Evaluating the feasibility of a national Multiple Sclerosis (MS) surveillance system

OBJECTIVE To test the feasibility of a MS surveillance system in the United States BACKGROUND Few surveillance systems exist for chronic, non-infectious diseases such as MS. Accessing medical records to identify MS cases is time consuming, costly, and potentially biased, as not all physicians grant access to their records. The Agency for Toxic Substances and Disease Registry in collaboration with MS researchers conducted a pilot study to test the feasibility of using existing databases to develop a national MS surveillance system. DESIGN/METHODS We looked at records from a hospital-based MS database (New York State Multiple Sclerosis Consortium [NYSMSC]) as well as administrative databases including Medicare, the Veterans Health Administration, the Veterans Benefits Administration, and a self-enrolled national MS registry (North American Research Committee on Multiple Sclerosis), for the period 2001-2005, to identify cases of MS. Records were matched by name (last, first, middle), sex, month and year of birth, and city of residence. RESULTS More than 30,000 likely MS cases were identified from the administrative databases, many in multiple databases. Almost 80% of cases from the hospital-based NYSMSC database were identified in at least one non-medical database. CONCLUSIONS Preliminary results show that existing administrative databases may be used to establish a national MS surveillance system. However, algorithms need to be developed to increase the accuracy of identifying “true” MS cases. Strategies need to be developed to increase the overall completeness of a national MS surveillance system.