Evaluation of children with a suspected bleeding disorder applying the Impact‐R [Cone and Plate(let) Analyzer]

• Published in: Journal of thrombosis and haemostasis Vol. 7; no. 12; pp. 1990 - 1996
• Main Authors: REVEL‐VILK, S., VARON, D., SHAI, E., AGMON, Y., HYAM, E., DAAS, N., MISKIN, H., WEINTRAUB, M.
• Format: Journal Article
• Language: English
• Published: Oxford, UK Blackwell Publishing Ltd 01-12-2009
• Subjects: Adolescent; bleeding disorder; Blood Coagulation Disorders - diagnosis; Blood Platelet Disorders; Child; Child, Preschool; Cohort Studies; Cross-Sectional Studies; Family Health; Female; Humans; Impact‐R method; Infant; Male; Mass Screening - methods; platelet dysfunction; Platelet Function Tests - instrumentation; Retrospective Studies; ROC Curve; screening; von Willebrand disease
• ISSN: 1538-7933; 1538-7836; 1538-7836
• DOI: 10.1111/j.1538-7836.2009.03628.x

Background: A convenient screening test for children with bleeding symptoms before more labor‐intensive diagnostic steps are taken would be of value. The Impact‐R was designed in an attempt to analyse platelet function under near physiological conditions. Results are presented as surface coverage (SC, %) and average size (AS, μm2). Objective: In this cross‐sectional retrospective study, we assessed the use of the Impact‐R in the evaluation of children with a suspected bleeding disorder (BD). Methods: The hospital charts of 110 children referred to the coagulation laboratory were reviewed for personal and family bleeding history (BH) as well as results of the laboratory evaluation. Results: A laboratory ‘diagnosable’ BD (LBD) was found in 23 children (21%, 95% CI 14–30%). A diagnosis of LBD was associated with the severity of bleeding but not with family BH. By receiver‐operating characteristic (ROC) curve analysis, the SC was superior to the AS for diagnosis of a LBD. The Impact‐R was abnormal in 43/97 children (44.3%, 95% CI 34–55%). The predictive values of a normal and abnormal Impact‐R were 96% (95% CI 92–97%) and 42% (95% CI 28–56%), respectively. When considering the personal and family BH, the post‐test probability for LBD after a normal Impact‐R was reduced from 20% to 3.5% (95% CI 2.5–7%). Conclusions: A normal Impact‐R test is highly effective in excluding LBDs. Yet, in case of an abnormal Impact‐R test, further testing is needed. An algorithm that includes the personal and family BH and the results of a screening test may improve the diagnostic process. Prospective studies are now needed to confirm these findings.