Article description

Summary:

• Heparin-induced thromocytopaenia (HIT) is an infrequent, antibody-mediated, prothrombotic complication of heart surgery with cardiopulmonary bypass (CPB).
• Diagnosis is complex involving a combination of clinical signs and laboratory detection of positive antibodies to the platelet factor 4/heparin complex using heparin-induced platelet antibodies (HIPA).
• The positive predictive value of HIPA is low in the setting of heart surgery and best utilised when there is the suspicion of HIT is high using a screening score.
• The aim of the study was to compare the sensitivities of 4T’s, modified 4T’s and Lilo-Le Louet (LLL) screening scores in patients who had undergone heart surgery and been tested for HIPA in diagnosing heparin-induced thromocytopaenia (HIT).
• Single centre, retrospective, observational study design.
• A 5-point level of confidence (LOC) score was used where A defined the highest confidence of HIT and E, the highest confidence of an alternative cause of thrombocytopenia. A narrow (A or B) and broader (A, B or C) definition of true HIT were explored.
• 6,053 patients undergoing surgery 266 (4%) underwent HIPA testing.
• 34 (13%) had a high likelihood of HIT, 16 (6%) were clinically dubious, and 216 (81%) had a high likelihood of another cause of thrombocytopenia.
• HIPA undertaken less that 5 days postoperatively were not associated with any high-likelihood HIT cases whether or not heparin had been administered preoperatively.
• Although traditional 4T’s, modified 4T’s, and LLL scores had statistically similar sensitivity for predicting HIT, the modified 4T’s and LLL had superior specificity (p < 0.001).

Conclusions:

• Traditional 4T’s score was inferior to both the LLL and modified 4T’s scores to screen for HIT within 2 weeks of CPB.