USING PRE-DIAGNOSTIC DATA FROM VETERINARY LABORATORIES TO DETECT DISEASE OUTBREAKS IN COMPANION ANIMALS
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree Doctor of Philosophy in the Graduate
School of The Ohio State University
By
Loren Eldon Shaffer, MPH
*****
The Ohio State University 2007
Dissertation Committee: Approved by Professor William J.A. Saville, Advisor
Professor Julie Funk
Professor Päivi Rajala-Schultz ______Advisor Professor Michael M. Wagner Graduate Program in Veterinary Preventive Medicine Professor Thomas E. Wittum
ABSTRACT
Emerging infectious diseases and the threat of bioterrorism have fostered a desire for improved timeliness of outbreak detection. Traditional disease reporting is reliant on confirmed diagnoses, often involving laboratory analysis that may require days to complete. Most emerging infectious and bioweapon pathogens are zoonotic organisms.
Detection of zoonotic outbreaks has often relied on the identification of human cases. We investigated how data from veterinary diagnostic laboratories (VDLs) might contribute to earlier outbreak detection efforts in Ohio.
We began by determining the representation of animal species in the data and evaluating the representation of human households. Companion animals comprised
98.1% of the total number of specimens submitted to a commercial, nation-wide VDL from clinics in Ohio in one year. Using estimates derived from a survey of pet owners, we determined that these data represented approximately 6.6% of Ohio households.
The value of microbiology test orders was determined by quantifying the representation and potential gain in timeliness from two VDL datasets. We also investigated the potential to determine estimated count values from historical records and detect significant increases in these values using statistical-based detection methods. The data represented specimens from mostly companion animals (85.0% and 74.3%) followed by horses (8.2% and 17.2%). We determined a potential gain of timeliness in outbreak
ii
detection of three to five days. We developed baselines of microorganism incidence and
total microbiology orders from the datasets and detected some of the clusters of
pathogen-specific isolates by analyzing the weekly totals of all microbiology orders.
We demonstrated how someone might use these data in a prospective system to
detect outbreaks of disease earlier than traditional methods. Case reviews from a pilot
system indicated the potential benefit to public health as well as veterinary community.
We concluded from these investigations that: 1) data from VDLs do possess certain qualities that validate their value for syndromic surveillance, 2) these data may be especially useful for surveillance in companion animals, and 3) earlier detection of certain disease outbreaks may be possible from a prospective system using VDL data.
iii
Dedicated to my wife Kelly for her support, partnership, and patience; and to our children
Taylor, Alex, Hunter, and Nate. May their dreams become their reality.
iv
ACKNOWLEDGMENTS
I wish to thank all of my advisors for their support, interest, and guidance during my program. To Drs. Julie Funk, Päivi Rajala-Schultz, and Tom Wittum, I greatly appreciate the insights and advice regarding veterinary issues that were outside of my experience. My deepest appreciation is extended to Dr. William Saville, for his guidance and mentorship in academic, political, and common sense matters. Thank you, Dr.
Saville, for your humble nature and curiosity that provided me the opportunity to explore something new to us both. I am also much indebted to Dr. Mike Wagner for his willingness to share his insights on public health surveillance, informatics, and the interactions that are necessary on many fronts to make things work. I am thankful for your honest and straightforward approach.
My thanks are also extended to Dr. Bob Campbell for his support of my decision
to enter this program while working for him full-time. Your confidence in me did not go
unnoticed and your words of encouragement seemed to come at just the right time.
I also offer my appreciation to Dr. Richard Bednarski, Bobbi Schmidt, and Fred
Marker at The Ohio State University Veterinary Teaching Hospital for providing me
some of the data resources critical to this research. Thanks also to Dr. Bill Wallen, Dr.
Dave Fisher, Dr. Jocelyn Johnsrude, Gary Watson, Robert Ledford, and Bill Davis at
IDEXX Laboratories, Inc. not only for providing me access to data resources but
v
especially for helping me to learn what to ask for and how to understand it. Dr. Garrick
Wallstrom at the RODS Laboratory and Steve DeFrancesco and Kevin Hutchison, both formerly of the RODS Laboratory, are also much due my gratitude for assisting me with both IT and statistical support and advice. For all those others not listed specifically that helped me during these years I am no less grateful. Your support and help came in many different forms. I am thankful for it all.
Lastly, I would humbly offer my gratitude to Rick and Pat Thrall. Although you acquired me as a son-in-law, you both accepted me as if I was your own. My own parents long since passed, I appreciate having you as my “Dad” and “Mom.” The pride you often expressed helped me many times to maintain my motivation.
vi
VITA
August 15, 1962 …………………………Born – Lancaster, Ohio
2003 …………………………………….. MPH, The Ohio State University
2000 - 2003 …………………………….. Preventive Medicine Officer United States Army Reserve Special Operations Command
2003 - 2004 …………………………….. Emergency Preparedness Manager Franklin County Board of Health Columbus, Ohio
2004 - present ………………………….. Early Event Surveillance Supervisor Ohio Department of Health Columbus, Ohio
PUBLICATIONS
1. L. E. Shaffer, “Syndromic Surveillance.” The Ohio Department of Health Infectious Disease Quarterly. 2005; 2(2): 1-4.
2. M. W. Wagner, L.E. Shaffer, and R. Shephard, “Biosurveillance Systems.” in Handbook of Biosurveillance. Elsevier Press, New York, NY, 2006.
3. R. Aryl, R. Shephard, and L. E. Shaffer, “Animal Health.” in Handbook of Biosurveillance. Elsevier Press, New York, NY, 2006.
4. L. E. Shaffer, S. A. Rowe, and D. E. Reed, “Early Detection of Influenza-like Illness: Developing a Multi-Variate Approach.” (Abstract) Advances in Disease Surveillance. 2007; 2(65): 67.
5. L. E. Shaffer, J. A. Funk, P. Rajala-Schultz, M. M. Wagner, T. E. Wittum, W. J. A. Saville. “Evaluation of Veterinary Diagnostic Laboratories as a Possible Data Source for Prospective Outbreak Surveillance.” (Abstract) Advances in Disease Surveillance. 2007; 2(66): 119.
vii
6. L. E. Shaffer, J. A. Funk, P. Rajala-Schultz, G. Wallstrom, M. M. Wagner, T. E. Wittum, W. J. A. Saville. “Early Outbreak Detection Using an Automated Data Feed of Test Orders from a Veterinary Diagnostic Laboratory.” Lecture Notes in Computer Science. 2007; 4506:1-10.
7. L. E. Shaffer, J. A. Funk, P. Rajala-Schultz, M. M. Wagner, T. E. Wittum, W. J. A. Saville. “Evaluation of Microbiology Orders from Two Veterinary Diagnostic Laboratories as Potential Data Source for Early Outbreak Detection.” Advances in Disease Surveillance. 2007; forthcoming.
8. L. E. Shaffer, J. A. Funk, P. Rajala-Schultz, M. M. Wagner, T. E. Wittum, W. J. A. Saville. “Clinical Rotation of Senior Veterinary Students as a Confounder for Outbreak Detection Using Microbiology Orders in a Veterinary Teaching Hospital.” Journal of Veterinary Medical Education. Under review.
9. L. E. Shaffer, J. A. Funk, P. Rajala-Schultz, M. M. Wagner, T. E. Wittum, W. J. A. Saville. “Contributing to a One-Medicine Approach: Cross-Species Disease Surveillance.” Public Health Reports. 2008; Invited paper in preparation.
FIELDS OF STUDY
Major Field: Veterinary Preventive Medicine
viii
TABLE OF CONTENTS
Page
Abstract ……………………………………………………………………….…………..ii
Dedication ………………………………………………………………………………..iv
Acknowledgments ………………………………………………………………………..v
Vita ……………………………………………………………………………………...vii
List of Tables …………………………………………………………………………....xii
List of Figures …………………………………………………………………………..xiv
Chapters:
1. Introduction …………………………………………………………………….....1
2. Literature Review ……………………………………………….………………...7
2.1 Bioterrorism ………………………………………………….……………….7 2.2 Emerging Infectious Diseases .………….…………………………………….8 2.3 Deficiencies in Detecting Outbreaks ………………………………………..10 2.4 Syndromic Surveillance ……………………………………….…………….13 2.4.1 Improving the Timeliness of Detection …………………………...16 2.4.2 Detection Methods ………..……………………………………….17 2.4.3 Limitations ………………………………………………………...19 2.5 “One-Medicine” and Animals as Sentinel Indicators ………………....…….21 2.6 Animal-based Syndromic Surveillance Initiatives …………………………..23 2.7 Summary …………………………………………………………………….25
3. Sentinel Surveillance of Human Households Using Companion Animals .….....41
3.1 Abstract ...……………………………………………………………………41 3.2 Introduction ………………………………………………………………….42
ix
3.3 Methods ……………………………………………………………………...43 3.3.1 American Veterinary Medical Association Survey …...... ….43 3.3.2 Data Sample ……………………………………………………….43 3.3.3 Calculations ………………………………………………………..44 3.4 Results ……………………………………………………………………….46 3.5 Discussion ….………………………………………………………………..46
4. Evaluation of Microbiology Orders from Two Veterinary Diagnostic Laboratories as Potential Data Sources for Early Outbreak Detection ………….53
4.1 Abstract ……………………………………………………………………...53 4.2 Introduction ……………………………………………………………….…54 4.3 Methods ……………………………………………………………………..57 4.3.1 Data Sample ……………………………………………………….57 4.3.2 Statistical Analysis ………………………………………………...57 4.4 Results ……………………………………………………………………….62 4.4.1 Descriptive Statistics ………………………………………………62 4.4.2 Baseline Modeling ………………………………………………...63 4.4.3 Detection Methods ………………………………………………...63 4.5 Discussion …………………………………………………………………...64
5. Clinical Rotation of Senior Veterinary Students as a Confounder for Outbreak Detection Using Microbiology Orders in a Veterinary Teaching Hospital ……………………………………………………………………….…82
5.1 Abstract …………………………………………………………………...…82 5.2 Introduction ………………………………………………………………….83 5.3 Methods ……………………………………………………………………...84 5.4 Results ……………………………………………………………………….85 5.5 Discussion …………………………………………………………………...86
6. Early Outbreak Detection Using Test Orders in an Automated Data Feed from a Veterinary Diagnostic Laboratory …………………………………94
6.1 Abstract ……………………………………………………………………...94 6.2 Introduction ……………………………………………………………….....95 6.3 Methods …………………………………………………………………...…96 6.3.1 Mapping Test Orders to Syndromic Categories ……………..…….97 6.3.2 Descriptive Statistical Analyses ………………………………...…98 6.3.3 Detection Method …………………………………………...….….98 6.4 Results …………………………………………………………………….....99 6.4.1 Evaluation of Data Transfer……………………………………..…99 6.4.2 Descriptive Statistics ……………………………………………..100 6.4.3 Aberration Detection ……………………………………………..100 6.4.4 Details About the Two Alerts Associated with Disease Activity …………………………………………….101 x
6.5 Discussion ………………………………………………………...... 102
7. Summary of Dissertation …..……………………………………….………….115 7.1 Summary of Conclusions ………………………………………….……….115 7.2 Future Studies ……………………………………………………………..117
Bibliography …………………………………………………….……………………..122
Appendices:
A Diseases notifiable to the OIE ...... 138 B Nationally notifiable infectious diseases ...... 144 C IDEXX Laboratories product code to syndrome category mapping ...... 148
xi
LIST OF TABLES
Table Page
2.1 Bacterial and viral disease-causing organisms considered to be potential bioweapon threat agents………………………………………………………….27
2.2 Description of select animal-based syndromic surveillance initiatives …………28
3.1 Estimating the number of Ohio households represented in IDEXX dataset (04/01/2004 – 03/31/2005) from the number of laboratory specimen accessions ………………………………………………………………………..48
4.1 Select characteristics of microbiology datasets for specimens originating in central Ohio, January 2001 – December 2003 ...... 70
4.2 Microorganism genera isolated from specimens originating in central Ohio, January 2001 – December 2003 ...... 71
4.3 Frequency of species from central Ohio providing specimens for microbiology analysis, January 2001 – December 2003 …...... 72
4.4 Clusters determined from IDEXX microbiology test results of specimens originating from animals treated at clinics located in central Ohio, 2003 ...... 73
4.5 Clusters determined from OSU microbiology test results of specimens originating from animals residing in central Ohio, 2003 ...... 74
4.6 Performance indicators for detection algorithms in identifying microorganism clusters from time series of aggregate microbiology orders …………………………………………………………………………….76
5.1 Average weekly values for microbiology orders made during clinical rotations at The Ohio State University Veterinary Teaching Hospital in 2003 and the OR of the proportion of isolates to orders being less than average during the first weeks of rotation compared to any other weeks of rotation ………………………………………………………………………..91
6.1 Syndrome category descriptions distributed to veterinarian sample for grouping laboratory products……………………………………………….106
xii
6.2 Alerts generated by pilot prospective surveillance system using laboratory orders from select providers in Ohio from September 1 through November 30, 2006 and summary of follow up findings with area veterinary providers …………………………………………………………..111
xiii
LIST OF FIGURES
Figure Page
3.1 Percentage of laboratory specimen accessions by genera received by IDEXX from clinics located in Ohio, April 1, 2004 – March 31, 2005 ………...49
3.2 Estimates of Ohio household representation by animals with consideration of possible over- and underestimation of AVMA Household Survey values …..50
4.1 Seven county region of central Ohio where microbiology specimens originated ...... 68
4.2 Counts of IDEXX and OSU microbiology orders originating from central Ohio, 2003 ...... 69
4.3 Counts of OSU microbiology orders and isolates for specimens received from patients residing in central Ohio during 2003 ...... 75
5.1 Counts of microbiology test orders and cultures that produced isolates originating from The Ohio State University Veterinary Teaching Hospital during 2003 ……………………………………………………………………...89
5.2 Proportion of cultures with isolates to microbiology orders by week originating from The Ohio State University Veterinary Teaching Hospital during 2003 ...... 92
6.1 Distribution of specimen origin for accessions submitted from veterinary clinics in Ohio to IDEXX from September 1 through November 30, 2006 compared to ZIP code population ……………………..…107
6.2 Counts of specimens received by IDEXX from veterinary clinics in Ohio from September 1 through November 30, 2006 …………………………….....108
6.3 Representation of animal species in prospective accession datasets received from IDEXX September 1 through November 30, 2006 ………………………109
6.4 Delay in receipt of daily records from IDEXX during prospective pilot ……...110
xiv
7.1 Determining the number of outbreaks required to occur during pilot surveillance system to achieve the desired level of power in evaluation measures ……………………………………………………………121
xv
CHAPTER 1
INTRODUCTION
Using infectious disease agents as weapons of war dates back at least to1346
when invading Tattar forces hurled plague victims over the walls of Kaffa (1). During
World War I, the Germans used Pseudomonas mallei (glanders) and Bacillus anthracis
(anthrax) to infect horses being shipped from the United States, Argentina, and Morocco
to Europe (2). Biologic agents have also been tools for terrorists. In 1984, followers of
Bhagwan Shree Rajneesh (the Rajneeshee) contaminated salad bars in The Dalles,
Oregon with Salmonella resulting in the illness of 751 people (3). In late 2001, terrorists
used the U.S. postal service to deliver spores of Bacillus anthracis (anthrax) to
unsuspecting recipients. Such attacks, especially occuring on the heels of the terrorist events of September 11, 2001, increased the attention on efforts to improve how quickly outbreaks of disease might be discovered (1, 4, 5).
