Browse the corpus

uman Lyme disease reporting to the Bureau of Health; this suggests that canine serosurveys may identify new areas of disease transmission. These are areas of low human population density, and repeat surveys may demonstrate the value of canine serosurveillance in detecting disease spread where human populations are low. Mapping of pooled data on a regional scale allows geographic patterns of disease to be viewed. Most notably, our data show a concentration of infected dogs in southern and coastal areas. Patterns of infection are suggested in inland areas as well. The significance of these patterns with respect to environmental variables favoring disease transmission is unknown but could be clarified by comparing prevalence rates with patterns of land use, deer herd density, habitat, and other ecologic attributes.

Lyme disease is the most commonly reported vectorborne disease in the United States; however, many experts believe that the number of cases is underreported. Lyme disease is often regarded as a routine condition or is frequently managed in high-volume settings (1). Few studies have assessed the accuracy of passive Lyme disease surveillance systems, but 1 study showed a 34% reporting rate (1). When tick identification services are offered, the identification data can show where disease vectors are found. In 1989, to determine the extent of the recently recognized infestation with Ixodes scapularis, the Maine Medical Center Research Institute's Vector-borne Disease Laboratory offered free tick identification to physicians, hospitals, veterinarians, and the general public. Since that time, >20,000 ticks, representing 14 species, have been identified. Testing has documented Borrelia burgorferi infection in I. scapularis from all Maine counties except 3.

itute's Vector-borne Disease Laboratory offered free tick identification to physicians, hospitals, veterinarians, and the general public. Since that time, >20,000 ticks, representing 14 species, have been identified. Testing has documented Borrelia burgorferi infection in I. scapularis from all Maine counties except 3. Mapping of ticks submitted for identification is subject to certain biases, which limits its utility for predicting human risk. Submission rates vary depending on population, education, and local concern, and results show little about disease transmission, particularly in disease-emergent areas where infection rates may lag behind tick distribution. The limitations of passive Lyme disease surveillance and tick identification that provide geographic information about risk can be largely overcome by using canine seroprevalence studies. Dogs are sensitive indicators because they have greater exposure to ticks. In disease-endemic areas, ≥50% of unvaccinated dogs have been reported to be infected (2,3). The prevalence of Lyme borreliosis in dogs correlates with infection in humans (4,5), as well as entomologic indicators of disease transmission (6). A newly available enzyme-linked immunosorbent assay (ELISA) kit (SNAP 3Dx, IDEXX Laboratories, Westbrook, ME, USA) is used widely by veterinarians in Maine to screen dogs for B. burgdorferi and heartworm infection. This test is used as part of a health screen during the heartworm testing season and can potentially generate large volumes of unbiased test data for public health application.

Dx, IDEXX Laboratories, Westbrook, ME, USA) is used widely by veterinarians in Maine to screen dogs for B. burgdorferi and heartworm infection. This test is used as part of a health screen during the heartworm testing season and can potentially generate large volumes of unbiased test data for public health application. The test kit detects antibodies directed against an invariable region (IR6) of the B. burgdorferi surface protein VlsE (Vmp-like sequence, Expressed) (2).The C6 ELISA test is not cross-reactive with antibodies induced by vaccination with either recombinant B. burdorferi outer-surface protein A (OspA) or whole-cell bacterin (2). This test has a very high accuracy rate, with 94.4% sensitivity and 99.6% specificity reported (7). In a clinical setting, when 18 dogs with known vaccination history were tested, the test results were 100% consistent with Western blot results (8). The Study One hundred sixty-four Maine clinics were contacted in February 2003 and invited to join the study; 69 of these agreed to participate. Clinics were instructed to record results of all IDEXX 3Dx Lyme disease tests that were conducted as part of a routine health screen, to record town of residence, and to record if a Lyme disease vaccine had ever been administered. Lyme disease vaccines can be highly effective (2); however, since vaccination rates are unevenly distributed, inclusion of vaccinated dogs would bias estimates of disease risk. This protocol was approved by the Maine Bureau of Health Institutional Review Board.

