Predictors of spontaneous pregnancy loss in single ventricle physiology

Background Pregnant patients with single ventricle (SV) physiology carry a high risk of spontaneous pregnancy loss (SPL), yet the clinical factors contributing to this risk are not well defined. Methods Single-centre retrospective study of pregnant patients with SV physiology seen in cardio-obstetrics clinic over the past 20 years with chart review of their obstetric history. Patients without a known pregnancy outcome were excluded. Univariable Bayesian panel-data random effects logit was used to model the risk of SPL. Results The study included 20 patients with 44 pregnancies, 20 live births, 21 SPL and 3 elective abortions. All had Fontan palliation except for two with Waterston and Glenn shunts. 10 (50%) had a single right ventricle (RV). 14 (70%) had moderate or severe atrioventricular valve regurgitation (AVVR). Atrial arrhythmias were present in 16 (80%), Fontan-associated liver disease (FALD) in 15 (75%) and FALD stage 4 in 9 (45%). 12 (60%) were on anticoagulation. Average first-trimester oxygen saturation was 93.8% for live births and 90.8% for SPL. The following factors were associated with higher odds of SPL: RV morphology (OR 1.72 (95% credible interval (CrI) 1.0008–2.70)), moderate or severe AVVR (OR 1.64 (95% CrI 1.003–2.71)) and reduced first-trimester oxygen saturation (OR 1.83 (95% CrI 1.03–2.71) for each per cent decrease in O2 saturation. Conclusion Pregnant patients with SV physiology, particularly those with RV morphology, moderate or severe AVVR, and lower first-trimester oxygen saturations, have a higher risk of SPL. Identifying these clinical risk factors can guide preconception counselling by the cardio-obstetrics team. Data are available upon reasonable request. The data generated and analysed during the current study are available from the corresponding author upon reasonable request. Researchers who wish to access the data should provide a detailed research proposal and demonstrate a commitment to maintaining confidentiality and using the data solely for the proposed research purposes. The data will be provided in a deidentified format to ensure participant privacy and comply with ethical guidelines.