Outbreak detection has typically occurred through reports of individual cases of
disease or the isolation of a disease-causing organism (6). Isolation of a pathogen is time
intensive and may do little to reduce the morbidity and economic losses caused by those
outbreaks of disease with shorter incubation time (7). These methods also tend to be quite
pathogen-specific and therefore limiting since many diseases associated with biological
1
weapons initially present with non-specific signs and symptoms. The term emerging infectious disease first appeared in the 1980’s, referring to those newly recognized,
clinically distinct diseases or known diseases that increase in incidence in a specific place
or population (8).
Most of the pathogens associated with bioweapons and emerging infectious
diseases are zoonotic (capable of infecting both humans and one or more species of lesser
animal) (8-11). Although humans serve as the primary reservoir for only 3% of all known
zoonotic pathogens (9), detection of outbreaks of zoonotic disease have often relied upon
identification of human cases (12, 13). Most current disease surveillance systems in
animals tend to be disease-specific and typically directed by regulatory programs (14).
Some efforts to detect outbreaks earlier in humans have used pre-diagnostic
indicators of disease in lieu of reports of cases. These indicators have included chief
complaints of emergency department patients (15, 16), sales of nonprescription
medication (17, 18), laboratory orders (19), self-reports of gastrointestinal and diarrheal
illness (20), grocery sales (21), and telephone triage logs (22). Early outbreak detection
methods developed for humans might provide for earlier detection of outbreaks in animal
populations (23, 24).
Earlier detection of an outbreak might also lead to earlier intervention efforts
resulting in decreased morbidity and mortality, and reduced economic impact (25-27).
Animals may exhibit signs of infection earlier than humans exposed at the same time to
an infectious agent (10, 28). Companion animals (e.g. dogs, cats) share much of the same
environment as their human owners and therefore may be exposed to many of the same
2
disease-causing organisms as well (29, 30). Hence, animals might serve as sentinels for
certain zoonotic disease outbreaks in humans. Early outbreak detection in animals might
therefore benefit public health as well as veterinary medicine.
Certain pre-diagnostic veterinary data sources appear to have similar characteristics of data used by human early outbreak detection systems. Test orders are
such a source. As opposed to a regulatory program, diagnostic laboratory orders are
generally the result of a clinical disease event (31, 32). Records of these orders include
information that identifies time and place that permits analysis to detect clusters. These records also precede diagnosis. Their true value for early outbreak detection is unknown.
The overarching goal of this work was to determine the value of these data and investigate the potential for using them for early outbreak detection.
3
REFERENCES
1. Venkatesh S, Memish ZA. Bioterrorism -- a new challenge for public health. Int J Antimicrob Agents 2003;21:200-6.
2. Federation of American Scientists. The Role of Disease Surveillance in the Watch for Agro-terrorism or Economic Sabotage. October 2001.
3. Torok TJ et al. A Large Community Outbreak of Salmonellosis Caused by Intentional Contamination of Restaurant Salad Bars. J Am Med Assoc 1997;278:389-95.
4. Huerta M, Leventhal A. The epidemiologic pyramid of bioterrorism. Isr Med Assoc J 2002;4:498-502.
5. Henning KJ. Syndromic Surveillance. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, D.C., National Academy Press.
6. Koo D, Parrish II RG. The Changing Health-Care Information Infrastructure in the United States: Opportunities for a New Approach to Public Health Surveillance. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford University Press, 2000:76-94.
7. Kaufmann AF, Meltzer MI, Schmid GP. The Economic Impact of a Bioterrorist Attack: Are Prevention and Postattack Intervention Programs Justifiable? Emerg Infect Dis 1997;3:83-94.
8. Institute of Medicine. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, DC, National Academy Press.
9. Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:983-9.
10. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category B and C agents. J Am Vet Med Assoc 2004;224:1096-104.
11. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category A agents. J Am Vet Med Assoc 2004;224:1084-95.
12. Meslin FX, Stohr K, Heymann D. Public health implications of emerging zoonoses. Rev Sci Tech 2000;19:310-7.
13. Childs J et al. Emerging Zoonoses. Emerg Infect Dis 1998;4:453-4.
4
14. Wurtz RM, Popovich ML. Animal Disease Surveillance: A Framework for Supporting Disease Detection in Public Health. March 2002. Tuscon, AZ, Scientific Technologies Corporation.
15. Tsui F-C et al. Technical Description of RODS: A Real-time Public Health Surveillance System. J Am Med Inform Assoc 2003;10:399-408.
16. Heffernan R et al. Syndromic Surveillance in Public Health Practice, New York City. Emerg Infect Dis 2004;10:858-64.
17. Magruder SF. Evaluation of Over-the-Counter Pharmaceutical Sales A a Possible Early Warning Indicator of Human Disease. Johns Hopkins APL Technical Digest 24[4], 349-353. 2003.
18. Das D et al. Monitoring Over-The-Counter Medication Sales Early Detection of Disease Outbreaks--New York City. MMWR Morb Mortal Wkly Rep 2005;54:41-6.
19. Bradley CA et al. BioSense: Implementation of a National Early Event Detection and Situational Awareness System. MMWR Morb Mortal Wkly Rep 2005;54:11- 9.
20. Wethington H , Bartlett P. The RUsick2 Foodborne Disease Forum for Syndromic Surveillance. Emerg Infect Dis 2004;10:401-5.
21. Fienberg SE, Shmueli G. Statistical issues and challenges associated with rapid detection of bio-terrorist attacks. Stat Med 2005;24:513-29.
22. Platt R et al. Syndromic Surveillance Using Minimum Transfer of Identifiable Data: the Example of the National Bioterrorism Syndromic Surveillance Demonstration Program. J Urban Health 2003;80:i25-i31.
23. Kruse H, Kirkemo A-M, Handeland K. Wildlife as Source of Zoonotic Infections. Emerg Infect Dis 2004;10:2067-72.
24. Green MS, Kaufman Z. Surveillance for Early Detection and Monioring of Infectious Disease Outbreaks Associated with Bioterrorism. Isr Med Assoc J 2002;4:503-6.
25. Food and Agriculture Organization of the United Nations. Manual on Livestock Disease Surveillance And Information Systems. Rome: 1999.
26. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality.
5
27. Shephard R, Aryel RM, Shaffer L. Animal Health. In: Wagner MM, Moore AW, Aryel RM, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:111-27.
28. Babin SM et al. Early detection of possible bioterrorist events using sentinel animals. The 131st Annual Meeting of the American Public Health Association. 2003.
29. Glickman LT et al. Purdue University-Banfield National Companion Animal Surveillance Program for Emerging and Zoonotic Diseases. Vector Borne Zoonotic Dis 2006;6:14-23.
30. Backer L et al. Pet dogs as sentinels for environmental contamination. Sci Total Environ 2001;274:161-9.
31. Buehler JW. Surveillance. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins, 1998:435-57.
32. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board.
6
CHAPTER 2
LITERATURE REVIEW
2.1 Bioterrorism:
In Autumn of 2001, the deliberate delivery of Bacillus anthracis spores via the
United States Postal Service resulted in heightened concerns about human exposure to
intentionally released disease pathogens (1, 2). Since preventing bioterrorism would be
extremely difficult, it became clear that the first evidence of such an attack might be
individual cases of disease (3). This prompted a new urgency for public health and the developers of disease surveillance systems; the early detection of disease outbreaks (4).
Many of the diseases that could result from pathogens known to be potential bioweapons have relatively short periods of incubation (Table 2.1). Near maximum mortality could
occur from a release of one of these diseases in as little as a few days (5). This would imply that detection of the outbreak would need to occur before tests used to confirm diagnoses were complete.
The presentation of these diseases is not unique in their early stages and could readily be misdiagnosed as some more commonly occurring ailments (6). It was also reasonable to conclude that endemic diseases might also be the weapon of choice in an attack (7). Although accepted as an essential element to preparing for a bioterrorist
7
attack, the methods for effective disease surveillance were not as clear. Nearly all of the
pathogens used as bioweapons are zoonotic. An important epidemiologic clue of a
bioterrorist attack may be an outbreak that occurs in humans and simultaneously, or is
preceded by one, in animals (8, 9). Using animals as sentinels for human disease has been
recognized as critical to efforts enabling the early warning of a bioterrorist attack (10,
11). However, there is paucity in the literature describing the development of a surveillance system to accomplish this in a timely and effective manner.
2.2 Emerging Infectious Diseases:
First appearing in literature during the late 1980’s, “emerging infectious disease”
primarily described diseases in humans that had appeared for the first time, increased in
incidence, or were reported in new areas or populations (12). Since that time, there has
been, on average, an occurrence of about one new emerging infectious disease every 8
months (13). By definition, a disease need not be novel to be emerging. Giardia is one
such example (14).
Prevalence of both giardiasis and cryptosporidiosis has increased over the past
few decades (15). Between 1999 and 2002, the number of cases of cryptosporidiosis
reported annually increased from 2,769 to 3,016 (16). One of the contributing factors to
this increase may be the lack of host specificity of Cryptosporidium parvum, the disease-
causing agent of cryptosporidiosis (17). Like Giardia, C. parvum infects a broad range of
mammalian species and provides for the uncertainty about how host specificity and the
possibility for zoonotic transmission might both fuel the re-emergence of each of these
diseases in humans (14, 17, 18).
8
Other examples include trichinellosis. Although considered a public health threat
for more than 150 years, a dramatic re-emergence of Trichinella has occurred over just
the past decade (19). Another is Toxoplasma gondii. The presence of T. gondii in meat
animals has decreased considerably over the past 20 years; however, toxoplasmosis remains one of the most common parasitic diseases worldwide in humans (20).
The majority of emerging infectious diseases are caused by zoonotic pathogens
(21-23). Over 75% of the outbreaks attributed to emerging infectious diseases that
occurred in the last two decades of the 20th century were the result of zoonotic pathogens
(24). Eleven out of the past 12 significant epidemics that occurred in humans were also caused by zoonotic pathogens (25). Indeed, pathogens with more than one species of host were two to four times more likely to result in an emerging disease than their single host counterparts (12, 24, 26). It should not be surprising that animal contact was cited as one of the major contributing factors to emerging infectious disease in a 2003 report published by the Institute of Medicine (22). The entire extent that animal contact is involved with the transfer of systemic disease to humans remains to be determined.