cord if a Lyme disease vaccine had ever been administered. Lyme disease vaccines can be highly effective (2); however, since vaccination rates are unevenly distributed, inclusion of vaccinated dogs would bias estimates of disease risk. This protocol was approved by the Maine Bureau of Health Institutional Review Board. Canine seroprevalence rates were calculated for minor civil divisions, including towns and unorganized townships. Rates were calculated only for divisions that had results of 10 or more tests. The relationships between the canine prevalence rates and human Lyme disease reports to the Bureau of Health (217 division-matched reports) and tick submissions to the Vector-borne Disease Laboratory (12,482 division-matched submissions) for the 2 years before this study, 2001–2002, were tested with Spearman rank correlation. Canine C6 antibodies persisted in experimentally infected, untreated dogs for ≥65 weeks, with no endpoint described (9); exposure status of the dogs in the present study could not be determined. Using data from 2 years allowed us to include sufficient numbers of human reports for meaningful statistic testing without sacrificing the ability to look at a "snapshot in time" of the Lyme disease spread.

weeks, with no endpoint described (9); exposure status of the dogs in the present study could not be determined. Using data from 2 years allowed us to include sufficient numbers of human reports for meaningful statistic testing without sacrificing the ability to look at a "snapshot in time" of the Lyme disease spread. Two maps were created. The first map (Figure 1) showed prevalence rates of minor civil divisions with ≥10 tests. The second map (Figure 2) showed pooled data from all divisions, including those with small sample sizes. For this map, an overlay of the state with 15-minute quadrangles was used. Each division from which data were collected was assigned to the quadrangle that contained the largest portion of its area. Seroprevalence rates for quadrangles were calculated by combining test results from all divisions within a quadrangle to find the average rate. Divisions were then assigned the average seroprevalence rate of their quadrangle for mapping. Quadrangles having a pooled total of <10 tests were not included in this map. Figure 1 Canine Lyme disease seroprevalence rates based on the IDEXX 3Dx test for minor civil divisions with ≥10 tests, Maine, 2003. Figure 2 Regional canine Lyme disease seroprevalence rates calculated from minor civil division pools created within 15-minute quadrangles, Maine, 2003.

Two maps were created. The first map (Figure 1) showed prevalence rates of minor civil divisions with ≥10 tests. The second map (Figure 2) showed pooled data from all divisions, including those with small sample sizes. For this map, an overlay of the state with 15-minute quadrangles was used. Each division from which data were collected was assigned to the quadrangle that contained the largest portion of its area. Seroprevalence rates for quadrangles were calculated by combining test results from all divisions within a quadrangle to find the average rate. Divisions were then assigned the average seroprevalence rate of their quadrangle for mapping. Quadrangles having a pooled total of <10 tests were not included in this map. Figure 1 Canine Lyme disease seroprevalence rates based on the IDEXX 3Dx test for minor civil divisions with ≥10 tests, Maine, 2003. Figure 2 Regional canine Lyme disease seroprevalence rates calculated from minor civil division pools created within 15-minute quadrangles, Maine, 2003. Test results from 9,511 dogs that had not been vaccinated for Lyme disease were submitted from 343 minor civil divisions. Tests were performed from March to July 2003. The overall seroprevalence rate was 8%. One hundred and eighty-three divisions met the criterion of a minimum sample size of 10 for calculating prevalence rates. At the division level, seroprevalence rates significantly correlated with the number of ticks submitted to the Maine Medical Center Research Institute's Vector-borne Disease Laboratory from 2001 to 2002 (r = 0.41, p<0.001), and human Lyme disease reports to the Bureau of Health (r = 0.15, p<0.05) from 2001 to 2002.

es. At the division level, seroprevalence rates significantly correlated with the number of ticks submitted to the Maine Medical Center Research Institute's Vector-borne Disease Laboratory from 2001 to 2002 (r = 0.41, p<0.001), and human Lyme disease reports to the Bureau of Health (r = 0.15, p<0.05) from 2001 to 2002. Regional seroprevalence rates were calculated for 65 quadrangles representing 297 minor civil divisions. Seroprevalence rates ranged from 0% to 47%. Rates were highest along southern coastal Maine (≤47%), with regional rates of 11% as far east as Columbia and along the mid-New Hampshire border as far north as Upton. Forty-four divisions with ≥10 tests had prevalence rates of 0%; 12 of these had ≥30 tests and 3 had ≥60. Conclusions This study demonstrates how canine serosurveys using the IDEXX 3Dx test can serve as an active surveillance system for potential human Lyme disease risk. This method overcomes the limitations of human Lyme disease reporting systems by relying on routine screening of populations of healthy dogs to calculate true seroprevalence rates. In this study, a large volume of data from across the state was generated for the most extensive and detailed measure of regional Lyme disease risk in Maine to date. In contrast, passive human Lyme disease surveillance during the previous 2 years yields cases from <90 towns, approximately two thirds of which had only 1 or 2 cases.

