Assisted reproductive technology (ART) is associated with an increased incidence of perinatal morbidity. Studies have shown that neonates conceived with ART are at increased risk of preterm delivery and low birth weight (LBW).1–3 These risks are largely attributable to multiple gestation; however, rates of preterm delivery and LBW are elevated even among singletons conceived with ART. It is not known which components of ART contribute to adverse perinatal outcomes.
For example, it is unknown whether obstetric outcomes differ after intracytoplasmic sperm injection (ICSI), in which a single spermatozoa is injected into each oocyte, and conventional in vitro fertilization (IVF), in which oocytes and sperm are mixed in a dish. Some studies have demonstrated decreased risks of prematurity and LBW with ICSI compared with conventional IVF, whereas others have not.3–10 Disparate findings may be the result of the fact that studies have not controlled for male factor infertility. Although ICSI was historically performed for male factor, it has increasingly been used for patients without male factor.11 In 2013, 69% of ART cycles used ICSI.12 This is an important distinction, because patients without male factor may have female infertility diagnoses associated with an increased baseline risk of morbidity. Given the increasing use of ICSI, additional investigation is warranted.
The objective of our study is to examine the association between ICSI and singleton adverse perinatal outcomes, specifically prematurity and LBW. We hypothesized that differences in these outcomes are driven by baseline characteristics rather than the fertilization method.
MATERIALS AND METHODS
We conducted a retrospective cohort study using the Society for Assisted Reproductive Technologies Clinical Outcomes Reporting System database from 2004 to 2013. This study was considered exempt by the Duke institutional review board because the participants were deidentified.
The Society for Assisted Reproductive Technologies Clinical Outcomes Reporting System database contains comprehensive data from greater than 90% of all clinics performing ART cycles in the United States. The data were collected through voluntary submission, verified by Society for Assisted Reproductive Technologies, and then reported to the Centers for Disease Control and Prevention in compliance with the Fertility Clinic Success Rate and Certification Act of 1992 (Public Law 102-493). The Society for Assisted Reproductive Technologies maintains Health Insurance Portability and Accountability Act–compliant business associates agreements with reporting clinics. In 2004, after a contract change with the Centers for Disease Control and Prevention, the Society for Assisted Reproductive Technologies gained access to the Society for Assisted Reproductive Technologies Clinical Outcomes Reporting System data system for the purposes of conducting research. The data in the Society for Assisted Reproductive Technologies Clinical Outcomes Reporting System are validated annually with select clinics having on-site visits for chart review based on an algorithm for clinic selection.13 During each visit, data reported by the clinic were compared with information recorded in patients' charts.13 In 2012, records for 2,045 cycles at 35 clinics were randomly selected for full validation along with 238 egg or embryo banking cycles. The full validation included review of 1,318 cycles for which a pregnancy was reported. Among the nondonor cycles, 331 were multiple fetus pregnancies. Ten of 11 data fields selected for validation were found to have discrepancy rates of 5% or less. The exception was the diagnosis field, which, depending on the diagnosis, had a discrepancy rate between 2.1% and 9.2%.13
We identified patients younger than 43 years of age who underwent their first, fresh, autologous IVF or ICSI cycle during the study period. Only singleton live births were included in our analysis. Exclusion criteria were the use of split ICSI, in which some oocytes were fertilized with ICSI and others with conventional IVF, or second-day ICSI, in which oocytes with failed conventional fertilization were then injected with sperm as a rescue attempt. Additional exclusion criteria were mixed sperm source or missing sperm data, preimplantation genetic diagnosis, missing birth weight data or other covariates, and the presence of more than one fetal heartbeat on ultrasonography.
The primary outcomes were preterm delivery, defined as delivery at less than 37 weeks of gestation, and LBW, defined as birth weight less than 2,500 g. Secondary outcomes were term LBW (LBW at 37 weeks of gestation or greater), preterm LBW (LBW at less than 37 weeks of gestation), very preterm delivery (delivery at less than 32 weeks of gestation), and very LBW (birth weight less than 1,500 g).
