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

Paper Details

Abstracts, authors, affiliations, and disclosures.

Monday, December 14, 2026
10:30 a.m. - 12:00 p.m.
MCCNO - 217-219
Treatment Considerations, Biological Therapies, Immunotherapy, Monoclonal Antibody Therapy, Biological Processes, Pathogenesis, Immune Mechanism, Metabolism, Diseases, Bleeding and Clotting, Platelet Disorders, Thrombocytopenias, Immunology, Autoimmune Disorders, Research, Clinical Research, Health Outcomes Research, Study Population, Human, Adult, Pediatric
PAPER

752 Use of the platelet variability index score to improve the diagnosis of immune thrombocytopenia

Syed Mahamad, MSc, Na Li2, Sameer Parpia, PhD3, Ishac Nazy, PhD4, John Kelton, MD5, Donald Arnold, MD6
1 McMaster University, Hamilton, Ontario, Canada
2 McMaster University, Hamilton, Ontario, Canada, University of Calgary, Calgary, Alberta, Canada
3 McMaster University, Hamilton, Ontario, Canada, McMaster University, Hamilton, Ontario, Canada
4 McMaster University, Hamilton, Ontario, Canada, McMaster University, Hamilton, Ontario, Canada, McMaster University, Hamilton, Ontario, Canada
5 McMaster University, Hamilton, Ontario, Canada, McMaster University, Hamilton, Ontario, Canada
6 McMaster University, Hamilton, Ontario, Canada, McMaster University, Hamilton, Ontario, Canada
Introduction
Misdiagnosis of immune thrombocytopenia (ITP) is common among patients presenting for hematology consultation. The platelet variability index (PVI) score, a measure of thrombocytopenia severity and platelet count fluctuations over time based on 3 or more sequential platelet count measurements, may help distinguish ITP from non-immune causes of thrombocytopenia (non-ITP). We conducted a multicenter validation study of the PVI score for the initial diagnosis of ITP.

Methods
We conducted a retrospective cohort study of consecutive adult outpatients referred to hematology for evaluation of thrombocytopenia from 9 hematology clinics in Canada. Eligible patients had a platelet count <100 x109/L and at least 3 sequential platelet count measurements leading up to the initial hematology consultation. The final diagnosis of ITP or non-ITP was ascertained clinically by the treating physician at the end of follow-up as per routine practice based on all available clinical and laboratory information. A subset of patients was classified as definite ITP if they had a platelet count increase to ≥50 x109/L and at least doubling from baseline within 4 weeks of receiving high-dose intravenous immunoglobulin or corticosteroids. Ambiguous diagnoses at the end of follow-up were adjudicated by an independent committee. The PVI score was calculated for each patient at the time of initial hematology consultation centrally and blinded to the final diagnosis. The primary analysis compared PVI scores for ITP vs. non-ITP patients. A sensitivity analysis compared PVI scores for definite ITP vs. non-ITP patients. Discrimination was assessed by the area under the receiver operating characteristic curve (AUC). Sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) for PVI scores were determined with 95% confidence intervals (CI). Based on established thresholds, a PVI score of 4 was associated with a moderate likelihood of ITP.

Results
Overall, 689 patients were enrolled. Of those, 30 were excluded because the final diagnosis was unresolved (n=19); a different platelet disorder (n=9); or mild thrombocytopenia only (n=2). Thus, the analysis included 659 patients; the median age was 62 years [interquartile range (IQR) 47, 73], and 46% were female. The final diagnosis was ITP (n=389, 59.0%; primary or secondary) or non-ITP [n=270, 41.0%; due to liver disease (n=116, 43.0%), myelodysplastic syndrome (n=28, 10.4%), malignancy (n=23, 8.5%), or other causes]. Of the 389 patients with ITP, 182 (46.8%) met our criteria for definite ITP. The median lowest platelet count values leading up to the initial hematology consultation were 35 x109/L and 68 x109/L for ITP and non-ITP patients, respectively. Patients had a median of 7 (IQR 4, 14) platelet count measurements for the calculation of PVI. The median duration of follow-up was 8.3 (IQR 1.2, 23.7) months. 

For patients diagnosed as ITP, the median PVI score at the time of the initial hematology consultation was 3 (IQR 1, 4) compared with 1 (IQR 0, 2) for patients diagnosed as non-ITP (P<0.001). The AUC of the PVI score was 0.72 (95% CI, 0.68-0.76), indicating moderate discrimination for the ITP diagnosis. A PVI score ≥4 at the initial hematology consultation had a specificity of 96.3% (95% CI, 93.3-98.0%), sensitivity of 31.6% (95% CI, 27.2-36.4%), PPV of 92.5% (95% CI, 86.7-95.9%), and NPV of 49.4% (95% CI, 45.2-53.7%). The frequency of ITP according to PVI score was 44.7% for PVI score 0–2; 74.8% for PVI score 3; 94.5% for PVI score 4; and 91.0% for PVI score 5–6.

For definite ITP, the PVI score had excellent discrimination (AUC, 0.90; 95% CI, 0.87-0.93), and a PVI score ≥4 had a specificity of 96.3% (95% CI, 93.3-98.0%), sensitivity of 59.3% (95% CI, 52.1-66.2%), PPV of 91.5% (95% CI, 85.1-95.3%), and NPV of 77.8% (95% CI, 73.1-82.0%). The frequency of definite ITP according to PVI scores was 11.7% for PVI score 0–2; 62.3% for PVI score 3; 93.6% for PVI score 4; and 90.1% for PVI score 5–6. 

Conclusions
The PVI score at the time of the initial hematology consultation can inform the likelihood of the ITP diagnosis among patients with thrombocytopenia. PVI had moderate discrimination for ITP and excellent discrimination for definite ITP. A PVI score ≥4 had high specificity and high PPV. Incorporating PVI into the initial hematology consultation has the potential to reduce diagnostic delays and improve clinical outcomes for patients with thrombocytopenia.
Syed Mahamad, MSc: Sobi - Research Funding