2026 ASH®
Annual Meeting & Exposition December 12-15, 2026 ▪ New Orleans, Louisiana

Paper Details

Abstracts, authors, affiliations, and disclosures.

Monday, December 14, 2026
06:00 p.m. - 08:00 p.m.
MCCNO - Hall F
Access, Representation, Population Health, Technology and Procedures, Blood Banking, Clinical Procedures, Treatment Considerations, Combination Therapy, Drug Development, Drug-Drug Interactions, Supportive Care, Therapy Sequence, Biological Therapies, Transfusion, Vaccines, Non-Biological Therapies, Opioids, Women’s Health, Adverse Events, Biological Processes, Hematopoiesis, Metabolism, Clinical Practice (Health Services and Quality), Patient Safety, Quality Improvement, Quality of Care, Systems-Based Hematology, Diseases, Bleeding and Clotting, Thromboembolism, Thrombotic Disorders, Genetic Disorders, Hemoglobinopathies, Sickle Cell Disease, Sickle Cell Trait, Thalassemia, Immunology, Immunodeficiency, Infectious Diseases, Bacterial, Research, Clinical Research, Clinical Trials, Epidemiology, Health Outcomes Research, Health Disparities Research, Patient-Reported Outcomes, Real-World Evidence, Registries, Study Population, Human, Adult, Pediatric, Young Adult
PAPER

4701 An algorithm for interpretation of hemoglobin fractionation to improve sickle cell disease diagnosis and phenotype classification

Julie Kanter, MD, Najibah Galadanci, PhD, MBBS, MPH2, Gerhard Hellemann, PhD3
1 University of Alabama at Birmingham, Birmingham, Alabama, United States
2 University of Alabama at Birmingham, Birmingham, Alabama, United States
3 University of Alabama at Birmingham, Birmingham, Alabama, United States

Background: Sickle cell disease (SCD) is a group of inherited hemoglobin (Hb) disorders caused by the presence of sickle Hb (HbS) either alone or in combination with another abnormal β-globin. Inheritance of a single HbS gene in combination with a normal adult hemoglobin (HbA), results in sickle cell trait (HbAS). Several laboratory methods diagnose SCD by separating and quantifying Hb fractions as percentages of total Hb, including high-performance liquid chromatography, isoelectric focusing, and cellulose acetate/citrate agar electrophoresis. Because SCD is clinically heterogenous and laboratory findings can vary even within the same genotype, interpretation of these tests can be challenging. Red blood cell (RBC) transfusions, which introduce donor HbA and disease-modifying therapies such as hydroxyurea that increase fetal hemoglobin (HbF) also add complexity as does co-inheritance of α-thalassemia, hereditary persistence of HbF other rare Hb variants and iron deficiency. Accurate interpretation of Hb analyses is necessary to ensure appropriate referral and management of individuals with SCD while avoiding misclassification of carriers as having SCD. To facilitate accurate diagnosis and SCD phenotype classification, our objective was to develop an algorithm to interpret Hb analysis and classify results in conjunction with the CDC-funded SCD surveillance program, ALSTATE.

Methods:   We evaluated Hb fraction results from all Hb analysis performed at the University of Alabama Birmingham over ten years (1/2016-6/2026).  The initial algorithm was designed using confirmed SCD phenotypes and iteratively refined by evaluating misclassified cases and adding decision rules to account for these cases (Shapire, 1990). Because published reference ranges describe expected Hb fractions rather than individual patient samples, algorithm development required an iterative approach.  Test sets were generated through stratified sampling from the complete dataset using inverse weighing by predicted diagnostic strata. An expert rater created gold standard classifications for each case, and these labeled test sets were used to train and refine the algorithm through successive iterations. Results were classified as HbSS (Homozygous sickle cell anemia), HbSC, HbS/β⁰ (HbS/β⁰-thalassemia) HbSβ⁺ (HbS/β+-thalassemia), HbAS (sickle cell trait), HbAC (HbC trait), HbCC (Hb C disease), or SCD-NOS (SCD with unknown phenotype) and HGB-P (not classified abnormal Hb (not SCD)).

Results:   A total of 58747 Hb analysis from 25,689 unique individuals were evaluated, with the number of analyses per individual ranging from 1 to 282.  Overall, 3543 (13.8%) individuals had more than one Hb analysis. Among the 25,689 individuals with available Hb fraction data, 22,925 (89%) had no detectable HbS, while 2764 (11%) had %HbS ranging from 0.2-97.3%.

The algorithm was serially re-configured to adequately account for treatment effects and improve classification of individuals with known SCD whose phenotype could not be determined (SCD NOS) and for persons with other abnormal Hb patterns (HGB-P).  Iterative resampling of testing sets with oversampling of cases most likely to be misclassified balanced the need of independence of the different test samples with efficient algorithm optimization.

Overall, the algorithm correctly classified an estimated 93% of samples with an estimated weighted positive predictive value of 91% and negative predictive value of 99%. Weights were derived empirically based on estimated population prevalence. Most classification errors involved overclassification as SCD NOS or misclassification of HbSB+.  As these errors were infrequent and did not result in missing anyone with SCD (i.e very low false negative rate), the results are encouraging. The development of this algorithm is ongoing with inclusion of additional data and active work on improving the decision rules. While the source dataset is large (25,785 participants) and likely to support generalizability, external validation in independent populations is needed because the algorithm was developed using data from a single location.

Conclusion:  There is a need to improve the diagnostic interpretation of laboratory tests used for hemoglobin fractionation. This algorithm may facilitate interpretation, improve identification and phenotype classification of individuals with SCD, and support timely referral for specialized care particularly in hospitals lacking a hematologist or SCD specialist.

Julie Kanter, MD: NIH - Research Funding; Sanofi - Research Funding; CDC - Research Funding; HRSA - Research Funding; BMS - Research Funding; Novo Nordisk - Research Funding; Beam Therapeutics - Research Funding; Terumo - Research Funding; Beam Therapeutics - Consultancy (Includes expert testimony); Disc Therapeutics - Consultancy (Includes expert testimony); Octapharma - Consultancy (Includes expert testimony); BMS - Consultancy (Includes expert testimony); Fulcrum - Consultancy (Includes expert testimony); Gamida Cell - Consultancy (Includes expert testimony); Merck - Consultancy (Includes expert testimony); Nuvamid - Consultancy (Includes expert testimony); Pfizer - Consultancy (Includes expert testimony); Agios - Consultancy (Includes expert testimony); Abt - Consultancy (Includes expert testimony)