Economic Impact of Lyme Disease

Xinzhi Zhang,* Martin I. Meltzer,* César A. Peña,†1 Annette B. Hopkins,†

Lane Wroth,‡ and Alan D. Fix†

To assess the economic impact of Lyme disease (LD), face protein A (rOspA) LD vaccine (LYMErix, SmithKline
the most common vectorborne inflammatory disease in the Beecham Biologicals, Rixensart, Belgium) for persons
United States, cost data were collected in 5 counties of the 15–70 years of age (11). However, in 2002, SmithKline
Maryland Eastern Shore from 1997 to 2000. Patients were withdrew the vaccine, citing low demand. Therefore, per-
divided into 5 diagnosis groups, clinically defined early-
sonal protection measures, early diagnosis, and early treat-
stage LD, clinically defined late-stage LD, suspected LD,
tick bite, and other related complaints. From 1997 to 2000, ment are extremely important in preventing and
the mean per patient direct medical cost of early-stage LD controlling LD.
decreased from $1,609 to $464 (p<0.05), and the mean per Since the first case reported in 1975 (12), LD has
patient direct medical cost of late-stage LD decreased from become the most common vectorborne inflammatory dis-
$4,240 to $1,380 (p<0.05). The expected median of all ease in the United States. Foci of LD are widely spread in
costs (direct medical cost, indirect medical cost, nonmed- the northeastern, mid-Atlantic, and north-central regions
ical cost, and productivity loss), aggregated across all diag- of the United States (13). Despite federal, state, and local
nosis groups of patients, was ≈$281 per patient. These efforts to prevent and control LD, total reported cases of
findings will help assess the economics of current and
LD increased almost 3-fold from 1991 to 2002 (Figure 1).
future prevention and control efforts.
In 2002, the Centers for Disease Control and Prevention
(CDC) received reports of 23,763 LD cases, 95% of which
yme disease (LD) is a multisystem, multistage, inflam-
L matory tickborne disorder caused by the spirochete
Borrelia burgdorferi. LD usually begins with an initial
were from Connecticut, Delaware, Maine, Maryland,
Massachusetts, Minnesota, New Hampshire, New Jersey,
New York, Pennsylvania, Rhode Island, and Wisconsin
expanding skin lesion, erythema migrans (EM), which (14). In Maryland, the overall incidence of LD was more
may be followed by musculoskeletal, neurologic, and car- than twice as high as the overall incidence of LD in the
diac manifestations in later stages of the disease (1–3). United States (13.0 vs. 6.3 cases per 100,000 population)
Enzyme-linked immunosorbent assay and Western blotting (13).
test are widely used to diagnose LD (4–6). LD is most Assessing the economic impact of LD will help assess
responsive to antimicrobial drugs in the early stage, while the economics of current and future prevention and control
further intensive therapy may be necessary in the late stage efforts. Although several studies of cost estimates of LD
(7,8). A variety of prevention and control procedures can have been published (e.g., 15), information on the econom-
be implemented to prevent and reduce LD incidence, ic impact of LD is limited. Therefore, we conducted a 4-
including, but not limited to, public education; personal year study to estimate the economic impact of LD on the
protection measures such as wearing protective clothing Maryland Eastern Shore.
(gloves, long clothes), checking one’s body daily for ticks,
avoiding tick-infested areas, and applying tick repellent Methods
(DEET, permethrin); host management; habitat modifica-
tion; and chemical control (9,10). In 1998, the Food and Study Population and Data
Drug Administration approved a recombinant outer-sur- This study was conducted in 5 counties (Caroline,
Dorchester, Kent, Queen Anne, and Talbot) on the
Maryland Eastern Shore, an area where LD is endemic
*Centers for Disease Control and Prevention, Atlanta, Georgia,
USA; †University of Maryland, Baltimore, Maryland, USA; and
‡Care First-Easton Branch (previously Delmarva Health Plan), 1Current affiliation: Maryland Department of Health and Mental
Easton, Maryland, USA Hygiene, Baltimore, Maryland, USA

