Rationalisation of a thrombophilia panel using laboratory medicine Delphi-like consensus evaluation and secondary artificial intelligence-based simulation assessment
-
Bernhard Strasser
,
Sebastian Mustafa
Abstract
Objectives
It is important to review laboratory test panels regularly and omit unnecessary tests. This avoids overdiagnosis and makes laboratory work more targeted. Artificial intelligence is increasingly being discussed as a possible aid in such decisions. The aim of the study was to revise an existing thrombophilia panel with the help of a modified Delphi consensus of laboratory physicians and to examine whether large language models (LLMs) can mimic such decision-making processes and serve as a support tool.
Methods
The study was conducted in two steps. First, six experts evaluated various thrombophilia parameters in three Delphi rounds, assessing technical reliability and clinical significance. Selected LLMs (Elicit, Consensus, and STORM) were then tested with questions. Their results were compared with the Delphi consensus. Agreement was calculated using percentage concordance and Cohen’s κ.
Results
PAI-1 genotyping, MTHFR genotyping, homocysteine and APC resistance were removed from the standard panel, and anti-annexin-V antibodies and anti-phosphatidylserine/prothrombin antibodies were completely eliminated. The reduced panel was incorporated into routine practice, with facultative parameters remaining available as second-line tests. Agreement between the LLMs and the experts was slight when using open prompts (κ ≈ 0.25), although, with specific questions, the agreement was higher (κ 0.50–0.52). However, the LLMs did not take into account analytical and technical aspects.
Conclusions
Thrombophilia panels should be reviewed regularly to avoid the application of unnecessary tests and ensure high diagnostic quality. The Delphi process is a suitable tool for this. LLMs can provide supporting information, but are currently no substitute for the experience and consensus of medical experts.
-
Research ethics: The local Institutional Review Board deemed the study exempt from review, as no patient-related data were analyzed. All participating experts provided informed consent prior to inclusion in the Delphi process. The study was conducted in accordance with the principles of the Declaration of Helsinki.
-
Informed consent: Informed consent was obtained from all individuals included in this study.
-
Author contributions: All authors made substantial contributions to the conception and design of the work; the first draft of the manuscript was written by B.S., and S.M., E.W. and J.S. reviewed the first draft. All authors have accepted responsibility for the entire content of this manuscript and approved its submission.
-
Use of Large Language Models, AI and Machine Learning Tools: None declared.
-
Conflict of interest: The authors state no conflict of interest.
-
Research funding: None declared.
-
Data availability: Not applicable.
References
1. Moore, GW. Thrombophilia screening: not so straightforward. Semin Thromb Hemost 2024;50:1131–52. https://doi.org/10.1055/s-0044-1786807.Search in Google Scholar PubMed
2. Middeldorp, S, Nieuwlaat, R, Baumann Kreuziger, L, Coppens, M, Houghton, D, James, AH, et al.. American society of hematology 2023 guidelines for management of venous thromboembolism: thrombophilia testing. Blood Adv 2023;7:7101–38. https://doi.org/10.1182/bloodadvances.2023010177.Search in Google Scholar PubMed PubMed Central
3. Middeldorp, S, Iorio, A. Thrombophilia testing: does practice follow guidelines? Blood Adv 2024;8:4948–9. https://doi.org/10.1182/bloodadvances.2024013478.Search in Google Scholar PubMed PubMed Central
4. Ashraf, N, Visweshwar, N, Jaglal, M, Sokol, L, Laber, D. Evolving paradigm in thrombophilia screening. Blood Coagul Fibrinolysis 2019;30:249–52. https://doi.org/10.1097/mbc.0000000000000809.Search in Google Scholar PubMed PubMed Central
