A machine-learning model for reducing misdiagnosis in heparin-induced thrombocytopenia: a prospective, multicenter, observational study

Nilius, Henning; Cuker, Adam; Haug, S.; Nakas, Christos T.; Studt, Jan‐Dirk; Tsakiris, Dimitrios Α.; Greinacher, Andreas; Méndez, Adriana; Schmidt, Adrian; Wuillemin, Walter A.; Gerber, Bernhard; Hovinga, Johanna A. Kremer; Vishnu, Prakash; Graf, Lukas; Kashev, Alexander; Sznitman, Raphael; Bakchoul, Tamam; Nagler, Michael

doi:10.1016/j.eclinm.2022.101745

Cited by 14 publications

(21 citation statements)

References 104 publications

(143 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…A major issue seen in the normal diagnostic approach for DITPs is that misdiagnosis is common due to lack of diagnostic data utilized [ 46 , 47 ]. A study by Nilius and others utilized ML algorithms that integrate clinical and laboratory information to diagnose HIT more accurately than the traditional approach by the American Society of Hematology (ASH) [ 15 ]. A total of 1393 patients with suspected HIT were included in their study.…”

Section: Resultsmentioning

confidence: 99%

“…While creating the model, the researchers accounted for the practical application of the model by ensuring that all variables used were readily available, timely, consistent, and easy to collect. Backwards stepwise selection was used to select the most important variables in the model [ 15 ].…”

Section: Resultsmentioning

confidence: 99%

“…The specificity was 95% for all models. These performance matrices outperformed the current metrics for the recommended algorithm for the diagnosis of HIT, and hence, a new algorithm was suggested by the authors ( Figure 4 ) [ 15 ].…”

Section: Resultsmentioning

confidence: 99%

“… Proposed diagnostic algorithm for heparin induced thrombocytopenia: TORADI−HITP (Nilius, H. et al, 2022) [ 15 ]. …”

Section: Figurementioning

confidence: 99%

See 3 more Smart Citations

Applications of Artificial Intelligence in Thrombocytopenia

et al. 2023

View full text Add to dashboard Cite

Thrombocytopenia is a medical condition where blood platelet count drops very low. This drop in platelet count can be attributed to many causes including medication, sepsis, viral infections, and autoimmunity. Clinically, the presence of thrombocytopenia might be very dangerous and is associated with poor outcomes of patients due to excessive bleeding if not addressed quickly enough. Hence, early detection and evaluation of thrombocytopenia is essential for rapid and appropriate intervention for these patients. Since artificial intelligence is able to combine and evaluate many linear and nonlinear variables simultaneously, it has shown great potential in its application in the early diagnosis, assessing the prognosis and predicting the distribution of patients with thrombocytopenia. In this review, we conducted a search across four databases and identified a total of 13 original articles that looked at the use of many machine learning algorithms in the diagnosis, prognosis, and distribution of various types of thrombocytopenia. We summarized the methods and findings of each article in this review. The included studies showed that artificial intelligence can potentially enhance the clinical approaches used in the diagnosis, prognosis, and treatment of thrombocytopenia.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

Section: Resultsmentioning

confidence: 99%

“… Proposed diagnostic algorithm for heparin induced thrombocytopenia: TORADI−HITP (Nilius, H. et al, 2022) [ 15 ]. …”

Section: Figurementioning

confidence: 99%

See 2 more Smart Citations

Applications of Artificial Intelligence in Thrombocytopenia

et al. 2023

View full text Add to dashboard Cite

show abstract

“…8,9 Our group recently published a machine learning-based algorithm for the diagnosis of HIT that was derived from a prospective, multicentre cohort of patients with suspected HIT in clinical practice. 10 The model was implemented as a web application to facilitate its use (https://torad i-hit.org).…”

mentioning

confidence: 99%

Next‐generation diagnostic instruments in haematology

Nagler

Nilius

2023

Br J Haematol

Self Cite

View full text Add to dashboard Cite

Gallo et al. assessed the impact of implementing a clinical decision support tool in a multi‐hospital setting. This is one of three important points to unlock the capabilities of the next‐generation of diagnostic instruments: (a) integrating diagnostic information from various sources, (b) ensuring accurate development and validation in well‐designed clinical studies and (c) seamlessly integrating them into clinical practice.Commentary on: Gallo et al. Clinical decision support to reduce unnecessary diagnostic testing for heparin‐induced thrombocytopenia. Br J Haematol 2023 (Online ahead of print). doi: 10.1111/bjh.18902.

show abstract

Accuracy of Diagnosing Heparin-Induced Thrombocytopenia

Larsen,

Nilius,

Studt

et al. 2024

JAMA Netw Open

Self Cite

View full text Add to dashboard Cite

ImportanceHeparin-induced thrombocytopenia (HIT) is a life-threatening condition that requires urgent diagnostic clarification. However, knowledge of the diagnostic utility of the recommended diagnostic tests is limited in clinical practice.ObjectiveTo evaluate the current diagnostic practice for managing the suspicion of HIT.Design, Setting, and ParticipantsThis prospective diagnostic study was conducted from January 2018 to May 2021 among consecutive patients with suspected HIT from 11 study centers in Switzerland, Germany, and the United States. Detailed clinical data and laboratory information were recorded. Platelet factor 4/heparin antibodies were quantified using an automated chemiluminescent immunoassay (CLIA). A washed-platelet heparin-induced platelet activation (HIPA) test was used as a reference standard to define HIT.ExposuresSuspicion of HIT.Main Outcomes and MeasuresThe primary outcome was the diagnostic accuracy of the 4Ts score, the CLIA, and the recommended algorithm serially combining both tests.ResultsOf 1448 patients included between 2018 and 2021, 1318 were available for the current analysis (median [IQR] age, 67 [57-75] years; 849 [64.6%] male). HIPA was positive in 111 patients (prevalence, 8.4%). The most frequent setting was intensive care unit (487 [37.0%]) or cardiovascular surgery (434 [33.0%]). The 4Ts score was low risk in 625 patients (46.8%). By 2 × 2 table, the numbers of patients with false-negative results were 10 (9.0%; 4Ts score), 5 (4.5%; CLIA), and 15 (13.5%; recommended diagnostic algorithm). The numbers of patients with false-positive results were 592 (49.0%; 4Ts score), 73 (6.0%; CLIA), and 50 (4.1%; recommended diagnostic algorithm), respectively.Conclusions and RelevanceIn this diagnostic study of patients suspected of having HIT, when the recommended diagnostic algorithm was used in clinical practice, antibody testing was required in half the patients. A substantial number of patients were, however, still misclassified, which could lead to delayed diagnosis or overtreatment. Development of improved diagnostic algorithms for HIT diagnosis should be pursued.

show abstract

A machine-learning model for reducing misdiagnosis in heparin-induced thrombocytopenia: a prospective, multicenter, observational study

Cited by 14 publications

References 104 publications

Applications of Artificial Intelligence in Thrombocytopenia

Applications of Artificial Intelligence in Thrombocytopenia

Next‐generation diagnostic instruments in haematology

Accuracy of Diagnosing Heparin-Induced Thrombocytopenia

Contact Info

Product

Resources

About