A study published in the journal Neurology suggests prevalence of multiple sclerosis (MS) in the US is 913,925—more than twice the previous estimate of 400,000 people in 1981. Because incidence of MS has not increased markedly, the researchers posit that increased prevalence reflects longer survival times for people with MS. However, they note that some increased incidence has been seen, especially in nonwhite populations, and that earlier radiologic diagnosis may also contribute to the larger number. The latter is of particular interest considering recent reports that as many as 20% of new MS diagnoses may be false positives.
This research used 5 insurance claims databases from 2008-2010 and a validated algorithm to identify prevalence within populations that were then combined to arrive at prevalence for 2010. That was then extrapolated to estimate prevalence in 2017. Companion articles on the background and methods for the study were published in the same issue of Neurology.
The research team, the United States Multiple Sclerosis Workgroup, considered multiple ways of assessing MS prevalence, exploring the strengths, weakness, opportunity, and threats of each to develop a 4-step method. The method consisted of examining public and private insurance claims database with an algorithm developed by the team to identify claims of persons with MS, prevalence within populations studied, and prevalence as a whole in the US. Of note, the algorithm was validated in the methods study that showed it had 86.6% to 96.0% sensitivity, 66.7%-99.0% specificity, and a positive predictive value of 95.4% to 99.0%. The algorithm required that any individual case include 3 or more MS-related claims from any combination of inpatient, outpatient, or disease-modifying treatment (DMT) claims within 1 year.
"This was an innovative, big data project that used information from more than 125 million health records," said Nick LaRocca, Vice President of Healthcare Delivery and Policy Research for the MS Society, and a co-author of the study. "Our hope is that the methods we used in this study can be applied to estimate prevalence of other medical disorders."