Skip Navigation


American Journal of Epidemiology Advance Access originally published online on March 10, 2007
American Journal of Epidemiology 2007 165(12):1380-1388; doi:10.1093/aje/kwm035
This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
165/12/1380    most recent
kwm035v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Farr, S. L.
Right arrow Articles by Jamieson, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Farr, S. L.
Right arrow Articles by Jamieson, D. J.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?

American Journal of Epidemiology Copyright © 2007 by the Johns Hopkins Bloomberg School of Public Health All rights reserved; printed in U.S.A.

ORIGINAL CONTRIBUTIONS

Pregnancy Loss among Pregnancies Conceived through Assisted Reproductive Technology, United States, 1999–2002

Sherry L. Farr1, Laura A. Schieve2 and Denise J. Jamieson1

1 Division of Reproductive Health, National Center for Chronic Disease Prevention and Health Promotion, US Centers for Disease Control and Prevention, Atlanta, GA
2 National Center on Birth Defects and Developmental Disabilities, US Centers for Disease Control and Prevention, Atlanta, GA

Correspondence to Dr. Sherry L. Farr, Division of Reproductive Health, National Center for Chronic Disease Prevention and Health Promotion, Centers for Disease Control and Prevention, 4770 Buford Highway, Mailstop K-34, Atlanta, GA 30341 (e-mail: SFarr{at}cdc.gov).

Received for publication July 14, 2006. Accepted for publication November 30, 2006.


   ABSTRACT
TOP
 ABSTRACT
INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Approximately 30% of pregnancies in the United States may end in miscarriage or stillbirth. Whether pregnancies conceived through assisted reproductive technology (ART) are at an increased risk of loss is inconclusive, and data on maternal age-, ART type-, and gestational age-specific risk of loss are limited. Data on 148,494 ART pregnancies conceived from 1999 through 2002 were analyzed by use of the Kaplan-Meier method to estimate risks of pregnancy loss after specified gestational ages (conditional risk) for 14 groups stratified by maternal age and ART procedure. Births, maternal deaths, and induced abortions were censored. The Kaplan-Meier estimate of total risk of pregnancy loss was 29% but ranged from 22% to 63% depending on patient age and ART procedure. By 6 weeks' gestation, 58% of all pregnancy losses occurred. Conditional risk of pregnancy loss ranged from 10% to 45% at 6 weeks' gestation and from 2% to 7% at the first trimester; it was less than 2% after 20 weeks' gestation. Results can be used to counsel ART patients and inform future research on the etiology of pregnancy loss.

abortion, spontaneous; fetal death; maternal age; reproductive techniques, assisted; stillbirth

Abbreviations: ART, assisted reproductive technology; CDC, US Centers for Disease Control and Prevention; ICSI, intracytoplasmic sperm injection; SART, Society for Assisted Reproductive Technology


   INTRODUCTION
TOP
 ABSTRACT
 INTRODUCTION
MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 
Pregnancy loss is estimated to occur in up to 31 percent of naturally conceived pregnancies detected by elevated beta human chorionic gonadotropin levels early in pregnancy (1). Whether pregnancies conceived through assisted reproductive technology (ART) are at an increased risk of loss compared with naturally conceived pregnancies (2–5) is inconclusive. Examining the associations among maternal age, type of ART procedure, and pregnancy loss at different gestational ages is important in understanding its etiology and in counseling pregnant women about their risk of pregnancy loss. Limited data exist on ART subgroup-specific risks of loss throughout pregnancy.

Pregnancies conceived through ART are closely monitored from gamete retrieval and fertilization to pregnancy outcome. A woman is seen by her ART provider after pregnancy is established, before beginning prenatal care with an obstetric provider. The ART provider directly collects data on fertilization, embryo transfer, and early ultrasounds and actively follows up with patients or their obstetric providers to obtain the pregnancy outcome and date. This information is routinely reported as part of the US Centers for Disease Control and Prevention (CDC) ART surveillance system. The purpose of this report is to estimate risk of pregnancy loss at different gestational ages for ART pregnancies, by ART procedure and maternal age, and to discuss the findings' relevance to ART and naturally conceived pregnancies.


