Effect of surveillance method on reported characteristics of Lyme disease, Connecticut, 1996-2007.

To determine the effect of changing public health surveillance methods on the reported epidemiology of Lyme disease, we analyzed Connecticut data for 1996-2007. Data were stratified by 4 surveillance methods and compared. A total of 87,174 reports were received that included 79,896 potential cases. Variations based on surveillance methods were seen. Cases reported through physician-based surveillance were significantly more likely to be classified as confirmed; such case-patients were significantly more likely to have symptoms of erythema migrans only and to have illness onset during summer months. Case-patients reported through laboratory-based surveillance were significantly more likely to have late manifestations only and to be older. Use of multiple surveillance methods provided a more complete clinical and demographic description of cases but lacked efficiency. When interpreting data, changes in surveillance method must be considered.