Abstract Number: PB1011
Meeting: ISTH 2020 Congress
Theme: Hemophilia and Rare Bleeding Disorders » Hemophilia - Clinical
Background: Hemophilia A is an inherited bleeding disorder. Because of rarity of the condition, large populations are needed to identify adequate numbers of patients. Administrative databases offer large populations for study, although validation of the coding accuracy is essential.
Aims: The aim of this study was to evaluate the accuracy of using a combination of International Classification of Diseases (ICD) diagnostic codes and coagulation factor level for identifying hemophilia A within administrative databases.
Methods: Our baseline population was people of all ages who were tested for factor VIII (FVIII) activity in the province of Alberta, Canada between 2004-2019. To calculate the validity of our algorithm (ICD codes for hemophilia A and FVIII activity < 0.4) , we aimed for a sample size which will allow us to detect high sensitivity and specificity. Medical charts of randomly selected patients were reviewed to confirm true cases of hemophilia A. Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV) were calculated.
Results: Of 11053 patients who had FVIII testing, 2114 were randomly sampled for chart review. The mean age was 39 years and 71% were female. According to chart review, 62 (2.9%) patients have a confirmed diagnosis of hemophilia A. With our predefined algorithm, 72 (3.4%) patients met the combined ICD code and laboratory criteria. The sensitivity, specificity, PPV and NPV of our algorithm were 93.5%, 99.3%, 80.5% and 99.8% respectively (Table 1). Our diagnostic algorithm has excellent discriminant properties (c-statistic 0.964; Figure 1) and yielded similar results between pediatric and adult populations.
Conclusions: When combined with low (<0.4) coagulation factor level, ICD diagnostic codes obtained from administrative data provide a highly sensitive and specific method to ascertain cases of hemophilia A .
| Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) | c-statistic (from ROC analysis) | |
| ICD code alone | 95.2 (85.6- 98.7) | 95.8 (94.8-96.6) | 40.4 (32.4-48.8) | 99.8 (99.5-99.9) | 0.96 |
| ICD code + low coagulation FVIII | 93.5 (83.5-97.9 ) | 99.3 (98.8-99.6) | 80.5 (69.1-88.5 ) | 99.8 (99.4 -99.9) | 0.964 |
[Diagnostic accuracy of ICD code alone vs ICD code + coagulation FVIII <0.4]
[ROC curve of diagnostic algorithm using ICD code + coagulation FVIII <0.4]
To cite this abstract in AMA style:
Alam AU, Karkhaneh M, Wu C, Sun H. Validation of an Algorithm to Identify Hemophilia A Patients in Administrative Database [abstract]. Res Pract Thromb Haemost. 2020; 4 (Suppl 1). https://abstracts.isth.org/abstract/validation-of-an-algorithm-to-identify-hemophilia-a-patients-in-administrative-database/. Accessed December 10, 2023.« Back to ISTH 2020 Congress
ISTH Congress Abstracts - https://abstracts.isth.org/abstract/validation-of-an-algorithm-to-identify-hemophilia-a-patients-in-administrative-database/