Statistical analyses were conducted using R 18.104.22.168 The two-sample Student t test (or Wilcoxon rank-sum test) and the χ2 test were used to analyze continuous and categorical variables, respectively. To reduce systematic differences in baseline characteristics between treatment groups in the observational data, we used propensity score matching. The propensity score was computed by accounting for the effect of age as a continuous variable, body mass index (BMI, calculated as weight (kg)/[height (m)]2), race, smoking status, gravidity, parity, infertility diagnoses as binary variables (male factor, unexplained, endometriosis, diminished ovarian reserve, polycystic ovarian syndrome, tubal factor, uterine factor, and other), sperm source (autologous or donor), method of sperm collection (surgical or nonsurgical), number of oocytes retrieved, and year of treatment as a categorical variable. Propensity scores were matched one to one using the nearest neighbor matching method. R2 and Cramer's V were used to assess the balance in baseline covariates for continuous variables and categorical variables, respectively, where values closer to zero represent more balance between treatment groups. The matched data set was used in logistic regression models to determine whether ICSI was associated with preterm delivery, LBW, term LBW, preterm LBW, very preterm delivery, and very LBW after controlling for posttreatment covariates including assisted hatching, number of embryos transferred, day of embryo transfer (blastocyst [day 5–6] or nonblastocyst [day 2–4]), number of embryos cryopreserved, neonatal sex, neonatal weight, and gestational age at birth. Subgroup analyses were performed after stratification by sperm source and female prognosis. A favorable fertility prognosis was defined as female age younger than 35 years with more than three oocytes retrieved. A P value of <.05 was considered statistically significant.
A total of 90,401 cycles were included in the analysis. Of these, ICSI was used in 60,719 (67.2%) and conventional IVF was used in 29,682 (32.8%) (Table 1). Male factor infertility was the most common infertility diagnosis reported for ICSI cycles (57.5%), but was relatively rare among conventional IVF cycles (9.2%). Conversely, unexplained infertility and tubal factor accounted for almost half of all conventional IVF cycles (49.1%) but less than one fourth of ICSI cycles (23.6%). Autologous sperm was used in more than 96% of both ICSI and IVF cycles. As expected, surgically obtained sperm accounted for 6.2% of ICSI cycles compared with only 0.1% of conventional IVF cycles.
After propensity score matching, 12,364 ICSI and 12,364 conventional IVF cycles were included in the analysis (Tables 1 and 2). The odds of preterm delivery were similar between ICSI and conventional IVF before and after adjustment for covariates (odds ratio [OR] 0.98, 95% CI 0.91–1.06, P=.61 and adjusted OR 1.02, 95% CI 0.89–1.18, P=.75) (Table 3). The odds of LBW were also similar between ICSI and conventional IVF before and after adjustment for covariates (OR 0.98, 95% CI 0.90–1.06, P=.56 and adjusted OR 0.92, 95% CI 0.78–1.10, P=.35, respectively) (Table 3). There were no significant treatment effects of ICSI on secondary outcome variables.
Multiple subgroup analyses were then performed using the propensity matched data set to isolate the effect of the fertilization method. First, the data were analyzed among patients who used donor sperm. There were no statistically significant differences in the selected perinatal outcomes between ICSI and conventional IVF groups after adjustment for covariates (Table 4). Patients were then stratified by the presence or absence of male factor infertility. Among those with male factor infertility, there were no significant differences in the selected perinatal outcomes between ICSI and conventional IVF (preterm delivery OR 0.77, 95% CI 0.54–1.12, P=.17; LBW OR 1.06, 95% CI 0.70–1.61, P=.79) (Table 5). Next, we examined outcomes among patients who used autologous sperm and had a favorable fertility prognosis based on female age younger than 35 years with more than three oocytes retrieved. Among the patients with a favorable prognosis, subgroup analyses were performed for those with male factor infertility only, unexplained infertility, and tubal factor only (Table 6). Among couples with a favorable prognosis and a diagnosis of male factor infertility only, there were no differences in the selected perinatal outcomes with ICSI or conventional IVF (preterm delivery OR 1.13, 95% CI 0.54–2.34, P=.75; LBW OR 0.79, 95% CI 0.35–1.78, P=.57). Findings were similar when analyses were restricted to patients with unexplained infertility (preterm delivery OR 0.98, 95% CI 0.70–1.36, P=.89; LBW OR 1.03, 95% CI 0.73–1.47, P=.86) or tubal factor infertility only (preterm delivery OR 1.14, 95% CI 0.79–1.65, P=.49; LBW OR 0.88, 95% CI 0.56–1.37, P=.56).
