Background: The McDonald criteria were introduced in 2001 as guidelines to facilitate early and accurate diagnosis of multiple sclerosis (MS). They were revised in 2005. Although validated in a number of research-focused clinical centres, their adequacy and utility in the general neurology setting is less certain.
Objective: In this study, we assessed new diagnoses of MS in our practice for compliance with both the original and the revised criteria.
Methods: We retrospectively identified new diagnoses of MS from 2001. Clinical notes and imaging were evaluated for compliance with McDonald criteria.
Results: Sixty-two patients were included: 53 with ;practice-definite' and nine with ;practice-possible' diagnoses of MS. At the time of diagnosis, 47% of the ;practice-definite' group fulfilled the 2001 criteria and 49% the revised criteria. Among patients not satisfying the criteria at time of diagnosis, 21% went on to fulfil the McDonald criteria over the 23-month follow-up.
Conclusions: There is a considerable gap between the clinical diagnosis of MS in a general neurology setting and compliance with the McDonald criteria. Failure to perform follow-up MRI on patients with clinically isolated syndromes is a sizeable factor in this diagnostic-gap and needs to be improved. In this setting, practical differences between the original and revised criteria appear to be small.