Evaluation of the impact of the 2008 revision to the national Lyme disease case definition, Pennsylvania

Background: In recent years, more than 3,000 cases of Lyme disease (LD) have been reported annually in Pennsylvania. LD became nationally notifiable in 1991, with case definition revisions in 1995, 1996, and 2008. In this study, we evaluate the impact the 2008 case definition changes will have on LD case counts and trends in Pennsylvania.

Methods: LD case reports from the period January through December 2008 were extracted from Pennsylvania's electronic disease surveillance system. Algorithms corresponding to the 1996 and 2008 CDC case definitions were written in SAS 9.1. Case classifications using the two definitions were compared.

Results: 7,294 potential LD cases were reported through July 2008; results will be updated to include all of 2008. The 2008 case definition resulted in a 26.3% decrease in confirmed cases. However, 672 (14.7%) of the 4,561 reports that were classified as non-cases using the 1996 definition shifted to probable (27) or suspect (645) cases using the 2008 definition. 3,667 reports (50.3%) remained non-cases under both definitions. An additional 222 (3.0%) “indeterminant” reports may have qualified as probable if physician diagnosis was available.

Conclusion: We expect that the 2008 case definition will substantially reduce the number of confirmed LD cases in Pennsylvania. Much of this decrease will be offset by cases newly classified as probable or suspect. Even if all 222 indeterminant reports were classified as probable, there would still be a 17.6% decline in counted cases using the 2008 definition. These changes will need to be considered when interpreting temporal LD trends.