Assessing Tick-Borne Disease Risk and Surveillance: Toward a Multi-Modal Approach to Diagnostic Positioning and Prediction.

Abstract

The true extent of tick-borne disease (TBD) incidence and risk among humans is largely unknown, posing significant public health challenges. This study offers an exploratory analysis of a multimodal dataset and is part of a larger ongoing project to determine if entomological data, canine serological reports, self-reported human tick bite encounters (TBEs), and/or associated TBD diagnoses can serve as proxies for human disease risk. Focusing on the United States (U.S.), it characterizes self-reported TBD diagnoses (specifically, anaplasmosis, ehrlichiosis, and Lyme disease), co-infections, and their frequency and distribution across U.S. counties in relation to the presence of other factors related to TBD risk. Survey data was used to construct a list of TBEs localizable to individual U.S. counties. National data regarding these counties—namely the presence of official Lyme Disease (LD) case reports from the Centers for Disease Control and Prevention, as well as the tick vectors I. scapularis and I. pacificus within a given county—were then linked with survey-reported TBEs, tabulated by diagnosis (including co-infections), to determine the distribution of county-level endpoints across diagnostic categories. In addition, data on the presence of positive serological diagnostic tests conducted in canines were considered due to their potential utility as a proxy for TBD and TBE risk. The final dataset contained 249 TBEs localized to a total of 144 counties across 30 states. Diagnostic categories included respondents with LD (n = 70) and those with anaplasmosis and ehrlichiosis diagnoses and co-infections (n < 20 per diagnostic category). TBEs also were indicated by respondents who did not report TBD diagnoses, with some indicating uncertainty. The distribution of respondent-reported TBEs varied between canine TBDs, with LD-positive respondents reporting noticeably larger proportions of TBEs in counties with canine LD and smaller proportions in counties with canine anaplasmosis, compared to respondents without an LD diagnosis; a notional logistic regression suggests these differences may be significant (canine LD: Odds Ratio [OR] = 6.04, p = 0.026) (canine anaplasmosis: OR = 0.50, p = 0.095). These results suggest that certain widely available diagnostic TBD data in animals (in this case, domesticated dogs) may be sensitive to differences in human TBD risk factors and thus may have utility as proxies in future research. In the absence of an available standardized, unified, and national TBD database, such proxies, along with relevant surveys and reports, may provide a much-needed working solution for scientists and clinicians studying TBDs.