y, a large volume of data from across the state was generated for the most extensive and detailed measure of regional Lyme disease risk in Maine to date. In contrast, passive human Lyme disease surveillance during the previous 2 years yields cases from <90 towns, approximately two thirds of which had only 1 or 2 cases. Canine seroprevalence rates were congruent with I. scapularis submissions and human Lyme disease reports during a 2-year period when dogs could have been infected, reinforcing the effectiveness of this method for predicting geographic human risk. One previous study has calculated canine seroprevalence rates in Maine (6), but a different assay technique was used (4), which limited our ability to compare those rates to those of the current study. In spite of substantial agreement between canine seroprevalence and rates of tick submissions, mapping of canine seroprevalence data shows high-risk foci in inland areas that were not previously identified by 14 years of tick submissions to the Vector-borne Disease Laboratory or from human Lyme disease reporting to the Bureau of Health; this suggests that canine serosurveys may identify new areas of disease transmission. These are areas of low human population density, and repeat surveys may demonstrate the value of canine serosurveillance in detecting disease spread where human populations are low.

eas. Patterns of infection are suggested in inland areas as well. The significance of these patterns with respect to environmental variables favoring disease transmission is unknown but could be clarified by comparing prevalence rates with patterns of land use, deer herd density, habitat, and other ecologic attributes. The widespread acceptance of the IDEXX 3Dx test facilitates the use of canine serosurveys for public health. In many Maine veterinary offices, virtually every dog tested for heartworm in the spring is tested for B. burgdorferi antibody; however, well below 100% of canine patients are vaccinated against Lyme disease. Test results can be collected opportunistically from collaborating veterinarians with minimal effort. Previous serosurveys have involved much more intensive effort because of the need for veterinarians to collect extra blood samples. The ease of data collection based on this manner of testing enhances real-time as well as long-term monitoring of disease. Furthermore, the large volumes of test results generated from routine B. burgdorferi screening, and the ability to collect information on dog residence, make large-scale studies of disease geography possible. That we did not exclude in our analyses dogs that have traveled suggests that caution should be used when considering the importance of low prevalence rates or prevalence rates calculated from low sample sizes. However, our finding of dozens of towns with 0% prevalence suggests that the effect of dogs that have traveled on calculated seroprevalence rates is small.

hat have traveled suggests that caution should be used when considering the importance of low prevalence rates or prevalence rates calculated from low sample sizes. However, our finding of dozens of towns with 0% prevalence suggests that the effect of dogs that have traveled on calculated seroprevalence rates is small. Suggested citation for this article: Stone EG, Lacombe EH, Rand PW. Antibody testing and Lyme disease risk. Emerg Infect Dis [serial on the Internet]. 2005 May [date cited]. http://dx.doi.org/10.3201/eid1105.040381 Acknowledgments We thank the epidemiology staff at the Maine Bureau of Health for sharing human Lyme disease case reports and 2 anonymous reviewers for their comments on earlier drafts of this manuscript. We also gratefully acknowledge the generous contribution of participating veterinarians and their staff. This research was supported by the Maine Bureau of Health's Division of Disease Control through the Centers for Disease Control and Prevention Epidemiology and Laboratory Capacity for Infectious Diseases Cooperative Agreement, Grant #CCU112431-07. The opinions expressed by authors contributing to this journal do not necessarily reflect the opinions of the Centers for Disease Control and Prevention or the institutions with which the authors are affiliated. Dr. Stone is a research associate of Maine Medical Center's Vector-borne Disease Laboratory and adjunct faculty in the Department of Animal and Veterinary Sciences at the University of Maine.

