BOSTON—An electronic version of the Expanded Disability Status Scale (EDSS), the eEDSS, an algorithm-based, tablet-driven tool, is superior to the standard paper version in providing consistent assessments, according to the results of a validation study presented at the 2014 Joint ACTRIMS–ECTRIMS Meeting.
“By reducing noise, the eEDSS should have a positive impact on the accuracy of individual assessments and statistical power in clinical trials,” said Marcus D’Souza, MD, from the Department of Neurology, University Hospital, Basel, Switzerland, and colleagues.
The EDSS is the most widely used disability rating scale in multiple sclerosis (MS), but has been criticized for its low reliability. To improve reliability, a standardized version of the EDSS (Neurostatus) was established during the last two decades and is used in most major clinical studies, including many of the pivotal trials for regulatory approval of MS drugs.
To further improve EDSS reliability, Dr. D’Souza and colleagues developed the electronic capturing and real-time feedback EDSS (eEDSS), an algorithm-based evaluation performed with a tablet computer (iPad). The eEDSS provides immediate feedback, based on the Neurostatus definitions, indicating inconsistencies in the scoring of parts of the neurologic examination, functional systems, ambulation score, and EDSS. The final decision on the correct scores remains with the physician, who can overrule the algorithm. To validate their tool, the researchers compared results obtained with the standardized EDSS paper version (pEDSS) and the eEDSS.
Three certified EDSS raters who received training in the use of the eEDSS examined 100 patients with MS with varying degrees of neurologic deficits. Each patient was assessed twice. The first rater began by using the pEDSS and then switched to the eEDSS after seeing a third of the study population.
In parallel, the second EDSS rater assessed the same patients on the same day using the eEDSS and switched accordingly. pEDSS and eEDSS were compared for the frequency of inconsistencies in the assignment of the functional systems, ambulation score, and EDSS using generalized mixed-effects models and the exact McNemar test. Calculation of functional system scores based on Neurostatus Subscores were then checked by a third EDSS rater who was blinded for type of assessment, EDSS examiner, and patient.
For all examined end points, errors were more likely when using the pEDSS, compared with using the eEDSS. Examiners were three times as likely to make an error of any type when using the pEDSS, compared with using eEDSS (odds ratio, 2.93). Functional systems errors were more likely when pEDSS was used, compared with eEDSS (odds ratio, 2.54).
Errors were much more likely to occur for patients with low EDSS scores (3.5 or lower) than for those with high EDSS scores (3.5 or higher), with an odds ratio of 5.32. Errors in calculation of functional systems after exclusion of the cerebral functional system were much more likely when pEDSS was used, compared with eEDSS (odds ratio, 3.107). Ambulation score errors and EDSS errors were rare, and the differences between the pEDSS and eEDSS in the rates of these errors were not statistically significant.
—Glenn S. Williams