Certainly, such association may not be as obvious as dermal infections and injuries received from contact with animals (27). Exotic pets in particular have been linked with both acute and chronic systemic illness, sometimes resulting in death (28, 29).
Contacts with animals in petting zoos, farms, and other animal exhibits have often
been associated with outbreaks of zoonotic disease in humans (30). Children are
especially at risk partly attributable to their curiosity and personal habits such as nail
biting and thumb sucking (28, 30). Immunocompromised people are also at increased risk
for contracting disease through animal contact. The number of people with weakened
9
immune systems has grown over the past several years because of increases in
chemotherapy, organ transplants, and human immunodeficiency virus (15).
Other domestic species, including those we keep as pets, should not be
overlooked as potential sources of emerging infectious diseases. A single study
conducted in the U.S. found 13.1% of owned cats were infected with C. parvum, Giardia,
Toxocara cati, Salmonella enterica, or Campylobacter jejuni (31). Another study, this
one in Sweden, associated cats as the intermediary host in spreading Salmonella
thyphimurium from wild birds to humans (32). Recently, it has been thought that cats might serve a similar role in the transmission of avian influenza viruses (33). Domestic poultry have also been considered (34) along with dogs (35) as concern mounts that these influenza viruses may cross species barriers and develop into our next pandemic.
Although the path of transmission remains debatable, all of these species are potential
host of at least certain influenza viruses making the development of a multi-species
surveillance system a desire and a challenge (36).
2.3 Deficiencies in Detecting Outbreaks:
Most of the surveillance systems currently in operation are considered inadequate
for detecting outbreaks of non-notifiable diseases or outbreaks resulting from the
intentional release of a disease pathogen (37-40). Indeed, many local public health
officials interviewed for a 2003 United States General Accounting Office report had
concerns that their existing surveillance systems were adequate to detect a bioterrorism
event (41). One study found that out of 1,099 infectious disease outbreaks investigated by
public health, 40% were caused by a pathogen with bioterrorism potential (42). However,
10
only 5% were discovered using existing surveillance systems. A 2000 report from the
Chemical and Biological Institute pointed out that most surveillance networks in place depend on an official diagnosis of disease (43). The pathogens most likely to be used as bioweapons result in disease with non-specific symptoms (37). Therefore, accurate diagnosis of an illness may depend upon the results of medical testing. Laboratory analysis could take days to complete. Additionally, new emerging pathogens may not have been previously identified further hindering reporting to public health authorities, as the pathogen is also not yet included on any reportable disease list (43). Chronic underreporting and time lags between diagnoses and reporting are noted deficiencies in these systems potentially resulting in delayed or missed outbreak identification (41, 44).
Underreporting has been considered by some to be the single most important constraint to the effectiveness of any disease surveillance system (45).
As noted previously, the majority of disease pathogens with bioweapon potential and those associated with most emerging diseases are zoonotic. In spite of this, outbreaks of zoonotic disease have typically relied on the identification of human cases (46, 47).
The possible economic implications drive surveillance for disease in agricultural animals rather than any zoonotic potential of the pathogen (48). Most outbreaks from zoonotic pathogens in animals do not have the economic implications or direct impact on humans as other diseases and therefore may not be considered as important (49). Hence, most disease surveillance in animals targets agricultural species and tends to be very disease specific. The detection of specific diseases is limited, among other things, by the ability
11
of farmers and veterinarians to recognize and associate the clinical signs with a notifiable
condition (50, 51). The disease specificity of such surveillance greatly reduces the ability
to detect outbreaks resulting from other than target diseases (49).
The National Surveillance Unit (NSU) of the United States Department of
Agriculture (USDA) identified surveillance for emerging diseases as a priority for the
National Animal Health Reporting System (NAHRS) (52). The NAHRS receives data from state veterinarians in participating states on the presence of confirmed clinical disease (53, 54). Diseases are limited to those that are reportable to the World
Organization for Animal Health (OIE) (Appendix A) in specific commercial species in
the United States including cattle, sheep, goats, equine, swine, poultry, and food fish.
Only six diseases on the OIE list also appear on the list of United States nationally
notifiable infectious diseases for humans (Appendix B) of which 38 out of 58 (65.5%) are
zoonotic. The NAHRS has been criticized as a passive, voluntary system without quality control, verification, or feedback (49). Passive collection of data, such as that utilized by the NAHRS, is limited by the inconsistency in collection for different diseases and among different states (55). The NAHRS provides little benefit as far as timeliness since a national summary report is compiled from data only monthly data (54).
The National Animal Health Laboratory Network (NAHLN) is another USDA project intended to provide for earlier detection and tracing of outbreaks (48). The
NAHLN, as part of a strategy to coordinate the Federal, State, and university laboratory resources, was promised to be “a cornerstone of animal health surveillance that will electronically connect surveillance data systems to laboratory diagnostics” (56, 57).
Initiated in 2002, the NAHLN included only 12 laboratories as of February 2006 (58).
12
Focus is again disease specific and limited to African swine fever, avian influenza,
classical swine fever, contagious bovine pleuropneumonia, exotic Newcastle disease, foot
and mouth disease, lumpy skin disease, and rinderpest in agricultural animals (47, 56).
The NAHLN has been considered to lack the capacity to deal with massive or multiple
outbreaks in the United States (59).
The lack of integration between human and animal disease surveillance has not gone unnoticed. At least one report published by the National Research Council recognizes the need for better integration of animal and public health surveillance in order to improve on the capabilities to rapidly detect outbreaks caused by zoonotic pathogens (60). The United Nations Food and Agriculture Organization, OIE, and the
World Health Organization went as far as to say that the weaknesses in disease detection have contributed to the spread of diseases of animal origin (61).
2.4 Syndromic Surveillance:
The application of syndromic surveillance methods has been recommended to detect novel and emerging zoonoses including those associated with potential
bioweapons (23). Where traditional surveillance has involved the manual reporting of
individual cases of disease or pathogen isolates, the availability of electronic data,
coupled with advances in information technology, have enabled improved surveillance of
infectious disease (62). One such approach is syndromic surveillance; the systematic and
ongoing collection, analysis, and interpretation of data that precede diagnosis and that can
signal a sufficient probability of an outbreak to warrant public health investigation (63).
Syndromic surveillance is not considered a replacement for traditional disease reporting
13
but rather supportive to aid in outbreak detection and monitoring of disease trends (64-
67). Syndromic surveillance has been described as “an investigational approach where
health department staff, assisted by automated data acquisition and generation of
statistical alerts, monitor disease indicators in real-time or near real-time to detect
outbreaks of disease earlier than would otherwise be possible with traditional public
health methods” (68). Systems that are based on this approach have also been referred to
as early warning, prodrome surveillance, outbreak detection, information system-based
sentinel surveillance, biosurveillance, health indicator surveillance, symptom-based surveillance, pre-diagnosis surveillance, non-traditional surveillance, enhanced
surveillance, and drop-in surveillance systems (63, 66).
Syndromic surveillance does not depend on specific, clinical, or laboratory-
defined disease but rather uses counts of health indicators, aggregated into categories
based on same or similar characteristics (e.g. disease symptoms), to detect aberrations in
trends that may indicate unusual incidence of disease. As a proxy to illness, there are many options to consider when choosing which health indicator/s to use for surveillance.
Data sources have included the verbalized complaints of emergency department patients
(69-71), sales of grocery items (72), sales of over-the-counter medicines (73-75), descriptions of symptoms from callers to nurse triage centers (76), self reports of gastrointestinal illness made via the internet (77), billing codes (74, 78, 79), temperature values (80), absenteeism (74), laboratory test orders (78), and news reports (81). The source of data is open to any indicator that may be present during the disease process.
Animal data has been identified as a potential and needed source of data for these early detection efforts (37, 82, 83).
14
No one single data source is sufficient for surveillance (64, 84). Outbreaks of
different disease will affect various data sources differently (85). For example,
surveillance of thermometer sales will probably do little to alert to an outbreak of disease
that does not present with fever. There are other inherent problems when using these
types of data for disease surveillance (67, 86). As the data are only proxy to a health
event, the specificity is presumed to be low. For this same reason, there is a high
probability of the data being influenced by non-health related factors. Since the majority
of these data are not associated to an individual level there can be difficulty in retracing
certain data aberrations.
The selection of data sources for syndromic surveillance are initially influenced
by evidence or the belief of the system developers that the source can provide an early
signal for the disease/s or condition/s of interest (85). It may be difficult to measure the
true value of non-traditional data sources, especially those that are not related to healthcare, as outbreaks of the diseases of interest are potentially rare (87). In these cases, system developers must rely on retrospective studies of other diseases with similar presentation or surveys to measure behaviors that could influence the data. The most valuable data sources will be those that are stored electronically, permit robust syndromic grouping, and are available in a timely fashion (88). Systems that require additional data entry and increase workloads were undesirable to data providers, especially for large- scale, sustained surveillance (43, 67, 84, 89). Using existing data that have already been collected for another purpose, are stored electronically, and can be transferred automatically are preferable as they do not depend on changes in workflow (37, 90).
15
Veterinary diagnostic laboratories (VDLs) are a potential source of data for
outbreak detection and considered important tools for surveillance in animals (48). There
are inherent biases associated with laboratory data that may affect the sensitivity and
specificity of analysis that developers need to consider (82, 91). One such bias has been
referred to as a hierarchy of scrutiny (92) and may actually serve to increase the
specificity of the system. A hierarchy of scrutiny extends from the bias that exists in
laboratory specimens only being submitted from clinically ill patients (93). Although
many cases of sub-clinical disease may be missed, the probability of the test order being
linked to disease is increased. These data are typically stored electronically, records are
created as part of the normal workflow, and large geographic areas are covered by a
single source. While there are indications of the value of medical laboratories for outbreak surveillance (94), there is paucity in the literature supporting the use of veterinary laboratories.
2.4.1 Improving the Timeliness of Detection:
Recent concerns about the intentional release of a disease pathogen are reflected in a new requirement for biosurveillance; very early detection of disease outbreaks (4, 95,
96). Earlier detection of a disease outbreak might lead to earlier intervention efforts and result in significant decreases in impact, including human (5, 97-99) and/or animal (83,
100, 101) mortality. Timeliness can be measured as the difference between the onset of
16
an outbreak and the discovery of the outbreak (96, 102). The Centers for Disease Control
and Prevention has identified good surveillance and quick response as the first major goal in modern strategies for preventing emerging infectious diseases (92).
The underlying objective of any system utilizing syndromic surveillance methods
is essentially the same; to identify illness clusters early, before diagnoses are confirmed
and reported by more traditional means (66, 94, 96, 103). Reliable detection of outbreaks
earlier is dependent upon timely data sources that will not only reflect the early characteristics of the outbreak but also be available for analysis almost immediately (68,
94, 104). Timeliness of outbreak detection is also influenced by the availability of
resources for alert follow up, recent user experience with unfounded alerts, and the
agency’s criteria for initiating investigations (103).
2.4.2 Detection Methods:
The statistical basis of early outbreak detection resides within signal detection
theory (96). Identification of clusters is accomplished by processing signals, often in the
form of counts of health indicators, using detection methods to determine the existence or
absence of a trigger event. An alert to a trigger event could focus the attention of
surveillance system users to a particular area and group resulting in earlier identification
of an outbreak of disease (96, 105). A trigger event might be an unusually high number of
laboratory orders submitted close in time from a common geographic area (106).
Although not proof of disease, the occurrence of such an event might alert disease
investigators to possible outbreaks before laboratory tests are completed.
17
In contrast to other tasks of epidemiologists, statistical analysis methods have not
been much involved in the detection of outbreaks and even less for prospective outbreak
detection (107). We can classify the majority of detection methods that have been
investigated into one of three categories: regression methods, time series methods, and
statistical process control methods.
Regression methods model a series of observed counts to estimate a prediction
interval. The upper limit of this interval serves as the threshold for alerting. A benefit of
using regression methods is that seasonal effects can be readily incorporated into the
equation (107, 108). The shortcomings of using regression methods is that they ignore any serial correlation between counts, may not work for surveillance that includes multiple pathogens with different seasonality (107), and the reliability decreases as count values approach zero (Garrick Wallstrom, University of Pittsburgh, personal communication, 2006). These methods require training data in the form of historical datasets of non-outbreak periods for accurate prediction of expected values (104). Other methods that do not require years of historical data often perform just as well (109, 110).
Time series methods do not require large training datasets. They address the limitation of serial correlation by weighting observations assigning greater weight to observations that are more recent. Some of these methods may be more difficult to integrate into an automated system and weighting of values may seriously affect the accuracy of forecasts when preceded by increased counts (107). This effect is potentially correctable by factoring a “guard band” (period between the baseline and the observed date) in the model (110). The guard band serves to separate the observed values from the baseline calculation and reduce the influence of recent increased counts.