   MATERIALS AND METHODS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
RESULTS
 DISCUSSION
 References
 
Annually, medical practices in the United States are required to report all attempted ART procedures to the CDC (6). ART is defined as any procedure that requires both oocytes and sperm be handled outside of the body and includes in vitro fertilization with transcervical embryo transfer, gamete intrafallopian transfer, zygote intrafallopian transfer, frozen embryo transfer, and donor embryo transfer. Data include patient demographics and reproductive history, type of ART procedure, and, if a pregnancy resulted, the maximum number of fetal heartbeats in early pregnancy, the pregnancy outcome (livebirth, spontaneous or therapeutic abortion, stillbirth, maternal death, unknown outcome), and the date of outcome.

More than 95 percent of procedures performed from 1999 to 2002 were reported to the CDC (7), and medical record review shows discrepancy rates consistently under 1 percent for outcome of treatment (pregnant vs. not pregnant) and pregnancy outcome.

From 1999 to 2002, there were 410,244 ART procedures in the United States reported to the CDC. For this analysis, we excluded treatments canceled prior to egg retrieval (n = 50,454), treatments with unsuccessful embryo transfers (n = 20,276), and treatments using zygote intrafallopian transfer (n = 2,727), gamete intrafallopian transfer (n = 1,717), or zygote or gamete intrafallopian transfer in combination with in vitro fertilization with transcervical embryo transfer (n = 461). There were 334,609 ART procedures with a successful transcervical embryo transfer. We further excluded observations indicating use of both donor and patient oocytes or embryos (n = 415), both freshly fertilized and frozen embryos (n = 601), a gestational carrier (n = 4,218), or those missing data on whether the treatment resulted in pregnancy (n = 225). Among the remaining 329,150 ART treatment procedures, 151,067 (46 percent) resulted in a biochemical pregnancy (elevated beta human chorionic gonadotropin level without visualization of a gestational sac on ultrasound) or clinical pregnancy (visualization of a gestational sac on ultrasound).

We excluded 1,521 (1.0 percent) pregnancies with missing or conflicting values for dates of oocyte retrieval, embryo transfer, ultrasound observation of fetal heartbeat, or pregnancy outcome and 1,052 (0.7 percent) pregnancies with missing data on potential confounders. The final data set for this analysis included 148,494 ART pregnancies.

Pregnancies were subdivided into 14 analytical groups on the basis of three criteria: 1) whether patient or donor oocytes were used; 2) whether embryos were freshly fertilized or frozen and thawed at the time of transfer; and 3) for women who used their own oocytes, the patient's age at the time of the ART procedure (<33, 33–34, 35–37, 38–40, 41–42, and >42 years). The CDC does not require clinics to report whether fetal losses in multigestational pregnancies were due to therapeutic reduction or spontaneous loss. Therefore, pregnancy loss was defined as loss of the entire pregnancy by either spontaneous abortion or stillbirth. For example, if three fetal hearts were detected on ultrasound, all three fetuses would have to be lost for the pregnancy to be recorded as a pregnancy loss. This definition of pregnancy loss reflects the probability that a woman will deliver at least one livebirth. Additionally, data were unavailable to link pregnancies from the same woman.

Using the Kaplan-Meier method, we computed estimates of probability of pregnancy loss at specific gestational weeks for the 14 analytical groups, conditional on the pregnancy's continuing to the beginning of the specified time. For instance, the probability of pregnancy loss at 6 weeks' gestation is the probability of pregnancy loss after 6 weeks' gestation among pregnancies with at least one viable fetus at 6 weeks' gestation. The time to event (gestational age at pregnancy loss) was calculated as the date of pregnancy loss minus the date of theoretical last menstrual period (14 days prior to the date of retrieval for procedures using freshly fertilized embryos and 17 days prior to the date of transfer for procedures using thawed embryos). Only 1 percent of pregnancies continued past 40 weeks' gestation. To increase the precision of estimates at gestational weeks 41–44, any pregnancy lasting beyond 286 days' gestation (40 weeks) was given a survival time of 287 days' (41 weeks') gestation.

