Poster Session I Thursday, February 1 • 10:30 AM - 12:00 PM • Trinity Exhibit Hall| Volume 218, ISSUE 1, SUPPLEMENT , S122-S123, January 01, 2018

180: Diagnostic precision of proposed criteria for research reporting of amniotic fluid embolism


      Amniotic fluid embolism (AFE) remains a leading cause of maternal and neonatal morbidity and mortality. Historically, case definitions have been non-specific leading to challenges in accurate diagnosis and reporting. Strict diagnostic criteria for the research reporting of AFE have recently been published to promote uniformity in data collection and facilitate the understanding of this condition. We sought to examine the diagnostic precision of recently published criteria for the research reporting of amniotic fluid embolism.

      Study Design

      Charts of patients submitted to the AFE registry between 2013 and 2017 were analyzed. A group of experts in critical care obstetrics conducted a standard, detailed clinical analysis of medical records to categorize these women into 3 groups: (1) Amniotic fluid embolism. (2) A different diagnosis (3) Indeterminate diagnosis. Charts were also analyzed separately to determine whether they met the recently published, objective criteria for the diagnosis of AFE. We then determined the correlation between standard expert record analysis and use of these strict, objective criteria.


      Records for 115 patients submitted to the registry as cases of AFE were analyzed. In 68 cases (59%) expert review resulted in the diagnosis of AFE. 26 cases (23%) had clear alternative diagnoses and the diagnosis in 21 cases (18%) was indeterminate. When subjected to published rigorous diagnostic criteria, 54 women were determined to have had a classic AFE. In all 54 cases, these women were also identified as classic AFE per standard expert review. 14 cases (22%) identified as AFE by expert review did not meet the strict diagnostic research criteria as a classic AFE, however were considered to be atypical cases of AFE.


      Strict diagnostic criteria for AFE were designed to reduce or eliminate the likelihood of including non-AFE cases in published data, a problem which has markedly hampered understanding of this condition in the past. These criteria were developed with the understanding that this methodology would exclude some atypical cases of AFE. Our data suggest that 100% of cases so identified will indeed have classic AFE, and that 22% of cases of probable AFE would be excluded from research consideration using these criteria.