Similarities and differences between WHO criteria and two other approaches for maternal near miss diagnosis

Abstract

Objectives To evaluate the similarities, differences and diagnostic aspects between World Health Organization (WHO) criteria and two other maternal near miss (MNM) diagnostic tools. methods A cross-sectional study was conducted from June 2011 to May 2012 in two reference maternity hospitals in Aracaju, Brazil. Prospective case identification and data collection were performed and patients were classified as an MNM case according to WHO, Waterstone and literature-based criteria. The diagnostic properties and concordance of literature-based and Waterstone criteria were calculated using WHO criteria as standard. results Of a total of 20 435 patients, 19 239 women did not have potentially life-threatening conditions, there were 17 maternal deaths, and 77 MNM cases based on the WHO criteria. Waterstone and literature-based criteria identified 404 and 959 MNM cases, respectively, most of them related to hypertensive disorders and haemorrhage. The sensitivity, specificity and accuracy in diagnosing MNM cases using Waterstone and literature-based criteria were above 90%, but Waterstone sensitivity was 48.1%. The similarities between the Waterstone and literature-based criteria were very weak compared to WHO criteria, with a positive percentage concordance below 9%. conclusions Although using WHO guidelines to detect MNM cases can be difficult when implemented in low-resource settings, the results from this study reinforce the importance of this tool in detecting the truly severe cases. Waterstone and literature-based criteria are not suitable for identifying indubitable MNM. However, they remain useful as a preliminary step to select potentially severe cases, mainly those related to hypertension and haemorrhage.