To cite this article: Pouplard C, Gueret P, Fouassier M, Ternisien C, Trossaert M, Ré gina S, Gruel Y. Prospective evaluation of the Ô4TsÕ score and particle gel immunoassay specific to heparin/PF4 for the diagnosis of heparin-induced thrombocytopenia. J Thromb Haemost 2007; 5: 1373-9.See also Arnold DM, Kelton JG. Testing for heparin-induced thrombocytopenia: are we there yet? This issue, pp 1371-2.Summary. Background: Heparin-induced thrombocytopenia (HIT) is a severe disease that is often difficult to diagnose. A clinical scoring system, the Ô4TsÕ score, has been proposed to estimate its probability before laboratory testing, and a particle gel immunoassay (H/PF4 PaGIA Ò ) has also been developed for rapid detection of HIT antibodies. Aim: To evaluate the performance of both methods when HIT is suspected clinically. Methods: Two hundred thirteen consecutive patients were included in four centers. The probability of HIT was evaluated using the 4Ts score blind to antibody test results. HIT was confirmed only when the serotonin release assay (SRA) was positive. Results: The risk of HIT was evaluated by the 4Ts score as low (LowR), intermediate (IR) or high (HR) in 34.7%, 60.6% and 4.7% of patients, respectively. The negative predictive value (NPV) of the 4Ts score was 100%, as the SRA was negative in all LowR patients. PaGIA Ò was negative in 176 patients without HIT (99.4%, NPV) and the negative likelihood ratio (LR-) was 0.05. PaGIA Ò was positive in 37 patients, including 21 with HIT (positive predictive value = 56.8%), with a positive LR of 11.4. A negative PaGIA Ò result decreased the probability of HIT in IR patients from 10.9% before assay to 0.6%, whereas a positive result did not substantially increase the likelihood for HIT. Conclusion: The use of the 4Ts score with PaGIA Ò appears to be a reliable strategy to rule out HIT.

Summary Cardiopulmonary bypass (CPB) induces the release of platelet factor 4 (PF4) and patients are at risk of heparin‐induced thrombocytopenia (HIT). This study was aimed to determine whether an abnormal evolution in platelet count (PC) after CPB is predictive of the development of HIT antibodies. Two abnormal PC patterns were defined: pattern P1, characterized by a decrease in PC following previous correction of thrombocytopenia occurring during CPB, and pattern P2, defined as a persistent low PC in the days following CPB. PC was evaluated for 10 d in 305 consecutive patients before and after CPB. Serotonin release assay (SRA) was carried out between days 8 and 10 to detect pathogenic heparin‐dependent antibodies. Moreover, antibodies to heparin–PF4 (H–PF4) complexes were assayed by enzyme‐linked immunosorbent assay. PC evolution after CPB was normal in 300 patients although antibodies to H–PF4 were frequently present (53·4%). Changes in PC were abnormal in five patients with pattern P1 (n = 4) or P2 (n = 1). As SRA was positive in four of the five cases, the positive predictive value of abnormal PC pattern for pathogenic HIT antibodies was 80%. Careful follow‐up of PC after CPB makes it possible to predict with high specificity (99%) for those patients who develop pathogenic HIT antibodies.

No abstract