Table 2.

Characteristics of included studies.

Author, year	Study design, location, study period, publication type	Study population	Intervention, comparator	Primary objective, secondary objective	Main results	Assessment of study
Murray et al. (2004)	RCT: UK. October 2001–April 2003. Full-length article.	First-time ART couples with no previous pregnancies (n = 272). Response rate: 200/272 = 73.5%. Mean age (female): 33.1–33.9 years. Mean years of infertility (couple): 3.4–4.0 years.	Group 2: Intervention (n = 66). Intervention: Single-page evidence-based document on twin pregnancy (including prematurity and disability) risks, both maternal and fetal/neonatal. Group 3: Intervention + discussion (n = 61). Comparator: Group 1 patients who received standard information provided to all patients (n = 62).	Primary objective: To evaluate hypothetical acceptability of eSET. eSET preference was defined as couples' acceptability for a policy of eSET. Secondary objective: To evaluate knowledge of risks associated with twin pregnancy.	Reported couples' joint decisions. Rates of eSET acceptability in Groups 1, 2 and 3 if patients believed that pregnancy rates declined with eSET: 27%, 30% and 32%, respectively (P = 0.39). Rates of eSET acceptability in Groups 1, 2 and 3 if patients believed that pregnancy rates were the same with DET and eSET: 82%, 83% and 87%, respectively (P = 0.76). Rates of eSET acceptability in Groups 1, 2 and 3 if cost of treatment was fixed regardless of number of cycles: 57%, 55% and 65%, respectively (P = 0.73). Knowledge that twins carry higher risks was high in all three groups with no significant differences (Group 1, 95%; Group 2, 98%; Group 3, 98%) (P = 0.26). Percentage of couples indicating that they would not mind having twins was high in all three groups with no significant differences (Group 1, 94%; Group 2, 95%; Group 3, 92%) (P = 0.70).	Strengths: RCT study. Intervention was pilot tested before the trial and derived from published studies on eSET outcomes. Able to assess impact of educational intervention on eSET preference alone. Weaknesses: 25% (72/272) of eligible couples could not be recruited. Hypothetical acceptability of eSET was examined. Single-centre study. Quality of study: Level I Risk for bias: Moderate
Newton et al. (2007)	Pre–post study: Canada. December 2003–December 2004. Full-length article.	ART couples after fresh transfer. n = 140. Men and women were interviewed separately. Response rate: women, 56% (79/140); men, 38% (53/140). Mean age (female): 33.3 years. Mean years of infertility (female): 3.0 years.	Intervention: During the interview, brief description of potential increase in complications (pre-eclampsia, low birth weight, postpartum depression) of multiple pregnancy, followed by low-, medium- and high-risk scenarios presented orally and on a card. Comparator: Patient responses to interview questions that preceded provision of information on complications.	Primary objective: To evaluate desire for different transfer options (including eSET) and twin pregnancies. eSET preference was defined as desire for different transfer options (including eSET).	Reported male and female preferences separately. Presentation of low-risk scenario decreased twin pregnancy desirability among men but not women (P < 0.001). Presentation of medium- and high-risk scenarios decreased twin pregnancy desirability among both men and women (P < 0.001). Presentation of low-, medium- and high-risk scenarios resulted in a significant increase in eSET desirability (P < 0.001) and a significant decrease in DET desirability (P < 0.001). Initial preference for DET > 75%. Perceived increase in chance of pregnancy with DET highly correlated to desire to transfer two embryos (P < 0.05). eSET preferences would decline if patients believed pregnancy rates would be lower with eSET.	Strengths: Examined patients' reported desirability of embryo transfer practices before and after provision of risk information, and evaluated male and female preferences separately. Able to assess independent effect of patient education on desire for eSET. Weaknesses: Hypothetical scenario provided to patients during interview process immediately after embryo transfer; hence, degree to which findings would apply to actual selection of number of embryos to transfer, and whether reported desirability would change over time, is unknown. Risk information was brief (approximately 125 words). Only legally married or cohabiting couples were included. Lesbian couples and non-English-speaking couples were excluded. Assessment of immediate response to hypothetical risk scenarios. Single-centre study. Self-reported. Small sample size. Quality of study: Level II-3 Risk for bias: High