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 653
RESEARCH

izations and emergency department visits; dates, results,

and costs of laboratory tests; and dates and costs of antimi-
crobial drug treatment. All abstracted information was kept
confidential. After 1999, an anonymous abstraction of
medical records was approved by the institutional review
board (IRB) and implemented, allowing inclusion of more
patients for all 4 study years, with the exclusion of the
records of those who had previously declined participa-
tion. All protocols of this study were approved by IRBs
from CDC, the state of Maryland, and the University of
Maryland. Those patients identified as having received an
LD vaccination were not included in this study.
Figure 1. Lyme disease (LD) cases reported to the Centers for
Case Definition
Disease Control and Prevention by state health departments in the
United States (1991–2002). Reported cases were defined accord- For the purpose of surveillance, a case of LD is defined
ing to the national surveillance definition. For the purpose of sur- as physician-diagnosed EM >5 cm or at least 1 late
veillance, a case of LD is defined as physician-diagnosed rheumatologic, neurologic, or cardiac manifestation with
erythema migrans >5 cm or >1 late rheumatologic, neurologic, or laboratory evidence of B. burgdorferi infection (16). These
cardiac manifestation with laboratory evidence of Borrelia burgdor-
criteria were developed as an epidemiologic case defini-
feri infection. Available from http://www.cdc.gov/ncidod/dvbid/
lyme/epi.htm (14). tion intended for surveillance purposes only. Although
such a standard may aid comparison across clinical studies
and facilitate development of research, exposure history
and clinical features are critical. For example, treating
(Table 1). The study population includes patients living in patients with seasonal (summer) musculoskeletal flulike
the 4 counties enrolled in Delmarva Health Plan (DHP, a symptoms in areas where LD is endemic may be clinically
managed healthcare organization) and non-DHP patients appropriate (12). Because the data for this study were col-
receiving health care from office-based physicians in Kent lected directly from healthcare organizations and physi-
County from 1997 to 2000. Eligible patients were identi- cians, we used a clinical definition of LD. This definition
fied through records of encounters for LD, tick bites, insect was based on physicians’ determination in the medical
bites, and serologic testing for LD antibodies. During 1997 record, according to patients’ clinical findings, tick expo-
and 1998, identified patients were contacted for informed sure, and other relevant details (e.g., laboratory results).
consent. Patients who indicated that they did not wish to In our study, LD patients were identified by using a
participate were excluded from our database. A cost and final diagnosis code in their medical records. LD patients
risk questionnaire (Appendix 1 available online at were then divided into 5 diagnosis groups: clinically
www.cdc.gov/ncidod/EID/vol12no04/05-0602_app1.htm) defined early-stage LD, clinically defined late-stage LD,
was sent to patients who gave informed consent. The suspected LD, tick bite, and other related complaints. Most
response rate of the survey was ≈22%. Interviewers then clinically defined early-stage LD patients had EM; some
reviewed patients’ charts and consulted relevant sources also had musculoskeletal flulike symptoms such as
(e.g., hospital, physician office, laboratory) to obtain the malaise, fatigue, headache, fever, and chills (12). In this
following information: patient demographics; insurance study, clinically defined late-stage LD patients included
coverage; diagnosis; symptoms; dates of onset and diagno- those with later manifestations (neurologic involvement,
sis; dates of tick bite exposure; dates and costs of primary cardiac involvement, and arthritis) and patients with
provider and consultant visits; dates and costs of hospital- chronic LD. The diagnosis groups of suspected LD, tick

654 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006
 Economic Impact of Lyme Disease

