School-age outcomes among IVF and ICSI-conceived children: a causal inference analysis using linked population-wide data

Abstract

Abstract Background Use of intracytoplasmic sperm injection (ICSI) continues to increase as the most common mode of oocyte insemination during in vitro fertilisation (IVF), sometimes in the absence of clear indications (i.e. male factor infertility). Several studies suggest an increased risk of congenital abnormalities after ICSI. The association between the ICSI technique and long-term childhood development remains unclear. Methods Our population-based study included singleton infants conceived via IVF and born between 2005 and 2013. The cohort included state-wide linked maternal and childhood administrative data from Victoria, Australia. The primary exposure was conception via ICSI (without severe male factor infertility), with those born following standard IVF as controls. Childhood development was examined using the Australian Early Development Census (AEDC), a broad assessment of childhood development across five domains of health and neurodevelopment performed in Australian schools every triennium at school entry (age 4–6 years). Our primary outcome used a validated global measure—developmental vulnerability—defined as scoring less than the 10th percentile in two or more of the five developmental domains (DV2). Causal inference methods were used to analyse observational data in a way that emulates a target randomised clinical trial. The adjustment variable set was determined a priori via a modified Delphi procedure. Given the use of observational data, there were missing data and inherent differences in the covariate profile between exposure cohorts. Multiple imputation, bootstrapping and doubly robust inverse probability weighted regression adjustment modelling was utilised to allow a causal interpretation of results. Results Our cohort (N = 3656) included 1489 IVF and 2167 ICSI-conceived children. We found no causal effect of ICSI on the risk of AEDC-defined developmental vulnerability at school-entry age compared with children conceived via standard IVF; adjusted risk difference − 1.11% (95% CI − 4.23 to 2.01%) and adjusted risk ratio 0.90 (95% CI 0.68 to 1.21). Conclusions Our findings suggest that the use of ICSI in IVF cycles without severe male factor infertility does not increase the risk of early childhood developmental vulnerability among children in their first year of school. These findings provide important reassurance for current and prospective parents and clinicians alike.

Keywords

• ICSI
• IVF
• Childhood development
• Causal analysis