User login
Key clinical point: Blood cell counts collected up to 5 years before chronic myeloid leukemia (CML) diagnosis could predict BCR-ABL1 test using machine learning methods. Such predictive models may enable early diagnosis of CML, and subsequently, earlier treatment initiation, leading to a better prognosis.
Major finding: The BCR-ABL1 positivity rate was 6.2%. The ability of machine learning models to predict CML diagnosis improved with the usage of blood cell counts closer to the time of diagnosis (at diagnosis: area under the curve [AUC], 0.87-0.96; 6 months to 1-year prediagnosis: AUC, 0.75-0.80; 2-5 years prediagnosis: AUC, 0.59-0.67).
Study details: This study included 1,623 patients with a BCR-ABL1 test and at least 6 consecutive prior years of differential blood cell counts between October 1999 and April 2020 from the Veterans Health Administration database.
Disclosures: No source of funding or author disclosures were reported.
Source: Hauser RG et al. Am J Clin Pathol. 2021 Jun 29. doi: 10.1093/ajcp/aqab086.
Key clinical point: Blood cell counts collected up to 5 years before chronic myeloid leukemia (CML) diagnosis could predict BCR-ABL1 test using machine learning methods. Such predictive models may enable early diagnosis of CML, and subsequently, earlier treatment initiation, leading to a better prognosis.
Major finding: The BCR-ABL1 positivity rate was 6.2%. The ability of machine learning models to predict CML diagnosis improved with the usage of blood cell counts closer to the time of diagnosis (at diagnosis: area under the curve [AUC], 0.87-0.96; 6 months to 1-year prediagnosis: AUC, 0.75-0.80; 2-5 years prediagnosis: AUC, 0.59-0.67).
Study details: This study included 1,623 patients with a BCR-ABL1 test and at least 6 consecutive prior years of differential blood cell counts between October 1999 and April 2020 from the Veterans Health Administration database.
Disclosures: No source of funding or author disclosures were reported.
Source: Hauser RG et al. Am J Clin Pathol. 2021 Jun 29. doi: 10.1093/ajcp/aqab086.
Key clinical point: Blood cell counts collected up to 5 years before chronic myeloid leukemia (CML) diagnosis could predict BCR-ABL1 test using machine learning methods. Such predictive models may enable early diagnosis of CML, and subsequently, earlier treatment initiation, leading to a better prognosis.
Major finding: The BCR-ABL1 positivity rate was 6.2%. The ability of machine learning models to predict CML diagnosis improved with the usage of blood cell counts closer to the time of diagnosis (at diagnosis: area under the curve [AUC], 0.87-0.96; 6 months to 1-year prediagnosis: AUC, 0.75-0.80; 2-5 years prediagnosis: AUC, 0.59-0.67).
Study details: This study included 1,623 patients with a BCR-ABL1 test and at least 6 consecutive prior years of differential blood cell counts between October 1999 and April 2020 from the Veterans Health Administration database.
Disclosures: No source of funding or author disclosures were reported.
Source: Hauser RG et al. Am J Clin Pathol. 2021 Jun 29. doi: 10.1093/ajcp/aqab086.