Livebirths and medically induced abortions are competing risks for pregnancy loss. However, 99 percent of pregnancy losses occurred by 23 weeks' gestation. The vast majority (98 percent) of livebirths occurred between 26 and 42 weeks' gestation. All medically induced abortions occurred before 22 weeks' gestation; however, this outcome constituted only 2.6 percent of all pregnancy losses. Because of the timing and number of these events, a competing risks analysis would give very similar results to treating these events as censored. Therefore, livebirths, medically induced abortions, maternal deaths during pregnancy, and pregnancies lost to follow-up were censored at the date of the occurrence or date of last known pregnancy status. For relative risk estimates, overall survival estimates from the Kaplan-Meier procedure for each of the 13 ART procedure groups were compared with the referent, pregnancies among women aged less than 33 years who used their own oocytes and freshly fertilized embryos. We computed 95 percent confidence intervals for relative risks by applying the delta method to the logarithm of the relative risk and using standard errors of the Kaplan-Meier estimates of risk (8).

It is difficult to assign a specific date of loss to biochemical pregnancies, because they represent a transient elevation in beta human chorionic gonadotropin without ultrasound confirmation of pregnancy. However, by 6 weeks' gestation, a clinical pregnancy should be detected on ultrasound; therefore, biochemical pregnancies with no reported date of loss (99 percent) were assumed to have ended in pregnancy loss at 6 weeks' gestation.

We also examined the risk of pregnancy loss and relative risks stratified by plurality of the pregnancy, defined as the maximum number of fetal hearts seen on ultrasound (one or ≥2). When the number of livebirths was greater than the maximum number of fetal hearts reported (n = 264), we corrected the number of fetal hearts to equal the number of infants born. Over 95 percent of clinical pregnancies reported an ultrasound procedure performed before 7 weeks' gestation. To avoid underestimating survival after visualization of a fetal heart, we calculated the risk of pregnancy loss after 7 weeks' gestation. In analyses among pregnancies with two or more fetal hearts, women aged 41–42 years and over 42 years using frozen embryos were combined because of low numbers.

We assessed potential confounders of the relation between the ART procedure and pregnancy loss (previous livebirths, year of procedure, previous ART procedures, use of intracytoplasmic sperm injection (ICSI), infertility diagnosis) by comparing within analytical groups the relative differences between unstandardized Kaplan-Meier estimates and Kaplan-Meier estimates standardized on each of these variables individually (9). Relative differences in Kaplan-Meier estimates did not exceed 2.5 percent for parity and 0.9 percent for year of procedure. Among freshly fertilized cycles, Kaplan-Meier estimates standardized for the previous number of ART procedures did not exceed 0.6 percent, and those standardized for ICSI did not exceed 1.1 percent. Kaplan-Meier estimates standardized for infertility diagnosis were 6.1 percent lower among pregnancies to women aged over 42 years using thawed embryos; however, the numbers were sparse, and the relative differences in all the other analytical groups did not exceed 1.3 percent. Therefore, these variables did not confound the association between analytical group and pregnancy loss, and we present unstandardized results. In separate analyses, we compared survival estimates stratified by number of previous ART cycles (zero vs. ≥1) and by use of ICSI.


   RESULTS
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
DISCUSSION
 References
 
Table 1 shows the distributions of maternal demographics and reproductive variables for the 148,494 ART pregnancies in this study. Maternal age was less than 33 years for a third of the pregnancies. Of the 85,481 (58 percent) pregnancies with data on race/ethnicity, 85 percent reported non-Hispanic White, and 15 percent reported Asian, Hispanic, African American, or another race/ethnicity. For nearly half (49 percent) of the pregnancies, this was the woman's first ART procedure.


View this table:
[in this window]
[in a new window]

  		TABLE 1. Distribution of women's demographic and reproductive health factors among 148,494 pregnancies conceived through assisted reproductive technology, United States, 1999–2002

 
Seventy-five percent of ART pregnancies used freshly fertilized embryos from the patient's oocytes, 11 percent used thawed embryos from the patient's oocytes, 11 percent used freshly fertilized embryos from donor oocytes, and 3 percent used thawed embryos from donor oocytes. Seventy percent of all pregnancies resulted in a livebirth, and 28 percent of procedures resulted in pregnancy loss.

