A comprehensive Bayesian analysis assessing the effectiveness of lymphocyte immunotherapy for recurrent spontaneous abortion

Abstract

Abstract Recurrent spontaneous abortion (RSA) affects 2%–5% of couples worldwide and remains a subject of debate regarding the effectiveness of lymphocyte immunotherapy (LIT) due to limited retrospective studies. We conducted a comprehensive Bayesian analysis to assess the impact of LIT on RSA. Using data from the Shenzhen Maternity and Child Healthcare Hospital (2001–2020, n = 2316), a Bayesian generalized linear model with predictive projection feature selection was employed. Our analysis revealed a significant improvement in live birth rates for RSA patients undergoing LIT. Notably, LIT had a greater impact compared to the other 85 factors considered. To mitigate research bias, we conducted a Bayesian meta-analysis combining our dataset with 19 previously reported studies (1985–2021, n = 4246). Additionally, we developed an empirical model highlighting the four key factors, which are the LIT result, age, paternal blood type, and anticardiolipin antibody. Younger age (19–27), paternal blood type B, and a positive anticardiolipin antibody (IgM) were associated with better therapeutic outcomes in LIT for RSA. These findings aid clinicians in identifying suitable candidates for LIT and improving treatment outcomes.