415 (26.0) 0.45
10,356 (51.6) 9,582 (47.8) 117 (0.6) 1,095 (5.5) 2,535 (12.7) 28 (0.14) 2,875 (14.3) 0.46
12,821 (51.6) 11,891 (47.8) 152 (0.6) 1,391 (5.6) 3,257 (13.2) 30 (0.12) 3,618 (14.6) 0.44
Note: Data are shown mean SD for continuous variables and n (%) for dichotomous variables. a Linear trends are assessed with linear regression using calendar year as the ordinal and independent variable. b Denominators are number of mothers for each year. c SGA was defined as being below the 10th percentile of the national reference. d LGA was defined as being above the 90th percentile of the national reference. e Excluding unknown cases for early neonatal death. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
P value for trenda
P value for trend restricting SET cyclesa
.46 < .001 < .001 .40
< .001 .07 .16 .19
< .001 < .001 < .001 .21
.73 .90 .41 .014
.062 .073 .28 < .001 < .001 .78 < .001 .36
< .001 < .001 .09 .19 < .001 .61 < .001 .55
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
6
TABLE 2
VOL. - NO. - / - 2015
Fertility and Sterility®
TABLE 3 Crude and adjusted OR of number of ET for perinatal outcomes stratified by fresh/frozen cycles. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend Still birth SET DET MET P value for trend PIH SET DET MET P value for trend GDM SET DET MET P value for trend Placenta previa SET DET MET P value for trend Placenta accreta SET DET MET P value for trend Placenta abruptio SET DET MET P value for trend pPROM SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend EPTBc SET DET MET P value for trend LBW SET DET MET P value for trend VLBW SET DET
Frozen cycles a
Crude OR (95% CI)
Adjusted OR (95% CI)
Crude OR (95% CI)
Adjusted OR (95% CI)b
Reference 16.6 (14.9–18.5) 23.6 (20.2–27.7) < .001
Reference 19.2 (17.2–21.5) 26.4 (22.2–31.3) < .001
Reference 15.1 (14.1–16.3) 20.1 (17.1–23.5) < .001
Reference 17.0 (15.8–18.4) 23.7 (19.9–28.0) < .001
Reference 1.1 (0.78–1.4) 1.2 (0.56–2.5) .62
Reference 0.95 (0.70–1.3) 0.99 (0.45–2.2) .80
Reference 1.2 (0.94–1.6) 0.31 (0.04–2.2) .41
Reference 1.1 (0.85–1.5) 0.26 (0.04–1.9) .87
Reference 0.84 (0.73–0.97) 0.64 (0.41–0.996) .003
Reference 0.77 (0.66–0.89) 0.58 (0.37–0.92) < .001
Reference 1.1 (1.004–1.2) 0.74 (0.49–1.1) .24
Reference 1.1 (0.997–1.2) 0.76 (0.50–1.2) .25
Reference 0.76 (0.61–0.95) 0.22 (0.07–0.69) < .001
Reference 0.76 (0.60–0.96) 0.31 (0.10–0.997) .003
Reference 0.69 (0.57–0.83) 0.29 (0.09–0.89) < .001
Reference 0.72 (0.59–0.88) 0.39 (0.12–1.2) < .001
Reference 0.97 (0.79–1.2) 1.04 (0.60–1.8) .86
Reference 0.89 (0.72–1.1) 0.91 (0.52–1.6) .31
Reference 0.74 (0.60–0.92) 0.52 (0.19–1.4) .002
Reference 0.71 (0.57–0.89) 0.48 (0.18–1.3) .001
Reference 0.70 (0.32–1.6) NA .20
Reference 0.72 (0.32–1.6) NA .27
Reference 1.4 (1.1–1.8) NA .08
Reference 1.5 (1.1–1.9) NA .04
Reference 0.98 (0.67–1.4) 1.5 (0.65–3.4) 0.68
Reference 1.1 (0.73–1.6) 1.95 (0.81–4.7) 0.31
Reference 0.96 (0.65–1.4) NA 0.44
Reference 0.90 (0.60–1.3) NA 0.33
Reference 1.9 (1.4–2.5) 1.4 (0.59–3.1) < .001
Reference 2.0 (1.5–2.8) 1.6 (0.66–3.7) < .001
Reference 1.9 (1.5–2.4) 0.66 (0.16–2.6) < .001
Reference 2.0 (1.5–2.5) 0.69 (0.17–2.8) < .001
Reference 1.8 (1.7–1.9) 1.9 (1.7–2.1) < .001
Reference 1.6 (1.57–1.71) 1.8 (1.6–2.0) < .001
Reference 1.6 (1.58–1.68) 2.0 (1.8–2.3) < .001
Reference 1.7 (1.62–1.74) 2.0 (1.8–2.3) < .001
Reference 2.2 (2.1–2.4) 3.1 (2.7–3.5) < .001
Reference 2.3 (2.1–2.4) 3.2 (2.8–3.6) < .001
Reference 2.4 (2.3–2.5) 2.9 (2.5–3.4) < .001
Reference 2.5 (2.4–2.6) 3.1 (2.6–3.6) < .001
Reference 1.9 (1.7–2.2) 2.6 (1.9–3.6) < .001
Reference 1.9 (1.6–2.2) 2.7 (1.9–3.7) < .001
Reference 1.8 (1.6–2.1) 1.6 (1.0–2.5) < .001
Reference 1.9 (1.7–2.2) 1.8 (1.1–2.9) < .001
Reference 2.8 (2.6–2.9) 3.9 (3.5–4.4) < .001
Reference 2.9 (2.7–3.0) 4.0 (3.5–4.5) < .001
Reference 3.2 (3.0–3.3) 4.3 (3.7–4.9) < .001
Reference 3.3 (3.1–3.4) 4.4 (3.8–5.1) < .001
Reference 2.3 (2.0–2.6)
Reference 2.3 (2.0–2.7)
Reference 2.2 (1.9–2.4)
Reference 2.3 (2.1–2.7)
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
7
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
TABLE 3 Continued. Fresh cycles Outcome MET P value for trend SGAd SET DET MET P value for trend LGAe SET DET MET P value for trend Early neonatal death SET DET MET P value for trend
Crude OR (95% CI)
Frozen cycles a
Adjusted OR (95% CI)
Crude OR (95% CI)
Adjusted OR (95% CI)b
3.7 (2.8–5.0) < .001
4.0 (3.0–5.4) < .001
2.1 (1.4–3.2) < .001
2.7 (1.8–4.0) < .001
Reference 1.6 (1.5–1.7) 2.2 (1.9–2.5) < .001
Reference 1.6 (1.5–1.7) 2.1 (1.8–2.5) < .001
Reference 1.8 (1.7–1.9) 2.3 (1.9–2.8) < .001
Reference 1.8 (1.6–1.9) 2.2 (1.8–2.7) < .001
Reference 0.77 (0.72–0.83) 0.61 (0.50–0.75) < .001
Reference 0.75 (0.70–0.81) 0.63 (0.50–0.78) < .001
Reference 0.71 (0.68–0.75) 0.71 (0.59–0.86) < .001
Reference 0.73 (0.69–0.77) 0.79 (0.65–0.96) < .001
Reference 1.4 (0.84–2.2) 4.5 (2.1–9.4) .003
Reference 1.2 (0.71–1.9) 3.8 (1.7–8.1) .03
Reference 1.04 (0.66–1.6) 0.72 (0.10–5.2) .99
Reference 1.1 (0.69–1.7) 0.87 (0.11–6.9) .77
Note: NA ¼ not available. a Adjusted for calendar year, maternal age, fertilization method, and embryo stage at transfer. b Adjusted for calendar year, maternal age, and embryo stage at transfer. c Gestational age at delivery <32 weeks. d SGA was defined as being below the 10th percentile of the national reference. e LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
singletons by two IVF treatments of SET showed better neonatal and maternal outcomes compared with those with two babies by one IVF treatment of DET. When it comes to perinatal outcome of singleton pregnancy from SET and DET, one study showed that singleton pregnancy from SET had a more positive outcome (13). However, some studies have stated that the outcomes between SET and DET were similar (14, 15). Notably, a study reported that vanishing twin pregnancies using DET had an adverse effect on perinatal outcome (14). In our study, DET was associated with a significantly higher risk for pPROM and placenta accreta in frozen cycles. Interestingly, DET was associated with a significantly decreased risk for PIH, GDM, and placenta previa. Those results may be due to DET and MET being performed in patients with fewer complications. In addition, our results demonstrated that maternal age and embryo stage at transfer played a role as effect modifiers for those outcomes. The risks of stillbirth, CS, PTB, LBW, SGA, and early neonatal death were significantly decreased after SET policy was launched. The rate of SGA showed a downward trend in both fresh and frozen ET, whereas the rate of LGA increased. Perinatal mortality also demonstrated a significantly decreasing trend in fresh cycles. These findings could be attributed to the decrease in multiple births. Indeed, these significant trends disappeared after restricting SET cycles. Interestingly, a significantly increasing trend was observed for PIH and GDM after adjusting for maternal age and restricting SET cycles, except for PIH in fresh cycles. These results suggest that the significant trend for those outcomes may not be due to an increase in maternal age or implementation of SET but rather are due to other unknown con-
founders. On the other hand, the rate of SGA by frozen ET was lower than that by fresh ET, while the rate of LGA by frozen ET was higher. Our results are in line with those from previous studies that reported that the birth weight of those born by frozen ET cycle was greater than that of those born by fresh ET and natural cycles (16, 17). Further, reports from Australia and New Zealand have shown that SET was associated with lower rates of fetal and neonatal mortality compared with DET (18, 19). We showed that the impact of SET policy is different in fresh and frozen cycles for several perinatal outcomes of CS, EPTB, LBW, VLBW, and SGA (Table 4). The pregnancy rate may be improved by performing frozen ET instead of fresh ET owing to the differences in hormonal environment and the receptivity of the endometrium (20). Indeed, frozen ET resulted in a higher pregnancy rate in Japan (7). Frozen ET is expected to be more widely adopted because it has been shown to result in a higher rate of pregnancy and positive perinatal outcomes compared with fresh ET (8, 21, 22). The rate of frozen ET has dramatically increased over the study period, accounting for over 70% of the cases treated with ART in 2012. Accordingly, SET with frozen ET has also increased (Table 2). Since our results demonstrate that the SET policy may have a different impact on perinatal outcomes in fresh and frozen cycles, DET or MET should be carefully practiced according to fresh/frozen status. Our results demonstrate that although a significant increase in SET use was observed from 52.2% in 2007 to 82.6% in 2012, singleton pregnancies gradually increased from 89.3% in 2007 to 96.0% in 2012, and the rate of multiple pregnancies (i.e., triplet) did not decrease so much (from 0.19% in 2007 to 0.05% in 2012) compared with the rate of VOL. - NO. - / - 2015
8
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Fertility and Sterility®
TABLE 4 Adjusted OR of effect of SET policy for selected pregnancy and birth outcomes stratified by fresh/frozen cycles.