bite, and other related complaints involved all patients Because of the potential complexity of accurately answer-
without a clear final diagnosis of LD. Suspected LD ing the question, we did not ask patients to estimate the
referred to patients who had some symptoms that could be time they lost from household production. We did, howev-
indicative of LD without further evidence and thus no er, ask patients if they paid anybody to do household tasks
definitive diagnosis of LD. Patients with tick bites without because their LD-related infirmities prevented them from
symptoms were placed in the tick bite group. The diagno- doing those tasks. For patients <15 years of age, we
sis group of other related complaints included all other assumed that their parents (usually the mother) had to take
diagnoses that were different from the above 4 diagnosis time off from their work to take care of them. Therefore,
groups, such as unknown insect bites and screening among their mothers’ values of lost days of work were included.
asymptomatic persons.
Analysis
Study Design We used the following formula to estimate the average
We calculated the following total costs of LD: 1) direct per capita cost of LD, i.e., the mean cost (direct medical
medical costs of LD diagnosis and treatment, 2) indirect costs, indirect medical costs, nonmedical costs, and pro-
medical costs, 3) nonmedical costs, and 4) productivity ductivity losses) aggregated across all diagnosis groups of
losses. Intangible costs (e.g., costs incurred because of patients:
pain and suffering) were not incorporated. Consumer price Expected mean cost of a LD outcome = Σ direct medical costs,
index (CPI) for medical care was used to adjust all medical indirect medical costs, nonmedical costs, and productivity losses (Mean cost of
payments into year 2000 dollars (17). For nonmedical outcome clinically defined early-stage LD, clinically defined late-stage LD, suspect-
costs and productivity losses, we adjusted costs by using ed LD, tick bite, and other related complaints × Probability of outcome clin-
the general CPI. We took a societal perspective, which ically defined early-stage LD, clinically defined late-stage LD, suspected LD, tick bite,
incorporates all costs and all benefits no matter who pays and other related complaints).
costs or who receives benefits. Because the distribution of cost data is often not nor-
Charges were used to estimate the direct medical cost. mal, we also calculated the medians of these costs and
To determine the direct medical costs associated with LD, used both mean and median to estimate the most likely per
we used charge data from both DHP and office-based capita cost of LD on the Maryland Eastern Shore. The
healthcare providers in Kent County. Direct medical costs median cost of an LD outcome was calculated by using the
of LD included costs (charges) of physician visits, consul- following formula:
tation, serology, procedure, therapy, hospitalization/emer- Expected median cost of a LD outcome = Σ direct medical
gency room (ER), and other related costs (Appendix 2 costs, indirect medical costs, nonmedical costs, and productivity losses (Median
available online at www.cdc.gov/ncidod/EID/vol12no04/ cost of outcome clinically defined early-stage LD, clinically defined late-stage
05-0602_app2.htm). LD, suspected LD, tick bite, and other related complaints ×Probability of out-
Indirect medical costs, nonmedical costs, and produc- come clinically defined early-stage LD, clinically defined late-stage LD, suspected LD,
tivity losses were all acquired from a patient questionnaire tick bite, and other related complaints)
used in 1997 and 1998. The questionnaire was sent to LD Differences between annual mean direct medical costs
patients with informed consent forms. Collection of these were analyzed by using 1-way analysis of variance fol-
data was restricted to those 2 years. In this study, indirect lowed by a Bonferroni test. Differences were considered
medical costs refer to extra prescription and nonprescrip- significant for p values <0.05. Additionally, we used a mul-
tion drug costs that patients paid out of pocket. tivariate linear regression model to estimate the relative
The patient’s questionnaire also collected information impact of a number of factors on the direct medical costs of
on nonmedical payments made for home or health aides LD. The ordinary linear regression (OLS) method was
and miscellaneous services, such as travel (transportation) applied by using SAS 8.2 (SAS Institute, Cary, NC, USA)
and babysitting. Each patient’s transportation costs to a and Stata SE (StataCorp LP, College Station, TX, USA).
physician’s office were estimated by using the US federal The dependent variable was total direct medical cost per
government reimbursement rate, multiplying the reported LD patient. We transformed total direct medical costs by
total travel miles per patient by $0.365/mile. Total travel using natural logarithms because the data were highly
mileage per patient was calculated by counting the number skewed. Independent variables of the equation included
of physician visits and multiplying total visits by the dis- cohort year, LD diagnosis groups, diagnostic and treatment
tance of a round trip to the physician’s office. procedures, and patient characteristics (e.g., sex, age). All
We used patient-reported time lost from work to esti- independent variables, except age, were binomial (yes = 1,
mate productivity losses due to LD on the basis of the no = 0). Baseline costs (i.e., the intercept term in the regres-
human capital method and valued the time lost by using sion equation) referred to those costs accrued by a woman
age- and sex-weighted productivity valuation tables (18). who had tick bite only (without EM symptoms) diagnosed

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 655
RESEARCH

in 1997 during an office visit. Such a patient had no hospi-

tal or ER stay, no serologic tests, no consultation from other
physicians, no antimicrobial drug therapy, and no other pro-
cedures outside a physician office and hospital/ER.
Additional direct medical costs were added or subtracted to
the baseline costs for each independent variable of interest
if significant (Appendix 3, available online at
http://www.cdc.gov/ ncidod/EID/vol12no04/05-
0602_app3.htm). We tested heteroscedasticity in Stata and
corrected mild heteroscedasticity by using “robust” and
“hc3” procedures. We also tested both linearity and multi-
collinearity in SAS and Stata.