This large national study demonstrated that the fertilization method is not associated with preterm delivery or LBW among neonates conceived with ART. Our findings differ from those of previous studies, which reported a lower risk of preterm delivery and LBW with ICSI compared with conventional IVF. For example, a meta-analysis of five studies demonstrated a reduced risk of preterm delivery with ICSI (adjusted OR 0.80, 95% CI 0.69–0.93).3 That meta-analysis was limited by the inclusion of both fresh and frozen embryo transfers as well as heterogeneity of included studies.
A large Australian study compared perinatal outcomes of 1,641 singleton births conceived with ART with 292,141 singleton births conceived spontaneously.15 Of the births conceived with ART, conventional IVF was used in 948 and ICSI was used in 693. Both ICSI and conventional IVF were associated with significantly increased odds of preterm birth and LBW compared with spontaneous conception. The effect sizes were smaller, however, for ICSI (preterm delivery OR 1.63, 95% CI 1.24–2.15; LBW OR 1.96, 95% CI 1.52–2.52) than conventional IVF (preterm delivery OR 2.20, 95% CI 1.79–2.70; LBW OR 3.07, 95% CI 2.55–3.69). The authors concluded that ICSI may have a weaker association with perinatal outcomes than conventional IVF. The study was limited, however, by lack of a direct comparison between ICSI and IVF and by not controlling for male factor infertility, an important confounder.
A similar study examined 4,886 singleton births from the Norwegian Medical Birth Register from 1999 through 2006.16 These authors demonstrated a 40% increased risk of iatrogenic preterm delivery with conventional IVF compared with ICSI after adjusting for fetal sex, small for gestational age, maternal age, parity, and smoking (OR 1.4, 95% CI 1.02–1.9). The risk of spontaneous preterm delivery was not substantially different between the two groups (OR 1.1, 95% CI 0.9–1.5). The study was limited by inclusion of fresh and frozen embryo transfers without adjustment for multiple cycles per woman. Like with the previous study, the analysis did not control for male factor infertility.
More recently, a study from the Centers for Disease Control and Prevention compared reproductive outcomes between ICSI and IVF cycles from 2008 to 2012.11 Among patients without male factor infertility, risks of multiple birth and LBW were significantly lower with ICSI than with IVF (relative risk 0.93, 95% CI 0.91–0.95 and relative risk 0.96, 95% CI 0.94–0.99, respectively). The inclusion of multiple births complicates the interpretation of this study's findings, because LBW is a known complication of multiple gestation. Additionally, this study was limited by the inclusion of multiple cycles per patient and the inability to control for important confounders such as BMI and smoking. Finally, the analyses did not adjust for the use of surgical sperm extraction.
The present study was strengthened by the inclusion of only first, fresh cycles, eliminating the need to adjust for multiple cycles per woman. Additionally, its large cohort size allowed us to perform subgroup analyses to specifically examine outcomes among patients with similar diagnoses. Performing these analyses with a propensity score matched data set improved our ability to isolate the effect of the ICSI procedure itself rather than differences in baseline patient characteristics.
Our study was limited by the possibility of input error as well as elective reporting of results to the Society for Assisted Reproductive Technologies . The analysis included only those patients undergoing first cycles; however, some clinics may not have correctly accounted for a patient's previous cycles. Moreover, we did not have details on the male fertility evaluation and cannot confirm whether all clinics adhere to a uniform diagnosis of male factor infertility. Additionally, the decision to perform ICSI is sometimes based on the quality of the fresh semen specimen provided at the time of oocyte retrieval. These men may have an underlying component of male factor infertility despite a previously normal semen analysis. Therefore, our study was subject to misclassification bias. Finally, we did not have data on other obstetric complications or birth defects because these are not consistently entered in the Society for Assisted Reproductive Technologies Clinical Outcomes Reporting System.