Outcome
Fresh cycles Adjusted OR (95% CI)a
Multiple pregnancy Before policy Reference After policy 0.42 (0.39–0.45) Still birth Before policy Reference After policy 0.52 (0.38–0.69) pPROM Before policy Reference After policy 0.85 (0.60–1.2) CS Before policy Reference After policy 0.80 (0.77–0.84) PTB Before policy Reference After policy 0.71 (0.67–0.76) EPTBd Before policy Reference After policy 0.72 (0.61–0.85) LBW Before policy Reference After policy 0.65 (0.62–0.69) VLBW Before policy Reference After policy 0.69 (0.59–0.81) e SGA Before policy Reference After policy 0.78 (0.73–0.84) LGAf Before policy Reference After policy 1.1 (1.06–1.24) Early neonatal death Before policy Reference After policy 0.54 (0.34–0.88)
Frozen cycles Adjusted OR (95% CI)b
P value for interactionc
Reference 0.50 (0.46–0.54)
.02
Reference 0.39 (0.30–0.51)
.14
Reference 0.79 (0.57–1.1)
.68
Reference 0.88 (0.84–0.92)
.03
Reference 0.77 (0.72–0.82)
.19
Reference 1.00 (0.84–1.2)
.01
Reference 0.74 (0.69–0.78)
.03
Reference 1.03 (0.86–1.2)
.003
Reference 0.90 (0.83–0.98)
.03
Reference 1.2 (1.08–1.23)
.85
Reference 1.1 (0.55–2.3)
.12
a
Adjusted for maternal age, fertilization method, and embryo stage at transfer. Adjusted for maternal age and embryo stage at transfer. P value for interaction between before/after SET policy and fresh/frozen status in adjusted model including maternal age and embryo stage at transfer. d Gestational age at delivery <32 weeks. e SGA was defined as being below the 10th percentile of the national reference. f LGA was defined as being above the 90th percentile of the national reference. b c
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
twin pregnancies (from 10.5% to 4.0%). A potential reason for the gradual decrease in multiple pregnancies may be that, when MET was conducted, low-grade embryos were more likely included for ET after 2007. Before SET policy was introduced by JSOG in 2008, the Japan Society for Reproductive Medicine recommended in 2007 that the first treatment cycle or high-grade blastocyst transfer in a patient under 35 years old should be SET and that the treatment cycle using especially high-grade blastocyst transfer in a patient under 40 years old should also be SET or DET. Thus, we speculate that when MET was conducted for women under 40 years old after 2007, low-grade embryos were more likely included for ET, which resulted in the gradual decrease in multiple pregnancies. Although our results demonstrated that perinatal outcomes after ART in Japan improved after introduction of SET policy, perinatal outcomes are still inferior to those of the non-IVF population. The rate of PTB of the whole pop-
ulation in Japan between 2007 and 2012 is steady at 5.7%– 5.8%, while that of LBW was 8.5% in boys and 10.8% in girls (23). In our study, the rate of PTB was 10.3%, and the rate of LBW was 17.8% in boys and 21.5% in girls in fresh cycles and 13.2% in boys and 16.3% in girls in frozen cycles, which was higher than those in all babies born in Japan. The main reason for these differences may be that the rate of multiple pregnancy was still higher in the ART population than in all populations in Japan. The rate of multiple pregnancy ranged from 10.7% in 2007 to 4.1% in 2012 in this study, while that of the general population was approximately 1.0%. In addition, advanced maternal age in ART may contribute to a worse outcome of ART in Japan (mean maternal age, 35 years in our sample vs. 31 years in the general population) (23). The strength of this study is that it included large sample sizes for each of the analyzed years, covering almost all cyclespecific ART data in Japan from 2007 to 2012. As the registration of data is mandatory, selection bias is not likely to have occurred. Nonetheless, there are some limitations. First, although the rate of SET significantly increased and DET and MET decreased after the introduction of the SET policy in 2008, other factors, such as improvements in embryo technique and perinatal care, may also affect perinatal outcomes. Indeed, the pregnancy rate slightly increased by approximately 1% from 2007 to 2012 in the group over 40 years old. Especially for the analysis shown in Table 4, we assumed that other factors except for the introduction of the SET policy would be similar before and after the SET policy. We conducted a sensitivity analysis further stratifying the number of embryos transferred in Table 4. Significant associations disappeared or were attenuated in many of perinatal outcomes such as multiple pregnancy, CS, PTB, LBW, VLBW, and LGA in fresh cycles and SGA, suggesting that the SET policy had a major impact on the perinatal outcomes through the implementation of SET. On the other hand, with respect to the perinatal outcomes of stillbirth and early neonatal death, ORs for the outcomes did not change (data not shown). For those outcomes, other factors may play a significant role. Second, multiple pregnancy was defined as two or more live births per one delivery in this study. If a different definition of multiple pregnancy was used, such as the number of gestational sacs and/or fetal heartbeats, another conclusion about perinatal outcomes may have been drawn. Notably, it was reported that vanishing twin during early pregnancy increases the risk for preterm delivery, LBW, and SGA (24– 26). To investigate this, we also performed a sensitivity analysis using different definitions of multiple pregnancy (i.e., based on the confirmation of fetal heartbeat or gestational sac). The results remained the same (data not shown). Hence, we can deduce that a different definition for multiple pregnancy will not have an impact on the current results. Third, the data did not include covariates such as patients' socioeconomic status, duration of infertility, number of repeated ART failures, and husband's age, which may affect results for residual confounding. Ideally, prospective studies including those covariates would be conducted to evaluate the impact of SET on perinatal outcomes.
VOL. - NO. - / - 2015
9
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
CONCLUSIONS
11.
On the basis of the data of the Japanese ART registry between 2007 and 2012, perinatal outcomes including PTB, LBW, SGA, perinatal mortality, and other pregnancy complications were significantly improved after SET policy was launched by JSOG in 2008. The results might be strongly associated with the decrease in the rate of preterm delivery and the increase in the average birth weight due to the decrease in multiple pregnancy, which could be attributed to the increased use of SET. The impact of SET policy on several perinatal outcomes was significantly different in fresh and frozen cycles, suggesting that DET or MET should be carefully conducted according to fresh/frozen status. Further study that includes additional covariates is necessary to evaluate the implementation of SET based on long-term accumulation of data for reviewing perinatal outcomes. Acknowledgments: The authors thank the JSOG for providing the data and all the participating clinics in the Japanese ART registry for their continuous support in data collection. We thank Dr. Julian Tang of the Department of Education for Clinical Research, National Center for Child Health and Development, for proofreading, editing, and rewriting part of this manuscript.
2. 3.
4. 5. 6. 7.
8.
9. 10.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
12.
Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. Prevention of twin pregnancies after IVF/ICSI by single embryo transfer. ESHRE Campus Course Report. Hum Reprod 2001;16:790–800. € T, Jablonowska B, Pinborg A, Strandell A, et al. Thurin A, Hausken J, Hillensjo Elective single-embryo transfer versus double-embryo transfer in in vitro fertilization. N Engl J Med 2004;351:2392–402. Maheshwari A, Griffiths S, Bhattacharya S. Global variations in the uptake of single embryo transfer. Hum Reprod Update 2011;17:107–20. Kajiwara K. Prevention of multiple pregnancy. J Jpn Soc Obstet Gynecol 2009;61:N335–9 (in Japanese). Yoshimura Y. Futuristics of reproductive medicine—–for a baby that is expected to be born. Jpn Shindan Chiryo sya 2010:26–9 (in Japanese). Takeshima K, Saito H, Nakaza A, Kuwahara A, Ishihara O, Irahara M, et al. Efficacy, safety, and trends in assisted reproductive technology in Japan— analysis of four-year data from the national registry system. J Assist Reprod Genet 2014;31:477–84. Ishihara O, Araki R, Kuwahara A, Itakura A, Saito H, Adamson GD. Impact of frozen-thawed single-blastocyst transfer on maternal and neonatal outcome: an analysis of 277,042 single-embryo transfer cycles from 2008 to 2010 in Japan. Fertil Steril 2014;101:128–33. ART registry of Japan. (in Japanese) Available at: http://plaza.umin.ac.jp/ jsog-art/. Accessed September 4, 2015. Itabashi K, Fujimura M, Kusuda S, Tamura M, Hayashi T, Takahashi T, et al. New standard of average size and weight of newborn in Japan. Jap J Pediat 2010;114:1271–93 (in Japanese).
20.
21.
22.
23.
24.
25.
26.
Guzoglu N, Kanmaz HG, Dilli D, Uras N, Erdeve O, Dilmen U. The impact of the new Turkish regulation, imposing single embryo transfer after assisted reproduction technology, on neonatal intensive care unit utilization: a single center experience. Hum Reprod 2012;27:2384–8. Sazonova A, K€allen K, Thurin-Kjellberg A, Wennerholm UB, Bergh C. Neonatal and maternal outcomes comparing women undergoing two in vitro fertilization (IVF) singleton pregnancies and women undergoing one IVF twin pregnancy. Fertil Steril 2013;99:731–7. De Sutter P, Delbaere I, Gerris J, Verstraelen H, Goetgeluk S, Van der Elst J, et al. Birthweight of singletons after assisted reproduction is higher after single- than after double-embryo transfer. Hum Reprod 2006;21: 2633–7. Sazonova A, K€allen K, Thurin-Kjellberg A, Wennerholm UB, Bergh C. Factors affecting obstetric outcome of singletons born after IVF. Hum Reprod 2011; 26:2878–86. Poikkeus P, Gissler M, Unkila-Kallio L, Hyden-Granskog C, Tiitinen A. Obstetric and neonatal outcome after single embryo transfer. Hum Reprod 2007; 22:1073–9. Nakashima A, Araki R, Tani H, Ishihara O, Kuwahara A, Irahara M, et al. Implications of assisted reproductive technologies on term singleton birth weight: an analysis of 25,777 children in the national assisted reproduction registry of Japan. Fertil Steril 2013;99:450–5. Pinborg A, Henningsen AA, Loft A, Malchau SS, Forman J, Andersen AN. Large baby syndrome in singletons born after frozen embryo transfer (FET): is it due to maternal factors or the cryotechnique? Hum Reprod 2014;29:618–27. Sullivan EA, Wang YA, Hayward I, Chambers GM, Illingnorth P, McBain J, et al. Single embryo transfer reduces the risk of perinatal mortality, a population study. Hum Reprod 2012;27:3609–15. Wang YA, Sullivan EA, Healy DL, Black DA. Perinatal outcomes after assisted reproductive technology treatment in Australia and New Zealand: single versus double embryo transfer. Med J Aust 2009;190:234–7. Roque M, Lattes K, Serra S, Sola I, Geber S, Carreras R, et al. Fresh embryo transfer versus frozen embryo transfer in in vitro fertilization cycles: a systematic review and meta-analysis. Fertil Steril 2013;99:156–62. 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. Maheshwari A, Pandey S, Shetty A, Hamilton M, Bhattacharya S. Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of frozen thawed versus fresh embryos generated through in vitro fertilization treatment: a systematic review and meta-analysis. Fertil Steril 2012;98: 368–77. Statistics Bureau, Ministry of Internal Affairs and Communications with the collaboration of Ministries and Agencies. Portal Site of Official Statistics of Japan. Demographic statistics. (in Japanese) Available at: http://www.e-stat.go. jp/SG1/estat/List.do?lid¼000001127058. Accessed July 6, 2015. Pinborg A, Lidegaard Ø, La Cour Freiesleben N, Andersen AN. Vanishing twins: a predictor of small-for-gestational age in IVF singletons. Hum Reprod 2007;22:2707–14. Shebl O, Ebner T, Sommergruber M, Sir A, Tews G. Birth weight is lower for survivors of the vanishing twin syndrome: a case-control study. Fertil Steril 2008;90:310–4. Luke B, Brown MB, Grainger D, Stern JE, Klein N, Cedars MI. The effect of early fetal losses on singleton assisted-conception pregnancy outcomes. Fertil Steril 2009;91:2578–85.