Results
From 1997 to 2000, we identified 3,415 LD-relevant
patients in the 5 counties studied on Maryland Eastern
Shore (Table 2). Among them, 10% had clinically defined
early-stage LD while almost 5% of all patients had clini- 1997 to $464 ($5–$5,338) in 2000 (p<0.05). The mean
cally defined late-stage LD. Of 284 patients who returned direct medical cost of clinically defined late-stage LD
a completed patient questionnaire, 59 patients had clinical- decreased from $4,240 ($275–$24,985) in 1997 to $1,380
ly defined early-stage LD; 25 patients had clinically ($45–$6,918) in 2000 (p<0.05).
defined late-stage LD. From 1997 to 2000, the mean cost of therapy of all
Table 3 provides cohort years, medians, means, and diagnosis groups decreased 75%, from $189 to $47, and
standard deviations of direct medical costs comparing the the mean cost of hospitalization/ER decreased 61%, from
different diagnosis groups. During the study’s time frame, $41 to $16 (Figure 2). During the same period, the mean
the mean (range) direct medical cost of clinically defined cost of an office visit, consultation, and serologic tests also
early-stage LD decreased from $1,609 ($95–$11,286) in decreased 20%, 15%, and 4%, respectively. Additionally,

656 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006
 Economic Impact of Lyme Disease

During the survey period, the mean productivity loss of

clinically defined early-stage LD was $411 in 1997 and
$88 in 1998, and the mean productivity loss of clinically
defined late-stage LD was $7,762 in 1997 and $9,108 in
1998. For all 3 types of costs shown in Table 4, a large dif-
ference was seen between mean and median values, with
the latter often less than half of the mean value, indicating
that a small number of LD patients account for a large por-
tion of total costs.
Using multivariate linear regression analysis, we found
that patients with clinically defined early- and late-stage
LD had direct medical costs that were ≈50% and 100%,
respectively, higher (p<0.001) relative to patients who
Figure 2. Distribution of elements of direct medical cost (US$) per only had tick bite, if the impact from other factors was not
Lyme disease (LD) patient in Maryland Eastern Shore
(1997–2000). Mean is based on direct medical costs of LD
considered (Table 5). Moreover, patients who were hospi-
patients. Direct medical costs were collected from medical record talized or made ER visits, who underwent serologic test-
abstraction (1997–2000). Direct medical costs of LD included ing, who needed therapy, who were referred for
costs of physician visits, consultation, serologic tests, procedure, consultation, and who had other procedures had substan-
therapy, hospitalization/emergency room, and other relevant tially (p<0.001) higher direct medical cost than those who
costs. All costs were converted to 2000 equivalent.
did not (Table 5). No cost difference was seen between
men and women. After controlling for other factors, direct
the proportion of patients within the highest percentile medical costs per LD patient in 2000 were lower than
(95th percentile for all 4 years) of therapy cost gradually those in 1997 (Table 5).
decreased from 8% in 1997 to 7% in 1998, to 4% in 1999, In year 2000 dollars, the expected mean total cost
and 3% in 2000 (data available upon request). attributable to LD was $1,965 per patient, and the expect-
A patient with clinically defined early-stage LD paid an ed median total cost attributable to LD was estimated at
average of $164 in 1997 and $307 in 1998 (in 2000 dol- $281 per patient (Figure 3). For LD patients at the clinical-
lars) for extra prescription and nonprescription drugs ly defined early stage, the median total cost was ≈$397
(Table 4). Those with clinically defined late-stage LD paid, (mean $1,310), whereas for patients at the clinically
for similar items, an average of $579 in 1997 and $389 in defined late stage, the median cost rose to $923 (mean
1998. The mean nonmedical cost for clinically defined $16,199). Suspected LD cases, tick bite cases, and other
early-stage LD was $109 in 1997 and $23 in 1998. For LD-related complaints had median costs of $238 (mean
patients with clinically defined late-stage LD, mean non- $461), $108 (mean $316), and $256 (mean $714), respec-
medical costs were $60 in 1997 and $6,703 in 1998. tively.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 657
RESEARCH