18
The problem of detecting certain increases is not unique to outbreak detection.
This issue is also a concern in certain industrial settings where statistical process control
methods have been used effectively (107). Some of these methods are sensitive to rapid
changes in either the count or baseline values (94, 107, 111, 112). While making them
desirable for disease outbreak detection, this characteristic may also increase their
vulnerability to other factors associated with the health indicator such as reporting
efficiency (107). Such sensitivity can be adjusted for by decreasing the variation of health
indicators used for surveillance (113), using another method to calculate the baseline (94,
111), or by applying a two-in-a-row rule that alerts only after the upper threshold is
breached in successive periods (112).
The ability to identify aberrations hinges on the ability to define the normal
occurrence of signals (i.e. the baseline) (114, 115). Such definition usually requires a reliable historical series of data (116). This can result in a substantial limitation for syndromic surveillance systems where historic records, able to be classified into syndrome categories, exists (84).
2.4.3 Limitations:
Analysis remains an intermediary step in the outbreak detection process; human
interpretation of the results is required to identify alerts that result from disease (94, 103,
107, 117, 118). Syndromic surveillance serves to aid the user by directing the epidemiologic investigation (117). Alerts provide the user an indication of an unusual trend that has occurred for some health indicator. As previously stated, these data may
19
well be influenced by events other than those related to disease. Therefore, it is critical to
distinguish which unusual trends are statistically significant compared to those that are
epidemiologically significant and reflective of disease in the population (93).
While some detection methods are better at detecting outbreaks of a certain type than others (110), the ability to account effectively for characteristics of the data is more important to system performance than the choice of method (119). Syndromic surveillance systems tend to be more sensitive compared to traditional disease reporting systems since the number of encounters included in surveillance is much higher (117).
While increased sensitivity (ability to correctly identify when an aberration exists) may be beneficial compared to the costs associated with a failure to recognize a bioterrorist event (5, 118), it comes with an increased chance for false positive alerts. Increasing the specificity (ability to correctly identify when an aberration does not exist) of the system reduces the potential for false positive alerts but results in a decrease in sensitivity and/or timeliness (120, 121). The usefulness of the system to the end user is the true test of balance in these attributes (93).
The size and timing of disease outbreaks that syndromic surveillance systems can
detect may be limited (120). Many individual cases with simultaneous onset might be
noticed through other means while outbreaks of only a few cases may be lost in the
aggregation of data. The true benefits of using syndromic surveillance systems will likely
depend on how effectively they can and are integrated into existing surveillance systems;
other syndromic as well as traditional (72, 120).
20
2.5 “One Medicine” and Animals as Sentinel Indicators:
In 1964, the late Dr. Calvin Schwabe shared his vision of veterinary and human
medicine working together to address disease and improve upon public health. Schwabe
coined the term “One Medicine” to describe this vision (122) inspired by his observations of Sudanese Dinka pastoralist healers treating both animals and humans for what were often common ailments. It is felt by some that effective surveillance of zoonotic
pathogens and control of emerging diseases that they may cause is outside the scope of
traditional medicine requiring integration across human and animal populations (12, 24,
26, 60). Such a holistic approach is lacking in contemporary veterinary and medical
communities. A 2000 report from the Chemical and Biological Arms Control Institute
notes the absence of strong links between public health and the veterinary community,
referring to it as an important disconnect (43).
Many species of animals might begin to show the signs of disease before humans
following near simultaneous infection (11, 123-126). Certain animals have served as
sentinels of select infectious disease in humans. Crows are a good example. In New York
State, a web-based dead bird surveillance project identified an increase in the density of
dead crow sightings and West Nile Virus positive dead birds 2 weeks prior to the
reporting of the first human case (127). Subsequent investigations support that avian morbidity and mortality surveillance provides information that is helpful in predicting onset in humans (128, 129), sometimes as much as 3 months in advance (130).
Other examples of animals as sentinels for infectious disease in humans include
pheasants and chickens for eastern equine and St. Louis encephalitis (131) and white-
tailed deer for Lyme disease (132). Animal-based surveillance is considered important for
21
identifying influenza viruses with the potential for causing disease in humans (133, 134).
The deaths of cats associated with the H5N1 strain of influenza A virus (avian influenza)
reinforces how influenza surveillance in certain animal populations might provide us
advance warning of pandemic influenza (135).
An example how the integration of veterinary medicine into public health might
be beneficial is provided by a case report from Minnesota. On August 25, 2000, the
Minnesota Board of Animal Health notified the Minnesota Department of Health of
Bacillus anthracis that was isolated from a steer. Health authorities interviewed the
family members and learned that about a month prior they slaughtered a downed cow
after their veterinarian gave approval then consumed some of the meat sometime
afterward. At least two family members reported developing gastrointestinal complaints
following consumption (136).
The earlier that a zoonotic disease is detected can equate to the earlier that action
can be taken to reduce the threat or impact to people (99). Enhanced disease surveillance in animals and a “One Medicine” approach as envisioned by Calvin Schwabe may lead to quicker response by both veterinary and public health authorities that greatly minimizes the impact of an outbreak of zoonotic disease. As suggested by the Minnesota case, integrated efforts could also identify cases that would otherwise go unknown. While the risk of infection was minimal in this example, a different pathogen could have resulted in a much different scenario.
22
2.6 Animal-based Syndromic Surveillance Initiatives:
A critical requirement for detecting disease outbreaks earlier in animal
populations is reliable surveillance (137). As discussed previously, traditional outbreak
surveillance in animals has been very disease specific and greatly influenced by potential
economic impact. Certain recent surveillance initiatives (Table 2.2) in animals have
attempted to increase the timeliness of outbreak detection by using other-than-diagnostic
reports or by collecting diagnostic reports in a more timely fashion.
The Center for Emerging Issues (CEI) within the United States Department of
Agriculture (USDA) maintains the VS Electronic Surveillance project. At the heart of this system is Pathfinder; a data mining tool that used to search the electronic open records of the internet. The CEI users determine words and word combinations of interest. Since searched sources include all those of the Worldwide Web, poor translation
of foreign sources tends to be a limitation (81). Another limitation is the sometimes
inaccurate reporting that occurs in the media where terminology may appear out of
proper context.
CEI uses data from this system to contribute to another syndromic surveillance
system located within USDA; the Offshore Pest Information System (OPIS). The OPIS
system was designed to enhance information sharing between Veterinary Services and
International Services within USDA (138). OPIS combines the CEI data from Pathfinder
with field reports from International Services of USDA to generate weekly reports for
USDA users.
23
The Rapid Syndrome Validation Project for Animals (RSVP-A) was an initiative
of the Kansas State University to have attending veterinarians determine specific syndromes from patient signs and upload them via internet connection or Palm® device.
Syndrome categories were non-neonatal diarrhea, neurologic dysfunction or inability to rise, abortion or birth defect, unexpected death, erosive or ulcerative lesions of the skin, mucosa, or coronet, and feed refusal or weight loss without clear explanation. The system included only cattle. Reporting limitations presented because of the provider dependent design and the frequent unavailability of service to support the wireless devices (139).
Other systems that are dependent on manual entries have also discovered them to be inadequate in the long term. The Petsavers Companion Animal Disease Surveillance system was an initiative of the British Small Animal Veterinary Association.
Investigators collected data from fifteen small animal veterinary practices in the form of regular surveys requiring written responses to questions about patients treated within a reporting period of up to four days. Some questions in the survey had non-response rates of 30%. The conclusion of the investigators was that a more robust technique for collection and preparation of data that is less time consuming and more accurate, is required (140).
Michigan State University conducted a similar project that included dairy farmers providing daily records of animal events and veterinarians recording diagnoses and treatments. While burdensome because of the manual reporting, weekly and monthly reports from the data provided back to the providers was determined to be useful in managing the health of herds. The usefulness of the reports served as a sufficient incentive for continued participation in the surveillance system (141).
24
2.7 Summary:
There are recently heightened concerns over an intentional release of a disease
pathogen. Many of the pathogens that are potential bioweapons are zoonotic, as are the
majority of emerging infectious diseases that have been identified in the past few
decades. An outbreak of infectious disease resulting from these circumstances could
spread rapidly through a population indicating the need for quick identification that could
result in earlier intervention. Most traditional surveillance systems utilized for people and
animals are inadequate for the early detection of emerging infectious diseases or a
bioterrorist event.
Syndromic surveillance systems use pre- and non-diagnostic data to detect
aberrant increases in signals that may indicate the existence of an outbreak of infectious disease. A potential for improving on the timeliness of outbreak detection is realized by considering signals that occur earlier in the disease process and including data from many sources. There are limitations in using these systems that need to be considered including the performance of the detection methods used and the size of outbreak that might be detected. Users of these systems should not rely solely on the statistical methods of these systems to investigate potential outbreaks but should also utilize other surveillance systems and information in conjunction.
A “One Medicine” vision embraces the idea of veterinary and human medicine
working together to address disease rather than species. Although the methods used by
syndromic surveillance systems were developed with human data, certain veterinary and
animal-related data might be used in the same way for surveillance in animal populations.
25
Earlier outbreak detection in animals expands on the idea of animals serving as sentinel indicators of disease in humans. Discovery of an outbreak of zoonotic disease in animals might help to reduce or prevent the occurrence of human cases.
Certain animal-based surveillance initiatives have explored how syndromic
methods might be used for animals. These initiatives have identified limitations that can
occur as well as important attributes that should be included. Additional investigation in
system design, detection methods, and system outputs is required to better maximize the
potential benefit from using syndromic surveillance methods to detect disease outbreaks
in animals earlier than possible from traditional surveillance methods.
26
Pathogen Disease Zoonotic Incubation Early Presentation in Humans Bacillus anthracis Anthrax Yes 1-6 days Fever, malaise, fatigue, cough, and mild chest discomfort
Brucella spp. Brucellosis Yes 5-60 days Fever, headache, myalgia, malaise, chills, sweats
Burkholderia mallei Glanders Yes 10-14 days Fever, rigors, sweats, myalgia, headache, pleuritic chest pain
Burkholderia Melioidosis Yes 10-14 days Fever, rigors, sweats, myalgia, headache, pleuritic psuedomallei chest pain
Yersina pestis Plague Yes 1-6 days High fever, chills, headache, malaise, cough
Francisella tularensis Tularemia Yes 3-5 days average Fever, chills, headache, malaise
Coxiella burnetii Q Fever Yes 2-14 days Fever, cough, pleuritic chest pain 27 Orthopoxvirus, Smallpox No 7-19 days Malaise, fever, rigors, vomiting, headache, backache variola
Group of eight Venezuelan Equine Yes 1-6 days Malaise, spiking fever, rigors, severe headache, mosquito-borne Encephalitis photophobia, myalgia alphaviruses
Select members of Viral Hemorrhagic Yes 2-21 days Flushing of chest and face, petechiae, bleeding, Arenaviridae, Fevers edema, malaise, myalgia, headache, vomiting Bunyaviridae, Filoviridae, and Flaviviridae families
Table 2.1: Bacterial and viral disease-causing organisms considered to be potential bioweapon threat agents. (adapted from USAMRIID’s Medical Management of Biological Casualties Handbook. February 2001. U.S. Army Medical Research Institute of Infectious Diseases, Fort Dietrick, Frederick, MD.)
Species System Agency Included System Design Pros & Cons VS Electronic Center for Emerging multiple searches internet open Pro - extensive number of sources Surveillance Issues, records for specified word and included (Pathfinder) USDA word combinations Con - subject to translation error and inaccurate reporting
Offshore Pest Veterinary and multiple field reports and results from Pro - enables increased communications Information International Pathfinder to create weekly reports between areas System Services, USDA Con - lack of timeliness and Pathfinder limitations
Rapid Kansas State University cattle veterinarians determine syndrome Pro - presumed increase in specificity 28 Syndrome and report via internet or wireless from provider-based reports Validation service at or near time of service Con - creates additional burden on Project for providers, unavailability of utility service, Animals limited to include a single species
Petsavers British Small Animal pets veterinarians submit written survey Pro - presumed increase in specificity Companion Veterinary Association of cases treated in practice during from provider-based reports Animal Disease multi-day period Con - creates additional burden on Surveillance providers and poor reporter compliance
Computerized Michigan State University dairy Farmers report daily animal events Pro - presumed increase in specificity Dairy Herd cattle and veterinarians report diagnoses from provider participants, provides Health Data and treatments feedback to reporters Base Con - creates additional burden on providers and farmers, limited to include a single species
Table 2.2: Description of select animal-based syndromic surveillance initiatives.