VOL. - NO. - / - 2015
10
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
VOL. - NO. - / - 2015
SUPPLEMENTAL TABLE 1 Pregnancy rate stratified by age from 2007 to 2012 in the Japanese Assisted Reproduction Technology registry.a
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
Age, y
2007
2008
2009
2010
2011
2012
Total
<30 30–34 35–39 R40
2,769/10,077 (27.5) 10,901/41,606 (26.2) 11,929/59,731 (20.0) 3,565/50,578 (7.0)
2,719/10,818 (25.1) 11,337/46,319 (24.5) 14,042/72,376 (19.4) 4,413/61,106 (7.2)
3,008/11,314 (26.6) 12,287/49,257 (24.9) 16,467/81,805 (20.1) 5,666/71,417 (7.9)
3,046/11,408 (26.7) 13,242/51,496 (25.7) 18,507/92,721 (20.0) 6,842/86,535 (7.9)
3,270/12,018 (27.2) 13,635/53,476 (25.4) 20,660/101,812 (20.2) 8,098/102,353 (7.9)
3,590/13,338 (26.9) 15,414/60,400 (25.5) 24,305/123,008 (19.8) 10,447/129,680 (8.1)
18,402/68,973 (26.7) 76,816/302,554 (26.7) 105,910/531,453 (19.9) 39,031/501,669 (7.8)
Note: Data in parentheses are percents. a The data originated from official annual reports by JSOG. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
Fertility and Sterility®
10.e1
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
SUPPLEMENTAL TABLE 2 Adjusted ORs of number of ET for selected perinatal outcomes stratified by fresh/frozen cycles and embryo stage at transfer. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend PIH SET DET MET P value for trend Placenta previa SET DET MET P value for trend Placenta accreta SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend LBW SET DET MET P value for trend SGAc SET DET MET P value for trend LGAd SET DET MET P value for trend
Frozen cycles
Early cleavage Adjusted OR (95% CI)a
Blastocyst Adjusted OR (95% CI)a
Early cleavage Adjusted OR (95% CI)b
Blastocyst Adjusted OR (95% CI)b
Reference 22.9 (19.6–26.8) 32.0 (25.9–39.4) < .001
Reference 15.5 (13.2–18.3) 17.4 (11.9–25.6) < .001
Reference 14.5 (11.8–17.7) 21.6 (16.3–28.7) < .001
Reference 17.5 (16.2–19.0) 21.9 (17.2–27.9) < .001
Reference 0.69 (0.58–0.82) 0.49 (0.29–0.82) < .001
Reference 1.1 (0.82–1.5) 1.3 (0.48–3.7) .46
Reference 1.0 (0.85–1.2) 0.54 (0.28–1.04) .45
Reference 1.1 (1.0–1.3) 1.1 (0.66–1.9) .03
Reference 0.90 (0.70–1.1) 0.78 (0.41–1.5) .27
Reference 0.87 (0.55–1.4) 2.2 (0.66–7.0) .99
Reference 1.1 (0.75–1.6) 0.75 (0.23–2.5) .86
Reference 0.56 (0.41–0.75) 0.30 (0.04–2.2) < .001
Reference 0.34 (0.098–1.2) NA NA
Reference 1.6 (0.55–4.9) NA NA
Reference 1.3 (0.75–2.1) NA NA
Reference 1.6 (1.2–2.1) NA NA
Reference 1.6 (1.5–1.7) 1.8 (1.6–2.0) < .001
Reference 1.7 (1.6–1.9) 1.7 (1.3–2.3) < .001
Reference 1.5 (1.4–1.6) 1.9 (1.6–2.3) < .001
Reference 1.8 (1.69–1.84) 2.1 (1.7–2.6) < .001
Reference 2.2 (2.0–2.4) 3.2 (2.7–3.7) < .001
Reference 2.5 (2.2–2.7) 2.7 (1.9–3.9) < .001
Reference 2.1 (1.9–2.3) 2.6 (2.1–3.3) < .001
Reference 2.6 (2.5–2.8) 3.1 (2.4–4.0) < .001
Reference 2.7 (2.6–2.9) 3.9 (3.4–4.4) < .001
Reference 3.1 (2.8–3.4) 3.8 (2.8–5.2) < .001
Reference 2.6 (2.3–2.8) 3.7 (3.0–4.6) < .001
Reference 3.5 (3.3–3.7) 4.7 (3.7–5.8) < .001
Reference 1.6 (1.5–1.7) 2.2 (1.8–2.5) < .001
Reference 1.5 (1.4–1.7) 1.9 (1.3–2.8) < .001
Reference 1.3 (1.2–1.5) 1.6 (1.2–2.2) < .001
Reference 1.9 (1.8–2.1) 2.9 (2.2–3.8) < .001
Reference 0.76 (0.69–0.83) 0.61 (0.48–0.77) < .001
Reference 0.74 (0.65–0.85) 0.74 (0.45–1.2) < .001
Reference 0.80 (0.72–0.88) 0.82 (0.63–1.1) < .001
Reference 0.70 (0.66–0.75) 0.72 (0.53–0.98) < .001
Note: NA ¼ not available. a Adjusted for calendar year, maternal age, and fertilization method. b Adjusted for calendar year and maternal age. c SGA was defined as being below the 10th percentile of the national reference. d LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
10.e2
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
Fertility and Sterility®
SUPPLEMENTAL TABLE 3 Adjusted ORs of number of embryo transfer for selected perinatal outcomes stratified by fresh/frozen cycles and age. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend PIH SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend LBW SET DET MET P value for trend LGAc SET DET MET P value for trend Early neonatal death SET DET MET P value for trend
Frozen cycles
Age <35 Adjusted OR (95% CI)a
Age‡35 Adjusted OR (95% CI)a
Age <35 Adjusted OR (95% CI)b
Age‡35 Adjusted OR (95% CI)b
Reference 24.4 (20.8–28.6) 32.3 (25.4–41.2) < .001
Reference 14.0 (12.0–16.3) 19.1 (15.0–24.3) < .001
Reference 21.2 (19.0–23.6) 32.5 (25.5–41.4) < .001
Reference 13.7 (12.4–15.1) 16.6 (13.1–21.1) < .001
Reference 1.03 (0.79–1.3) 0.83 (0.40–1.7) .93
Reference 0.71 (0.53–0.84) 0.51 (0.28–0.93) < .001
Reference 1.2 (1.03–1.4) 1.1 (0.59–2.0) .03
Reference 1.1 (0.94–1.2) 0.63 (0.36–1.1) .9
Reference 2.1 (1.9–2.2) 2.3 (1.9–2.7) < .001
Reference 1.5 (1.4–1.6) 1.6 (1.4–1.8) < .001
Reference 1.9 (1.8–2.1) 2.4 (1.9–2.9) < .001
Reference 1.53 (1.46–1.6) 1.9 (1.6–2.2) < .001
Reference 2.9 (2.6–3.2) 3.9 (3.2–4.9) < .001
Reference 1.9 (1.8–2.1) 2.6 (2.2–3.2) < .001
Reference 3.2 (2.9–3.4) 4.2 (3.3–5.3) < .001
Reference 2.1 (2.0–2.3) 2.5 (2.0–3.1) < .001
Reference 3.7 (3.4–4.0) 5.0 (4.2–6.0) < .001
Reference 2.3 (2.2–2.5) 3.3 (2.8–3.8) < .001
Reference 4.1 (3.8–4.3) 6.2 (5.0–7.6) < .001
Reference 2.8 (2.6–2.9) 3.4 (2.7–4.1) < .001
Reference 0.66 (0.59–0.76) 0.51 (0.36–0.72) < .001
Reference 0.81 (0.74–0.88) 0.72 (0.55–0.94) < .001
Reference 0.63 (0.57–0.69) 0.66 (0.48–0.91) < .001
Reference 0.80 (0.75–0.85) 0.88 (0.69–1.1) < .001
Reference 2.4 (0.93–6.1) 14.5 (5.5–40.3) < .001
Reference 0.84 (0.48–1.5) 0.94 (0.21–4.3) .62
Reference 1.4 (0.72–2.6) NA NA
Reference 0.86 (0.44–1.7) 1.5 (0.17–1.7) .83
Note: NA ¼ not available. a Adjusted for calendar year, embryo stage at transfer, and fertilization method. b Adjusted for calendar year and embryo stage at transfer. c LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
10.e3
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
Impact of single embryo transfer policy on perinatal outcomes in fresh and frozen cycles—analysis of the Japanese Assisted Reproduction Technology registry between 2007 and 2012 Kazumi Takeshima, M.D.,a Seung Chik Jwa, M.D., Ph.D., M.P.H.,a,b Hidekazu Saito, M.D., Ph.D.,a Aritoshi Nakaza, B.S.,b Akira Kuwahara, M.D., Ph.D.,c Osamu Ishihara, M.D., Ph.D.,d Minoru Irahara, M.D., Ph.D.,c Fumiki Hirahara, M.D., Ph.D.,e Yasunori Yoshimura, M.D., Ph.D.,f and Tetsuro Sakumoto, M.D., Ph.D.b a Division of Reproductive Medicine, Center of Maternal-Fetal, Neonatal and Reproductive Medicine, National Center for Child Health and Development, Tokyo; b SORA no MORI clinic, Okinawa; c Department of Obstetrics and Gynecology, School of Medicine, University of Tokushima, Tokushima; d Department of Obstetrics and Gynecology, Saitama Medical University, Saitama; e Department of Obstetrics and Gynecology, Yokohama City University Graduate School of Medicine, Hukuura, Kanazawa-ku, Yokohama City, Kanagawa; and f Yoshimura Bioteic Literacy Institute, Tokyo, Japan
Objective: To investigate whether the introduction of single embryo transfer (SET) policy in Japan has improved perinatal outcomes. Design: A retrospective cohort study. Setting: Not applicable. Patient(s): A total of 140,718 live births and 510 stillbirths (after 22 weeks of gestation) conceived by assisted reproductive technology in Japan between 2007 and 2012 were reviewed. Intervention(s): None. Main Outcome Measure(s): Preterm birth (PTB), low birth weight (LBW), very low birth weight (VLBW), small for gestational age (SGA), large for gestational age (LGA), perinatal mortality, and other pregnancy complications. Result(s): The rate of SET increased significantly from 52.2% in 2007 to 82.6% in 2012, while the rate of multiple pregnancy decreased significantly from 10.7% to 4.1% over the same period. The rates of PTB, LBW, and SGA decreased significantly, while that of LGA increased. Perinatal mortality decreased from 0.70% to 0.40% in fresh cycles, while that of frozen cycles did not change. Double ET or more was associated with a significantly increased risk for multiple pregnancy, placenta accreta, preterm premature rupture of membrane, cesarean section (CS), PTB, LBW, SGA, and early neonatal death compared with SET. Compared with before the SET policy was launched, the risks of multiple pregnancy, CS, early PTB before 32 weeks, LBW, VLBW, and SGA were significantly decreased after the policy was launched, with significant interactions of fresh/frozen status. Conclusion(s): The results suggest that the SET policy improved perinatal outcomes in Japan. The impact of SET policy was different in fresh and frozen cycles for several perinatal outcomes. Use your smartphone (Fertil SterilÒ 2015;-:-–-. Ó2015 by American Society for Reproductive Medicine.) to scan this QR code Key Words: Assisted reproductive technology, perinatal outcome, preterm delivery, low birth and connect to the weight, single embryo transfer Discuss: You can discuss this article with its authors and with other ASRM members at http:// fertstertforum.com/takeshimak-set-policy-perinatal-outcomes/
discussion forum for this article now.*
* Download a free QR code scanner by searching for “QR scanner” in your smartphone’s app store or app marketplace.
