Key points
Affiliates
Michael Benatar1, Joanne Wuu2, Sharon Usher1, and Kevin Ward3
- Department of Neurology, School of Medicine, Emory University
- Section of Neurostatistics, Department of Neurology, School of Medicine, Emory University
- Department of Epidemiology, Rollins School of Public Health, Emory University
Summary
In preparation for the development of the National ALS Registry, four pilot studies were conducted. This publication reports the findings of one such pilot study seeking to evaluate the accuracy in which existing data sources can identify ALS cases within the state of Georgia. Data were obtained from Medicare, Medicaid, Veterans Administration, Emory Healthcare, two community neurologists, the ALS Association, and death records to estimate the positive predictive value (PPV) of the ICD-9/ICD-10 codes to accurately identify ALS cases. This study concluded that existing data sources are useful for identifying cases of ALS/MND. Moreover, predictive algorithms may help identify a large proportion of ALS cases without needing to verify the diagnosis via medical chart review, as has been done with existing registries.