When HIT Testing Hurts- the Financial and Clinical Repercussions of Heparin Induced Thrombocytopenia Testing: A Single Center Analysis

Introduction: Thrombocytopenia is a common inpatient hematologic concern. Most often, thrombocytopenia is a multifactorial phenomenon, without one clear cause. Heparin Induced Thrombocytopenia (HIT) is a rare but serious hematologic consequence of exposure to heparin. In many hospitalized patients, heparin is routinely used as venous thromboembolism (VTE) prophylaxis or for therapeutic anticoagulation.

The incidence of HIT is approximately 0.5-1.0% in patients exposed to unfractionated heparin (UFH). Though rare, there seems to be over-testing of HIT antibody in the inpatient setting. As such, there is potential for high cost with low yield and false positives that alter patient management, resulting in prolonged hospital stays and further complications in patient care.

Methods: This was a retrospective, observational study reviewing the indications, results, and patient characteristics of cases in which HIT testing was ordered. Indications for testing and influence of results on management were determined based on clinical documentation. The tests of interest included HIT antibody testing and Seratonin Release Assay (SRA). We analyzed how many times a HIT antibody was positive with a negative result in the SRA. We also determined whether there was an increased length of stay in patients who had HIT testing ordered. The study population included patients who were admitted to University of Florida Shands Medical Center, an academic institution in Jacksonville, FL in 2022. All patients were >18 years of age.

Results: 190 patients were identified, for whom 190 HIT antibody tests were ordered and 75 SRAs. All patients who had a SRA performed had exposure to low dose heparin or therapeutic high dose heparin. Out of 190 patients, HIT testing was positive in 30 patients (15.8%). SRA testing was performed on 75 patients despite only 30 positive HIT tests. The true positives were 10 patients who had positive SRA confirmatory testing out of 190 patients with suspected HIT. Based on the results of confirmatory testing, the false positive fraction of HIT testing was 66%.

When analyzing the 190 patients who underwent HIT testing, the average length of stay was 14 days. On average, a hospital stay in the US is 5.5 days and evaluation of HIT results in 8.5 additional days of hospital admission at our institution. The prolonged length of stay for HIT testing is likely impacted by awaiting SRA results along with other factors.

Conclusion:

This analysis examines the relative over-testing for HIT at a single institution. Compared with the known incidence of positive HIT tests, there is a significant number of positive HIT antibody tests at our hospital. The data further reveals a high probability of a false positive HIT antibody test results in the hospitalized patient, possibly due to immune reactivity. Of 190 patients suspected with HIT, only 10 patients had a positive confirmatory SRA, indicating that only 5% of suspected patients truly had HIT. This demonstrates that 95% of the time HIT assays ordered did not provide clinically useful information. From the time work up was initiated to the time of discharge, the average stay length was 14 days in a patient suspected with HIT. Cost burden of HIT testing was evaluated. Each HIT panel is $153.83 and each confirmatory SRA test is approximately $250, at a commercial laboratory. Taken together, the total average cost per patient with suspected HIT is $413.83 in diagnostic testing. Prolonged hospital stay contributes to approximately $25,000 increase in hospital stay per patient being evaluated for HIT. This does not include the cost of further work up or interventions. Changing to an alternate anticoagulant such as Argatroban increased costs by nearly $1,500 per case and this number increased further when adding the daily coagulation panel required while on this drug.

In other studies, this cost has been estimated to add $90,000 to the hospital stay. Prolonged hospital stays are also associated with unintended incidents such as GI bleeding leading to blood transfusions, invasive testing (such as endoscopy/colonoscopy) and increased morbidity to the patient. We will use this information to implement a tool in our Electronic Medical Record (EMR) that will require answering the 4T score in order to establish high pre-test probability, prior to ordering HIT antibody testing at the University of Florida in Jacksonville.