Received May 27, 2015; revised September 13, 2015; accepted October 5, 2015. K.T. has nothing to disclose. S.C.J. has nothing to disclose. H.S. has nothing to disclose. A.N. has nothing to disclose. A.K. has nothing to disclose. O.I. has nothing to disclose. M.I. has nothing to disclose. F.H. has nothing to disclose. Y.Y. has nothing to disclose. T.S. has nothing to disclose. K.T. and S.C.J. should be considered similar in author order. This study was supported by the grant of National Center for Child Health and Development (24-6) and a grant of the Ministry of Health, Labour and Welfare (no. H25-001). Reprint requests: Kazumi Takeshima, M.D., Division of Reproductive Medicine, Center of Maternal-Fetal, Neonatal and Reproductive Medicine, National Center for Child Health and Development, 2-10-1, Okura, Setagaya-ku, Tokyo, 157-8535, Japan (E-mail: [email protected]). Fertility and Sterility® Vol. -, No. -, - 2015 0015-0282/$36.00 Copyright ©2015 American Society for Reproductive Medicine, Published by Elsevier Inc. http://dx.doi.org/10.1016/j.fertnstert.2015.10.002 VOL. - NO. - / - 2015
1
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
S
ince the first IVF baby was born in 1978 (1), assisted reproductive technology (ART) has been widely performed globally. However, along with ART treatment, the rate of multiple pregnancy and poor perinatal outcome is on the rise. To that end, single embryo transfer (SET) has been introduced (2, 3) and advocated as a preventive measure in several countries (4). In Japan, multiple pregnancy caused by ART has also been a concern. In 1996, the Japan Society of Obstetrics and Gynecology (JSOG) recommended that the number of transferred embryos should be confined to three or less. Although the rate of high-order multiple pregnancy (triplet or more) has decreased since then, the real number of twin pregnancies has increased (5, 6). Subsequently, in 2008, the JSOG recommended that SET should normally be performed to prevent multiple pregnancy, while double embryo transfer (DET) is allowed in women who are over 35 years old and/or have experienced failure of implantation for more than two consecutive attempts. As a result, the rate of SET being performed between 2007 and 2010 surged from 49.9% to 73.0%, while the rate of multiple pregnancy decreased from 11.5 % to 4.8% (7). In view of these facts, it is worthwhile to investigate whether perinatal outcomes, including preterm delivery and low birth weight (LBW) rates, are affected by the introduction of the SET policy from 2008 onward. In the present study, we analyzed the data of the national ART registry from 2007 to 2012 to evaluate whether the SET policy improved perinatal outcomes including perinatal mortality, multiple pregnancy, preterm birth (PTB), LBW, small for gestational age (SGA), large for gestational age (LGA), and other pregnancy complications. Since it was suggested that frozen-thawed embryo transfer (frozen ET) has better perinatal outcome than fresh embryo transfer (fresh ET) (8), we also investigated whether the impact of SET policy on perinatal outcomes is different in fresh and frozen cycles.
MATERIALS AND METHODS Data Source
All data analyzed were obtained from the ART online registry provided by the JSOG. Since 2007, the JSOG has required all institutes that perform ART to register cycle-specific information including patient's age, number of ETs, and perinatal outcomes with an online system. Over 99% of registered facilities record their treatment data in the registry every year. The data were provided by the JSOG upon approval by the registration and research subcommittee of the JSOG ethics committee. Further, the Institutional Review Board at the National Center for Child Health and Development approved this study.
Patients The number of registered institutions in the database each year between 2007 and 2012 was 606, 609, 625, 591, 586, and 589, respectively. Pregnancy rate stratified by age from the data announced officially by JSOG are shown as Supplemental Table 1 (available online) (9). The pregnancy
rate decreased as age rose, and the trend was particularly prominent in patients over 40 years old. There was no trend in pregnancy rate from 2007 to 2012 in the groups under 40 years old. However, the pregnancy rate slightly increased by approximately 1% in these 6 years in the group over 40 years old. Those with live birth at 22–41 of gestational weeks and stillbirth after 22 weeks of gestation were analyzed. There were 144,808 eligible cycles for live births with known gestational age at delivery and birth weight. Among them, treatment cycle of gamete intrafallopian transfer (n ¼ 36), unknown fertilization method (n ¼ 465), cancellation cycles for ET (n ¼ 252), unknown number of ETs (n ¼ 1), cycles using frozen oocytes (n ¼ 57), and cases with missing or incomplete records (n ¼ 3,279) were excluded. There were 510 eligible stillbirth cases after 22 weeks of gestation. Consequently, the total number of analyzed subjects was 141,228. Since donor gametes or embryos are not allowed to be used in ART in Japan, all of the embryos transferred were autologous.
Perinatal Outcomes Our primary outcomes are gestational age and birth weight. Gestational age was divided into three groups: delivery at over 37 gestational weeks (term delivery), delivery before 37 weeks (PTB), and delivery before 32 weeks (early preterm birth [EPTB]). In terms of birth weight, we classified the newborns into three groups: newborns of over 2,500 g (normal birth weight), newborns of 1,500–2,499 g (LBW), and newborns of less than 1,500 g (very low birth weight [VLBW]). Further, according to the new standard of average size and weight of newborns of the Japan Pediatric Society, we also classified birth weight into SGA defined as below the 10th percentile of standard and LGA defined as above the 90th percentile (10). As secondary outcomes, stillbirth, early neonatal death within 7 days after birth, multiple pregnancy, cesarean section (CS), and other pregnancy complications including pregnancy-induced hypertension (PIH), gestational diabetes mellitus (GDM), placenta previa, placenta accreta, placenta abruption, and preterm premature rupture of membrane (pPROM) were included for the analysis. Perinatal mortality was calculated as incidence of stillbirth or early neonatal death among known cycles for each year. Multiple pregnancy was defined based on the number of live births.
Statistical Analyses We evaluated whether there is a linear trend for each variable over the years 2007–2012 by calculating linear regression using each variable as a dependent variable and the calendar year as the independent and ordinal variable. If a variable was continuous or dichotomous, it was put into the regression model directly as a dependent variable. If a variable consisted of more than two categories, indicator variables were created and those indicator variables were separately put into the regression model. For perinatal outcomes, whether there is a linear trend from year 2007 to 2012 was separately evaluated according to fresh/frozen status. Further, we evaluated whether these trends were changed after restricting SET VOL. - NO. - / - 2015
2
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Fertility and Sterility® cycles. Crude and adjusted odds ratios (ORs) of DET and multiple embryo transfer (MET, i.e., three or more ET) compared with SET were calculated by logistic generalized estimating equations, with robust variance estimation to adjust the effect of clustering of births among the mothers stratified by fresh and frozen cycles. Priori covariates for adjusted analysis were maternal age (categorized into <30, 30–34, 35–39, R40 years), years of cycles reported, and embryo stage at transfer (early cleavage/blastocyst stage). Since fertilization method (i.e., IVF or intracytoplasmic sperm injection [ICSI]) was only available in fresh cycles, it was adjusted in the analysis of fresh cycles. Trends across the number of embryos transferred (i.e., SET, DET, and MET) for perinatal outcomes were tested by treating SET, DET, and MET as ordinal variables in the logistic regression model. We also investigated interactions between embryo stage at transfer, maternal age (before/after 35 years), and number of ETs for perinatal outcomes, and significant interactions were observed for the outcomes of multiple birth, PIH, placenta previa, placenta accreta, CS, PTB, LBW, SGA, and LGA between embryo stage at transfer and number of ETs and for the outcomes of multiple birth, PIH, CS, PTB, LBW, LGA, and early neonatal death between maternal age and number of ETs. Thus, we considered embryo stage at transfer and age to serve as effect modifiers for those outcomes and additionally stratified by embryo stage at transfer and age. Finally, since the SET policy was announced by JSOG in April 2008, we divided samples into before/after the SET policy was announced based on the date of ET and evaluated whether risks for those perinatal outcomes were changed before/after the SET policy was
launched. We also investigated whether the effect of the SET policy on perinatal outcomes differed by fresh/frozen status by adding an interaction term between number of ETs and fresh/frozen status in adjusted model. All analyses were performed with the STATA/SE statistical package, version 12.1 (Stata Corp.). Two-tailed P values < .05 were considered as statistically significant.
RESULTS Table 1 shows the characteristics of cycles reported in each year from 2007 to 2012. During that period, the reported number of cycles dramatically increased from 14,991 cycles in 2007 to 33,066 cycles in 2012, of which fresh cycles increased moderately from 7,797 cycles to 9,126 cycles, while frozen cycles increased more drastically from 7,194 cycles to 23,940 cycles, resulting in the proportion of fresh cycles significantly decreasing from 2007 to 2012. There was a significant upward trend in terms of the age of subjects both in fresh and frozen cycles (P for trends < .001). In particular, the rate of over 35 years old increased from nearly 50% in 2007 to 60% in 2012. In terms of the fertilization method of fresh ET, the rate of IVF significantly decreased from 49.2% in 2007 to 45.9% in 2012, while that of ICSI and split fertilization (IVFþICSI) increased slightly. After the introduction of SET policy by the JSOG in 2008, the rate of SET significantly increased during the analyzed period from 52.2% in 2007 to 82.6% in 2012. On the other hand, the rate of DET significantly decreased from 38.1% in 2007 to 17.0% in 2012. Similarly, the rate of three or more
TABLE 1 Characteristics of cycles transferred among ART patients stratified by year, 2007–2012 (n [ 141,228 cycles). Characteristic Maternal age <30 30–34 35–39 R40 Fresh or frozen Fresh Frozen Fertilization method in fresh cyclea IVF ICSI Split (IVFþICSI) No. of ETs One Two Three or more Embryo stage at transfer Early cleavage Blastocyst No. of single and multiple pregnancies Single Twin Triplet
2007 (n [ 14,991)
2008 (n [ 17,956)
2009 (n [ 22,366)
2010 (n [ 24,832)
2011 (n [ 28,017)
2012 (n [ 33,066)
34.4 3.8 1,519 (10.1) 6,008 (40.1) 6,118 (40.8) 1,346 (9.0)
34.6 3.8 1,696 (9.5) 6,793 (37.8) 7,678 (42.8) 1,789 (10.0)
34.8 3.8 2,010 (9.0) 8,049 (36.0) 9,811 (43.9) 2,496 (11.2)
35.0 3.8 2,063 (8.3) 8,609 (34.7) 11,136 (44.9) 3,024 (12.2)
35.2 3.9 2,266 (8.1) 9,166 (32.7) 12,825 (45.8) 3,760 (13.4)
35.4 3.9 2,493 (7.5) 10,501 (31.8) 15,208 (46.0) 4,864 (14.7)
7,797 (52.0) 7,194 (48.0)
7,634 (42.5) 10,322 (57.5)
8,603 (38.5) 13,763 (61.5)
8,557 (34.5) 16,275 (65.5)
8,728 (31.2) 19,289 (68.9)
9,126 (27.6) 23,940 (72.4)
3,834 (49.2) 3,349 (43.0) 614 (7.9)
3,800 (49.8) 3,233 (42.4) 601 (7.9)
4,207 (48.9) 3,661 (42.6) 735 (8.5)
4,011 (46.9) 3,743 (43.7) 803 (9.4)
3,961 (45.4) 3,885 (44.5) 882 (10.1)
4,190 (45.9) 4,026 (44.2) 910 (10.0)
7,825 (52.2) 5,718 (38.1) 1,448 (9.7)
12,347 (68.8) 5,199 (29.0) 410 (2.3)
16,930 (75.7) 5,246 (23.5) 190 (0.9)
19,567 (78.8) 5,114 (20.6) 151 (0.6)
22,642 (80.8) 5,232 (18.7) 143 (0.5)
27,316 (82.6) 5,620 (17.0) 130 (0.4)
7,400 (49.4) 7,591 (50.6)
7,132 (39.7) 10,824 (60.3)
8,471 (37.9) 13,895 (65.1)
8,413 (33.9) 16,419 (66.1)
9,085 (32.4) 18,932 (67.6)
10,464 (31.7) 22,602 (68.4)
13,382 (89.3) 1,580 (10.5) 29 (0.19)
16,824 (93.7) 1,122 (6.3) 10 (0.06)
21,193 (94.8) 1,160 (5.2) 13 (0.06)
23,634 (95.2) 1,182 (4.8) 16 (0.06)
26,840 (95.8) 1,156 (4.1) 21 (0.07)
31,722 (96.0) 1,327 (4.0) 17 (0.05)
Note: Data are shown mean SD for continuous variables and n (%) for dichotomous variables. a Denominators are numbers of fresh cycles in each year. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
3
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION ETs decreased from 9.7% in 2007 to 2.3% in 2008 and remained below 1% after 2009. Accordingly, singleton pregnancies gradually increased from 89.3% in 2007 to 96.0% in 2012, and twin pregnancies decreased from 10.5% in 2007 to 4.0% in 2012 (P for trend < .001), while multiple pregnancies (triplet or more) decreased from 0.19% in 2007 to 0.06% in 2008 and became almost stable after 2008. Table 2 shows perinatal outcomes and pregnancy complications for fresh ET and frozen ET stratified by calendar year. For the gestational weeks, the rates of PTB and LBW decreased both in fresh and frozen cycles (P for trend < .001). The mean birth weight in frozen ET was heavier than that in fresh ET in all the analyzed years. The rate of SGA decreased (11.3% to 8.4% in fresh ET, 6.9% to 5.6% in frozen ET, from 2007 to 2012), while that of LGA increased (7.3% to 8.6% in fresh ET, 10.7% to 13.2% in frozen ET, from 2007 to 2012; P for trend < .001). In frozen ET, the rate of SGA was lower, while that of LGA was higher compared with those in fresh ET throughout the analyzed years. The rate of stillbirth did not change over the study period in either fresh or frozen cycles, while the rate of early neonatal death decreased from 0.21% in 2007 to 0.11% in 2012 in fresh cycles with significant trend (P for trend ¼ .004) but did not change in frozen cycles, resulting in a significant decreasing trend in perinatal mortality in fresh cycles. For pregnancy complications, PIH and GDM significantly increased from 2007 to 2012, while other pregnancy complications of placenta previa, placenta accreta, placenta abruption, and pPROM did not demonstrate significant trend over the study period. These trends for PIH and GDM remained significant after adjusting for maternal age and restricting SET cycles, except for PIH in fresh cycles. CS decreased significantly from 35.6% in 2007 to 32.8% in 2012 in fresh cycles but did not change in frozen cycles. When we restricted SET cycles, significant trends disappeared or were attenuated in the outcomes of PIH, mode of deliveries, gestational weeks at birth, birth weight, SGA, LGA, early neonatal death, and perinatal mortality in fresh cycles and vaginal deliveries, gestational weeks at birth, birth weight, and SGA in frozen cycles, suggesting that these significant trends in perinatal outcomes were related to the increased rate of SET over time. In Table 3, crude and adjusted ORs of DET and MET compared with SET for perinatal outcomes are presented. In fresh DET and MET, the adjusted OR of multiple pregnancy was 19.2 (95% confidence interval [CI], 17.2–21.5) and 26.4 (95% CI, 22.2–31.3) with significant trend for number of ET (P for trend < .001). Similar significant results were observed in frozen cycles. MET was associated with a significantly higher risk for early neonatal death in fresh cycles. On the other hand, DET was associated with a significantly higher risk for pPROM in both fresh and frozen cycles but a significantly lower risk for placenta previa in frozen cycles only. Interestingly, both DET and MET demonstrated a significantly decreased risk for PIH and GDM in fresh cycles. For delivery outcomes of PTB, EPTB, LBW, VLBW, SGA, and CS, DET and MET demonstrated a significantly increased risk compared with SET and a significantly decreased risk for LGA in both fresh and frozen cycles with significant trend.