Over the past 2 decades, the range of Ixodes scapularis, the deer tick, vector of Lyme disease, anaplasmosis, babesiosis, and deer tick virus infections, has expanded in northern New England. Because Lyme disease and anaplasmosis affect humans and dogs (Canis lupus familiaris), serosurveys of canids have proved useful for monitoring emergence of these infections. Sample selection may be confounded when dogs that are remotely exposed, vaccinated, or treated with topical acaricides are included. In recent years, however, the advent of a multitarget, in-clinic test kit (SNAP 4Dx; IDEXX Laboratories, Westbrook, ME, USA) has increased the scope and efficiency of these serosurveys. The SNAP 4Dx tests for heartworm antigen and antibodies to Borrelia burgdorferi, Anaplasma phagocytophilum, and Ehrlichia canis on 3 drops of blood. Its sensitivity and specificity for antibodies against B. burgdorferi and A. phagocytophilum exceed 98% (1,2).

has increased the scope and efficiency of these serosurveys. The SNAP 4Dx tests for heartworm antigen and antibodies to Borrelia burgdorferi, Anaplasma phagocytophilum, and Ehrlichia canis on 3 drops of blood. Its sensitivity and specificity for antibodies against B. burgdorferi and A. phagocytophilum exceed 98% (1,2). In Maine, deer ticks were first reported at a coastal site in 1988 and have since spread inland (3). Lyme disease has become a major public health problem; reported human cases reached 169 per 100,000 population in 1 mid-coastal county in 2008. Human cases of anaplasmosis and babesiosis are also being reported (4). In 1990, we conducted a statewide serosurvey to map B. burgdorferi–positive dogs and to correlate their distribution with reported human cases. Four percent of 828 samples were seropositive for B. burgdorferi, 89% of which were from dogs residing within 20 miles of the coast. No positivity was found among 102 dogs in the northern half of the state (5). Given the widespread acceptance of SNAP 4Dx tests by Maine veterinarians, we resurveyed dogs statewide in 2007 for exposure to B. burgdorferi and A. phagocytophilum. Data from questionnaires to veterinarians and dog owners enabled assessment of the influence of the use of Lyme vaccines and topical acaricides on canine serologic test results.

ceptance of SNAP 4Dx tests by Maine veterinarians, we resurveyed dogs statewide in 2007 for exposure to B. burgdorferi and A. phagocytophilum. Data from questionnaires to veterinarians and dog owners enabled assessment of the influence of the use of Lyme vaccines and topical acaricides on canine serologic test results. The Study From 87 veterinary clinics solicited in 2007, we selected 47 on the basis of their size and geographic distribution. Each was supplied with 15–30 SNAP 4Dx kits (contributed by IDEXX Laboratories). Veterinarians were instructed to obtain samples from all dogs routinely tested for heartworm. In northern areas, where heartworm is rarely tested for, they were asked to collect samples from dogs undergoing surgery. They recorded each dog’s age, town of residence, Lyme disease vaccination status (ever or never vaccinated), and the test results. Each dog owner completed a form (99.6% response rate) to describe the dog, its function, history of unexplained lameness, travel history (town, state, visited within the past year), history of tick infestation, and use of tick control products (yes or no).

nation status (ever or never vaccinated), and the test results. Each dog owner completed a form (99.6% response rate) to describe the dog, its function, history of unexplained lameness, travel history (town, state, visited within the past year), history of tick infestation, and use of tick control products (yes or no). We summarized test results to town and county levels. We used Spearman rank correlation tests to examine associations between canine seropositivity, human Lyme disease cases reported to the Maine Center for Disease Control and Prevention (Augusta, ME, USA) (4), and the number of deer ticks submitted to our laboratory in 2007. We used B. burgdorferi and A. phagocytophilum test results and questionnaire responses to cross-tabulate responses and calculate the likelihood (odds ratios) of B. burgdorferi and A. phagocytophilum positivity as a function of risk factors by using χ2 tests of association. We considered differences significant at p<0.05. Analyses were conducted by using SAS version 9.2 for Windows (SAS, Cary, NC, USA).

s to cross-tabulate responses and calculate the likelihood (odds ratios) of B. burgdorferi and A. phagocytophilum positivity as a function of risk factors by using χ2 tests of association. We considered differences significant at p<0.05. Analyses were conducted by using SAS version 9.2 for Windows (SAS, Cary, NC, USA). Of 1,087 dogs tested across Maine’s 16 counties, 12.7% were B. burgdorferi–positive and 7.1% were A. phagocytophilum–positive (Table 1); 1.9% were co-infected. The distribution of all dogs seropositive for either pathogen is shown by town in Figure 1. At the county level, canine B. burgdorferi seropositivity among unvaccinated dogs correlated positively with the number of human Lyme disease cases reported for 2007(ρSpearman = 0.84; p<0.0001) and the number of deer ticks submitted to our laboratory for identification (ρSpearman = 0.63; p = 0.009). In Figure 2, which shows statewide distributions by county north to south, only unvaccinated dogs are included in B. burgdorferi–positive data shown. Dogs had been exposed to A. phagocytophilum in all but 2 northern counties. At the town level, remarkably higher levels of canine A. phagocytophilum seropositivity were found in southern coastal Cape Elizabeth (Cumberland County) (76.5%, n = 17) and York (York County) (58.0%, n = 19) than in towns in their immediate vicinity.

