OC 69.5 - An integrative approach to diagnose heparin-induced thrombocytopenia: development, validation, and implementation of a multivariable prediction model

Tuesday, July 12, 2022

3:45 PM – 4:00 PM

Location: ExCel Center, ICC Capital Suite Room 1

Presentation Format: In-Person

Co-Author(s)

HN

Henning Nilius

PhD Student & Resident
Department of Clinical Chemistry, Inselspital, Bern University Hospital, and University of Bern, Bern, Switzerland
Bern, Bern, Switzerland
Adam Cuker, MD, MS

Associate Professor
Department of Medicine and Department of Pathology & Laboratory Medicine, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA
Philadelphia, Pennsylvania, United States
SH

Sigve Haug

Professor
University of Bern
Bern, Bern, Switzerland
CN

Christos Nakas

Professor
University of Bern, University of Thessaly
Volos, Thessaloniki, Greece
JS

Jan-Dirk Studt

Senior Physician
Division of Medical Oncology and Hematology, University Hospital Zurich
Zurich, Zurich, Switzerland
DT

Dimitrios A. Tsakiris

Professor
Basel University Hospital
Basel, Basel-Stadt, Switzerland
AG

Andreas Greinacher

MD, Head of
Institut für Transfusionsmedizin, Universitätsmedizin Greifswald, Germany
Greifswald, Mecklenburg-Vorpommern, Germany
AM

Adriana Mendez

Senior Physician
Department of Laboratory Medicine, Kantonsspital Aarau
Aarau, Aargau, Switzerland
AS

Adrian Schmidt

Consultant
City Hospital Waid and Triemli
Zurich, Zurich, Switzerland
WW

Walter A. A. Wuillemin

Professor
Division of Hematology and Central Hematology Laboratory, Cantonal Hospital of Lucerne and University of Bern
Lucerne, Luzern, Switzerland
BG

Bernhard Gerber

Senior Physician
Clinic of Hematology, Oncology Institute of Southern Switzerland
Bellinzona, Ticino, Switzerland
JK

Johanna A. Kremer-Hovinga

Professor
Centre of Laboratory Medicine Section Hemostasis Inselspital Bern
Bern, Bern, Switzerland
PV

Prakash Vishnu

Professor
CHI Franciscan Medical Group
Seattle, Washington, United States
LG

Lukas Graf

Senior Physician
Center for Laboratory Medicine, Hemophilia- and Hemostasis Center St. Gallen
St. Gallen, Sankt Gallen, Switzerland
AK

Alexander Kashev

Research associate
University of Bern
Bern, Bern, Switzerland
RS

Raphael Sznitman

Professor
University of Bern
Bern, Bern, Switzerland
Tamam Bakchoul, MD

Medical Director
Center for Clinical Transfusion Medicine, University Hospital of Tübingen
Tübingen, Baden-Wurttemberg, Germany

Presenting Author(s)

Michael Nagler, MD, PhD, MSc

Deputy Head
Department of Clinical Chemistry, Inselspital, Bern University Hospital, and University of Bern, Bern, Switzerland
Bern, Bern, Switzerland

Background: Interpreting clinical information and laboratory test results to diagnose heparin-induced thrombocytopenia (HIT) is challenging, particularly for the inexperienced.

Aims: We aimed to develop, validate, and implement an easy-to-use multivariable diagnostic prediction model integrating clinical information and laboratory test results.

Methods: We conducted a prospective cohort study including 1’448 patients with suspected HIT from 11 study centers. Detailed clinical information was collected, and various immunoassays were conducted. The washed platelet heparin-induced platelet activation assay (HIPA) served as the reference standard. Using 75% of the patients, five different machine-learning models were trained per immunoassay and internally validated on the remaining 25%.

Results: Sufficient sample material was available in 1’393 individuals. HIPA was positive in 119 patients (prevalence 8.5%). The following variables were selected for the final models: (1) immunoassay test result (either ELISA, chemiluminescent immunoassay [CLIA], or particle-gel immunoassay [PaGIA]), (2) platelet nadir, (3) unfractionated heparin use, (4) CRP, (5) timing of thrombocytopenia, and (6) the likelihood of other causes of thrombocytopenia. The c-statistic of all models in the validation dataset was 0.99 (95%CI: 0.97, 1.00). Compared to the currently recommended diagnostic algorithm (4Ts score, immunoassay), the number of false-negative patients was reduced by 64.3% (CLIA), 66.6% (PaGIA), and 45.5% (ELISA). False-positive individuals were reduced by 29.2% (CLIA), 72.1% (PaGIA), and 53.1% (ELISA).

Conclusion(s): We developed, validated, and implemented an accurate and easy-to-use decision support tool for the diagnosis of HIT (http://www.toradi-hit.org), which has the potential to reduce overtreatment and delayed diagnosis.