Integration of a Heparin-Induced Thrombocytopenia Order-Set (HITOS): A Retrospective Study

Introduction: Heparin-induced thrombocytopenia (HIT) may occur after exposure to heparin containing products. When HIT is suspected, early recognition and diagnosis is crucial due to the risk for developing a life-threatening thrombotic event that can lead to substantial morbidity and even death.However, making the diagnosis of HIT is challenging due to a high false positive rate associated with most widely used screening tests for HIT. Therefore, clinical assessment of the likelihood of HIT is essential to help guide the decision to order a HIT laboratory test and to make changes in anticoagulation management. In 2014, we integrated a new custom HIT Order-Set (HITOS) with the EPIC electronic medical record (EMR) system. HITOS requires a clinician seeking to order a HIT laboratory test to input data to calculate a 4Ts pretest likelihood score for HIT. Recommendations are then automatically provided for proceeding or not with the HIT test and for anticoagulant management. Specifically, for patients with a low likelihood of HIT, HITOS recommends continuing heparin without HIT testing. For patients with an intermediate or high likelihood, HITOS recommends discontinuing heparin in favor of an alternative anticoagulant and proceeding with HIT testing. Our study goals were to evaluate the effect of HITOS on patterns of HIT test ordering and on anticoagulation management in patients with suspected HIT.

Methods: We performed a retrospective review of all inpatient charts in which an order for the laboratory test for HIT (heparin-PF4 antibody enzyme immunoassay) was recorded at Dartmouth Hitchcock Medical Center from December 2012 to December 2016. HITOS was launched in December 2014. Subjects were excluded from the study if their charts contained incomplete data, external laboratory results, previous documentation of a HIT diagnosis or duplicate testing . Data were extracted from the EMR for each study subject and included demographic information, clinical data, HIT test results and details of anticoagulation management before and after HIT laboratory testing. In the pre-HITOS group, the 4Ts score was applied retrospectively to assess the likelihood of HIT and was calculated independently by two study members. A score of ≤ 3 defined low likelihood for HIT; 4-5 points, intermediate; 6-8 points, high. The 4Ts score was obtained from the EMR in the post-HITOS group. Before and after HITOS comparisons were analyzed using the chi-square test for independence with a = 0.05.

Results: A total of 465 EMR were reviewed; 401 subjects were included in the final analysis. The pre and post-HITOS groups included 213 and 189 subjects, respectively, with a similar median age (66 years, range 21 to 94 and 65 years, range 23 to 94). Pre-HITOS, a HIT test was ordered in 54% of low, 39% of intermediate and 7% of high likelihood cases. Post-HITOS, a HIT test was ordered in 30%, 57% and 14% of the respective cases (p <0.001 for each pre/post comparison). When there was at least an intermediate clinical likelihood for HIT, appropriate discontinuation of heparin at time of HIT testing occurred more frequently in the post-HITOS group (74% vs 66 %, respectively; p <0.001), and more patients in the post-HITOS group were accorded a clinical diagnosis of HIT than in the pre-HITOS group (11.1% vs 5.6%, respectively; p <0.05) group.

Conclusions: Implementation of HITOS was associated with a decrease in HIT laboratory test ordering if the clinical likelihood for HIT was low and was associated with a higher rate of appropriate heparin discontinuation when the clinical likelihood of HIT was intermediate or high. Our results suggest that integration of HITOS may help refine the diagnostic process in patients with suspected HIT, though additional study will be required to evaluate the cost-effectiveness and effects on morbidity and mortality of this approach.