d been exposed to A. phagocytophilum in all but 2 northern counties. At the town level, remarkably higher levels of canine A. phagocytophilum seropositivity were found in southern coastal Cape Elizabeth (Cumberland County) (76.5%, n = 17) and York (York County) (58.0%, n = 19) than in towns in their immediate vicinity. Table 1 Canine seroprevalence of Borrelia burgdorferi and Anaplasma phagocytophilum and Lyme diseases vaccination status, Maine, USA, 2007 County No. dogs tested B. burgdorferi* A. phagocytophilum* Lyme disease vaccination status No. negative No. (%) positive No. negative No. (%) positive No. reporting No. (%) vaccinated Aroostook 59 56 3 (5.1) 59 0 59 9 (15.3) Piscataquis 46 44 2 (4.3) 43 3 (6.5) 44 23 (52.3) Somerset 57 55 2 (3.5) 54 3 (5.3) 55 33 (60.0) Penobscot 77 73 4 (5.2) 73 4 (5.2) 75 30 (40.0) Franklin 78 73 5 (6.4) 78 0 78 38 (48.7) Washington 38 35 3 (7.9) 37 1 (2.6) 37 6 (16.2) Hancock 54 46 8 (14.8) 53 1 (1.9) 54 24 (44.4) Oxford 76 66 10 (13.2) 75 1 (1.3) 76 41 (53.9) Waldo 62 57 5 (8.1) 60 2 (3.2) 62 38 (61.3) Kennebec 120 106 14 (11.7) 114 6 (5.0) 119 82 (68.9) Knox 87 67 20 (23.0) 83 4 (4.6) 81 44 (54.3) Lincoln 91 75 16 (17.6) 85 6 (6.6) 85 63 (74.1) Androscoggin 62 53 9 (14.5) 60 2 (3.2) 62 27 (43.5) Sagadahoc 24 22 2 (8.3) 23 1 (4.2) 24 21 (87.5) Cumberland 91 78 13 (14.3) 72 19 (20.9) 88 62 (70.5) York 65 42 22 (34.4) 41 24 (36.9) 59 34 (57.6) Total 1,087 948 138 (12.7) 1,010 77 (7.1) 1,058 575 (54.3) *Tested by using SNAP 4Dx test kit (IDEXX Laboratories, Westbrook, ME, USA).

53 9 (14.5) 60 2 (3.2) 62 27 (43.5) Sagadahoc 24 22 2 (8.3) 23 1 (4.2) 24 21 (87.5) Cumberland 91 78 13 (14.3) 72 19 (20.9) 88 62 (70.5) York 65 42 22 (34.4) 41 24 (36.9) 59 34 (57.6) Total 1,087 948 138 (12.7) 1,010 77 (7.1) 1,058 575 (54.3) *Tested by using SNAP 4Dx test kit (IDEXX Laboratories, Westbrook, ME, USA). Figure 1 Towns where dogs were tested for seropositivity to Borrelia burgdorferi (A) and Anaplasma phagocytophilum (B) in a statewide serosurvey of domestic dogs, Maine, USA, 2007. Figure 2 A) Canine seroprevalence for Anaplasma phagocytophilum and, in dogs never vaccinated against Lyme disease, for Borrelia burgdorferi in Maine counties arranged north to south, 2007. B) Maine counties, with the 10 tick-abundant counties used in the analyses shaded in gray. Counties: 1, Aroostook; 2, Piscataquis; 3, Somerset; 4, Penobscot; 5, Franklin; 6, Washington; 7, Hancock; 8, Oxford; 9, Waldo; 10, Kennebec; 11, Knox; 12, Lincoln; 13, Androscoggin; 14, Sagadahoc; 15, Cumberland; 16, York.

