Abstract
Introduction
Studies have shown that medical centers are faced with a major challenge in making an accurate diagnosis of heparin -induced thrombocytopenia (HIT). Some centers have decreased overdiagnosis with the implementation of anticoagulation stewardship programs, collaboration between pharmacy and hospital laboratory, and mandating clinicians to calculate the 4T score prior to testing. A high frequency of PF4-ELISA testing in patients with low 4T scores has been observed in multiple medical centers, including our institution. Education of ordering clinicians has been suggested to reduce inappropriate testing.
In this study, we determined current trends of PF4-ELISA testing in our institution after the introduction of a HIT program that included ongoing education, feedback, and continuous clinical audit of PF4-ELISA utilization.
Methods
A HIT Program was developed over a 6-year period. This program consisted of: i) providing department-wide lectures to residents, fellows, and attendings on HIT testing and management including 4T score calculation; ii) providing individual feedback from the Anticoagulation and Bleeding Management service to the ordering clinician when a false positive PF4-ELISA test was ordered inappropriately; iii) providing interactive lectures based on real life case scenarios to internal medicine residents on calculating 4T scores.
To assess the impact of the education provided on PF4-ELISA testing trends, we conducted a prospective cohort review of all adult patients (age≥18) who had a PF4-ELISA ordered over a 3-month period (the last quarter of the academic year, March - June 2018) at our academic medical center (an 800-bed hospital). The Hematology Quality Improvement (QI) fellow and attending received a list of all PF4-ELISA tests ordered from the laboratory technician daily. All PF4-ELISA tests requested prior to 12pm are resulted on the same day in the afternoon. 4T scores were independently assessed prior to the results being available to avoid bias. Testing trends, 4T scores, ordering services, and documentation of 4T scores by ordering clinicians were assessed.
Results
72 PF4-ELISA tests were ordered during the study period. Prospectively calculated 4T scores by investigators revealed 60 low-risk (83.3%), 9 intermediate-risk (12.5%), and 3 high-risk (4.16%). 1 patient had a positive serotonin-release assay (SRA); optical density (O.D.) of the PF4-ELISA was 2.499 and 4T score revealed intermediate risk. The 4T score was documented by the ordering team in only 7 cases; 5 of these revealed discordant calculation between the ordering clinician and the Hematology QI team. All documented 4T scores were by Internal Medicine services. In all discordant cases, ordering clinicians' scores were higher than scores calculated by the Hematology QI team. The majority of PF4-ELISA testing was ordered by the intensive care units (ICUs) (n=32, 44.44%), followed by medical (n=20, 27.77%) and surgical teams (n=20, 27.77%)
Conclusion
Our study revealed that calculation of 4T scores remains poor and is not universally applied, suggesting that clinician education alone is insufficient. We also observed divergent 4T scores with the ordering clinician calculating a higher 4T score compared to the Hematology QI team. As such, introducing mandatory 4T score calculation prior to PF4-ELISA testing may not be helpful as ordering clinicians can bypass the restriction through inaccurate 4T score calculation.
The majority of PF4-ELISA testing was performed in the ICU setting. This is concerning as studies have shown that thrombocytopenia is a common laboratory finding in up to 50% of ICU patients. This seems to suggest the difficulties that ordering clinicians have in distinguishing the multiple causes of thrombocytopenia in ICU patients.
Further study is required to assess if initiating an automatic non-malignant hematology consult for each PF4 testing would be an effective addition to our continued attempts to decrease inappropriate testing for HIT.
No relevant conflicts of interest to declare.
