Patterns Testing for Tick-Borne Diseases and Implications for Surveillance in the Southeastern US

IMPORTANCE Tick-borne diseases (TBD), including spotted fever group rickettsiosis (SFGR), ehrlichiosis, and, increasingly, Lyme disease, represent a substantial public health concern throughout much of the southeastern United States. Yet, there is uncertainty about the epidemiology of these diseases because of pitfalls in existing diagnostic test methods. OBJECTIVE To examine patterns of diagnostic testing and incidence of TBD in a large, academic health care system. DESIGN, SETTING, AND PARTICIPANTS This cross-sectional study included diagnostic test results for TBD at UNC Health, a large academic health care system with inpatient and outpatient facilities, from January 1. 2017, to November 30. 2020. Participants included all individuals seeking routine care at UNC Health facilities who had testing for SFGR, ehrlichiosis, or Lyme disease performed during the study period. MAIN OUTCOMES AND MEASURES Rates of test positivity, testing completeness, and incidence of TBD. RESULTS During the 4-year study period, 11367 individuals (6633 [58.4%] female; 10 793 [95%] non-Hispanic individuals and 8850 [77.9%] White individuals; median [IQR] age, 53 [37-66] years) were tested for TBD. Among the 20 528 diagnostic tests performed, 47 laboratory-confirmed, incident cases of SFGR, 27 cases of ehrlichiosis, and 76 cases of Lyme were confirmed, representing incidence rates of 4.7%, 7.1%, and 0.7%, respectively. However, 3984 of SFGR tests (79.3%) and 3606 of Ehrlichia tests (74.3%) lacked a paired convalescent sample. Of 20 528 tests, there were 11977 tests (58.3%) for Lyme disease from 10 208 individuals, 5448 tests (26.5%) for SFGR from 4520 individuals, and 3103 tests (15.1%) for ehrlichiosis from 2507 individuals. Most striking, testing for ehrlichiosis was performed in only 55% of patients in whom SFGR was ordered, suggesting that ehrlichiosis remains underrecognized. An estimated 187 incident cases of SFGR and 309 of ehrlichiosis were potentially unidentified because of incomplete testing. CONCLUSIONS AND RELEVANCE In this cross-sectional study, most of the patients suspected of having TBD did not have testing performed in accordance with established guidelines, which substantially limits understanding of TBD epidemiology. Furthermore, the data revealed a large discrepancy between the local burden of disease and the testing performed. These findings underscore the need to pursue more robust, active surveillance strategies to estimate the burden of TBD and distribution of causative pathogens.