Introduction: Thrombocytopenia is a common complication seen in critically ill patients. Elucidating a specific cause is often difficult as there are many confounding contributory factors. Mechanical circulatory support (MCS) is prothrombotic often requiring anticoagulation with heparin. However, MCS also serves as an independent source of consumptive thrombocytopenia which makes the diagnosis of heparin-induced thrombocytopenia (HIT) challenging. The incidence of HIT in cardiac surgery patients is as high as 5% and represents a significant cause of both morbidity and mortality. There are many validated diagnostic scoring tools to risk stratify those with clinical suspicion of HIT, but none were developed specifically for MCS patients. The utility of these tools such as the 4T score and modified 4T scores require validation in this patient population. Methods: This study aimed to determine the utility of two diagnostic scoring tools in HIT risk stratification for patients requiring MCS. This was a single center, retrospective cohort study that included MCS patients with a positive heparin induced platelet antibody between August 1, 2022 and July 31, 2024. The primary outcome was the negative predictive value of the 4T score and the modified 4T score. Results: A total of 64 patients were included for analysis; the most frequently used MCS device was the Impella with 42 (66%), followed by ECMO with 30 (47%). Twenty patients were supported by multiple devices. The median duration from heparin initiation to platelet nadir was 7.5 days, and the median platelet nadir was 52,000/uL. The negative predictive value for patients deemed low risk by the scoring tools was 93% and 94% for the 4T score and m4T score, respectively. For patients deemed intermediate risk, the NPV for the 4T score was 74% and for the m4T score was 80%. Conclusions: In this retrospective study, the utility of both the 4T score and modified 4T score to rule out HIT was high for mechanical circulatory support patients deemed low or intermediate risk. Future studies with larger sample sizes and a prospective design are needed to confirm these findings.
Dillon et al. (Sun,) studied this question.