Objective: To assess the contribution of maternal blood detection of IGFBP-1 for the diagnosis of amniotic-fluid embolism in clinical daily practice.
Design: A retrospective multicentre cohort study.
Setting: Three tertiary care obstetric units in France.
Sample: Data of 86 women for whom amniotic-fluid embolism had been suspected and maternal serum detection of IGFBP-1 had been performed between 2011 and 2019 were analysed.
Methods: The criteria defined by the United Kingdom Obstetric Surveillance System (UKOSS) were used for the retrospective diagnosis of amniotic-fluid embolism. The more structured definition proposed by the Society for Maternal-Fetal Medicine and the Amniotic Fluid Embolism Foundation (SMFM) was also used as secondary endpoint.
Main outcome measures: Agreements between biological and clinical assessments were tested. The performance of blood detection of IGFBP-1 for the diagnosis of amniotic-fluid embolism according to the UKOSS criteria, and to the SMFM definition, was also assessed.
Results: There was only slight agreement between clinical and laboratory diagnosis of amniotic-fluid embolism (Cohen's Kappa coefficient: 0.04). Blood detection of IGFBP-1 had a sensitivity of 16%, a specificity of 88%, a positive and a negative likelihood ratio of 1.3 and 0.95, respectively, and a positive and a negative predictive value of 58 and 50%, respectively, for the diagnosis of amniotic-fluid embolism based on the UKOSS criteria. The use of the more structured SMFM definition of amniotic-fluid embolism did not substantially change the results.
Conclusion: These results question the usefulness of blood detection of IGFBP-1 for the early diagnosis of amniotic-fluid embolism in daily clinical practice.
Tweetable abstract: This retrospective multicentre study questions the contribution of IGFBP-1 detection for the diagnosis of AFE.
Keywords: Amniotic-fluid embolism; diagnostic; pregnancy; serum marker.
© 2021 John Wiley & Sons Ltd.