REFERENCES
1. Venkatesh S, Memish ZA. Bioterrorism -- a new challenge for public health. Int J Antimicrob Agents 2003;21:200-6.
2. Huerta M, Leventhal A. The epidemiologic pyramid of bioterrorism. Isr Med Assoc J 2002;4:498-502.
3. Henderson DA. Bioterrorism as a Public Health Threat. Emerg Infect Dis 1998;4:488-92.
4. Lober WB et al. Roundtable on Bioterrorism Detection: Infomration System- based Surveillance. J Am Med Inform Assoc 2002;9:105-15.
5. Kaufmann AF, Meltzer MI, Schmid GP. The Economic Impact of a Bioterrorist Attack: Are Prevention and Postattack Intervention Programs Justifiable? Emerg Infect Dis 1997;3:83-94.
6. Kortepeter M. USAMRIID's Medical Management of Biological Casualties Handbook. Fort Detrick, Frederick, MD: U.S. Army Medical Research Institute of Infectious Diseases, 2001.
7. Federation of American Scientists. The Role of Disease Surveillance in the Watch for Agro-terrorism or Economic Sabotage. October 2001.
8. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category B and C agents. J Am Vet Med Assoc 2004;224:1096-104.
9. Pavlin JA. Epidemiology of Bioterrorism. Emerg Infect Dis 1999;5:528-30.
10. Ashford DA et al. Biological terrorism and veterinary medicine in the United States. J Am Vet Med Assoc 2000;217:664-7.
11. Conti L. Petborne Zoonoses: Detection and Surveillance Challenges. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press.
12. Cleaveland S, Laurenson MK, Taylor LH. Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:991-9.
13. Karesh WB et al. Wildlife Trade and Global Disease Emergence. Emerg Infect Dis 2005;11:1000-2.
14. Thompson RCA. Giardiasis as a re-emerging infectious disease and its zoonotic potential. Int J Parasitol 2000;30:1259-67.
29
15. Robertson ID et al. The role of companion animals in the emergence of parasitic zoonoses. Intl J Parasitol 2000;30:1369-77.
16. Hlavsa MC, Watson JC, Beach MJ. Cryptosporisiosis Surveillance--United States, 1999-2002. MMWR Surveill Summ 2005;54.
17. Saini PK, Ransom G, McNamara AM. Emerging public health concerns regarding cryptosporidiosis. J Am Vet Med Assoc 2000;217:658-63.
18. Fayer R, Morgan U, Upton SJ. Epidemiology of Cryptosporidium: transmission, detection and identification. Int J Parasitol 2000;30:1305-22.
19. Murrell KD, Pozio E. Trichinellosis: the zoonosis that won't go quietly. Int J Parasitol 2000;30:1339-49.
20. Tenter AM, Heckeroth AR, Weiss LM. Toxoplasma gondii: from animals to humans. Int J Parasitol 2000;30:1217-58.
21. Meslin FX, Stohr K, Heymann D. Public health implications of emerging zoonoses. Rev Sci Tech 2000;19:310-7.
22. Institute of Medicine. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, DC, National Academy Press.
23. Kruse H, Kirkemo A-M, Handeland K. Wildlife as Source of Zoonotic Infections. Emerg Infect Dis 2004;10:2067-72.
24. Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:983-9.
25. Fiala J. Veterinary roles must expand to protect public health, reports say. DVM: The Newsmagazine of Veterinary Medicine [Sept], 21-24. 2005.
26. Woolhouse MEJ. Population biology of emerging and re-emerging pathogens. Trends Microbiol 2002;10:S3-S7.
27. Chretien JH, Garagusi VF. Infections associated with pets. Am Fam Physician 1990;41:831(15).
28. Rosen T, Jablon J. Infectious threats from exotic pets: dermatological implications. Dermatol Clin 2003;21:229-36.
29. Riley PY, Chomel BB. Hedgehog Zoonoses. Emerg Infect Dis 2005;11:1-5.
30
30. Bender JB, Shulman SA. Reports of zoonotic disease outbreaks associated with animal exhibits and availability of recommendations for preventing zoonotic disease transmission from animals to people in such settings. J Am Vet Med Assoc 2004;224:1105-9.
31. Hill SL et al. Prevalence of enteric zoonotic organisms in cats. J Am Vet Med Assoc 2000;216:687-92.
32. Tauni MA, Österlund A. Outbreak of Salmonella typhimurium in cats and humans associated with infection in wild birds. J Small Anim Pract 2000;41:339-41.
33. Tansey B. Debate over pets' role in spread of avian flu: No cases passed yet by dogs or cats, but experts want study. June 12, 2006. San Francisco, CA.
34. Kaye D, Pringle CR. Avian Influenza Viruses and their Implication for Human Health. Clin Infect Dis 2005;40:108-12.
35. Crawford PC et al. Transmission of Equine Influenza Virus to Dogs. 10.1126/science.1117950. September 29, 2005. Available at: www.scienceexpress.org.
36. Webster RG. The importance of animal influenza for human disease. Vaccine 2002;20:S16-S20.
37. Green MS, Kaufman Z. Surveillance for Early Detection and Monioring of Infectious Disease Outbreaks Associated with Bioterrorism. Isr Med Assoc J 2002;4:503-6.
38. Council of State and Territorial Epidemiologists. National Assessment of Epidemiologic Capacity in Public Health: Findings and Recommendations. March 2003.
39. Inglesby TV, Grossman R, O'Toole T. A Plague on Your City: Observations from TOPOFF. Clin Infect Dis 2001;32.
40. Committee on Science and Technology for Countering Terrorism. Making the Nation Safer: The Role of Science and Technology in Countering Terrorism. 2002. Washington, D.C., National Research Council.
41. Heinrich J. SARS Outbreak: Improvements to Public Health Capacity Are Needed for Responding to Bioterrorism and Emerging Infectious Diseases. [GAO-03-769T]. May 7, 2003. Committee on Energy and Commerce, House of Representatives, Subcommittee on Oversight and Investigations. 2003.
42. Ashford DA et al. Planning against Biological Terorism: Lessons from Outbreak Investigations. Emerg Infect Dis 2003;9:515-9.
31
43. Moodie M et al. Biological Terrorism in the United States: Threat, Preparedness, and Response. [200*1999*00132]. November 2000. Chemical and Biological Arms Control Institute.
44. Heinrich J. Infectious Diseases: Gaps Remain in Surveillance Capabilities of State and Local Agencies. [GAO-03-1176T]. September 24, 2003. Select Committee on Homeland Security, Subcommittee on Emergency Preparedness and Response. 2003.
45. Zepeda C, Salman MD. Planning Survey, Surveillance, and Monitoring Systems - - Roles and Requirements. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:35-43.
46. Childs J et al. Emerging Zoonoses. Emerg Infect Dis 1998;4:453-4.
47. Kelsey H. Improvements in methodologies for tracking infectious disease needed. The Newsbulletin . January 13, 2005. Los Alamos National Laboratory. Available at: www.lanl.gov/orgs/pa/newsbulletin/2005/01/13/includes/text02.inc.
48. Conner CF. Review of efforts to protect the agricultural sector and food supply from a deliberate attack with a biological agent, a toxin or a disease directed at crops and livestock. July 20, 2005. Bio-security and Agro-terrorism. 2005.
49. Wurtz RM, Popovich ML. Animal Disease Surveillance: A Framework for Supporting Disease Detection in Public Health. March 2002. Tuscon, AZ, Scientific Technologies Corporation.
50. Christensen J. Application of Surveillance and Monitoring Systems in Disease Control Programs. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:15-34.
51. Kelly TK et al. The Office of Science and Technology Policy Blue Ribbon Panel on The Threat of Biological Terrorism Directed Against Livestock. March 2004. Arlington, VA, RAND Science and Technology.
52. Engle M. Improving Emerging Disease Surveillance in the Pork Industry. Swine Health Report [Fall], 4-5. 2004. Bowling Green, KY, National Institute for Animal Agriculture.
53. Kanof M. Emerging Infectious Diseases: Review of State and Federal Disease Surveillance Efforts. [GAO-04-877]. September 30, 2004. Committee on Governmental Affairs, Permanent Subcommittee on Investigations. 2004. Available at: www.gao.gov/cgi-bin/getrpt?GAO-04-877.
32
54. United States Department of Agriculture Animal and Plant Health Inspection Service. National Animal Health Reporting System [Web Page]. Accessed 2006 Jun 22. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nahrs/.
55. Salman MD. Surveillance and Monitoring Systems for Animal Health Programs and Disease Surveys. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:3-13.
56. United States Department of Agriculture National Veterinary Services Laboratories. The National Animal Health Laboratory Network: A State and Federal Partnership to Safeguard Animal Health [Web Page]. Accessed 2006 Apr 22. Available at: http://nahln.vs/.
57. Baca R, Cassidy K. NAHLN: A Standards-based Approach for Laboratory Diagnostic Data. NAHSS Outlook . October 2004. Fort Collins, CO, United States Department of Agriculture.
58. Mark L. Strengthening Veterinary Diagnostic Lab Network is Critical. February 24, 2006. Richmond, VA, United States Animal Health Association.
59. Kearney B. Strengthening Safeguards Against Disease Outbreaks. In Focus 5[2]. 2005. Washington, D.C., The National Academy of Sciences. Available at: http://infocusmagazine.org/5.2/env_animal_health.html.
60. National Research Council. Animal Health at the Crossroads: Preventing, Detecting, and Diagnosing Animal Diseases. July 2005. Washington, D.C., The National Academy of Sciences.
61. Center for Infectious Disease Research and Policy. Agencies launch warning system for animal diseases [Web Page]. 2006 Jul 25; Accessed 2006 Jul 27. Available at: www.cidrap.umn.edu/cidrap/content/influenza/avianflu/news/jul2506warning.htm l.
62. Koo D, Parrish II RG. The Changing Health-Care Information Infrastructure in the United States: Opportunities for a New Approach to Public Health Surveillance. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford University Press, 2000:76-94.
63. Sosin DM. Draft Framework for Evaluating Syndromic Surveillance Systems. J Urban Health 2003;80:i8-i13.
33
64. Chin N. Dr. James Buehler On the Promise of Syndromic Surveillance. Homeland First Response [May/June]. 2004. Jems Communications.
65. Begier EM et al. The National capitol Region's Emergency Department Syndromic Surveillance System: Do Chief Complaint and Discharge Diagnosis Yield Different Results? Emerg Infect Dis 2003;9:393-6.
66. Henning KJ. What is Syndromic Surveillance? MMWR Morb Mortal Wkly Rep 2004;53:7-11.
67. Henning KJ. Syndromic Surveillance. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, D.C., National Academy Press.
68. Buehler JW et al. Framework for Evaluating Public Health Surveillance Systems for Early Detection of Outbreaks. MMWR Recomm Rep 2004;53.
69. Dafni UG et al. Algorithm for Statistical Detection of Peaks--Syndromic Surveillance System for the Athens 2004 Olympic Games. MMWR Morb Mortal Wkly Rep 2004;53:86-94.
70. Tsui F-C et al. Technical Description of RODS: A Real-time Public Health Surveillance System. J Am Med Inform Assoc 2003;10:399-408.
71. Heffernan R et al. Syndromic Surveillance in Public Health Practice, New York City. Emerg Infect Dis 2004;10:858-64.
72. Fienberg SE, Shmueli G. Statistical issues and challenges associated with rapid detection of bio-terrorist attacks. Stat Med 2005;24:513-29.
73. Das D et al. Monitoring Over-The-Counter Medication Sales Early Detection of Disease Outbreaks--New York City. MMWR Morb Mortal Wkly Rep 2005;54:41-6.
74. Witt CJ. Electronic Surveillance System for the Early Notification of Community-based Epidemics (ESSENCE). December 11, 2003. Department of Defense Global Emerging Infections System.
75. Magruder SF. Evaluation of Over-the-Counter Pharmaceutical Sales A a Possible Early Warning Indicator of Human Disease. Johns Hopkins APL Technical Digest 24[4], 349-353. 2003.
76. Platt R et al. Syndromic Surveillance Using Minimum Transfer of Identifiable Data: the Example of the National Bioterrorism Syndromic Surveillance Demonstration Program. J Urban Health 2003;80:i25-i31.
34
77. Wethington H, Bartlett P. The RUsick2 Foodborne Disease Forum for Syndromic Surveillance. Emerg Infect Dis 2004;10:401-5.
78. Bradley CA et al. BioSense: Implementation of a National Early Event Detection and Situational Awareness System. MMWR Morb Mortal Wkly Rep 2005;54:11- 9.
79. Centers for Disease Control and Prevention. Syndrome Definitions for Diseases Associated with Critical Bioterrorism-associated Agents. October 23, 2003.
80. Chapman WW, Dowling JN, Wagner MM. Fever detection from free-text clinical records for biosurveillance. J Biomed Inform 2004;37:120-7.
81. Johnson CL. Electronic Surveillance for Emerging Diseases . National Veterinary Electronic Surveillance Meeting. 2004.
82. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board. Available at: www.porkboard.org/docs/mengle.pdf.