For perinatal outcomes with significant interaction between embryo stage at transfer and number of ETs, adjusted ORs additionally stratified by embryo stage are shown in Supplemental Table 2 (available online). Significant ORs of DET and MET for multiple pregnancy were attenuated in blastocyst ET in fresh cycles but were still significant. A significantly decreased risk for PIH in fresh cycles was observed in early cleavage ET, whereas in blastocyst ET, an increased risk was observed, although it was not significant. A significantly decreased risk of DET in frozen cycles for placenta previa and a significantly increased risk for placenta accreta were observed only in blastocyst ET. Similarly, for perinatal outcomes with significant interaction between age (before/after 35 years) and number of ETs, adjusted ORs additionally stratified by age are shown in Supplemental Table 3 (available online). Significant ORs of DET and MET for multiple pregnancy were attenuated after the age of 35 years both in fresh and frozen cycles but were still significant. A significantly decreased risk of DET and MET for PIH in fresh cycles was observed only in women after the age of 35 years, while a significantly increased risk of DET for PIH in frozen cycles was observed only in women before 35 years of age. For the delivery outcomes of CS, LBW, and SGA, adjusted ORs of DET and MET were attenuated in women older than 35 years in both fresh and frozen cycles. Table 4 shows adjusted ORs of the effect of SET policy for selected perinatal outcomes. After the SET policy was introduced in April 2008, SET increased from 54.2% to 78.9%, while DET and MET decreased from 34.7% to 20.5% and 8.4% to 0.70%, respectively. After SET policy was announced, adjusted ORs for multiple pregnancy were significantly decreased in both fresh and frozen cycles (adjusted OR in fresh cycles, 0.42; 95% CI, 0.39–0.45; adjusted OR in frozen cycles, 0.50; 95% CI, 0.46–0.54, respectively). A similar significantly decreased risk was observed in stillbirth, CS, PTB, LBW, and SGA in both fresh and frozen cycles. For EPTB, VLBW, and early neonatal death, adjusted ORs were significantly decreased only in fresh cycles after SET policy was announced. P values for interaction between before/after SET policy and fresh/frozen status were significant for the outcomes of multiple birth, CS, EPTB, LBW, VLBW, and SGA, suggesting that the effect of the SET policy on those perinatal outcomes is significantly different by fresh/frozen status.
DISCUSSION Our study demonstrated that the introduction of SET policy by the JSOG in 2008 resulted in significant improvements in many perinatal outcomes including multiple pregnancy, PTB, LBW, SGA, perinatal mortality, and other pregnancy complications over the study period. Our results suggest that introduction of SET policy may have a different impact on several perinatal outcomes in fresh and frozen cycles. To date, many studies have reported on the improvement of perinatal outcome by SET. A study from Turkey reported that mandatory SET resulted in a decrease in multiple pregnancy and admissions to the neonatal intensive care unit (11). According to Sazonova et al. (12), women with two VOL. - NO. - / - 2015
4
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
VOL. - NO. - / - 2015
TABLE 2 Perinatal outcomes stratified by calendar year and fresh/frozen cycles, 2007–2012. Outcome
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
.16 .012 < .001 .85 .55 .34 .81
.76 .39 < .001 .22 .81 .86 .27
< .001 < .001 < .001 < .001 < .001 < .001 .01
.011 .45 < .001 .81 .23 .37 .38
< .001 < .001 < .001 < .001
.07 .10 .25 .13
4,693 (49.6) 4,699 (49.7) 62 (0.7) 786 (8.4) 806 (8.6) 10 (0.11) 1,021 (10.8) 0.40
.017 .012 .84 < .001 .001 .004 < .001 .002
.09 .12 .09 .046 .41 .26 < .001 .21
n ¼ 19,289 62 (0.32) 657 (3.4) 238 (1.23) 122 (0.63) 83 (0.43) 36 (0.19) 71 (0.37) n ¼ 19,227
n ¼ 23,940 78 (0.33) 873 (3.7) 324 (1.35) 204 (0.85) 89 (0.37) 41 (0.17) 107 (0.45) n ¼ 23,862
.58 < .001 < .001 .25 .15 .94 .11
.99 < .001 < .001 .30 .014 .72 .021
11,097 (57.7) 7,754 (40.3) 376 (2.0)
13,628 (57.1) 9,776 (41.0) 458 (1.9)
< .001 .60 < .001
.32 < .001 < .001
n ¼ 7,797 38 (0.49) 139 (1.8) 29 (0.14) 72 (0.92) 3 (0.04) 17 (0.22) 35 (0.45) n ¼ 7,759
n ¼ 7,634 26 (0.34) 142 (1.9) 32 (0.42) 74 (0.97) 6 (0.08) 20 (0.26) 28 (0.37) n ¼ 7,608
n ¼ 8,603 37 (0.43) 164 (1.9) 40 (0.46) 78 (0.91) 9 (0.10) 20 (0.23) 25 (0.29) n ¼ 8,566
n ¼ 8,557 45 (0.53) 168 (2.0) 72 (0.84) 65 (0.76) 2 (0.02) 35 (0.41) 35 (0.41) n ¼ 8,512
n ¼ 8,728 34 (0.39) 200 (2.3) 108 (1.24) 78 (0.89) 8 (0.09) 22 (0.25) 35 (0.40) n ¼ 8,694
n ¼ 9,126 26 (0.28) 197 (2.2) 139 (1.52) 93 (1.02) 7 (0.08) 26 (0.28) 34 (0.37) n ¼ 9,100
4,576 (59.0) 2,760 (35.6) 423 (5.5) 38.2 2.2 6,709 (86.5) 895 (11.5) 155 (2.0) n ¼ 8,675 2,781 556 6,424 (74.1) 2,012 (23.2) 239 (2.8)
4,802 (63.1) 2,597 (34.1) 209 (2.8) 38.4 2.2 6,712 (88.2) 771 (10.1) 125 (1.6) n ¼ 8,161 2,839 537 6,457 (79.1) 1,519 (18.6) 185 (2.3)
5,534 (64.6) 2,761 (32.2) 271 (3.2) 38.4 2.1 7,645 (89.3) 785 (9.2) 136 (1.6) n ¼ 9,060 2,855 526 7,274 (80.3) 1,587 (17.5) 199 (2.2)
5,478 (64.4) 2,774 (32.6) 260 (3.1) 38.4 2.1 7,649 (89.9) 718 (8.4) 145 (1.7) n ¼ 8,960 2,868 517 7,303 (81.5) 1,470 (16.4) 187 (2.1)
5,648 (65.0) 2,901 (33.4) 145 (1.7) 38.5 2.1 7,880 (90.6) 697 (8.0) 117 (1.4) n ¼ 9,061 2,891 506 7,516 (83.0) 1,385 (15.3) 160 (1.8)
5,937 (65.2) 2,984 (32.8) 179 (2.0) 38.5 2.0 8,251 (90.7) 708 (7.8) 141 (1.6) n ¼ 9,454 2,891 506 7,866 (83.2) 1,399 (14.8) 189 (2.0)
4,485 (51.7) 4,120 (47.5) 70 (0.8) 972 (11.3) 628 (7.3) 18 (0.21) 1,307 (15.1) 0.70
4,136 (50.7) 3,973 (48.7) 52 (0.6) 766 (9.4) 669 (8.3) 17 (0.21) 1,762 (21.6) 0.53
4,624 (51.0) 4,392 (48.5) 44 (0.5) 853 (9.5) 732 (8.1) 16 (0.18) 2,017 (22.3) 0.59
4,530 (50.6) 4,364 (48.7) 66 (0.7) 788 (8.9) 763 (8.6) 11 (0.12) 2,036 (22.7) 0.64
4,605 (50.8) 4,403 (48.6) 53 (0.6) 735 (8.2) 791 (8.8) 8 (0.09) 983 (10.9) 0.46
n ¼ 7,194 23 (0.32) 213 (3.0) 32 (0.44) 51 (0.71) 15 (0.21) 7 (0.10) 32 (0.44) n ¼ 7,171
n ¼ 10,322 46 (0.45) 289 (2.8) 59 (0.57) 83 (0.80) 46 (0.45) 26 (0.25) 32 (0.31) n ¼ 10,276
n ¼ 13,763 35 (0.25) 416 (3.0) 73 (0.53) 83 (0.60) 43 (0.31) 26 (0.19) 39 (0.28) n ¼ 13,728
n ¼ 16,275 60 (0.37) 532 (3.3) 155 (0.95) 113 (0.69) 45 (0.28) 32 (0.20) 40 (0.25) n ¼ 16,215
3,784 (52.8) 2,931 (40.9) 456 (6.4)
5,912 (57.5) 4,102 (39.9) 262 (2.6)
7,933 (57.8) 5,432 (39.6) 363 (2.6)
9,413 (58.1) 6,338 (39.1) 464 (2.9)
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
2010
P value for trend restricting SET cyclesa
2008
2011
2012
5
Fertility and Sterility®
Fresh Pregnancy complications Still birth PIH GDM Previa Placental accreta Placenta abruption pPROM Delivery outcomes Mode of delivery Vaginal CS Unknown Gestational weeks at birth (wk)b R37 3236 <32 Neonates Birth weight, g R2,500 1,500–2,499 <1,500 Sex Male Female Unknown SGAc LGAd Early neonatal deathe Unknown Perinatal mortality, %e Frozen Pregnancy complications Still birth PIH GDM Previa Placental accreta Placenta abruption pPROM Delivery outcomes Mode of delivery Vaginal CS Unknown
2009
P value for trenda
2007
Continued. Outcome
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Gestational weeks at birthb R37 3236 <32 Neonates Birth weight, g R2,500 1,500–2,499 <1,500 Sex Male Female Unknown SGAc LGAd Early neonatal deathe Unknown Perinatal mortality, %e
2007
2008
2009
2010
2011
2012
38.4 2.2 6,259 (87.3) 805 (11.2) 107 (1.5) n ¼ 7,887 2,901 538 6,343 (80.4) 1,417 (18.0) 127 (1.6)
38.5 2.1 9,239 (90.0) 898 (8.7) 139 (1.3) n ¼ 10,863 2,961 519 9,186 (84.6) 1,511 (13.9) 166 (1.5)
38.6 2.1 12,402 (90.3) 1,138 (8.3) 188 (1.4) n ¼ 14,418 2,977 517 12,350 (85.7) 1,851 (12.8) 217 (1.5)
38.5 2.2 14,606 (90.1) 1,350 (8.3) 259 (1.6) n ¼ 16,977 2,967 528 14,472 (85.2) 2,201 (13.0) 304 (1.8)
38.5 2.1 17,342 (90.2) 1,594 (8.3) 291 (1.5) n ¼ 20,055 2,975 516 17,295 (86.2) 2,424 (12.1) 336 (1.7)
38.5 2.1 21,565 (90.4) 1,942 (8.1) 355 (1.5) n ¼ 24,864 2,980 519 21,443 (86.2) 3,002 (12.1) 419 (1.7)
4,105 (52.1) 3,738 (47.4) 44 (0.6) 544 (6.9) 836 (10.7) 6 (0.08) 1,685 (21.4) 0.38
5,710 (52.6) 5,089 (46.9) 64 (0.6) 632 (5.9) 1,293 (12.0) 17 (0.16) 3,191 (29.4) 0.58
7,540 (52.3) 6,795 (47.1) 84 (0.6) 834 (5.8) 1,785 (12.5) 10 (0.07) 3,922 (27.2) 0.31
8,889 (52.4) 7,989 (47.1) 99 (0.6) 1,045 (6.2) 2,115 (12.5) 15 (0.09) 4,415 (26.0) 0.45
10,356 (51.6) 9,582 (47.8) 117 (0.6) 1,095 (5.5) 2,535 (12.7) 28 (0.14) 2,875 (14.3) 0.46
12,821 (51.6) 11,891 (47.8) 152 (0.6) 1,391 (5.6) 3,257 (13.2) 30 (0.12) 3,618 (14.6) 0.44
