Using Electronic Health Records to Enhance Lyme Disease Surveillance: Protocol for the SubLyme Network

Area of Science:

• Public Health
• Epidemiology
• Health Informatics

Background:

• Lyme disease is a prevalent vector-borne illness in the U.S.
• Traditional surveillance methods have limitations in tracking Lyme disease burden and evaluating interventions.
• The Centers for Disease Control and Prevention (CDC) established the Surveillance Based Lyme Disease Network (SubLyme) to enhance surveillance and research using electronic health record (EHR) data.

Purpose of the Study:

• To develop and validate scalable computable phenotypes (CPs) for identifying Lyme disease cases within EHR data.
• To estimate Lyme disease incidence and describe its distribution across key demographics.
• To support secondary objectives including distinguishing acute from disseminated disease and identifying clinical manifestations.

Main Methods:

• SubLyme involves five health systems across high-risk Lyme disease regions in the U.S.
• The network evaluates the validity of EHR-based CP definitions by comparing them against manually abstracted medical charts.
• Performance metrics include sensitivity, specificity, positive predictive value, and negative predictive value, using 500 manually reviewed charts per site as the gold standard.

Main Results:

• Over 4.6 million patients were analyzed, identifying 332,256 with at least one Lyme disease element in their EHR.
• Demographic data showed 55.6% female, 87.9% White, and 90.8% non-Hispanic patients.
• The majority of identified cases had only a medication order (53.4%) or a test order (35.8%), with fewer having combined orders.

Conclusions:

• The SubLyme network is poised to advance Lyme disease surveillance through multi-system EHR data utilization.
• Developing novel surveillance methods is crucial given the increasing incidence and geographic expansion of Lyme disease.
• EHR-based surveillance offers a promising approach to overcome current challenges and expedite Lyme disease research.