83. Shephard R, Aryel RM, Shaffer L. Animal Health. In: Wagner MM, Moore AW, Aryel RM, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:111-27.
84. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10.
85. Zeng X, Wagner M. Modeling the Effects of Epidemics on Routinely Collected Data. J Am Med Inform Assoc 2002;9:S17-S22.
86. Duchin JS. Epidemiological Response to Syndromic Surveillance Signals. J Urban Health 2003;80:i115-i116.
87. Johnson HA, Wagner MM, Saladino RA. A New Method for Investigating Non- traditional Biosurveillance Data: Studying Behaviors Prior to Emergency Department Visits. 2005. Available at: www.rods.pitt.edu.
88. Mandl KD et al. Implementing Syndromic Surveillance: A Practical Guide Informed by the Early Experience. J Am Med Inform Assoc 2004;11:141-50.
89. Gobar GM, Case JT, Kass PH. Program for surveillance of causes of death of dogs, using the internet to survey small animal veterinarians. J Am Vet Med Assoc 1998;213:251-6.
35
90. Tsui F-C et al. Data, Network, and Application: Technical Description of the Utah RODS Winter Olympic Biosurveillance System. Proc AMIA Ann Symp. 2002.
91. Power C. Passive Animal Disease Surveillance in Canada: A Benchmark. Proceedings of a CAHNet Workshop. November 1999. Canadian Food Inspection Agency.
92. Doherr MG, Audige L. Monitoring and surveillance for rare health-related events: a review from the veterinary perspective. Philos Trans R Soc Lond B Biol Sci 2001;356:1097-106.
93. Buehler JW. Surveillance. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins, 1998:435-57.
94. Hutwagner LC et al. Using Laboratory-Based Surveillance Data for Prevention: An Algorithm for Detecting Salmonella Outbreaks. Emerg Infect Dis 1997;3:395-400.
95. Wagner MM, Shaffer L, Shephard R. Functional Requirements for Biosurveillance. In: Wagner MM, Moore A, Aryel R, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:51-64.
96. Wagner MM et al. The Emerging Science of Very Early Detection of Disease Outbreaks. J Pub Health Manag Pract 2001;7:51-9.
97. Mitchell J. State Strategies for Fully Integrating Public Health into Homeland Security. November 23, 2005. Washington, D.C., NGA Center for Best Practices.
98. Dato V, Wagner MM, Fapohunda A. How Outbreaks of Infectious Disease are Detected: A Review of Surveillance Systems and Outbreaks. Public Health Rep 2004;119:464-71.
99. Hartl G, Meslin F. Launch of global early warning system for animal diseases transmissable to humans. July 24, 2006. July 27, 2006.
100. Food and Agriculture Organization of the United Nations. Manual on Livestock Disease Surveillance and Information Systems. Rome: 1999.
101. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality.
36
102. Centers for Disease Control and Prevention. Updated Guidelines for Evaluating Public Health Surveillance Systems: Recommendations from the Guidelines Working Group. MMWR Recomm Rep 2001;50.
103. Buehler JW et al. Syndromic Surveillance and Bioterrorism-related Epidemics. Emerg Infect Dis 2003;9:1197-204.
104. Tsui F-C et al. Value of ICD-9-Coded Chief Complaints for Detection of Epidemics. Proceedings of the AMIA Annual Symposium. 711-715. 2001.
105. U.S. Department of the Interior. An Early Detection System for Highly Pathogenic H5N1 Avian Influenza in Wild Migratory Birds U.S. Interagency Strategic Plan [Web Page]. 2006 Mar 14; Accessed 2006 Aug 24. Available at: www.doi.gov/issues/birdflu_strategicplan.pdf.
106. National Center for Infectious Diseases. Recognition of Illness Associated with the Intentional Release of a Biologic Agent. MMWR Morb Mortal Wkly Rep 2001;50.
107. Farrington P, Andrews N. Outbreak Detection: Application to Infectious Disease Surveillance. In: Brookmeyer R, Stroup DF, eds. Monitoring the Health of Populations: Statistical Principles and Methods for Public Health Surveillance. New York, NY: Oxford Univeristy Press, 2004:203-31.
108. Serfling RE. Methods for Current Statistical Analysis of Excess Pneumonia- Influenza Deaths. Public Health Rep 1963;78:494-506.
109. Hutwagner L et al. Comparing Aberration Detection Methods with Simulated Data. Emerg Infect Dis 2005;11:314-6.
110. Buckeridge DL et al. Algorithms for rapid outbreak detection: a research synthesis. J Biomed Inform 2005;38:99-113.
111. O'Brien SJ, Christie P. Do CuSums have a role in routine communicable disease surveillance. Public Health 1997;111:255-8.
112. Lucas JM. Counted Data CUSUM's. Technometrics 1985;27:129-44.
113. Janes GR et al. Descriptive Epidemiology: Analyzing and Interpreting Surveillance Data. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford Univeristy Press, 2000:112-67.
114. Wong W-K et al. Rule-Based Anomaly Pattern Detection for Detecting Disease Outbreaks. 2002. American Association for Artificial Intelligence.
37
115. Reis BY, Mandl KD. Time series modeling for syndromic surveillance. Med Inform Decis Making 2003;3.
116. Focks D. Mathematical Models and Predictors of Disease Outbreaks. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washnigton, DC, National Academy Press.
117. Pavlin JA. Investigation of Disease Outbreaks Detected by "Syndromic" Surveillance Systems. J Urban Health 2003;80:i107-i114.
118. Hutwagner L et al. The Bioterrorism Preparedness and Response Early Aberration Reporting System (EARS). J Urban Health 2003;80:i89- i96.
119. Siegrist D et al. Evaluation of Algorithms for Outbreak Detection Using Clinical Data from Five U.S. Cities. October 15, 2004. Defense Advanced Research Projects Agency.
120. Stoto MA, Schonlau M, Mariano LT. Syndromic Surveillance: Is It Worth the Effort? Chance 2004;17:19-24.
121. Reingold A. If Syndromic Surveillance Is the Answer, What Is the Question? Biosecur Bioterror 2003;1:77-81.
122. Schwabe CW. Veterinary Medicine and Human Health. Baltimore, MD: Williams and Wilkins, 1984.
123. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category A agents. J Am Vet Med Assoc 2004;224:1084-95.
124. Babin SM et al. Early detection of possible bioterrorist events using sentinel animals. The 131st Annual Meeting of the American Public Health Association. 2003.
125. Mauer WA, Kaneene JB. Integrated Human-Animal Disease Surveillance. Emerg Infect Dis 2005;11:1490-1.
126. Backer L et al. Pet dogs as sentinels for environmental contamination. Sci Total Environ 2001;274:161-9.
127. Eidson M et al. Dead Crow Densities and Human Cases of West Nile Virus, New York State, 2000. Emerg Infect Dis 2001;7:662-4.
38
128. Centers for Disease Control and Prevention. Epidemic/Epizootic West Nile Virus in the United States: Guidelines for Surveillance, Prevention, and Control. 2003. Ft. Collins, CO.
129. Guptill S et al. Early-Season Avian Deaths from West Nile Virus as Warnings of Human Infection. Emerg Infect Dis 2003;9:483-4.
130. Eidson M et al. Dead Bird Surveillance as an Early Warning System for West Nile Virus. Emerg Infect Dis 2001;7:631-5.
131. Morris CD et al. Comparison of chickens and pheasants as sentinels for eastern equine encephalitis and St. Louis encephalitis viruses in Florida. J Am Mosq Control Assoc 1994;10:545-8.
132. Gill JS et al. Serologic Surveillance for the Lyme Disease Spirochete, Borrelia burgdorferi, in Minnesota by Using White-Tailed Deer as Sentinel Animals. J Clin Microbiol 1994;32 :444-51.
133. Fouchier RAM, Osterhaus ADME, Brown IH. Animal influenza virus surveillance. Vaccine 2003;21:1754-7.
134. Glickman LT et al. Purdue University-Banfield National Companion Animal Surveillance Program for Emerging and Zoonotic Diseases. Vector Borne Zoonotic Dis 2006;6:14-23.
135. Yingst SL, Saad MD, Felt S. Qinghai-like H5N1 from Domestic Cats, Northern Iraq. Emerg Infect Dis 2006;12:1295-7.
136. Kassenborg H et al. Human Ingestion of Bacillus Anthracis-Contaminated Meat -- Minnesota, August 2000. MMWR Morb Mortal Wkly Rep 2000;49.
137. Salman MD, Stark KDC, Zepeda C. Quality assurance applied to animal disease surveillance systems. Rev Sci Tech 2003;22:689-96.
138. Johnson C. The Role of the Center for Emerging Issues in Coordinating International Safeguarding Information via the Offshore Pest Information System. NAHSS Outlook [Aug]. 2005. Fort Collins, CO, United States Department of Agriculture.
139. DeGroot B. The Rapid Syndrome Validation Project for Animals - Augmenting Contact with the Network of Accredited Veterinarians. NAHSS Outlook [April]. 2005. United States Department of Agriculture. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook.
140. Robotham J, Green LE. Pilot study to investigate the feasibility of surveillance of small animals in the UK. J Small Anim Pract 2004;45:213-8.
39
141. Bartlett PC et al. Development of a computerized dairy herd health data base for epidemiologic research. Prev Vet Med 1986;4:3-14.
40
CHAPTER 3
SENTINEL SURVEILLANCE OF HUMAN HOUSEHOLDS
USING COMPANION ANIMALS
3.1 Abstract:
Based on the quantity of time and common space that people share with their pets,
these animals might prove beneficial as a sentinel of infectious disease in humans. These
animals could develop clinical disease from infection earlier than their human owners
might. Detecting an outbreak first in a population of pets may provide for earlier
intervention efforts that limit or prevent an outbreak spreading to humans. To what extent
companion animals may be proxy for their human caretakers is unknown. We used
results from the 2002 American Veterinary Medical Association Household Pet Survey to estimate the number of Ohio households that owned dogs, cats, horses, or birds that provided laboratory specimens to IDEXX Laboratories in a one-year period. Dogs were the genus representing the greatest number of human households in Ohio during the study period (4.8% to 8.8%). The results provide for an estimation of human representation to consider when contemplating using these types of data for outbreak surveillance.
41
3.2 Introduction:
Animals might be used as sentinels in syndromic surveillance efforts designed to
detect emerging diseases (1-5). Detection of disease events in animals might provide
advance warning of disease in humans and potentially enable us to take action earlier that
could prevent transmission between populations or mitigate the risk from exposure.
Agricultural animals are a focus of many prior investigations since they represent a large
economic base in the United States as well as a possible means of introduction for an
infectious agent through consumption (6, 7). Certain wildlife species have become the
focus of surveillance given concerns over avian influenza becoming the next pandemic
and the belief that waterfowl might serve as the vector for introduction of the virus into
the United States (8).
Of all animals, companion animals are of particular interest because of their large
numbers and close and frequent contact with their human owners. Because companion
animals often range outside of the home, they sometimes can act as the intermediary for
the transfer of pathogens from wildlife and agricultural animals to humans (9, 10). How
well companion animals could serve as proxy for humans is a consideration for
surveillance system developers. The extent to which humans and animals transfer disease
pathogens is only one aspect of this consideration. Knowing how well animals represent the population of humans in the same area is another that may provide us with helpful information for estimating prevalence of disease, extent of exposure to pathogens, and sensitivity of detection. We explored the animal-to-human representation by estimating the number of households that would be included on average by a surveillance system using data from a veterinary diagnostics laboratory.
42
3.3 Methods:
3.3.1 AVMA Household Pet Survey:
During the first week of January 2002, the American Veterinary Medical
Association mailed the Household Pet Survey to 80,000 randomly selected households in the United States (11). This survey asked a number of questions regarding animals owned anytime during 2001 including veterinary care, number of animals owned, and household demographics. The response rate for this survey was 67.8%. The survey reported the average number of animals by species per household, and the rate of annual veterinary
visits per animal by genera. Responses were used to determine the proportion of
veterinary visits that included laboratory submissions. Results for dogs and cats were
reported at the state level while estimates for horses and pet birds were reported by
regions of the nation. Responses for horses did not include farm or ranch operations.
There were 3,440 (4.3% of mail sample) surveys mailed to households in Ohio of which
2,495 (4.6% of respondent sample) were returned; a response rate of 72.5%. Responses
were representative of family household design (e.g., single person, married couple, or
retired couple), type of residence (e.g., single family home, duplex, or apartment), home
ownership, household income, stage of life, and household size according to the AVMA
surveyors.
3.3.2 Data Sample:
We conducted a retrospective study of a dataset provided by IDEXX
Laboratories, Inc. (Westbrook, Maine) that contained all test orders received between
April 1, 2005 and March 31, 2006 for specimens originating from patients treated at
43
clinics located in Ohio. IDEXX is a commercial veterinary diagnostics laboratory that
provides services to private clinicians throughout the United States including Ohio.
