Exploring an alternative approach to Lyme disease surveillance in Maryland

Abstract

In Maryland, Lyme disease (LD) is a reportable disease and all laboratories and healthcare providers are required to report to the local health department. Given the volume of LD reports and effort required for investigation, surveillance for LD is burdensome and subject to underreporting. We explored the utility of International Classification of Diseases, 9th Revision, Clinical Modification (administrative) codes for use with LD surveillance. We aimed to collect the administrative codes for a 10% sample of 2009 LD reports (n = 474) from 292 facilities stratified by case classification (confirmed, probable, suspected and not a case). Sixty-three per cent (n = 184) of facilities responded to the survey, and 341 different administrative codes were obtained for 91% (n = 430) of sampled reports. The administrative code for Lyme disease (088.81) was the most commonly reported code (133/430 patients) among sampled reports; while it was used for 62 of 151 (41%) confirmed cases, it was also used for 48 of 192 (25%) not a case reports (sensitivity 41% and specificity 73%). A combination of nine codes was developed with sensitivity of 74% and specificity of 37% when compared to not a case reports. We conclude that the administrative code for LD alone has low ability to identify LD cases in Maryland. Grouping certain codes improved sensitivity, but our results indicate that administrative codes alone are not a viable surveillance alternative for a disease with complex manifestations such as LD.

Keywords: Lyme disease; administrative codes; surveillance.

FIGURE 1

Sampling scheme for analysis of administrative codes

FIGURE 2

Sensitivity and specificity of administrative code combinations for identifying confirmed and probable cases (%)