Technical note Thromboelastography: a reliable test? S. Vig, A. Chitolie, D. H. Bevan, A. Halliday and J. Dormandy (Received 12 September 2000; revised 12 March 2001; accepted 19 March 2001) The thromboelastograph (TEG), a measure of global haemostasis, is routinely used during cardiac and hepatic surgery to optimize blood product selection and usage. It has recently been suggested that it may also be a useful tool to screen patients with hypercoagulable states. Limited published data on performance characteristics has led to speculation regarding its consistency and, therefore, validity of the results. This study was designed to assess the effect of stability of blood samples prior to testing, repeated sampling, intra- and inter-assay variability using the native, celite, tissue factor (TF) and Reopro-modi®ed TEG. Analysis of native and celite samples after storage over 90 min showed a period of instability up to 30 min. Thereafter, all parameters between 30 and 90 min were stable [ P not signi®cant (NS)]. When the same sample was repeatedly assayed, both native and celite TEG parameters showed a signi®cant change towards hypercoagulability ( P < 0.01), whereas the TF and Reopro-modi®ed TEG showed no change. Intra- and inter-assay variability on samples tested after 30 min showed excellent reproducibility for all parameters ( P NS). The data suggest that the TEG is a useful tool in haemostasis but requires a formal standard operating procedure to be adopted that takes into account the initial period of sample instability. Blood Coagul Fibrinolysis 12:555±561 # 2001 Lippincott Williams & Wilkins. Keywords: thromboelastography, performance data, stability, hypercoagulability Introduction Hartert ®rst described the thromboelastograph (TEG) as a global test of blood coagulation over 50 years ago [1]. The instrument allows a rapid assessment of coagulation and ®brinolysis from a single sample of native whole blood, anticoagulated citrated recalci®ed whole blood or activated whole blood. As the TEG produced an overall impression of coagulation rather than precise quantitative analy- sis of individual contributing factors, it remained largely a research tool in haematology. However, the role of the TEG is more diverse, with some haematologists advocating it as a screening tool for hypercoagulable defects [2]. The TEG has been investigated extensively by anaesthetists in cardiac and hepatic surgery [3±5]. Where used clinically, it has allowed appropriate selection and usage of blood and blood products during and following cardiac bypass and hepatic transplantation, resulting in a signi®cant reduction in the number of blood trans- fusions [6±12]. Its popularity has also led to an increase in the number of publications over the past 10 years. The principle components of the TEG are a cylindrical cup and a pin (Fig. 1). A warmed cup oscillates for 10 s through an angle of 48 459 with a pin freely suspended in the cup by a torsion wire. Blood is added and a torque is ®rst transmitted as the clot forms linking the cup and pin together, increasing as the clot strengthens and decreasing as the clot lyses. The clot's physical properties, i.e. rate of formation, clot strength and stability, are depen- dent on the interaction of ®brinogen, platelets and plasma proteins, and this process produces a char- acteristic trace that historically has been visually classi®ed as normal, hypocoagulable or hypercoagul- able. The CTEG 1 model 3000 (Haemoscope Corp, Skokie, Illinois, USA) now consists of a bench-top instrument comprising a dual-channel Coagulation Analyser (TEG 1 ) and TEG 1 analytical software. S. Vig, A. Halliday and J. Dormandy are with the Department of Vascular Surgery, and A. Chitolie and D. H. Bevan are with the Department of Haematology, St George's Hospital, London, UK. Address correspondence to Dr S. Vig, Department of Vascular Surgery, St George's Hospital, Blackshaw Road, London SW17 OQT, UK. Tel: 44 20 8725 2877; fax: 44 20 8725 2490; e-mail svig@doctors.org.uk Blood Coagulation and Fibrinolysis 2001, Vol 12, No 7 555 Blood Coagulation and Fibrinolysis 2001, 12:555±561 0957±5235 # 2001 Lippincott Williams & Wilkins