Impact of frozen embryo transfer on placenta accreta and institutional variations.

RESEARCH QUESTION: Does programmed cycle frozen embryo transfer (PC-FET) increase the risk of placenta accreta spectrum (PAS) compared with natural cycle FET (NC-FET), and does this risk vary according to facility-level utilization of PC-FET?

DESIGN

This nationwide, cross-sectional study analysed 38,973 frozen embryo transfer deliveries identified in the Japanese perinatal registry database (2020-2022), linked to the assisted reproductive technology registry database (2019-2022). The primary exposure was PC-FET versus NC-FET. Facilities were stratified into tertiles based on the proportion of PC-FET (low, medium and high). The primary outcome was PAS, defined as either clinical or pathological. Morbid PAS was defined as PAS accompanied by haemorrhage >2000 ml or managed with balloon tamponade, compression sutures, embolization, or peripartum hysterectomy. OR and 95% CI were estimated with stabilized inverse probability of treatment weighting and multivariable regression.

RESULTS

PAS occurred in 3.6% and 0.7% of PC-FET and NC-FET deliveries, respectively (OR 5.02, 95% CI 3.87-6.53). Morbid PAS occurred in 1.5% and 0.3% of PC-FET and NC-FET deliveries, respectively (OR 5.36, 95% CI 3.51-8.20). Facilities with a high proportion of PC-FET had a higher risk of PAS (OR 1.36, 95% CI 1.16-1.58) and morbid PAS (OR 1.34, 95% CI 1.07-1.70) compared with facilities with a low proportion of PC-FET.

CONCLUSIONS

This study provides robust evidence that PC-FET increases the risk of PAS. Further research should explore the association between facility variations and the incidence of PAS.