Ryan et al. (2007)	Pre–post study on educational intervention and retrospective cohort study of mSBT clinic policy: USA. September 2004–June 2005. Full-length article.	Educational campaign included all new ART couples. n = 120. Response rate for completion of both pre- and post-education questionnaires: 110/120 = 92%. Mean age (female): 32.4 years. Mean years of infertility (couple): 2.8 years. Concurrent mSBT policy was targeted at couples at high risk for twins: first-time ART IVF patients aged < 38 years with at least seven embryos, no previous failed cycles and at least one good-quality blastocyst. n = 355 transfer cycles. Mean age: 33 years. Mean years of infertility: not reported.	Intervention: Multifaceted intervention. One-page educational description of comparative risks of twins versus singletons and discussion with physician; concurrent implementation of mSBT policy for couples at high risk for twin pregnancies. Comparator: Survey responses prior to educational campaign; cycle outcomes prior to mSBT policy.	Primary objective: To evaluate MPR. Secondary objective: Knowledge of twin risks, desired number of embryos transferred. eSET preference was defined as preference for desired number of embryos transferred.	Reported couples' joint decisions. Knowledge of twin risks improved after educational intervention (from 61% correct responses about risks prior to intervention to 93% correct responses after intervention; P < 0.05). More patients preferred SET after the educational campaign (61%) than before the campaign (22%) (P < 0.001). Before and after the educational campaign, 75% of patients would choose SET only if pregnancy rates were equivalent to or better than DET. After the educational campaign, 25% of the patients would continue to choose DET over SET even if the pregnancy rates were equivalent.	Strengths: Compared responses before and after intervention. Used evidence-informed education. Able to assess independent effects of education component on knowledge of twin risks and preference for SET. Weaknesses: Cannot separately evaluate the impact of the educational intervention on MPR as all patients at high risk for twins were subject to clinic's mandatory mSBT policy. No adjustment for confounders. Single-centre study. Quality of study: Level II-3 Risk for bias: Moderate
Arny et al. (2010)	Retrospective cohort study: USA. January 2005–December 2009. Conference abstract.	Cycle-specific analysis of ART patients. Blastocyst transfers performed for cycles with more than six top-quality embryos available (1/1/07–6/8/08) or more than four top-quality embryos available (6/9/08–12/31/09). n = 677 autologous cycles, patients aged < 35 years. Mean age and mean years of infertility were not reported.	Intervention: Multifaceted intervention. eSET with blastocyst transfers for selected patients (initiated in 2007) and enhanced patient education regarding twin pregnancy risks (initiated in January 2009). Education included mandatory lecture, written information on twin pregnancy risks, and sessions with practitioners. Clinic-specific graphs showing association between number of embryos transferred, pregnancy rates and twin rates were provided. Comparator: Cleavage-stage transfers performed during 2005–2006 (with no patient education or policy) and 2007–2008 (policy but no patient education).	Primary objective: To evaluate use of eSET. Secondary objective: To evaluate PR and twin pregnancies.	eSET rates increased from 15.7% in 2005–2006 (37/236) to 31.2% in 2007–2008 (88/282) to 46.5% (74/159) in 2009 (P < 0.05). The increase in eSET rates from 31.2% to 46.5% after enhanced patient education was significant (P < 0.0013). Twin pregnancy rates decreased significantly from 35% in 2005–2006 (42/120) to 29.9% in 2007–2008 (43/144) to 20.5% in 2009 (18/88) (P = 0.02). The decrease in twin pregnancy rates from 29.9% to 20.5% after enhanced patient education was not significant. No significant change in pregnancy rate over all time periods.	Strengths: Compared effectiveness of intervention on rates of eSET, pregnancies, twins and triplets. Able to assess independent effects of education component on eSET rates and twin rates after initiation of single blastocyst policy. Weaknesses: Reports combined eSET rates for cleavage and blastocyst transfers. No information on patient characteristics. Single-centre study. Quality of study: Level II-2 Risk for bias: High