The total risk of pregnancy loss for all pregnancies was 29 percent (table 2). The risk of pregnancy loss ranged from 22 percent among pregnancies to women under 33 years using their own oocytes and freshly fertilized embryos to 63 percent among pregnancies to women older than 42 years using their own oocytes and freshly fertilized embryos. Among pregnancies using patient oocytes, the risk of pregnancy loss increased with increasing maternal age. Patient age at the time of the procedure was not associated with risk of loss among pregnancies using donor oocytes (data not shown). For pregnancies among women under 41 years of age and those using donor oocytes, freshly fertilized embryos had a lower risk of pregnancy loss than thawed embryos. However, among pregnancies to women over 41 years of age, lower risk of loss was seen for thawed compared with freshly fertilized embryos (41–42 years: 43 vs. 51 percent; >42 years: 51 vs. 63 percent, thawed vs. freshly fertilized, respectively). For thawed embryo procedures, the age of the women when oocyte retrieval occurred was unavailable; the analysis is based on the age at the time of embryo transfer.


View this table:
[in this window]
[in a new window]

  		TABLE 2. Risk of subsequent pregnancy loss, relative risks, and 95% confidence intervals by oocyte source, embryo type, and woman's age among 148,494 pregnancies conceived through assisted reproductive technology, United States, 1999–2002

 
Compared with pregnancies to women aged under 33 years using patient oocytes and freshly fertilized embryos, the relative risks ranged from 1.1 to 2.9, depending on the patient's age and ART procedure (table 2). Conditional risk of pregnancy loss decreased rapidly during the first trimester, from a high of 45 percent at 6 weeks among women aged more than 42 years using fresh embryos to between 2 and 7 percent for all groups at 12 weeks. By 20 weeks' gestation, the risk of subsequent loss was below 2 percent for all groups.

Figures 1, 2, and 3 show the probability of livebirth conditional on the pregnancy's achieving a specified time (conditional survival probabilities) from gestational weeks 6 through 24 for the 14 analytical groups. Pregnancies using patient oocytes and freshly fertilized embryos had a greater range of conditional survival probabilities at 6 weeks' gestation (55–90 percent) compared with pregnancies using patient oocytes and thawed patient embryos (68–87 percent). However, for all groups, survival probabilities are above 95 percent at 16 weeks' gestation and above 98 percent by 20 weeks' gestation.


Figure 1
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 1. Probability of a livebirth through week 41 of gestation by gestational week and maternal age among 111,407 pregnancies conceived through assisted reproductive technology by use of freshly fertilized embryos and patient oocytes, United States, 1999–2002.

 

Figure 2
View larger version (10K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 2. Probability of a livebirth by gestational week and maternal age among 16,758 pregnancies conceived through assisted reproductive technology by use of frozen and thawed embryos and patient oocytes, United States, 1999–2002.

 

Figure 3
View larger version (8K):
[in this window]
[in a new window]
[Download PowerPoint slide]
  		FIGURE 3. Probability of a livebirth by gestational week and embryo type among 20,329 pregnancies conceived through assisted reproductive technology by use of donor oocytes, United States, 1999–2002.

 
Among pregnancies using freshly fertilized embryos, overall survival estimates for women who had never had a previous ART treatment were slightly higher compared with pregnancies to women with one or more previous ART treatments. However, the absolute differences in survival estimates did not exceed 3.3 (Kaplan-Meier estimates = 50.6 percent vs. 47.3 percent, no previous ART treatment vs. one or more previous ART treatments, respectively, for women aged 41–42 years). Pregnancies that used ICSI had slightly lower overall survival probabilities than did pregnancies that did not use ICSI. The greatest absolute difference in survival estimates between procedures using ICSI and those not using ICSI was 4.6 (Kaplan-Meier estimates = 34.1 vs. 38.7, ICSI vs. no ICSI, respectively, for women aged more than 42 years); differences were less than 2.9 for all other analytical groups. A sensitivity analysis treating medically induced abortions as pregnancy losses lowered weekly Kaplan-Meier survival probabilities by less than 0.02 for all analytical groups.