Discussion costs, and productivity losses. In 2002, 23,763 LD cases

Previous studies of the economic impact of LD were were reported to CDC. Hence, the estimated nationwide
often based on numerous assumptions and experts’ sug- annual economic impact of LD and relevant complaints
gestions (e.g., Maes et al. [15]). Only a few studies provid- was ≈$203 million (in 2002 dollars). However, since LD
ed cost estimates of LD based on data collected from the cases reported on the basis of the surveillance case defini-
field (e.g., Fix et al. [19], Strickland et al. [20]). Even in tion are believed to be underreported (13,21), this nation-
those studies, however, cost estimates only related to wide estimate is likely to be low.
direct medical charges or certain diagnosis or treatment We found that the average cost per LD case decreased
procedures. By combining data from medical records with over the study period. In LD-endemic areas, personal pro-
results from a patient survey, this study more comprehen- tection measures are frequently emphasized and insecti-
sively documents the economic impact of LD from a soci- cides are widely used (22). Persons in LD-endemic areas
etal perspective. likely visit physicians more frequently whenever they have
To approximate the annual economic impact of LD an exposure or an insect bite, and physicians attending
nationwide, we extrapolated our results to the total number patients from an LD-endemic area likely order serologic
of LD cases reported nationwide. In this study, the annual testing for possible LD patients and provide prompt treat-
total direct medical cost of LD cases on Maryland Eastern ment. However, our current evidence was limited in that
Shore was $1,455,081; 490 cases were in the clinically we were only able to find a decrease in per capita cost
defined early or late stage of LD. Total indirect medical within diagnosis groups (e.g., clinically defined early- and
costs, nonmedical costs, and productivity losses were late-stage LD), but we could not find a shift in the number
$436,949; 84 cases were clinically defined early- or late- of cases from late to early stage. Therefore, we don’t know
stage LD. Therefore, in general, an LD patient (clinically what caused the decrease in average cost per LD case.
defined early or late stage) costs $2,970 in direct medical This study has certain limitations. First, we used clini-
costs plus $5,202 in indirect medical costs, nonmedical cal case definition (physician determination) instead of

658 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006
 Economic Impact of Lyme Disease

and prevention (24). More research on the social behavior

of LD patients and economic evaluation of LD prevention
interventions is needed.

Acknowledgments
We thank David T. Dennis and G. Thomas Strickland for
their help and support.

This project was supported by CDC.

Dr Zhang is a health services researcher and health econo-

Figure 3. Expected mean (median) cost per Lyme disease (LD) mist with CDC. His research interests include economic evalua-
patient in Maryland Eastern Shore by using LD outcome tree. tion of disease prevention, public health intervention, medical
Direct medical costs collected from medical record abstraction technology, and strategic development of public health planning
(1997-2000). Indirect medical costs, nonmedical costs, and pro-
and emergency preparedness.
ductivity losses were acquired from patient questionnaire
(1998–1999). The mean (median) of all costs was aggregated
across all diagnostic groups of patients. Percentages refer to prob-
References
abilities of outcome of a possible LD case (clinically defined early-
stage LD, clinically defined late-stage LD, suspected LD, tick bite, 1. Steere AC, Malawista SE, Snydman DR, Shope RE, Andiman WA,
and other related complaints). Total percentages do not add to Ross MR, et al. Lyme arthritis: an epidemic of oligoarticular arthritis
100% because of rounding. All costs were converted to 2000 in children and adults in three Connecticut communities. Arthritis
equivalent. Rheum. 1977;20:7–17.
2. Bujak DI, Weinstein A, Dornbush RL. Clinical and neurocognitive
features of the post Lyme syndrome. J Rheumatol. 1996;23:1392–7.
surveillance case definition of LD because of limited data. 3. Shadick NA, Phillips CB, Logigian EL, Steere AC, Kaplan RF,
Thus, we may have overestimated the number of LD cases. Berardi VP, et al. The long-term clinical outcomes of Lyme disease.
As a result of case definition, our estimation of cost not A population-based retrospective cohort study. Ann Intern Med.
only included the cost of LD (clinically defined early- and 1994;121:560–7.
4. Johnson BJ, Robbins KE, Bailey RE, Cao BL, Sviat SL, Craven RB,
late-stage LD) but also the costs of LD-relevant complaints et al. Serodiagnosis of Lyme disease: accuracy of a two-step approach
(suspected LD, tick bite, and other related complaints). using a flagella-based ELISA and immunoblotting. J Infect Dis.
Second, medical charges used in our study may not reflect 1996;174:346–53.
the true cost. Third, our results are likely to underestimate 5. Dressler F, Whalen JA, Reinhardt BN, Steere AC. Western blotting in
the serodiagnosis of Lyme disease. J Infect Dis. 1993;167:392–400.
the costs per case because some of the costs were not 6. Tugwell P, Dennis DT, Weinstein A, Wells G, Shea B, Nichol G, et al.
included. Costs that were omitted included any costs Laboratory evaluation in the diagnosis of Lyme disease. Ann Intern
incurred by a patient beyond the study period. Likewise, Med. 1997;127:1109–23.
Steer et al. reported that ≈7% of LD cases remained 7. Steere AC, Levin RE, Molloy PJ, Kalish RA, Abraham JH III, Liu
NY, et al. Treatment of Lyme arthritis. Arthritis Rheum.
asymptomatic within the 20-month study (23). These 1994;37:878–88.
asymptomatic patients may have costs beyond the study. 8. Wormser GP, Nadelman RB, Dattwyler RJ, Dennis DT, Shapiro ED,
Public health surveillance and administration costs and Steere AC, et al. Practice guidelines for the treatment of Lyme dis-
intangible costs (e.g., costs incurred because of pain and ease. The Infectious Diseases Society of America. Clin Infect Dis.
2000;31(Suppl 1):S1–14.
suffering) were also not incorporated in the study. Fourth, 9. Benenson AS. Control of communicable diseases manual. 16th ed.
because of the large variance between mean and median Washington: American Public Health Association; 1995.
costs, using mean cost to estimate national impact could be 10. Hayes EB, Maupin GO, Mount GA, Piesman J. Assessing the preven-
an overestimation. Finally, this study is also limited in that tion effectiveness of local Lyme disease control. J Public Health
Manag Pract. 1999;5:84–92.
we only had information for indirect medical costs, non- 11. Centers for Disease Control and Prevention. Recommendations for
medical costs, and productivity losses from ≈8% of total the use of Lyme disease vaccine: recommendations of the Advisory
patients in the study. Therefore, the results from survey Committee on Immunization Practices (ACIP). MMWR Morb Mortal
data were extrapolated to represent the whole study popu- Wkly Rep. 1999;48(RR-7):1–17, 21–5.
12. Malawista SE. Lyme disease. In: Goldman L, Bennett JC, editors.
lation. This method may have biased our results. Cecil textbook of medicine. 21st ed. Philadelphia: W.B. Saunders
LD is the most common vectorborne zoonotic inflam- Company; 2000. p. 1757–61.
matory disease in the United States. The longterm seque- 13. Centers for Disease Control and Prevention. Lyme disease—United
lae of LD are debilitating to patients and costly to society. States, 2000. MMWR Morb Mortal Wkly Rep. 2002;51;29–31.
14. Centers for Disease Control and Prevention. Notice to readers: final
The emergence of LD and previous experience predict the 2002 reports of notifiable diseases. MMWR Morb Mortal Wkly Rep.
feasibility of public health interventions for LD control 2003;52:741–50.

Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006 659
RESEARCH

15. Maes E, Lecomte P, Ray N. A cost-of-illness study of Lyme disease 21. Meek JI, Roberts CL, Smith EV Jr, Cartter ML. Underreporting of
in the United States. Clin Ther. 1998;20:993–1008. Lyme disease by Connecticut physicians, 1992. J Public Health
16. Centers for Disease Control and Prevention. Case definitions for Manag Pract. 1996;2:61–5.
infectious conditions under public health surveillance. MMWR Morb 22. Barbour AG, Fish D. The biological and social phenomenon of Lyme
Mortal Wkly Rep. 1997;46(RR-10):20–1. disease. Science. 1993;260:1610–6.
17. US Department of Labor. Consumer price index. [cited 2004 June 23. Steer AC, Sikand VK, Schoen RT, Nowakowski J. Asymptomatic
16]. Available from http://www.bls.gov/cpi/home.htm infection with Borrelia burgdorferi. Clin Infect Dis. 2003;37:528–32.
18. Haddix AC, Teutsch SM, Corso PS, editors. Prevention effectiveness: 24. Steere AC, Coburn J, Glickstein L. The emergence of Lyme disease
a guide to decision analysis and economic evaluation. 2nd ed. New [review]. J Clin Invest. 2004;113:1093–101.
York: Oxford University Press; 2003. p. 70–1.
19. Fix AD, Strickland GT, Grant J. Tick bites and Lyme disease in an
Address for correspondence: Xinzhi Zhang, Centers for Disease Control
endemic setting: problematic use of serologic testing and prophylac-
tic antibiotic therapy. JAMA. 1998;279:206–10. and Prevention, 1600 Clifton Rd NE, Mailstop D59, Atlanta, GA 30333,
20. Strickland GT, Karp AC, Mathews A, Peña CA. Utilization and cost USA; fax: 404-371-5445; email: [email protected]
of serologic tests for Lyme disease in Maryland. J Infect Dis.
1997;176:819–21.

Search

660 Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 12, No. 4, April 2006