Characterization of analytical errors in thromboelastography interpretation

Interpretation of Thromboelastography (TEG) curve involves correlating patient's clinical profile with TEG parameters and the tracing, keeping in mind the potential sources of errors, and hence requires expertise. We aimed to analyse the analytical errors in TEG interpretation due to paucity of literature in this regard.

The retrospective study was conducted in an apex trauma center in North India. Five months of data was reviewed by two laboratory physicians, with differences resolved by consensus. Cases with pre-analytical errors, missing data and TEG runs lasting <10 ​min were excluded. The analytical errors were classified into: preventable, potentially preventable, non-preventable, and non-preventable but care could have been improved.

Out of 440 TEG tracings reviewed, 70 were excluded. An analytical error was present in 60/370 (16.2%) tracings. There were six types analytical errors, of which, tracings of severe hypocoagulable states showing k-time ​= ​0 (33.3%) was the commonest, followed by tracings with spikes at irregular intervals (30%). Of all the analytical errors, 29/60 (48.2%) were preventable and 5/60 (8.3%) were potentially preventable.

Analytical variables that lead to errors in TEG interpretation were identified in about one-sixth of the cases and almost half of them were preventable. Awareness about the common errors amongst clinicians and laboratory physicians is critical to prevent treatment delay and safeguard patient safety.