Note: Data are shown mean SD for continuous variables and n (%) for dichotomous variables. a Linear trends are assessed with linear regression using calendar year as the ordinal and independent variable. b Denominators are number of mothers for each year. c SGA was defined as being below the 10th percentile of the national reference. d LGA was defined as being above the 90th percentile of the national reference. e Excluding unknown cases for early neonatal death. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
P value for trenda
P value for trend restricting SET cyclesa
.46 < .001 < .001 .40
< .001 .07 .16 .19
< .001 < .001 < .001 .21
.73 .90 .41 .014
.062 .073 .28 < .001 < .001 .78 < .001 .36
< .001 < .001 .09 .19 < .001 .61 < .001 .55
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
6
TABLE 2
VOL. - NO. - / - 2015
Fertility and Sterility®
TABLE 3 Crude and adjusted OR of number of ET for perinatal outcomes stratified by fresh/frozen cycles. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend Still birth SET DET MET P value for trend PIH SET DET MET P value for trend GDM SET DET MET P value for trend Placenta previa SET DET MET P value for trend Placenta accreta SET DET MET P value for trend Placenta abruptio SET DET MET P value for trend pPROM SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend EPTBc SET DET MET P value for trend LBW SET DET MET P value for trend VLBW SET DET
Frozen cycles a
Crude OR (95% CI)
Adjusted OR (95% CI)
Crude OR (95% CI)
Adjusted OR (95% CI)b
Reference 16.6 (14.9–18.5) 23.6 (20.2–27.7) < .001
Reference 19.2 (17.2–21.5) 26.4 (22.2–31.3) < .001
Reference 15.1 (14.1–16.3) 20.1 (17.1–23.5) < .001
Reference 17.0 (15.8–18.4) 23.7 (19.9–28.0) < .001
Reference 1.1 (0.78–1.4) 1.2 (0.56–2.5) .62
Reference 0.95 (0.70–1.3) 0.99 (0.45–2.2) .80
Reference 1.2 (0.94–1.6) 0.31 (0.04–2.2) .41
Reference 1.1 (0.85–1.5) 0.26 (0.04–1.9) .87
Reference 0.84 (0.73–0.97) 0.64 (0.41–0.996) .003
Reference 0.77 (0.66–0.89) 0.58 (0.37–0.92) < .001
Reference 1.1 (1.004–1.2) 0.74 (0.49–1.1) .24
Reference 1.1 (0.997–1.2) 0.76 (0.50–1.2) .25
Reference 0.76 (0.61–0.95) 0.22 (0.07–0.69) < .001
Reference 0.76 (0.60–0.96) 0.31 (0.10–0.997) .003
Reference 0.69 (0.57–0.83) 0.29 (0.09–0.89) < .001
Reference 0.72 (0.59–0.88) 0.39 (0.12–1.2) < .001
Reference 0.97 (0.79–1.2) 1.04 (0.60–1.8) .86
Reference 0.89 (0.72–1.1) 0.91 (0.52–1.6) .31
Reference 0.74 (0.60–0.92) 0.52 (0.19–1.4) .002
Reference 0.71 (0.57–0.89) 0.48 (0.18–1.3) .001
Reference 0.70 (0.32–1.6) NA .20
Reference 0.72 (0.32–1.6) NA .27
Reference 1.4 (1.1–1.8) NA .08
Reference 1.5 (1.1–1.9) NA .04
Reference 0.98 (0.67–1.4) 1.5 (0.65–3.4) 0.68
Reference 1.1 (0.73–1.6) 1.95 (0.81–4.7) 0.31
Reference 0.96 (0.65–1.4) NA 0.44
Reference 0.90 (0.60–1.3) NA 0.33
Reference 1.9 (1.4–2.5) 1.4 (0.59–3.1) < .001
Reference 2.0 (1.5–2.8) 1.6 (0.66–3.7) < .001
Reference 1.9 (1.5–2.4) 0.66 (0.16–2.6) < .001
Reference 2.0 (1.5–2.5) 0.69 (0.17–2.8) < .001
Reference 1.8 (1.7–1.9) 1.9 (1.7–2.1) < .001
Reference 1.6 (1.57–1.71) 1.8 (1.6–2.0) < .001
Reference 1.6 (1.58–1.68) 2.0 (1.8–2.3) < .001
Reference 1.7 (1.62–1.74) 2.0 (1.8–2.3) < .001
Reference 2.2 (2.1–2.4) 3.1 (2.7–3.5) < .001
Reference 2.3 (2.1–2.4) 3.2 (2.8–3.6) < .001
Reference 2.4 (2.3–2.5) 2.9 (2.5–3.4) < .001
Reference 2.5 (2.4–2.6) 3.1 (2.6–3.6) < .001
Reference 1.9 (1.7–2.2) 2.6 (1.9–3.6) < .001
Reference 1.9 (1.6–2.2) 2.7 (1.9–3.7) < .001
Reference 1.8 (1.6–2.1) 1.6 (1.0–2.5) < .001
Reference 1.9 (1.7–2.2) 1.8 (1.1–2.9) < .001
Reference 2.8 (2.6–2.9) 3.9 (3.5–4.4) < .001
Reference 2.9 (2.7–3.0) 4.0 (3.5–4.5) < .001
Reference 3.2 (3.0–3.3) 4.3 (3.7–4.9) < .001
Reference 3.3 (3.1–3.4) 4.4 (3.8–5.1) < .001
Reference 2.3 (2.0–2.6)
Reference 2.3 (2.0–2.7)
Reference 2.2 (1.9–2.4)
Reference 2.3 (2.1–2.7)
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
7
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
TABLE 3 Continued. Fresh cycles Outcome MET P value for trend SGAd SET DET MET P value for trend LGAe SET DET MET P value for trend Early neonatal death SET DET MET P value for trend
Crude OR (95% CI)
Frozen cycles a
Adjusted OR (95% CI)
Crude OR (95% CI)
Adjusted OR (95% CI)b
3.7 (2.8–5.0) < .001
4.0 (3.0–5.4) < .001
2.1 (1.4–3.2) < .001
2.7 (1.8–4.0) < .001
Reference 1.6 (1.5–1.7) 2.2 (1.9–2.5) < .001
Reference 1.6 (1.5–1.7) 2.1 (1.8–2.5) < .001
Reference 1.8 (1.7–1.9) 2.3 (1.9–2.8) < .001
Reference 1.8 (1.6–1.9) 2.2 (1.8–2.7) < .001
Reference 0.77 (0.72–0.83) 0.61 (0.50–0.75) < .001
Reference 0.75 (0.70–0.81) 0.63 (0.50–0.78) < .001
Reference 0.71 (0.68–0.75) 0.71 (0.59–0.86) < .001
Reference 0.73 (0.69–0.77) 0.79 (0.65–0.96) < .001
Reference 1.4 (0.84–2.2) 4.5 (2.1–9.4) .003
Reference 1.2 (0.71–1.9) 3.8 (1.7–8.1) .03
Reference 1.04 (0.66–1.6) 0.72 (0.10–5.2) .99
Reference 1.1 (0.69–1.7) 0.87 (0.11–6.9) .77
Note: NA ¼ not available. a Adjusted for calendar year, maternal age, fertilization method, and embryo stage at transfer. b Adjusted for calendar year, maternal age, and embryo stage at transfer. c Gestational age at delivery <32 weeks. d SGA was defined as being below the 10th percentile of the national reference. e LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
singletons by two IVF treatments of SET showed better neonatal and maternal outcomes compared with those with two babies by one IVF treatment of DET. When it comes to perinatal outcome of singleton pregnancy from SET and DET, one study showed that singleton pregnancy from SET had a more positive outcome (13). However, some studies have stated that the outcomes between SET and DET were similar (14, 15). Notably, a study reported that vanishing twin pregnancies using DET had an adverse effect on perinatal outcome (14). In our study, DET was associated with a significantly higher risk for pPROM and placenta accreta in frozen cycles. Interestingly, DET was associated with a significantly decreased risk for PIH, GDM, and placenta previa. Those results may be due to DET and MET being performed in patients with fewer complications. In addition, our results demonstrated that maternal age and embryo stage at transfer played a role as effect modifiers for those outcomes. The risks of stillbirth, CS, PTB, LBW, SGA, and early neonatal death were significantly decreased after SET policy was launched. The rate of SGA showed a downward trend in both fresh and frozen ET, whereas the rate of LGA increased. Perinatal mortality also demonstrated a significantly decreasing trend in fresh cycles. These findings could be attributed to the decrease in multiple births. Indeed, these significant trends disappeared after restricting SET cycles. Interestingly, a significantly increasing trend was observed for PIH and GDM after adjusting for maternal age and restricting SET cycles, except for PIH in fresh cycles. These results suggest that the significant trend for those outcomes may not be due to an increase in maternal age or implementation of SET but rather are due to other unknown con-
founders. On the other hand, the rate of SGA by frozen ET was lower than that by fresh ET, while the rate of LGA by frozen ET was higher. Our results are in line with those from previous studies that reported that the birth weight of those born by frozen ET cycle was greater than that of those born by fresh ET and natural cycles (16, 17). Further, reports from Australia and New Zealand have shown that SET was associated with lower rates of fetal and neonatal mortality compared with DET (18, 19). We showed that the impact of SET policy is different in fresh and frozen cycles for several perinatal outcomes of CS, EPTB, LBW, VLBW, and SGA (Table 4). The pregnancy rate may be improved by performing frozen ET instead of fresh ET owing to the differences in hormonal environment and the receptivity of the endometrium (20). Indeed, frozen ET resulted in a higher pregnancy rate in Japan (7). Frozen ET is expected to be more widely adopted because it has been shown to result in a higher rate of pregnancy and positive perinatal outcomes compared with fresh ET (8, 21, 22). The rate of frozen ET has dramatically increased over the study period, accounting for over 70% of the cases treated with ART in 2012. Accordingly, SET with frozen ET has also increased (Table 2). Since our results demonstrate that the SET policy may have a different impact on perinatal outcomes in fresh and frozen cycles, DET or MET should be carefully practiced according to fresh/frozen status. Our results demonstrate that although a significant increase in SET use was observed from 52.2% in 2007 to 82.6% in 2012, singleton pregnancies gradually increased from 89.3% in 2007 to 96.0% in 2012, and the rate of multiple pregnancies (i.e., triplet) did not decrease so much (from 0.19% in 2007 to 0.05% in 2012) compared with the rate of VOL. - NO. - / - 2015
8
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:41 am ce S
Fertility and Sterility®
TABLE 4 Adjusted OR of effect of SET policy for selected pregnancy and birth outcomes stratified by fresh/frozen cycles.