Information in the dataset included accession numbers that identified unique specimens submitted from individual animals. We determined from this dataset the number of times that dogs, cats, horses, and pet birds from Ohio that provided specimens to IDEXX for testing during the one-year study period by counting the unique accession numbers for each genus.
3.3.3 Calculations:
We estimated the number of human households represented in the IDEXX dataset
using the steps outlined in Table 3.1. We divided the number of accessions for each genus in the IDEXX dataset by the genus-specific proportion of veterinary visits that
resulted in a laboratory submission from the AVMA survey to estimate the number of
visits to veterinarians utilizing IDEXX services that occurred during the study period:
IDEXX _# accessions Acolumn )_( # veterinary visits Ccolumn )_(_ = Eq. (3.1) # laboratory orders /_ veterinary visit Bcolumn )_(_
44
The number of visits, divided by the rate of annual veterinary visits for each genus from
the AVMA survey, estimated the number of individual animals seen by veterinarians who
used IDEXX in Ohio during the study period:
# veterinary visits Ccolumn )_(_ animals Ecolumn )_(# = Eq. (3.2) annual visits animal Dcolumn )_(/_#
The number of households represented was then determined by dividing the number of
animals by the average number of animals in each animal-owned household as
determined by the AVMA survey:
animals Ecolumn )_(# # households Gcolumn )_( = Eq. (3.3) animals animal owning __/# household Fcolumn )_(
We then compared the number of Ohio households represented to the estimated
total number of Ohio households from the 2002 United States Census (12) to estimate the
percentage of all Ohio households included in the IDEXX data. In order to consider how
error in the AVMA survey might influence these estimates, we repeated the calculations using the AVMA estimates (i.e. number of animals per household, number of annual
veterinary visits, and percentage of visits that involved a laboratory submission) plus and
minus 10% of their reported value.
45
3.4 Results:
IDEXX Laboratories received 208,346 specimens from Ohio clinics during the study period, 150,309 (72.1%) were from dogs, 48,276 (23.2%) from cats, 5,402 (2.6%) from horses, and 350 (0.2%) from pet birds (Figure 3.1). We estimated, using the
methods previously described, that these laboratory submissions originated from
1,225,405 veterinary visits. These were representative of 425,323 visits for dogs, 357,600
visits for cats, 64,540 visits for horses, and 8,030 visits for pet birds. The number of
households represented by these visits included an estimated 283,549 dog-owning,
162,546 cat-owning, 30,734 horse-owning, and 2,591 pet bird-owning households (Table
3.1). The percentage represented by each genus of animal was 6.6% (dogs), 3.8% (cats),
0.7% (horses), and 0.1% (pet birds) of all households in Ohio. The range of
representation for each genus determined from our arbitrarily assigned +/-10% of error was 4.8% to 8.8% (dogs), 2.9% to 5.1% (cats), 0.5% to 1.0% (horses), and <0.1% to
0.1% (pet birds) (Figure 3.2).
3.5 Discussion:
According to the AVMA, the number of animals per genus-specific owning
household was 2.2 (cats), 2.1 (birds), and 3.1 (horses) compared to 1.5 for dogs (11). A limitation of this estimate was the lack of knowledge regarding which households owned
more than a single genus of animal. Actual representation might exceed the upper
estimate if the number of households owning only cats, horses, or pet birds were added to the value for dog-owning households.
46
The AVMA estimates that 35.9% (dogs), 30.3% (cats), 4.1% (pet birds), and
1.3% (horses) of Ohio households own companion animals (11). Our estimates of household representation included only those animals providing specimens to one
veterinary diagnostic laboratory. These values determine what percentage of households,
on average, that might be expected to be included in a system using data from this single
source. They also provide an estimate of the average distribution of animal genera.
Detection of outbreaks might be provided not only from unexpected increases in specific
tests ordered but also from significant changes in either the distribution of theses genera
or the percentage of total households represented.
Animal ownership was representative of Ohio’s population according to the
AVMA survey report. Bias may exist if we consider the representation of animal owners
that provided for veterinary care, laboratory tests in particular, of their pets (13, 14). Our analysis did not provide any indication how much of this type of bias is present or how much it would affect the sensitivity of a detection system. Although reflective of a segment of the larger biased subpopulation of pet owners, the estimates provided an indication of the proportion included from this single source.
Enhancing outbreak detection capabilities by including surveillance of animal
populations continues to gain interest. Our results provide an estimate for consideration
by those investigating companion animals as a sentinel population for these efforts. The
question of adequacy remains a determination to be made by system developers and will
likely be influenced by their disease/s of interest.
47
A B C D E F G % of visits Est. # of Est. # of Est. # of # of lab # of annual vet # of resulting vet visits animals households accessions visits/animal* animals/household* in lab (A/B) (C/D) (E/F) testing* +10% error 20.5 733,215 2.1 349,150 1.7 205,383 Dogs 150,309 18.6 808,113 1.9 425,323 1.5 283,549 -10% error 16.7 900,054 1.7 529,444 1.4 378,174
Cats +10% error 14.9 324,000 1.1 294,546 2.4 122,728 48,276 13.5 357,600 1.0 357,600 2.2 162,546
48 -10% error 12.2 395,705 0.9 439,673 2.0 219,836
+10% error 10.2 52,961 1.0 52,961 2.3 23,027 Horses 5,402 9.3 58,086 0.9 64,540 2.1† 30,734 -10% error 8.4 64,310 0.8 80,387 1.9 42,309
+10% error 24.0 1,459 0.2 7,292 3.4 2,145 Pet 350 21.8 1,606 0.2 8,030 3.1† 2,591 Birds -10% error 19.6 1,786 0.2 8,929 2.8 3,189 * U.S. Pet Ownership and Demographics, American Veterinary Medical Association, 2002 †rate based on northeast central region of U.S. that includes Ohio
Table 3.1: Estimating the number of Ohio households represented in IDEXX dataset (04/01/2004 – 03/31/2005) from the number of laboratory specimen accessions.
Pet Birds, 0.2 Unknown, 1.4 Horses, 2.6 Other, 0.5
Cats, 23.2
Dogs, 72.1
Figure 3.1: Percentage of laboratory specimen accessions by genera received by IDEXX from clinics located in Ohio, April 1, 2004 – March 31, 2005.
49
10.0
9.0 8.8
8.0
7.0 6.6 10% over estimate
6.0 no error
5.1 5.0 4.8 10% under estimate 50 4.0 3.8
2.9 % of Ohio Households % 3.0
2.0
1.0 0.7 1.0 0.5 0.0 0.1 0.1 0.0 Dogs Cats Horses Pet Birds
Figure 3.2: Estimates of Ohio household representation by animals with consideration of possible over- and underestimation of AVMA Household Survey values. Total Ohio household estimate of 4,293,649 from the 2002 U.S. Census.
REFERENCES
1. Eidson M et al. Dead Bird Surveillance as an Early Warning System for West Nile Virus. Emerg Infect Dis 2001;7:631-5.
2. Eidson M et al. Dead Crow Densities and Human Cases of West Nile Virus, New York State, 2000. Emerg Infect Dis 2001;7:662-4.
3. Gill JS et al. Serologic Surveillance for the Lyme Disease Spirochete, Borrelia burgdorferi, in Minnesota by Using White-Tailed Deer as Sentinel Animals. J Clin Microbiol 1994;32:444-51.
4. Morris CD et al. Comparison of chickens and pheasants as sentinels for eastern equine encephalitis and St. Louis encephalitis viruses in Florida. J Am Mosq Control Assoc 1994;10:545-8.
5. Torok TJ et al. A Large Community Outbreak of Salmonellosis Caused by Intentional Contamination of Restaurant Salad Bars. J Am Med Assoc 1997;278:389-95.
6. Lynn T. Coordinating Zoonotic Disease Surveillance: Partnering Agriculture and Public Health. NAHSS Outlook [June]. 2005. United States Department of Agriculture. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook.
7. Conner CF. Review of efforts to protect the agricultural sector and food supply from a deliberate attack with a biological agent, a toxin or a disease directed at crops and livestock. Bio-security and Agro-terrorism. July 20, 2005.
8. U.S. Department of the Interior. An Early Detection System for Highly Pathogenic H5N1 Avian Influenza in Wild Migratory Birds U.S. Interagency Strategic Plan [Web Page]. 2006 Mar 14; Accessed 2006 Aug 24. Available at: www.doi.gov/issues/birdflu_strategicplan.pdf.
9. Tauni MA, Österlund A. Outbreak of Salmonella typhimurium in cats and humans associated with infection in wild birds. J Small Anim Pract 2000;41:339-41.
10. Yingst SL, Saad MD, Felt S. Qinghai-like H5N1 from Domestic Cats, Northern Iraq. Emerg Infect Dis 2006;12:1295-7.
11. Amercian Veterinary Medical Association. U.S. Pet Ownership and Demographics Sourcebook. 2002. Schaumburg, IL, American Veterinary Medical Association.
12. U.S. Census Bureau. American Factfinder [Web Page]. Accessed 2006 Aug 29. Available at: http://factfinder.census.gov.
51
13. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10.
14. United States Department of Agriculture National Veterinary Services Laboratories. The National Animal Health Laboratory Network: A State and Federal Partnership to Safeguard Animal Health [Web Page]. Accessed 2006 Apr 22. Available at: http://nahln.vs/.
52
CHAPTER 4
EVALUATION OF MICROBIOLOGY ORDERS FROM TWO VETERINARY
DIAGNOSTIC LABORATORIES AS POTENTIAL DATA SOURCES FOR
EARLY OUTBREAK DETECTION
4.1 Abstract:
Animals continue to be recognized as a potential source of data for surveillance related to emerging infectious diseases, bioterrorism preparedness, pandemic influenza preparedness, and other zoonotic diseases. Detection of disease outbreaks in animals though remains mostly dependent upon systems that are disease specific and not very timely. Most zoonotic disease outbreaks are detected only after they have spread to humans. Others have suggested the use of syndromic surveillance methods (outbreak surveillance using pre-diagnostic data) in animals as a possible solution. We examined microbiology orders from two veterinary diagnostics laboratories (VDL) as a possible data source for early outbreak detection by establishing the species representation in the data, quantifying the potential gain in timeliness, demonstrating a method to determine estimated counts, and studying how statistical detection methods perform in identifying
53
clusters that might be indicators of an outbreak. The results indicated that VDL microbiology orders might be a useful source of data for a surveillance system designed to detect outbreaks of disease earlier than traditional reporting systems.
4.2 Introduction:
Emerging infectious diseases are newly recognized, clinically distinct diseases, or known diseases that are increasing in incidence in a given place or specific population
(5). More than 35 such diseases have been reported in humans between 1980 and 2003
(16). Many emerging pathogens are zoonotic (capable of infecting both humans and one or more species of lesser animal) (6, 7, 17). Indeed, diseases caused by organisms that can infect multiple species are two to four times more likely of being an emerging infectious disease than those that are specific to a single host (6, 17, 18). Although humans serve as the main reservoir for only 3% of all zoonotic pathogens (17), discovery of zoonotic disease outbreaks has often relied on the identification of human cases rather than surveillance in animals (7, 8). Detection of a zoonotic disease outbreak in an animal population first could possibly result in control efforts that prevent or greatly reduce human morbidity (8, 19). Improved surveillance systems and methods in animals might specifically benefit pandemic influenza preparedness (20-22), bioterrorism defense (23-
26), and response to other public health threats (12, 13).
Disease surveillance continues to develop as a core veterinary activity (27); however, regulatory programs and efforts to eradicate specific diseases have typically directed disease surveillance in animals (28). Therefore, these activities tend to be very disease specific and limited in their ability to detect other diseases, especially those with
54
non-specific presentation (28-31). Analysis of pre-diagnostic data may provide a solution
to the disease-specific limitations of current surveillance systems used for animals (32).
Several recently developed surveillance systems that use pre-diagnostic data for detecting disease outbreaks earlier in humans now exist. These systems do not focus on any specific disease, but rather capture and analyze incidence of non-specific, health-related events that may be indicators of disease. A key feature of many of these systems is the use of pre-existing electronic data found in registration, accounting, or inventory records.
Using data routinely collected for other purposes eliminates the need for additional entry and reduces the associated cost of more labor-intensive surveillance efforts.
Certain criteria for evaluating data to be used for surveillance systems are
recommended (4, 33-35). The Centers for Disease Control and Prevention have identified
the ability to provide baseline information on incidence trends and geographic
distribution as a prerequisite to detecting new or re-emerging infectious disease threats
(36). The baseline becomes especially important to determining when counts are
abnormally elevated. Making accurate interpretations from the results of detection
analyses is difficult without first establishing what is normal (32, 37, 38). Baselines help
to determine the noise in the data and provide for establishing expected values required in the analyses. Such indices are important to validate the predictive models used by detection systems to determine abnormal patterns of distribution or counts (32, 39, 40).