Table 3 presents the risk of pregnancy loss stratified by pregnancy plurality for the 117,730 (79 percent) pregnancies that survived to ultrasound visualization of at least one fetal heartbeat and 7 weeks' gestation. The cumulative risk of pregnancy loss among pregnancies with one fetal heartbeat was 14 percent; it was 5.0 percent for pregnancies with two or more fetal heartbeats. For pregnancies with one fetal heartbeat, the patterns of risk by procedure and maternal age were similar to the overall patterns presented in table 2 and ranged from 10 percent to 37 percent. Among pregnancies with at least two fetal heartbeats, pregnancy loss after 7 weeks' gestation ranged from 4 percent to 14 percent. Among pregnancies with at least two fetal heartbeats, using patient oocytes and thawed embryos, there was no apparent association with maternal age. However, for some analytical groups, the numbers are low and the estimates are imprecise. Approximately 4 percent of pregnancies with two or more fetal heartbeats on ultrasound visualization using donor oocytes or embryos resulted in pregnancy loss.


View this table:
[in this window]
[in a new window]

  		TABLE 3. Risk of pregnancy loss after 7 weeks' gestation, relative risks, and 95% confidence intervals by pregnancy plurality,* oocyte source, embryo type, and woman's age among pregnancies conceived through assisted reproductive technology, United States, 1999–2002

 

   DISCUSSION
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
References
 
This large, comprehensive study on risk of pregnancy loss throughout pregnancy was based on data from 148,494 ART pregnancies, the vast majority of pregnancies conceived through ART in the United States between 1999 and 2002. These findings suggest that the age of the oocyte plays a large role in determining pregnancy success early in pregnancy. Among pregnancies using patient oocytes and freshly fertilized embryos, women over the age of 42 years had approximately three times the risk of pregnancy loss compared with women aged under 33 years. Among these same analytical groups, 90 percent of viable pregnancies at 7 weeks' gestation in women under 33 years of age progressed to a livebirth, while pregnancies among women over 42 years of age do not reach a 90 percent conditional survival probability until 11 weeks' gestation.

Previously, Schieve et al. (3) studied spontaneous abortion in the same surveillance system. The current analysis updated and expanded the analysis of Schieve et al. in several ways: 1) Data were from 1999 to 2002 (previous data were from 1996 to 1998); 2) biochemical pregnancies were included; 3) spontaneous abortions and stillbirths were assessed rather than spontaneous abortion only; and 4) conditional probability of pregnancy loss throughout pregnancy was examined, rather than cumulative risk only. Results in the current study support those of the previous study, with a greater risk of complete pregnancy loss for older women, pregnancies using thawed embryos, and singleton versus multigestational pregnancies.

The overall risk of pregnancy loss in this study was 29 percent, higher than other studies on ART pregnancies that found risks of clinical pregnancy loss between 15 percent and 22 percent (2, 3, 5). However, biochemical pregnancies in this analysis accounted for 54 percent of pregnancy losses. The Kaplan-Meier risk of pregnancy loss among clinical pregnancies in our data set was 16 percent. The probability of early pregnancy loss, defined as loss of a pregnancy by 7 weeks' gestation, was 18.6 percent in our data, similar to the 19 percent risk in a separate cohort of in vitro fertilization pregnancies (10).

The presence of a fetal heartbeat on ultrasound visualization is sometimes used as an indicator of low risk of subsequent pregnancy loss (11). However, estimates of risk of pregnancy loss after confirmation of a fetal heartbeat vary from 3 percent to 12 percent for couples with and without fertility issues (11–15). In this analysis, the probability of loss was 11 percent overall and 14 percent for singleton gestational pregnancies surviving to 7 weeks' gestation. These results are similar to those of Spandorfer et al. (15), who reported risk of loss at 12 percent overall and 16 percent for singleton pregnancies in an in vitro fertilization with a transcervical embryo transfer population. However, these estimates are slightly higher than those among naturally conceived pregnancies. This may be due to inherent differences among ART pregnancies, differences in maternal demographics of the study populations, or timing of ultrasound examinations. A higher median age of study participants (35 years) and a lower gestational age at detection of fetal heartbeat (95 percent by 7 weeks' gestation) may account for the higher risk of pregnancy loss after detection of a fetal heartbeat in this study compared with studies examining naturally conceived pregnancies. However, these results show that no single embryonic or fetal milestone is an indicator of a positive prognosis for women of all ages and types of ART procedures. In addition to the duration of the pregnancy and stage of the fetus, physicians and health-care providers should take into account the individual characteristics of the woman when evaluating and informing patients about their risk of pregnancy loss.

