Abstract
AbstractObjectivePoint-of-care tests like FIBTEM A5 have been proposed to guide the treatment of patients who might have low fibrinogen concentrations (≤2 g/L). The aim of this study was to describe fibrinogen concentrations according to previously proposed FIBTEM A5 cut-off points in blood samples collected from women during postpartum haemorrhage.Design and settingProspective multicentre cohort studyPopulation511 women sustaining postpartum haemorrhage. A total of 637 blood samples were drawn during haemorrhage.MethodsClauss fibrinogen concentrations and ROTEM® FIBTEM A5 values were studied to assess the diagnostic properties of previously proposed FIBTEM A5 cut-off points for the detection of low fibrinogen concentrations.Main outcome measuresYouden index, sensitivity and specificityResultsOf 511 women with a median total volume of blood loss of 1500 mL (IQR 1200 to 2000) 31 women (6%) developed Clauss fibrinogen concentrations below 2 g/L. Using FIBTEM A 5 cut-off of ≤ 7 mm: 48% of cases with Clauss fibrinogen ≤2 g/L were missed (FIBTEM A5 > 7mm), and of the 28 samples with FIBTEM A5 ≤ 7mm, 12 (43%) samples had Clauss fibrinogen >2 g/L. Using FIBTEM A5 cut-off of ≤12mm: 13% of cases with Clauss fibrinogen ≤2 g/L were missed and of the 145 samples with a FIBTEM A5 ≤12 mm, 118 had Clauss fibrinogen >2 g/L, resulting in false positive selection of 81% of women. Using FIBTEM A5 ≤ 15 mm: 97% (30/31) of the samples with Clauss fibrinogen ≤2 g/L were accurately selected; yet 89% (248/278) of samples that were selected had a fibrinogen concentration of >2 g/L. Based on the Youden index, the optimal cut-off point in our cohort was a FIBTEM A5 of 12mm with sensitivity 87% and specificity 81%.ConclusionsOur findings suggest that if FIBTEM A5 lower than 12 mm would have been used to detect women with fibrinogen concentrations below 2 g/L in order to treat them with fibrinogen concentrate, 87 % of the women with fibrinogen below 2 g/L would correctly have received fibrinogen. However, most women (81%) receiving fibrinogen concentrate would not have needed it, because they had plasma fibrinogen concentrations above 2 g/L.Fundingno funding was obtained to conduct this research
Publisher
Cold Spring Harbor Laboratory