Hope and Rombauts (2010)	RCT: Australia. February–October 2007. Full-length article.	First-time ART couples. n = 215. Response rate: 131/215 = 61%. Mean age (female): 33.6 years (DVD), 34.7 years (brochure). Mean years of infertility (female): 2.5–2.7 years.	Intervention: Educational DVD containing information on twin pregnancy risks. DVD was 12 min long and included interview with embryologist, obstetrician, and two mothers of twins (one with uncomplicated delivery, the other with preterm delivery) and images of children. Comparator: Brochure containing identical factual information on twin pregnancy risks and challenges as included in DVD. Baseline pre-intervention data also used for comparison with data from DVD and brochure groups.	Primary objective: To evaluate preference for eSET. Secondary objective: To evaluate changes in knowledge of twin pregnancy risks. eSET preference was defined as preference for eSET.	Reported couples' joint decisions. Compared with brochure group, DVD group were more likely to agree that twins are usually smaller (P = 0.01), more likely to require neonatal intensive care unit admission (P = 0.044), more likely to have learning disabilities (P = 0.003), less likely to breast feed (P = 0.003), and more likely to have heart abnormalities (P = 0.004). Compared with baseline data, after intervention, a significant increase in acceptability of eSET in brochure group (39.2% to 66.7%) and DVD group (32.6% to 82.6%) but change significantly more pronounced in DVD group (P = 0.014). DVD group were significantly more likely to prefer eSET compared with brochure group among individuals who were initially undecided about the number of embryos to transfer (86% versus 42%; P = 0.0003). Both groups had improved knowledge of twin risks after the intervention.	Strengths: RCT study with staff blinded to randomization. Able to assess independent effect of education component. Weaknesses: Recruitment bias from excluding non-English-speaking couples and those who did not have access to a DVD player. Sample size not large enough to evaluate statistical significance of actual number of embryos transferred (only 68/100 couples completing study had more than one embryo available). Single-centre study. Did not meet intention-to-treat analysis criteria as not all participants randomized completed the study. Quality of study: Level I Risk for bias: High
Griffin et al. (2012)	Pre–post study: USA. September 2008–October 2009. Full-length article.	Couples undergoing first ART treatment, aged 21–38 years. Clinic had equivalent pregnancy success rates from SET and DET. n = 163. Response rate: 163/500 = 32.6%. Mean age (female): 32.1 years. Mean years of infertility: not reported.	Intervention: Educational handout describing maternal and fetal risks of twin gestation was assessed alone and in three different payment scenarios (patient pays $15,000 for each cycle, insurance covers up to two cycles, insurance covers unlimited cycles). Comparator: Survey responses prior to reviewing educational handout.	Primary objective: To evaluate patients' preference of SET versus DET in the context of three different payment scenarios. eSET preference was defined as preference for SET.	Reported couples' joint decisions. Among couples who thought they would conceive with IVF, after education alone, preference for singletons increased significantly from 37% to 67%, and preference for twins decreased significantly from 20% to 12% (P < 0.001). Patients who had no stated preference declined from 43% to 21% (P < 0.001). Education regarding risks of multiple gestations significantly increased preference for SET for all insurance scenarios (P < 0.001). More patients would choose SET in the context of unlimited insurance coverage for ART versus having to pay out of pocket (80% versus 61%) (P < 0.001).	Strengths: Anonymous survey. Used logistic regression to determine whether patient characteristics were associated with preference for SET. Able to assess impact of educational handout alone and in each payment scenario. Weaknesses: Excluded couples with previous ART. Hypothetical scenarios. Low response rate. Single-centre study. Quality of study: Level II-3 Risk for bias: High

ART, assisted reproductive technology; BT, blastocyst transfer; CPR, clinical pregnancy rate; DET, double embryo transfer; eSBT, elective single blastocyst transfer; eSET, elective single embryo transfer; ICSI, intra-cytoplasmic sperm injection; mSBT, mandatory single blastocyst transfer; MPR, multiple pregnancy rate; PGD, pre-implantation genetic diagnosis; RCT, randomized controlled trial; SBT, single blastocyst transfer; SET, single embryo transfer.