The average age among women in our data set was 35 years, the majority with information on race and ethnicity were White, and all couples had previous infertility issues. Generalizing results from this population to naturally conceived pregnancies is difficult. Wilcox et al. (1) found a 31 percent risk of spontaneous abortion (both unrecognized and recognized) among women with elevated beta human chorionic gonadotropin levels, as compared with a 29 percent overall risk in this study. However, comparison of the two studies may be limited by differences in beta human chorionic gonadotropin testing. Wilcox et al. used a highly sensitive assay with a lower level of detectability than that typically found in clinical assays; data from the ART surveillance system are based on various commercially available assays performed as part of clinical care. Analyses by Wang et al. (5) showed the risk of spontaneous abortion among pregnancies lasting at least 6 weeks to be 16 percent and 14 percent in two cohorts of women with naturally conceived pregnancies. The conditional risk of pregnancy loss at 6 weeks' gestation among women under 38 years of age using their own oocytes and freshly fertilized embryos ranged from 10 percent to 14 percent in this analysis, similar to a 12 percent risk of subsequent loss at 6 weeks' gestation found in a health-care cohort of clinical pregnancies (16). The majority of women in these studies were White and college educated. However, the similarity between results suggests that the current analysis may be informative for naturally conceived pregnancies among women of similar age, race/ethnicity, and socioeconomic status.

ART pregnancies are monitored very closely in early pregnancy with frequent ultrasound examinations. It is likely that many nonviable pregnancies (i.e., gestational sac without a fetus or fetus without a heartbeat) would be detected earlier than would be clinically apparent through spontaneous vaginal bleeding leading to an ultrasound confirmation of loss. Earlier detection of a nonviable pregnancy may mean more accurate and earlier reporting of date of loss in ART pregnancies compared with women with naturally conceived pregnancies that are not monitored as closely.

This analysis must be considered in the context of several limitations. An exact date of loss is often not available for biochemical pregnancies. However, by 6 weeks' gestation, a clinical pregnancy should be visible on ultrasound examination. Therefore, biochemical pregnancies without a date of outcome were given a failure time of 6 weeks' gestation, and this is the limit of our gestational age-specific assessment potential. Even with this limitation, this analysis provides estimates of risk of biochemical pregnancy and cumulative risk of pregnancy loss from fertilization, which are more informative than risk estimates including only clinical pregnancies.

Because more than 50 percent of the sample had undergone a previous ART procedure, more than 30 percent had experienced a spontaneous abortion in the past, and the data set covered 4 years, it is probable that some of the pregnancies in this data set are from the same woman. Data are not available to link such pregnancies; however, when analyses were stratified by the number of previous ART treatments, the absolute differences in overall survival probabilities did not exceed 3.3 for any analytical group. Additionally, the age of the patient when oocytes were retrieved for procedures using thawed embryos is not collected by CDC. Our findings are based on patient age at the time of embryo transfer, which may partially account for the decreased risk of pregnancy loss with use of thawed embryos among women over 40 years of age.

Over 23 percent of multigestational pregnancies that progressed to a livebirth delivery experienced loss of at least one fetus during the pregnancy. However, the ART surveillance system does not include data on whether these fetal losses were medically induced or spontaneous and, therefore, we cannot speculate on risk of fetal loss for multigestational pregnancies that eventually lead to a livebirth.

With these findings, clinicians may adequately evaluate a patient's risk of pregnancy loss based on the patient's age, ART procedure, pregnancy plurality, and gestational age of the fetus. Additionally, these findings may inform reproductive health epidemiologists on population risks of pregnancy loss and contribute to future research on its etiology. Based on the comparable results to past studies among naturally conceived pregnancies, with caution, these results may also provide information for naturally conceived pregnancies among similar populations of women.


   ACKNOWLEDGMENTS
 
The ART data used for this study were collected by use of the Society for Assisted Reproductive Technology (SART) ART reporting system, developed by SART in 1986 and jointly supported by SART, the American Society for Reproductive Medicine, and the CDC.

The authors thank SART and the American Society of Reproductive Medicine, with whose contributions this work has been possible. They would also like to thank Victoria Wright and Dr. Maurizio Macaluso for help with data validation, Dr. John Karon for statistical assistance, and Drs. Lee Warner and Jennita Reefhuis for their helpful comments on the manuscript.