In summary, the method of fertilization is not associated with preterm delivery or LBW among neonates conceived with ART. Previously reported differences in outcomes were likely secondary to baseline characteristics rather than the fertilization procedure itself. These data can be used to reassure patients who are deciding between ICSI and conventional IVF.
1. Perri T, Chen R, Yoeli R, Merlob P, Orvieto R, Shalev Y, et al. Are singleton assisted reproductive technology pregnancies at risk of prematurity? J Assist Reprod Genet 2001;18:245–9.
2. McDonald SD, Han Z, Mulla S, Murphy KE, Beyene J, Ohlsson A, et al. Preterm birth and low birth weight among in vitro fertilization singletons: a systematic review and meta-analyses. Eur J Obstet Gynecol Reprod Biol 2009;146:138–48.
3. Pinborg A, Wennerholm UB, Romundstad LB, Loft A, Aittomaki K, Söderstrom-Anttila V, et al. Why do singletons conceived after assisted reproduction technology have adverse perinatal outcome? Systematic review and meta-analysis. Hum Reprod Update 2013;19:87–104.
4. Fedder J, Loft A, Parner ET, Rasmussen S, Pinborg A. Neonatal outcome and congenital malformations in children born after ICSI with testicular or epididymal sperm: a controlled national cohort study. Hum Reprod 2013;28:230–40.
5. Rajesh H, Yap HA, Wu YJ. Pregnancy outcomes from in-vitro fertilisation and intracytoplasmic sperm injection: a comparison. Singapore Med J 2006;47:309–14.
6. Ombelet W, Cadron I, Gerris J, De Sutter P, Bosmans E, Martens G, et al. Obstetric and perinatal outcome of 1655 ICSI and 3974 IVF singleton and 1102 ICSI and 2901 IVF twin births: a comparative analysis. Reprod Biomed Online 2005;11:76–85.
7. 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.
8. Romundstad LB, Romundstad PR, Sunde A, von During V, Skjaerven R, Gunnell D, et al. Effects of technology or maternal factors on perinatal outcome after assisted fertilisation: a population-based cohort study. Lancet 2008;372:737–43.
9. Nelson SM, Lawlor DA. Predicting live birth, preterm delivery, and low birth weight in infants born from in vitro fertilisation: a prospective study of 144,018 treatment cycles. PLoS Med 2011;8:e1000386.
10. Eindhoven SC, van Uitert EM, Laven JS, Willemsen SP, Koning AH, Eilers PH, et al. The influence of IVF/ICSI treatment on human embryonic growth trajectories. Hum Reprod 2014;29:2628–36.
11. Boulet SL, Mehta A, Kissin DM, Warner L, Kawwass JF, Jamieson DJ. Trends in use of and reproductive outcomes associated with intracytoplasmic sperm injection. JAMA 2015;313:255–63.
12. Assisted reproductive technology 2013 report. Available at: http://www.cdc.gov/ART
. Retrieved July 16, 2016.
13. Center for Disease Control and Prevention, American Society for Reproductive Medicine, and Society for Assisted Reproductive Technology. 2012 assisted reproductive technology success rates: national summary and fertility clinic reports. Atlanta (GA): Centers for Disease Control and Prevention; 2014.
14. R Core Team. R: a language and environment for statistical computing. V Available at: https://www.R-project.org/
. Retrieved February 1, 2016.
15. Marino JL, Moore VM, Willson KJ, Rumbold A, Whitrow MJ, Giles LC, et al. Perinatal outcomes by mode of assisted conception and sub-fertility in an Australian data linkage cohort. PLoS One 2014;9:e80398.
16. Morken NH. Preterm delivery in IVF versus ICSI singleton pregnancies: a national population-based cohort. Eur J Obstet Gynecol Reprod Biol 2011;154:62–6.