2007. B) Maine counties, with the 10 tick-abundant counties used in the analyses shaded in gray. Counties: 1, Aroostook; 2, Piscataquis; 3, Somerset; 4, Penobscot; 5, Franklin; 6, Washington; 7, Hancock; 8, Oxford; 9, Waldo; 10, Kennebec; 11, Knox; 12, Lincoln; 13, Androscoggin; 14, Sagadahoc; 15, Cumberland; 16, York. Overall, 54.3% of the dogs had been vaccinated against Lyme disease. Never-vaccinated dogs were 1.5× as likely to be seropositive for B. burgdorferi than were vaccinated dogs (15.3% vs. 9.9%; p = 0.008) (Table 2). Vaccine use was higher in 10 southern counties where Lyme disease has become endemic (Figure 2, panel B) than in the 6 northern counties where it is emerging (63.9% vs. 42.7%; p<0.0001) and correlated positively with the number of deer ticks submitted to our laboratory for identification in 2007 (n = 16 counties, ρSpearman = 0.63; p = 0.009). Two thirds of respondents said that their dogs had traveled out of town; however, no associations were found between B. burgdorferi or A. phagocytophilum seropositivity and the dog’s travel history. Three of 59 dogs in the northernmost county of Maine were B. burgdorferi–positive, 1 of which had never traveled beyond its home town. Eighty-three percent of owners reported using acaricides. Despite the effective protection reported for topical acaricides (6,7), no difference in seropositivity between treated and untreated dogs was evident on the basis of their reported use (Table 2). Unexplained lameness was 1.5× more likely in dogs that were only A. phagocytophilum–positive than in those only B. burgdorferi–positive (40.0% vs. 26.5%; p<0.06).

ed for topical acaricides (6,7), no difference in seropositivity between treated and untreated dogs was evident on the basis of their reported use (Table 2). Unexplained lameness was 1.5× more likely in dogs that were only A. phagocytophilum–positive than in those only B. burgdorferi–positive (40.0% vs. 26.5%; p<0.06). Table 2 Risk factors vs. canine seroprevalence of Anaplasma phagocytophilum and Borrelia burgdorferi, Maine, USA, 2007* Variable No. dogs Borrelia burgdorferi Anaplasma phagocytophilum No. (%) positive OR (95% CI) p value† No. (%) positive OR (95% CI) p value† Lyme vaccine Yes 575 57 (9.9) 49 (8.5) No 483 75 (15.3) 1.5 (1.1–2.1) 0.008 23 (4.8) 0.6 (0.3–0.9) 0.02 Travel‡ All dogs None 357 49 (13.7) 27 (7.6) >1 730 89 (12.2) 0.9 (0.6–1.3) NS 50 (6.9) 0.9 (0.6–1.5) NS Unvaccinated dogs None 163 21 (12.9) 8 (4.9) >1 320 53 (16.6) 1.3 (0.8–2.3) NS 15 (4.7) 0.9 (0.4–2.3) NS Tick control products All dogs No 182 20 (11.0) 7 (3.9) Yes 899 115 (12.9) 1.2 (0.7–2.0) NS 66 (7.4) 2.0 (0.9–4.4) 0.08 Unvaccinated dogs No 124 13 (10.5) 3 (2.4) Yes 350 59 (16.9) 1.7 (0.9–3.3) 0.09 18 (5.1) 2.2 (0.6–7.6) NS History of unexplained lameness No 877 97 (11.1) 42 (4.8) Yes 191 40 (20.9) 2.1 (1.4–3.2) 0.0002 28 (15.2) 3.6 (2.1–5.9) <0.0001 *A total of 1,087 dogs were tested. OR, odds ratio; CI, confidence interval; NS, not significant. †Significance based on Pearson χ2 statistic with 1 degree of freedom. ‡Trips away from town of residence.