The findings and conclusions in this paper are those of the authors and do not necessarily represent the views of the US Centers for Disease Control and Prevention.

Conflict of interest: none declared.


   References
TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 References
 

  1. Wilcox AJ, Weinberg CR, O'Connor JF, et al. Incidence of early loss of pregnancy. N Engl J Med (1988) 319:189–94.[Abstract]
  2. Pezeshki K, Feldman J, Stein DE, et al. Bleeding and spontaneous abortion after therapy for infertility. Fertil Steril (2000) 74:504–8.[CrossRef][ISI][Medline]
  3. Schieve LA, Tatham L, Peterson HB, et al. Spontaneous abortion among pregnancies conceived using assisted reproductive technology in the United States. Obstet Gynecol (2003) 101:959–67.[Abstract/Free Full Text]
  4. Simon C, Landeras J, Zuzuarregui JL, et al. Early pregnancy losses in in vitro fertilization and oocyte donation. Fertil Steril (1999) 72:1061–5.[CrossRef][ISI][Medline]
  5. Wang JX, Norman RJ, Wilcox AJ. Incidence of spontaneous abortion among pregnancies produced by assisted reproductive technology. Hum Reprod (2004) 19:272–7.[Abstract/Free Full Text]
  6. Fertility Clinic Success Rate and Certification Act of 1992 (FCSRCA). Public law no. 102-493. 2006. 10-24-1992. Ref type: bill/resolution.
  7. Wright VC, Schieve LA, Reynolds MA, et al. Assisted reproductive technology surveillance—United States, 2002. MMWR Surveill Summ (2005) 54:1–24.[Medline]
  8. Bishop YMM, Fienberg SE, Holland PW. Discrete multivariate analysis: theory and practice. (1975) Cambridge, MA: MIT Press.
  9. Amato DA. A generalized Kaplan-Meier estimator for heterogeneous populations. Commun Stat Theory Methods (1988) 17:263–86.
  10. Winter E, Wang J, Davies MJ, et al. Early pregnancy loss following assisted reproductive technology treatment. Hum Reprod (2002) 17:3220–3.[Abstract/Free Full Text]
  11. Simpson JL, Mills JL, Holmes LB, et al. Low fetal loss rates after ultrasound-proved viability in early pregnancy. JAMA (1987) 258:2555–7.[Abstract]
  12. Molo MW, Kelly M, Balos R, et al. Incidence of fetal loss in infertility patients after detection of fetal heart activity with early transvaginal ultrasound. J Reprod Med (1993) 38:804–6.[ISI][Medline]
  13. Frates MC, Benson CB, Doubilet PM. Pregnancy outcome after a first trimester sonogram demonstrating fetal cardiac activity. J Ultrasound Med (1993) 12:383–6.[Abstract]
  14. Keenan JA, Rizvi S, Caudle MR. Fetal loss after early detection of heart motion in infertility patients. Prognostic factors. J Reprod Med (1998) 43:199–202.[ISI][Medline]
  15. Spandorfer SD, Davis OK, Barmat LI, et al. Relationship between maternal age and aneuploidy in in vitro fertilization pregnancy loss. Fertil Steril (2004) 81:1265–9.[CrossRef][ISI][Medline]
  16. Goldhaber MK, Fireman BH. The fetal life table revisited: spontaneous abortion rates in three Kaiser Permanente cohorts. Epidemiology (1991) 2:33–9.[Medline]

Add to CiteULike CiteULike   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us    What's this?



This Article
Right arrow Abstract Freely available
Right arrow FREE Full Text (PDF) Freely available
Right arrow All Versions of this Article:
165/12/1380    most recent
kwm035v1
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in ISI Web of Science
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Add to My Personal Archive
Right arrow Download to citation manager
Right arrow Search for citing articles in:
ISI Web of Science (1)
Right arrowRequest Permissions
Right arrow Disclaimer
Google Scholar
Right arrow Articles by Farr, S. L.
Right arrow Articles by Jamieson, D. J.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Farr, S. L.
Right arrow Articles by Jamieson, D. J.
Social Bookmarking
Add to CiteULike   Add to Connotea   Add to Del.icio.us  
What's this?