5. Wang, J, Wang, C, Chen, N, Shu, C, Guo, X, He, Y, et al.. Association between the plasminogen activator inhibitor-1 4G/5G polymorphism and risk of venous thromboembolism: a meta-analysis. Thromb Res 2014;134:1241–8. https://doi.org/10.1016/j.thromres.2014.09.035.Search in Google Scholar PubMed
6. Moll, S, Varga, EA. Homocysteine and MTHFR mutations. Circulation 2015;132:e6–9. https://doi.org/10.1161/circulationaha.114.013311.Search in Google Scholar
7. Zeng, J, Zeng, Q. Correlations between methylenetetrahydrofolate reductase gene polymorphisms and venous thromboembolism: a meta-analysis of 99 genetic association studies. Eur J Prev Cardiol 2019;26:120–34. https://doi.org/10.1177/2047487318799467.Search in Google Scholar PubMed
8. Arnold, J, Holmes, Z, Pickering, W, Farmer, C, Regan, L, Cohen, H. Anti-beta 2 glycoprotein 1 and anti-annexin V antibodies in women with recurrent miscarriage. Br J Haematol 2001;113:911–4. https://doi.org/10.1046/j.1365-2141.2001.02812.x.Search in Google Scholar PubMed
9. Tonello, M, Bison, E, Cattini, MG, Pontara, E, Iaccarino, L, Denas, G, et al.. Anti-Phosphatidyl-Serine/prothrombin antibodies (aPS/PT) in isolated lupus anticoagulant (LA): is their presence linked to dual test positivity? Clin Chem Lab Med 2021;59:1950–3. https://doi.org/10.1515/cclm-2021-0692.Search in Google Scholar PubMed
10. Hoffer, EP, James, CA, Wong, A, Ranji, S. Artificial intelligence and medical diagnosis: past, present and future. Diagnosis. 2025;12:564–9. https://doi.org/10.1515/dx-2025-0111.Search in Google Scholar PubMed
11. Bridges, JM. Computerized diagnostic decision support systems – a comparative performance study of Isabel Pro vs. ChatGPT4. Diagnosis (Berl) 2024;11:250–8. https://doi.org/10.1515/dx-2024-0033.Search in Google Scholar PubMed
12. Lippi, G, Jackson, B, Plebani, M. Improving diagnosis in health care: laboratory medicine. Diagnosis 2025;12:570–7. https://doi.org/10.1515/dx-2025-0052.Search in Google Scholar PubMed
13. Ambasta, A, Holroyd-Leduc, JM, Pokharel, S, Mathura, P, Shih, AW, Stelfox, HT, et al.. Re-Purposing the ordering of routine laboratory tests in hospitalized medical patients (RePORT): protocol for a multicenter stepped-wedge cluster randomised trial to evaluate the impact of a multicomponent intervention bundle to reduce laboratory test over-utilization. Implement Sci 2024;19:45. https://doi.org/10.1186/s13012-024-01376-6.Search in Google Scholar PubMed PubMed Central
14. Dube, K, Wu, B. Supporting clinical laboratory test-ordering protocol specification, execution and management: an event-condition-action rule and database approach. Health Inf J 2001;7:20–8. https://doi.org/10.1177/146045820100700105.Search in Google Scholar
15. Plebani, M, Cadamuro, J, Vermeersch, P, Jovicic, S, Ozben, T, Trenti, T, et al.. A vision to the future: value-based laboratory medicine. Clin Chem Lab Med 2024;62:2373–87. https://doi.org/10.1515/cclm-2024-1022.Search in Google Scholar PubMed
16. Salinas, M, Torreblanca, R, Sanchez, E, Blasco, A, Flores, E, Lopez-Garrigos, M. Managing laboratory test ordering: a challenge in the new laboratory medicine model. Adv Lab Med 2024;5:236–47. https://doi.org/10.1515/almed-2024-0085.Search in Google Scholar PubMed PubMed Central
17. Hanly, JG, Smith, SA, Anderson, D. Inhibition of annexin V binding to cardiolipin and thrombin generation in an unselected population with venous thrombosis. J Rheumatol 2003;30:1990–3.Search in Google Scholar
18. Varga, EA, Sturm, AC, Misita, CP, Moll, S. Cardiology patient pages. Homocysteine and MTHFR mutations: relation to thrombosis and coronary artery disease. Circulation 2005;111:e289–93. https://doi.org/10.1161/01.CIR.0000165142.37711.E7.Search in Google Scholar PubMed