nexplained lameness No 877 97 (11.1) 42 (4.8) Yes 191 40 (20.9) 2.1 (1.4–3.2) 0.0002 28 (15.2) 3.6 (2.1–5.9) <0.0001 *A total of 1,087 dogs were tested. OR, odds ratio; CI, confidence interval; NS, not significant. †Significance based on Pearson χ2 statistic with 1 degree of freedom. ‡Trips away from town of residence. Conclusions This study demonstrates that risk of contracting Lyme disease has reached northernmost Maine and that anaplasmosis is now being transmitted to dogs throughout the lower half of the state. The study expands on nationwide SNAP 4Dx data documenting B. burgdorferi and A. phagocytophilum positivity in the southern half of the state (8). In southern coastal Maine, overabundant white-tail deer, appropriate habitat, and maritime climate all contribute to high densities of I. scapularis ticks (3) and consequent disease transmission; thus, the remarkably high level of A. phagocytophilum seroreactivity observed in the southern coastal towns of Cape Elizabeth and York calls for further work to understand the dynamics of the intense local emergence of this pathogen. The higher level of unexplained lameness in A. phagocytophilum–positive dogs than in B. burgdorferi–positive dogs is consistent with findings by Beall et al. (9), who reported a 2.6× greater incidence of A. phagocytophilum seroreactivity than B. burgdorferi seroreactivity among 32 lame, non–co-infected dogs in Minnesota who were suspected of having either disease. The lameness also reflects the high percentage of B. burgdorferi positivity among asymptomatic dogs (10). That B. burgdorferi and A. phagocytophilum seropositivity rates were essentially identical between dogs who had a history of travel and those who did not lessens concern about travel as a confounding variable, an exposure difficult to interpret in any event, given the spotty distribution of ticks even where Lyme disease is endemic (11).

i and A. phagocytophilum seropositivity rates were essentially identical between dogs who had a history of travel and those who did not lessens concern about travel as a confounding variable, an exposure difficult to interpret in any event, given the spotty distribution of ticks even where Lyme disease is endemic (11). In a recent study, Hamer et al. (12) reported that a serosurvey of canines for B. burgdorferi is ineffective in a region that includes areas with little B. burgdorferi transmission, and less informative than analysis of ticks removed from dogs. The authors referred to the confounding influence of tick chemoprophylaxis. Our inability to detect an effect of topical acaricides may reflect their ubiquitous use for flea control and a lack of information on the frequency of their use. Although the widespread use of protective measures now complicates interpretation of serosurveys of canines, in selected dogs the availability of a reliable, multitarget test that is used routinely nationwide (8) remains a valuable and cost-effective method for documenting transmission of the agents of Lyme borreliosis and anaplasmosis, particularly in areas where disease is emerging. Suggested citation for this article: Rand PW, Lacome EH, Elias SP, Cahill BK, Lubelczyk CB, Smith RP. Multitarget test for emerging Lyme disease and anaplasmosis in a serosurvey of dogs, Maine, USA. Emerg Infect Dis [serial on the Internet]. 2011 May [date cited]. http://dx.doi.org/10.3201/eid1705.100408

In a recent study, Hamer et al. (12) reported that a serosurvey of canines for B. burgdorferi is ineffective in a region that includes areas with little B. burgdorferi transmission, and less informative than analysis of ticks removed from dogs. The authors referred to the confounding influence of tick chemoprophylaxis. Our inability to detect an effect of topical acaricides may reflect their ubiquitous use for flea control and a lack of information on the frequency of their use. Although the widespread use of protective measures now complicates interpretation of serosurveys of canines, in selected dogs the availability of a reliable, multitarget test that is used routinely nationwide (8) remains a valuable and cost-effective method for documenting transmission of the agents of Lyme borreliosis and anaplasmosis, particularly in areas where disease is emerging. Suggested citation for this article: Rand PW, Lacome EH, Elias SP, Cahill BK, Lubelczyk CB, Smith RP. Multitarget test for emerging Lyme disease and anaplasmosis in a serosurvey of dogs, Maine, USA. Emerg Infect Dis [serial on the Internet]. 2011 May [date cited]. http://dx.doi.org/10.3201/eid1705.100408 Acknowledgments We thank Ramaswamy Chandrashekar for assistance with the multitarget test, the staffs of the 47 veterinary clinics for participating in this statewide serosurvey, and Linda Siddons for invaluable data entry. This study was supported by a grant from the Elmina B. Sewall Foundation.

Suggested citation for this article: Rand PW, Lacome EH, Elias SP, Cahill BK, Lubelczyk CB, Smith RP. Multitarget test for emerging Lyme disease and anaplasmosis in a serosurvey of dogs, Maine, USA. Emerg Infect Dis [serial on the Internet]. 2011 May [date cited]. http://dx.doi.org/10.3201/eid1705.100408 Acknowledgments We thank Ramaswamy Chandrashekar for assistance with the multitarget test, the staffs of the 47 veterinary clinics for participating in this statewide serosurvey, and Linda Siddons for invaluable data entry. This study was supported by a grant from the Elmina B. Sewall Foundation. Dr Rand is co-director of the Maine Medical Center Research Institute Vector-borne Disease Laboratory. His primary research interests include the epidemiology and ecology of vector-borne diseases.