Outcome
Fresh cycles Adjusted OR (95% CI)a
Multiple pregnancy Before policy Reference After policy 0.42 (0.39–0.45) Still birth Before policy Reference After policy 0.52 (0.38–0.69) pPROM Before policy Reference After policy 0.85 (0.60–1.2) CS Before policy Reference After policy 0.80 (0.77–0.84) PTB Before policy Reference After policy 0.71 (0.67–0.76) EPTBd Before policy Reference After policy 0.72 (0.61–0.85) LBW Before policy Reference After policy 0.65 (0.62–0.69) VLBW Before policy Reference After policy 0.69 (0.59–0.81) e SGA Before policy Reference After policy 0.78 (0.73–0.84) LGAf Before policy Reference After policy 1.1 (1.06–1.24) Early neonatal death Before policy Reference After policy 0.54 (0.34–0.88)
Frozen cycles Adjusted OR (95% CI)b
P value for interactionc
Reference 0.50 (0.46–0.54)
.02
Reference 0.39 (0.30–0.51)
.14
Reference 0.79 (0.57–1.1)
.68
Reference 0.88 (0.84–0.92)
.03
Reference 0.77 (0.72–0.82)
.19
Reference 1.00 (0.84–1.2)
.01
Reference 0.74 (0.69–0.78)
.03
Reference 1.03 (0.86–1.2)
.003
Reference 0.90 (0.83–0.98)
.03
Reference 1.2 (1.08–1.23)
.85
Reference 1.1 (0.55–2.3)
.12
a
Adjusted for maternal age, fertilization method, and embryo stage at transfer. Adjusted for maternal age and embryo stage at transfer. P value for interaction between before/after SET policy and fresh/frozen status in adjusted model including maternal age and embryo stage at transfer. d Gestational age at delivery <32 weeks. e SGA was defined as being below the 10th percentile of the national reference. f LGA was defined as being above the 90th percentile of the national reference. b c
Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
twin pregnancies (from 10.5% to 4.0%). A potential reason for the gradual decrease in multiple pregnancies may be that, when MET was conducted, low-grade embryos were more likely included for ET after 2007. Before SET policy was introduced by JSOG in 2008, the Japan Society for Reproductive Medicine recommended in 2007 that the first treatment cycle or high-grade blastocyst transfer in a patient under 35 years old should be SET and that the treatment cycle using especially high-grade blastocyst transfer in a patient under 40 years old should also be SET or DET. Thus, we speculate that when MET was conducted for women under 40 years old after 2007, low-grade embryos were more likely included for ET, which resulted in the gradual decrease in multiple pregnancies. Although our results demonstrated that perinatal outcomes after ART in Japan improved after introduction of SET policy, perinatal outcomes are still inferior to those of the non-IVF population. The rate of PTB of the whole pop-
ulation in Japan between 2007 and 2012 is steady at 5.7%– 5.8%, while that of LBW was 8.5% in boys and 10.8% in girls (23). In our study, the rate of PTB was 10.3%, and the rate of LBW was 17.8% in boys and 21.5% in girls in fresh cycles and 13.2% in boys and 16.3% in girls in frozen cycles, which was higher than those in all babies born in Japan. The main reason for these differences may be that the rate of multiple pregnancy was still higher in the ART population than in all populations in Japan. The rate of multiple pregnancy ranged from 10.7% in 2007 to 4.1% in 2012 in this study, while that of the general population was approximately 1.0%. In addition, advanced maternal age in ART may contribute to a worse outcome of ART in Japan (mean maternal age, 35 years in our sample vs. 31 years in the general population) (23). The strength of this study is that it included large sample sizes for each of the analyzed years, covering almost all cyclespecific ART data in Japan from 2007 to 2012. As the registration of data is mandatory, selection bias is not likely to have occurred. Nonetheless, there are some limitations. First, although the rate of SET significantly increased and DET and MET decreased after the introduction of the SET policy in 2008, other factors, such as improvements in embryo technique and perinatal care, may also affect perinatal outcomes. Indeed, the pregnancy rate slightly increased by approximately 1% from 2007 to 2012 in the group over 40 years old. Especially for the analysis shown in Table 4, we assumed that other factors except for the introduction of the SET policy would be similar before and after the SET policy. We conducted a sensitivity analysis further stratifying the number of embryos transferred in Table 4. Significant associations disappeared or were attenuated in many of perinatal outcomes such as multiple pregnancy, CS, PTB, LBW, VLBW, and LGA in fresh cycles and SGA, suggesting that the SET policy had a major impact on the perinatal outcomes through the implementation of SET. On the other hand, with respect to the perinatal outcomes of stillbirth and early neonatal death, ORs for the outcomes did not change (data not shown). For those outcomes, other factors may play a significant role. Second, multiple pregnancy was defined as two or more live births per one delivery in this study. If a different definition of multiple pregnancy was used, such as the number of gestational sacs and/or fetal heartbeats, another conclusion about perinatal outcomes may have been drawn. Notably, it was reported that vanishing twin during early pregnancy increases the risk for preterm delivery, LBW, and SGA (24– 26). To investigate this, we also performed a sensitivity analysis using different definitions of multiple pregnancy (i.e., based on the confirmation of fetal heartbeat or gestational sac). The results remained the same (data not shown). Hence, we can deduce that a different definition for multiple pregnancy will not have an impact on the current results. Third, the data did not include covariates such as patients' socioeconomic status, duration of infertility, number of repeated ART failures, and husband's age, which may affect results for residual confounding. Ideally, prospective studies including those covariates would be conducted to evaluate the impact of SET on perinatal outcomes.
VOL. - NO. - / - 2015
9
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
CONCLUSIONS
11.
On the basis of the data of the Japanese ART registry between 2007 and 2012, perinatal outcomes including PTB, LBW, SGA, perinatal mortality, and other pregnancy complications were significantly improved after SET policy was launched by JSOG in 2008. The results might be strongly associated with the decrease in the rate of preterm delivery and the increase in the average birth weight due to the decrease in multiple pregnancy, which could be attributed to the increased use of SET. The impact of SET policy on several perinatal outcomes was significantly different in fresh and frozen cycles, suggesting that DET or MET should be carefully conducted according to fresh/frozen status. Further study that includes additional covariates is necessary to evaluate the implementation of SET based on long-term accumulation of data for reviewing perinatal outcomes. Acknowledgments: The authors thank the JSOG for providing the data and all the participating clinics in the Japanese ART registry for their continuous support in data collection. We thank Dr. Julian Tang of the Department of Education for Clinical Research, National Center for Child Health and Development, for proofreading, editing, and rewriting part of this manuscript.
2. 3.
4. 5. 6. 7.
8.
9. 10.
13.
14.
15.
16.
17.
18.
19.
REFERENCES 1.
12.
Steptoe PC, Edwards RG. Birth after the reimplantation of a human embryo. Lancet 1978;2:366. Prevention of twin pregnancies after IVF/ICSI by single embryo transfer. ESHRE Campus Course Report. Hum Reprod 2001;16:790–800. € T, Jablonowska B, Pinborg A, Strandell A, et al. Thurin A, Hausken J, Hillensjo Elective single-embryo transfer versus double-embryo transfer in in vitro fertilization. N Engl J Med 2004;351:2392–402. Maheshwari A, Griffiths S, Bhattacharya S. Global variations in the uptake of single embryo transfer. Hum Reprod Update 2011;17:107–20. Kajiwara K. Prevention of multiple pregnancy. J Jpn Soc Obstet Gynecol 2009;61:N335–9 (in Japanese). Yoshimura Y. Futuristics of reproductive medicine—–for a baby that is expected to be born. Jpn Shindan Chiryo sya 2010:26–9 (in Japanese). Takeshima K, Saito H, Nakaza A, Kuwahara A, Ishihara O, Irahara M, et al. Efficacy, safety, and trends in assisted reproductive technology in Japan— analysis of four-year data from the national registry system. J Assist Reprod Genet 2014;31:477–84. Ishihara O, Araki R, Kuwahara A, Itakura A, Saito H, Adamson GD. Impact of frozen-thawed single-blastocyst transfer on maternal and neonatal outcome: an analysis of 277,042 single-embryo transfer cycles from 2008 to 2010 in Japan. Fertil Steril 2014;101:128–33. ART registry of Japan. (in Japanese) Available at: http://plaza.umin.ac.jp/ jsog-art/. Accessed September 4, 2015. Itabashi K, Fujimura M, Kusuda S, Tamura M, Hayashi T, Takahashi T, et al. New standard of average size and weight of newborn in Japan. Jap J Pediat 2010;114:1271–93 (in Japanese).
20.
21.
22.
23.
24.
25.
26.
