Discussion

We demonstrate that in Monmouth County, New Jersey, ehrlichiosis infections from A. americanum ticks should be occurring, at a minimum, one half as often as Lyme disease (e.g., 1 ehrlichiosis case for every 2 Lyme disease cases). This rate of occurrence is clearly not the case (Figure 2), and these numbers suggest that ≈99% of potential ehrlichiosis infections are not recognized. It is possible that not all persons who become infected in Monmouth are county residents (and so their case would be recorded elsewhere) making the total number of Monmouth County–derived infections likely to be somewhat higher than the observed 2 cases. However, even if one assumes all ehrlichiosis reported for the entire state of New Jersey to originate in Monmouth County (64 cases in 2014),^[22] these values still indicate a substantial discrepancy between numbers of observed and expected cases.

When selecting values for the parameters used in our calculations, every opportunity to be conservative was taken to avoid biasing estimates of relative risk. For example, infection (I_x ) probabilities were higher for Lyme disease than ehrlichiosis (39.87% vs. 11.7% for adult ticks and 23.3% vs. 9.04% for nymphs). Although reported B. burgdorferi infection rates in I. scapularis adults frequently range from 40% to 50% in hyperendemic areas, several studies have reported lower rates.^[9,31] Our weighted infection prevalence estimate of 9.98% (encompassing both E. chaffeensis and E. ewingii across adults and nymphs) is probably lower than the actual value, given that many studies have reported infection prevalence in the range of 5%–15% for E. chaffeensis alone and in some locations twice that rate.^[3,32] Last, use of passive surveillance numbers (C_x ) probably underestimates the actual risk for exposure to A. americanum ticks; residents who recognize the tick species may be less likely to bring in A. americanum ticks to the passive surveillance program, because we only test I. scapularis (as stated on our website), and taking the 10-year average for relative abundances (2006–2015) does not account for the recent surge in A. americanum tick submissions from 2012 onward (Figure 1). In light of these considerations, the actual risk for A. americanum tick–associated ehrlichioses in Monmouth County may be much higher than we have estimated.

One caveat should be noted. Differences in PCR sensitivity between B. burgdorferi and Ehrlichia sp. assays could affect our ability to compare infection rates, although to mitigate this problem as much as possible, we relied on established primers and checked that infection rates were within ranges reported by other studies as described previously. Thus, any difference in our ability to detect pathogens between the 2 tick species is unlikely to alter our conclusions.

One possible explanation for the lower-than-expected number of reported ehrlichiosis cases is a lack of awareness about ehrlichial disease on the part of the public and physicians, leading to misdiagnosis and underreporting. The infection tends to manifest as a general influenza-like illness, and onset of a rash is rare,^[12] so persons may be less likely to visit a doctor unless more severe symptoms emerge or they are aware of a recent tick bite and the presence of tickborne diseases in the area. Awareness of non–Lyme disease tickborne illnesses is startlingly low, even in parts of the country where ehrlichiosis cases outnumber Lyme disease cases.^[33–35] One study found that >50% of respondents in the United States had heard of Lyme disease, whereas only 1.4% had heard of ehrlichiosis.^[35] As a consequence of these factors, when active screening for ehrlichial infections is performed, the resulting case rates are often much higher than those reported to governmental agencies.^[3,36,37] For example, Olano et al.^[37] found that the incidence of ehrlichiosis observed when actively screening patients was as much as 2 orders of magnitude higher than the passively reported incidence.

Another explanation could be the existence of asymptomatic infections. Several studies of E. chaffeensis antibody seroprevalence in adults found that most of those carrying the antibodies had no recollection of a symptomatic infection.^[38–40] Further, a study screening blood samples from children from the Southeast United States for E. chaffeensis revealed that many more children had been exposed than showed clinical signs.^[41] Most documented cases of E. chaffeensis and E. ewingii infection come from older adults,^[13,15] so when younger adults and children become infected, they may be less likely to have symptoms or seek treatment,^[41] accounting for the large number of unreported infections. Other vectorborne diseases, such as West Nile virus, transmitted by mosquitoes, also demonstrate large numbers of asymptomatic infections and increased severity among the older population.^[42] However, because of cross-reactivity in the serologic test for E. chaffeensis and the lack of specific testing for E. ewingii, one could argue that many presumed asymptomatic cases of E. chaffeensis were actually incidences of infection with E. ewingii, which tends to be more mild.^[16] Although our study considered E. chaffeensis and E. ewingii interchangeably, if we repeat our calculations by using E. chaffeensis infection rates alone, we would expect Monmouth County to have seen 84 cases of E. chaffeensis in 2014, when only 2 were reported. Therefore, 97.6% of E. chaffeensis infections are potentially going unnoticed (versus 99% of ehrlichiosis infections overall), which is still a troubling discrepancy.

Across the entire United States, the number of Lyme disease cases dwarfs ehrlichiosis cases. During 2004–2013, the Centers for Disease Control and Prevention annually reported 279–528 Lyme disease cases/1 million persons.^[2] In contrast, during that same period, annual ehrlichiosis cases ranged from 3.3 to 26 cases/1 million persons.^[43] If our numbers can be extrapolated to other areas in the country, this implies that, on a national scale, potentially thousands of ehrlichiosis cases are going undiagnosed. The A. americanum tick appears to be expanding into geographic areas where it did not occur previously^[6,32] and is becoming more abundant within its existing range.^[9,21,44] Because ehrlichiosis cases have steadily increased since becoming reportable,^[13,15] the spread of A. americanum ticks and the emergence of ehrlichiosis as a human pathogen in the United States may parallel increases in I. scapularis tick populations and the emergence of Lyme disease that occurred 30 years prior.^[14] Even if most unrecognized infections are mild or asymptomatic, these could still have consequences for public health; for example, blood donors who are unknowingly infected could pass the infection to immunocompromised patients,^[45] or prescription of sulfa drugs for unrelated ailments could result in worsened disease presentation.^[46]

Our findings indicate a need to increase public education efforts about the risks for acquiring tickborne diseases other than Lyme disease in the United States, and in particular, to expand prevention awareness of medically important tick species other than I. scapularis. Fortunately, many of these diseases, including ehrlichiosis, can be prevented in the same manner as Lyme disease (e.g., avoidance of tick bites) and are treated similarly to Lyme disease;^[12,47] also, the number of ehrlichiosis cases peaks during spring and summer months, corresponding with peak Lyme disease transmission months.^[13,15] Consequently, refocusing existing public health education efforts to encompass the full spectrum of tickborne diseases could be accomplished without changing much of its content, although some attention should be given to the risks imposed by viruses and protozoa as well as unique characteristics of the questing behavior of A. americanum ticks. To better inform these educational efforts and more accurately assess tickborne disease risk, more research into diseases other than Lyme disease is required.

Humans can alter their environment in many ways that affect disease transmission, from localized changes affecting tick habitat and host abundance, to larger changes affecting the planet’s climate. As these changes continue to occur, the study of vector and pathogen distributions and abundance will be critical to understanding the potential risk to humans posed by emerging pathogens. Our calculations imply that infections with E. chaffeensis and E. ewingii are underrecognized, at least in Monmouth County, New Jersey, if not throughout a larger portion of the United States where A. americanum ticks are abundant or becoming abundant. Additional effort is needed to determine the causes for this apparent discrepancy and to characterize the actual prevalence of ehrlichiosis in the human population, as well as to raise awareness about the risk for exposure to these pathogens in areas where A. americanum ticks are common.