• C M Chan, P J Corso, X Sun, P C Hill, and A F Shorr.
• Washington Hospital Center, Washington, DC 20010, USA.
• Thromb Haemostasis. 2011 Nov 1;106(5):934-8.

AbstractThe poor accuracy of the enzyme immune assay (EIA) contributes to the diagnostic challenge of heparin-induced thrombocytopenia (HIT) following cardiac surgery. We sought to determine if adjusting the threshold optical density (OD) defining a positive EIA improves the test's accuracy in subjects with an OD>0.40. We retrospectively analysed the results from both EIA and confirmatory serotonin release assays (SRAs) in cardiac surgery patients with EIA OD of >0.4. Employing the SRA as the standard, we compared the area under the receiver-operating characteristic (AUROC) curves of various OD measurements for identifying HIT. We examined baseline clinical variables associated with a positive SRA in the setting of a positive HIT EIA (OD >0.4). We then used logistic regression to identify baseline clinical variables independently associated with a positive SRA given a positive EIA. The cohort included 99 subjects with positive EIAs and 35% had positive SRAs. An OD>0.40 had moderate utility as a screening test for a positive SRA (AUROC: 0.68; 95% CI: 0.55-0.80). Increasing the OD threshold did not improve the HIT EIA's screening utility. Clinical variables independently associated with a positive SRA if the EIA were positive included female gender, absence of diabetes, and use of cardiopulmonary bypass. A relatively modest elevation in the OD measurement, when it is already known to be greater than 0.4, does not reliably exclude the potential for a positive SRA in this setting.

39 keywords...

hide…