Guzoglu N, Kanmaz HG, Dilli D, Uras N, Erdeve O, Dilmen U. The impact of the new Turkish regulation, imposing single embryo transfer after assisted reproduction technology, on neonatal intensive care unit utilization: a single center experience. Hum Reprod 2012;27:2384–8. Sazonova A, K€allen K, Thurin-Kjellberg A, Wennerholm UB, Bergh C. Neonatal and maternal outcomes comparing women undergoing two in vitro fertilization (IVF) singleton pregnancies and women undergoing one IVF twin pregnancy. Fertil Steril 2013;99:731–7. De Sutter P, Delbaere I, Gerris J, Verstraelen H, Goetgeluk S, Van der Elst J, et al. Birthweight of singletons after assisted reproduction is higher after single- than after double-embryo transfer. Hum Reprod 2006;21: 2633–7. Sazonova A, K€allen K, Thurin-Kjellberg A, Wennerholm UB, Bergh C. Factors affecting obstetric outcome of singletons born after IVF. Hum Reprod 2011; 26:2878–86. Poikkeus P, Gissler M, Unkila-Kallio L, Hyden-Granskog C, Tiitinen A. Obstetric and neonatal outcome after single embryo transfer. Hum Reprod 2007; 22:1073–9. Nakashima A, Araki R, Tani H, Ishihara O, Kuwahara A, Irahara M, et al. Implications of assisted reproductive technologies on term singleton birth weight: an analysis of 25,777 children in the national assisted reproduction registry of Japan. Fertil Steril 2013;99:450–5. Pinborg A, Henningsen AA, Loft A, Malchau SS, Forman J, Andersen AN. Large baby syndrome in singletons born after frozen embryo transfer (FET): is it due to maternal factors or the cryotechnique? Hum Reprod 2014;29:618–27. Sullivan EA, Wang YA, Hayward I, Chambers GM, Illingnorth P, McBain J, et al. Single embryo transfer reduces the risk of perinatal mortality, a population study. Hum Reprod 2012;27:3609–15. Wang YA, Sullivan EA, Healy DL, Black DA. Perinatal outcomes after assisted reproductive technology treatment in Australia and New Zealand: single versus double embryo transfer. Med J Aust 2009;190:234–7. Roque M, Lattes K, Serra S, Sola I, Geber S, Carreras R, et al. Fresh embryo transfer versus frozen embryo transfer in in vitro fertilization cycles: a systematic review and meta-analysis. Fertil Steril 2013;99:156–62. 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. Maheshwari A, Pandey S, Shetty A, Hamilton M, Bhattacharya S. Obstetric and perinatal outcomes in singleton pregnancies resulting from the transfer of frozen thawed versus fresh embryos generated through in vitro fertilization treatment: a systematic review and meta-analysis. Fertil Steril 2012;98: 368–77. Statistics Bureau, Ministry of Internal Affairs and Communications with the collaboration of Ministries and Agencies. Portal Site of Official Statistics of Japan. Demographic statistics. (in Japanese) Available at: http://www.e-stat.go. jp/SG1/estat/List.do?lid¼000001127058. Accessed July 6, 2015. Pinborg A, Lidegaard Ø, La Cour Freiesleben N, Andersen AN. Vanishing twins: a predictor of small-for-gestational age in IVF singletons. Hum Reprod 2007;22:2707–14. Shebl O, Ebner T, Sommergruber M, Sir A, Tews G. Birth weight is lower for survivors of the vanishing twin syndrome: a case-control study. Fertil Steril 2008;90:310–4. Luke B, Brown MB, Grainger D, Stern JE, Klein N, Cedars MI. The effect of early fetal losses on singleton assisted-conception pregnancy outcomes. Fertil Steril 2009;91:2578–85.
VOL. - NO. - / - 2015
10
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
VOL. - NO. - / - 2015
SUPPLEMENTAL TABLE 1 Pregnancy rate stratified by age from 2007 to 2012 in the Japanese Assisted Reproduction Technology registry.a
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
Age, y
2007
2008
2009
2010
2011
2012
Total
<30 30–34 35–39 R40
2,769/10,077 (27.5) 10,901/41,606 (26.2) 11,929/59,731 (20.0) 3,565/50,578 (7.0)
2,719/10,818 (25.1) 11,337/46,319 (24.5) 14,042/72,376 (19.4) 4,413/61,106 (7.2)
3,008/11,314 (26.6) 12,287/49,257 (24.9) 16,467/81,805 (20.1) 5,666/71,417 (7.9)
3,046/11,408 (26.7) 13,242/51,496 (25.7) 18,507/92,721 (20.0) 6,842/86,535 (7.9)
3,270/12,018 (27.2) 13,635/53,476 (25.4) 20,660/101,812 (20.2) 8,098/102,353 (7.9)
3,590/13,338 (26.9) 15,414/60,400 (25.5) 24,305/123,008 (19.8) 10,447/129,680 (8.1)
18,402/68,973 (26.7) 76,816/302,554 (26.7) 105,910/531,453 (19.9) 39,031/501,669 (7.8)
Note: Data in parentheses are percents. a The data originated from official annual reports by JSOG. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
Fertility and Sterility®
10.e1
ORIGINAL ARTICLE: ASSISTED REPRODUCTION
SUPPLEMENTAL TABLE 2 Adjusted ORs of number of ET for selected perinatal outcomes stratified by fresh/frozen cycles and embryo stage at transfer. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend PIH SET DET MET P value for trend Placenta previa SET DET MET P value for trend Placenta accreta SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend LBW SET DET MET P value for trend SGAc SET DET MET P value for trend LGAd SET DET MET P value for trend
Frozen cycles
Early cleavage Adjusted OR (95% CI)a
Blastocyst Adjusted OR (95% CI)a
Early cleavage Adjusted OR (95% CI)b
Blastocyst Adjusted OR (95% CI)b
Reference 22.9 (19.6–26.8) 32.0 (25.9–39.4) < .001
Reference 15.5 (13.2–18.3) 17.4 (11.9–25.6) < .001
Reference 14.5 (11.8–17.7) 21.6 (16.3–28.7) < .001
Reference 17.5 (16.2–19.0) 21.9 (17.2–27.9) < .001
Reference 0.69 (0.58–0.82) 0.49 (0.29–0.82) < .001
Reference 1.1 (0.82–1.5) 1.3 (0.48–3.7) .46
Reference 1.0 (0.85–1.2) 0.54 (0.28–1.04) .45
Reference 1.1 (1.0–1.3) 1.1 (0.66–1.9) .03
Reference 0.90 (0.70–1.1) 0.78 (0.41–1.5) .27
Reference 0.87 (0.55–1.4) 2.2 (0.66–7.0) .99
Reference 1.1 (0.75–1.6) 0.75 (0.23–2.5) .86
Reference 0.56 (0.41–0.75) 0.30 (0.04–2.2) < .001
Reference 0.34 (0.098–1.2) NA NA
Reference 1.6 (0.55–4.9) NA NA
Reference 1.3 (0.75–2.1) NA NA
Reference 1.6 (1.2–2.1) NA NA
Reference 1.6 (1.5–1.7) 1.8 (1.6–2.0) < .001
Reference 1.7 (1.6–1.9) 1.7 (1.3–2.3) < .001
Reference 1.5 (1.4–1.6) 1.9 (1.6–2.3) < .001
Reference 1.8 (1.69–1.84) 2.1 (1.7–2.6) < .001
Reference 2.2 (2.0–2.4) 3.2 (2.7–3.7) < .001
Reference 2.5 (2.2–2.7) 2.7 (1.9–3.9) < .001
Reference 2.1 (1.9–2.3) 2.6 (2.1–3.3) < .001
Reference 2.6 (2.5–2.8) 3.1 (2.4–4.0) < .001
Reference 2.7 (2.6–2.9) 3.9 (3.4–4.4) < .001
Reference 3.1 (2.8–3.4) 3.8 (2.8–5.2) < .001
Reference 2.6 (2.3–2.8) 3.7 (3.0–4.6) < .001
Reference 3.5 (3.3–3.7) 4.7 (3.7–5.8) < .001
Reference 1.6 (1.5–1.7) 2.2 (1.8–2.5) < .001
Reference 1.5 (1.4–1.7) 1.9 (1.3–2.8) < .001
Reference 1.3 (1.2–1.5) 1.6 (1.2–2.2) < .001
Reference 1.9 (1.8–2.1) 2.9 (2.2–3.8) < .001
Reference 0.76 (0.69–0.83) 0.61 (0.48–0.77) < .001
Reference 0.74 (0.65–0.85) 0.74 (0.45–1.2) < .001
Reference 0.80 (0.72–0.88) 0.82 (0.63–1.1) < .001
Reference 0.70 (0.66–0.75) 0.72 (0.53–0.98) < .001
Note: NA ¼ not available. a Adjusted for calendar year, maternal age, and fertilization method. b Adjusted for calendar year and maternal age. c SGA was defined as being below the 10th percentile of the national reference. d LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
10.e2
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S
Fertility and Sterility®
SUPPLEMENTAL TABLE 3 Adjusted ORs of number of embryo transfer for selected perinatal outcomes stratified by fresh/frozen cycles and age. Fresh cycles Outcome Pregnancy complications Multiple pregnancy SET DET MET P value for trend PIH SET DET MET P value for trend Delivery outcomes CS SET DET MET P value for trend PTB SET DET MET P value for trend LBW SET DET MET P value for trend LGAc SET DET MET P value for trend Early neonatal death SET DET MET P value for trend
Frozen cycles
Age <35 Adjusted OR (95% CI)a
Age‡35 Adjusted OR (95% CI)a
Age <35 Adjusted OR (95% CI)b
Age‡35 Adjusted OR (95% CI)b
Reference 24.4 (20.8–28.6) 32.3 (25.4–41.2) < .001
Reference 14.0 (12.0–16.3) 19.1 (15.0–24.3) < .001
Reference 21.2 (19.0–23.6) 32.5 (25.5–41.4) < .001
Reference 13.7 (12.4–15.1) 16.6 (13.1–21.1) < .001
Reference 1.03 (0.79–1.3) 0.83 (0.40–1.7) .93
Reference 0.71 (0.53–0.84) 0.51 (0.28–0.93) < .001
Reference 1.2 (1.03–1.4) 1.1 (0.59–2.0) .03
Reference 1.1 (0.94–1.2) 0.63 (0.36–1.1) .9
Reference 2.1 (1.9–2.2) 2.3 (1.9–2.7) < .001
Reference 1.5 (1.4–1.6) 1.6 (1.4–1.8) < .001
Reference 1.9 (1.8–2.1) 2.4 (1.9–2.9) < .001
Reference 1.53 (1.46–1.6) 1.9 (1.6–2.2) < .001
Reference 2.9 (2.6–3.2) 3.9 (3.2–4.9) < .001
Reference 1.9 (1.8–2.1) 2.6 (2.2–3.2) < .001
Reference 3.2 (2.9–3.4) 4.2 (3.3–5.3) < .001
Reference 2.1 (2.0–2.3) 2.5 (2.0–3.1) < .001
Reference 3.7 (3.4–4.0) 5.0 (4.2–6.0) < .001
Reference 2.3 (2.2–2.5) 3.3 (2.8–3.8) < .001
Reference 4.1 (3.8–4.3) 6.2 (5.0–7.6) < .001
Reference 2.8 (2.6–2.9) 3.4 (2.7–4.1) < .001
Reference 0.66 (0.59–0.76) 0.51 (0.36–0.72) < .001
Reference 0.81 (0.74–0.88) 0.72 (0.55–0.94) < .001
Reference 0.63 (0.57–0.69) 0.66 (0.48–0.91) < .001
Reference 0.80 (0.75–0.85) 0.88 (0.69–1.1) < .001
Reference 2.4 (0.93–6.1) 14.5 (5.5–40.3) < .001
Reference 0.84 (0.48–1.5) 0.94 (0.21–4.3) .62
Reference 1.4 (0.72–2.6) NA NA
Reference 0.86 (0.44–1.7) 1.5 (0.17–1.7) .83
Note: NA ¼ not available. a Adjusted for calendar year, embryo stage at transfer, and fertilization method. b Adjusted for calendar year and embryo stage at transfer. c LGA was defined as being above the 90th percentile of the national reference. Takeshima. SET policy improves perinatal outcomes. Fertil Steril 2015.
VOL. - NO. - / - 2015
10.e3
FLA 5.4.0 DTD FNS29913_proof 4 November 2015 5:42 am ce S