19. Simone, B, De Stefano, V, Leoncini, E, Zacho, J, Martinelli, I, Emmerich, J, et al.. Risk of venous thromboembolism associated with single and combined effects of factor V Leiden, Prothrombin 20210A and methylenetethraydrofolate reductase C677T: a meta-analysis involving over 11,000 cases and 21,000 controls. Eur J Epidemiol 2013;28:621–47. https://doi.org/10.1007/s10654-013-9825-8.Search in Google Scholar PubMed PubMed Central
20. Di Minno, MN, Tremoli, E, Coppola, A, Lupoli, R, Di Minno, G. Homocysteine and arterial thrombosis: challenge and opportunity. Thromb Haemost 2010;103:942–61. https://doi.org/10.1160/th09-06-0393.Search in Google Scholar PubMed
21. Chiasakul, T, Bauer, KA. The dos, don’ts, and nuances of thrombophilia testing. Hematology Am Soc Hematol Educ Program 2023;2023:593–9. https://doi.org/10.1182/hematology.2023000491.Search in Google Scholar PubMed PubMed Central
22. Huang, G, Wang, P, Li, T, Deng, X. Genetic association between plasminogen activator inhibitor-1 rs1799889 polymorphism and venous thromboembolism: evidence from a comprehensive meta-analysis. Clin Cardiol 2019;42:1232–8. https://doi.org/10.1002/clc.23282.Search in Google Scholar PubMed PubMed Central
23. Herskovits, AZ, Morgan, EA, Lemire, SJ, Lindeman, NI, Dorfman, DM. An improved algorithm for activated protein C resistance and factor V Leiden screening. Am J Clin Pathol 2013;140:379–86. https://doi.org/10.1309/ajcpo6vup3mzeywu.Search in Google Scholar
24. Favaloro, EJ, Arunachalam, S, Dean, E, Salwa, M, Ahuja, M, Connelly, L, et al.. Activated protein C resistance testing: an update from Australasia/Asia-Pacific. Int J Lab Hematol 2025;47:720–9. https://doi.org/10.1111/ijlh.14447.Search in Google Scholar PubMed PubMed Central
25. Rodeghiero, F, Tosetto, A. Activated protein C resistance and factor V Leiden mutation are independent risk factors for venous thromboembolism. Ann Intern Med 1999;130:643–50. https://doi.org/10.7326/0003-4819-130-8-199904200-00004.Search in Google Scholar PubMed
26. Tripodi, A, Chantarangkul, V, Martinelli, I, Bucciarelli, P, Mannucci, PM. A shortened activated partial thromboplastin time is associated with the risk of venous thromboembolism. Blood 2004;104:3631–4. https://doi.org/10.1182/blood-2004-03-1042.Search in Google Scholar PubMed
27. Siriez, R, Dogne, JM, Gosselin, R, Laloy, J, Mullier, F, Douxfils, J. Comprehensive review of the impact of direct oral anticoagulants on thrombophilia diagnostic tests: practical recommendations for the laboratory. Int J Lab Hematol 2021;43:7–20. https://doi.org/10.1111/ijlh.13342.Search in Google Scholar PubMed
28. Carobene, A, Cadamuro, J, Frans, G, Goldshmidt, H, Debeljak, Z, Bruyne, SD, et al.. EFLM checklist for the assessment of AI/ML studies in laboratory medicine: enhancing general medical AI frameworks for laboratory-specific applications. Clin Chem Lab Med 2026;64:27–40. https://doi.org/10.1515/cclm-2025-0841.Search in Google Scholar PubMed
29. Trulson, I, Holdenrieder, S, Hoffmann, G. Using machine learning techniques for exploration and classification of laboratory data. J Lab Med 2024;48:203–14. https://doi.org/10.1515/labmed-2024-0100.Search in Google Scholar
30. Yesil, MR, Talli, I, Pelloso, M, Cosma, C, Pangrazzi, E, Plebani, M, et al.. Impact of analytical bias on machine learning models for sepsis prediction using laboratory data. Clin Chem Lab Med 2025;63:2022–30. https://doi.org/10.1515/cclm-2025-0491.Search in Google Scholar PubMed
31. Niu, Z, Kuang, X, Chen, J, Cai, X, Zhang, P. The application and challenges of ChatGPT in laboratory medicine. Adv Lab Med/Avances en Medicina de Laboratorio 2025;6:385–96. https://doi.org/10.1515/almed-2025-0080.Search in Google Scholar PubMed PubMed Central
32. Sele, D, Chugunova, M. Putting a human in the loop: increasing uptake, but decreasing accuracy of automated decision-making. PLoS One 2024;19:e0298037. https://doi.org/10.1371/journal.pone.0298037.Search in Google Scholar PubMed PubMed Central
Supplementary Material
This article contains supplementary material (https://doi.org/10.1515/dx-2025-0144).
© 2026 Walter de Gruyter GmbH, Berlin/Boston