Representativeness and timeliness establish, in part, the quality of the data, one of
the criteria important to building a successful system. Representativeness describes how
well records in the system describe the population and indicates the potential of
accurately determining the distribution of cases by time and place (41). The presence of a
55
species may be a more important measure of representativeness for early outbreak surveillance (35). If detecting emerging diseases in pets is the goal of the system then it follows that the data need to include information for companion animals. The availability of data reflects the potential gain in terms of timeliness (42), the time from the disease event to the time the event is discovered (3, 34). Timeliness has become a major objective
of surveillance systems used to detect outbreaks of infectious disease (35). This potential gain establishes the value of data for earlier detection of disease outbreaks compared to
traditional disease reporting and detection systems.
Although scarcer in the veterinary community, sources of data exist that appear to
be similar to those used by early outbreak detection systems for humans. We considered
veterinary diagnostic laboratories (VDLs) one of those sources. These facilities typically
maintain electronic records of test orders and results that include animal species, date,
and geographical references (e.g. ZIP code). We hypothesized that:
1. VDL microbiology orders would possess qualities of
representativeness and timeliness required for use in an early outbreak
detection system,
2. baselines for microbiology orders and isolated microorganisms could
be determined from historic records, and
3. examination of microbiology order counts using detection algorithms
could identify pattern changes in the time-series that might indicate a
possible outbreak.
56
4.3 Methods:
4.3.1 Data Sample:
We conducted a retrospective study using datasets provided by IDEXX
Laboratories (Westbrook, Maine), a commercial veterinary diagnostics laboratory, and
The Ohio State University Veterinary Teaching Hospital (OSU) in Columbus, Ohio. The
datasets contained microbiology orders and results of tests on specimens originating from
patients in a seven county area of central Ohio received between January 2001 and
December 2003 (Figure 4.1). Origins of patients providing specimens were determined
using the ZIP code for either the submitting clinic (IDEXX) or patient’s residence (OSU).
Other information contained in each dataset included a unique accession number for each
specimen, date the laboratory received the specimen, animal species, date of test result,
and species of microorganism isolated.
4.3.2 Statistical Analysis:
To evaluate the quality of the data, we studied the datasets with descriptive statistics
using EpiInfo v3.3.2 (Centers for Disease Control and Prevention,
http://www.cdc.gov/epiinfo/). We determined the frequency of animal species in the
microbiology records to evaluate representativeness. Turnaround time, the time between the laboratory receiving a specimen and recording the test results, measured potential timeliness by averaging the difference between the dates for each record. We developed weekly time series of counts to study the baseline occurrence of isolates by grouping the
57
records according to the genus of the microorganism. Weeks consisted of seven days
beginning on Sunday. Weeks 1, 53, and 105 were the seven-day periods that included
January 1.
Serfling’s regression method, using Excel 2003 (Microsoft Corporation), provided
the method for establishing temporal clusters by developing models for the baseline occurrence of specific microorganism genera. The Serfling method combines a linear
term with sine and cosine terms to describe any seasonal change (36),
t t ˆ tY ++= 2sin( + 2cos() πβπββα ) , Eq (4.1) t st 52 c 52
where α is the intercept value, βt is the linear coefficient, and βs and βc are the model
coefficients for the sine and cosine terms, respectively, that describe any seasonal effect
at week t. The Serfling method requires data from non-epidemic periods to develop a base model. Information was not available to distinguish epidemic from non-epidemic weeks so we used a procedure similar to the one described by Tsui et al.(37) to remove counts that possibly represented epidemic weeks. The first 104 weeks of data (January
2001 – December 2002) were used to build the regression model which was applied to weeks 105-156 (January – December 2003). The procedure involved four steps:
1. Calculate an initial regression model for the first 104-week series using
equation (4.1).
58
2. Remove those counts with standardized residual values greater than
1.645, representing a one-tailed upper 95% CI.
3. Calculate a second regression model using equation (4.1) for the
remaining data to obtain a predicted value curve for the third year’s
data series.
4. Identify clusters in weeks 105 through 156 characterized by weekly
counts exceeding a threshold established by the upper 95% confidence
interval for two or more consecutive weeks.
We chose this method because of the seasonal variance in counts that were
expected for many enteric organisms such as Escherichia coli and Staphylococcus. As the frequency of observed counts in the time series approach zero, the accuracy of the
Serfling model becomes more unreliable (Garrick Wallstrom, University of Pittsburgh, personal communication, 2006) so we examined only the more frequently occurring microorganisms from each dataset. Temporal clusters, identified in this way, served as the gold standard (38) for studying the detection methods. We examined the 2003 weekly time-series of total microbiology orders for the IDEXX and the OSU datasets (Figure 4.2)
using three detection methods; Serfling regression, Exponentially Weighted Moving
Average (EWMA), and Cumulative Sum (CuSum) to determine if microorganism- specific clusters could be identified.
59
The Serfling regression method used the first 104 weeks of counts of orders as
previously described to build the expected values model for weeks 105 – 156. We
defined alerts as those weeks where the count value exceeded a threshold established by
the upper limit of a 95% CI.
EWMA is a simple time-series method and found to be particularly useful when
data are sparse (39). EWMA is
YS −= nntt −ωωσμ ))2(()( , Eq (4.2)
where St is the value at time t, Yt is the smoothed value of the actual count, μn the mean
of the baseline, σ n the standard deviation of the mean of the baseline, and ω the
smoothing constant with a value between zero and 1. A smoothing constant equal to 1.0
approximates a Shewart chart (39). Higher values of this constant are appropriate to detect sudden changes while a lower value is recommended for outbreaks with more gradual progression (40). We used ω = 0.4. The smoothed value is
t = ωXY t + − ω)1( Yt−1 , Eq (4.3)
where X t is the actual count for time t. EWMA provides more weight to more recent observations. Therefore, recent outbreaks could affect the accuracy of this model in
estimating counts (41). We studied the EWMA method, using a workbook developed by
Burkom (40) in Excel 2003 (Microsoft Corporation), with three baseline periods: 2
60
weeks, 3 weeks, and 5 weeks. Incorporating a lag time (i.e. buffer) in the baseline of
EWMA provides for increased sensitivity to more gradually increasing counts (39). This
buffer serves to avoid contamination of the baseline by a recent outbreak signal. We
studied each baseline with no buffer and a buffer of 1 week. The threshold was set at two
standard deviations above the expected value.
The third algorithm studied was a CuSum model. CuSum is sensitive to small
shifts in deviations from the mean and also detects shifts in the mean more quickly than
some other statistical process control methods (31, 41). CuSum value is
St = { t−1 + XS − μ + σ ))((,0max σ nnnt }, Eq (4.4)
where X t is the actual count for time t, μn the mean of the baseline counts, and σ n the standard deviation of the baseline counts. The sensitivity of CuSum to small changes also makes it susceptible to changes in reporting efficiency (41). We studied CuSum, using a workbook also developed by Burkom (40) in Excel 2003 (Microsoft Corporation), with baselines equal to two, three, and five weeks. We examined each baseline with no buffer and a buffer of 1 week. An alert occurred when the CuSum value exceeded one, approximately two standard deviations greater than the expected value determined from the baseline.
We calculated the Sensitivity, Predictive Value Positive (PVP), and Predictive
Value Negative (PVN) for each model. Sensitivity was determined by the number of cluster periods alerted to by the detection algorithm divided by the total number of
61
clusters. PVP, the probability of an alert identifying a cluster, was the number of true alerts (those alerts for weeks associated with one or more clusters) divided by the total number of alerts. PVN, the probability of no cluster when there was no alert, was the number of weeks with no alerts divided by the number of weeks not associated with any clusters.
4.4 Results:
4.4.1 Descriptive Statistics:
From January 2001 through December 2003, the number of accessions was 4,378
(OSU) and 9,287 (IDEXX). Of these, 1,840 (OSU) and 9,278 (IDEXX) accessions had at least one genus of microorganism isolated from them. Counts of unique isolates were
3,101 (OSU) and 10,761 (IDEXX). Median turnaround time (time between specimen receipt and test result) was three days (IDEXX, range 0-42 days) and five days (OSU, range 0-67 days). Seventy-nine unique genera of microorganisms were in the IDEXX dataset and 48 in the OSU dataset for this period. Table 4.1 summarizes the descriptive characteristics for each dataset. Escherichia coli and Staphylococcus were the most frequently isolated organisms in each dataset (Table 4.2). Both datasets consisted mainly of specimens from companion animal species: 59.0% canine and 21.6% feline (IDEXX), and 56.6% canine and 16.8% feline (OSU) (Table 4.3). Neither dataset contained records for any identified wildlife species.
62
4.4.2 Baseline Modeling:
We identified eight (IDEXX, Table 4.4) and three (OSU, Table 4.5) temporal
clusters from the microorganism-specific time series. Cluster periods ranged from two to
seven weeks with counts ranging from 18 to 172 cases (IDEXX) and nine to 15 cases
(OSU). There were two occasions (one for each dataset) where clusters of two unique
microorganism genera occurred during the same period. A pattern of increased orders,
not associated with an increase in the number of isolates, was present in the OSU time
series (Figure 4.3). The occurrence of these instances appeared to correlate with the
schedule for senior student clinical rotation for that year.
4.4.3 Detection Methods:
The detection methods studied performed markedly different between the two
datasets (Table 4.6). The sensitivity of methods when used with the IDEXX data ranged
from 0.50 (Serfling, EWMA five-week baseline with and without buffer, and EWMA
three-week baseline with buffer) to one (CuSum, two-week baseline, no buffer). Two
CuSum models, both with five-week baselines, had PVP=1. The lower range value was
0.64 (CuSum, two-week baseline, no buffer). All methods had similar PVN values (range
0.62 to 0.68). Eleven of the 13 detection method models studied using the OSU dataset
had sensitivity values of zero (i.e. no clusters detected). The upper sensitivity range was
0.67 (CuSum, two- and three-week baselines, one-week buffer). These models were also the only two with non-zero PVP values, 0.17 (CuSum, two-week baseline, one-week buffer) and 0.20 (CuSum, three-week baseline, one-week buffer). All methods had PVN values equal to 0.91 or 0.92.
63
4.5 Discussion:
There is an expectation of bias in data from laboratories toward patients that are
clinically ill (42, 43). This might limit the capability to detect sub-clinical occurrences or
outbreaks of disease that present with milder signs. At the same time, specificity might
also increase from the tiered structure of scrutiny (18) that begins with the animal owner,
includes the veterinarian, and ends at the laboratory (13).
Datasets from both laboratories contained records of specimens from companion,
agricultural, and exotic species. Companion animals (e.g. canine, feline, and pet birds) were the more frequent group with test orders, 82.4% (IDEXX) and 74.0% (OSU).
Equine species were also a frequent provider of samples, 8.2% (IDEXX) and 17.2%
(OSU). Although sometimes labeled as agricultural (19), horses differ from most other
species of agricultural animal since they generally are kept individually or in small groups for pleasure and/or show rather than in herds for consumption (44). Veterinary care of dogs, cats, pet birds, and horses is routinely more individually based as opposed to herd animals (e.g. cows, pigs, and sheep) (16) where diagnosis of disease in the unit
(i.e. herd) does not require testing every member (45). Economics may also influence decisions where agricultural animals are concerned where the expense to diagnose and treat may affect profit. The animal owner might request an attending veterinarian forego the additional expense of tests and rely on a diagnosis made from clinical presentation as a way to reduce the cost of veterinary care.
None of the clusters identified included any agricultural species, unless we
include equine in this group. This may have been a result of aggregating all microbiology orders, regardless of species, for analysis or because of the lower representation of
64
specimens from agricultural species in the datasets. The number needed to detect disease is less than that needed for surveillance to estimate prevalence (46). The lower frequency of agricultural species in the datasets should not automatically preclude them as a valuable source of data for early outbreak detection in these populations. Further investigation may better evaluate the adequacy of agricultural animal representation in these data as it pertains to outbreak detection efforts.
The average turnaround time for microbiology tests in these two datasets, three days (IDEXX) and five days (OSU), indicated a potential gain in timeliness that might be possible using microbiology order-based analysis compared to analysis using date of test result. The gain in timeliness makes the assumption that outbreak discovery occurs at the same time results are known. This may not always be the reality. Indeed, outbreak discovery might be delayed for some time until individual results from the various sample-submitting clinics are aggregated for analysis. Therefore, the actual gain in timeliness might be greater than that indicated by turnaround time. Based on previous models estimating these impacts from outbreaks of select disease a gain of three to five days could be substantial for reducing mortality and cost (8).
Examination indicated several periods of increased counts (clusters) that may have resulted from disease outbreaks. The majority of pathogens identified by the laboratories were not reportable in animals. Therefore, no registry existed to validate these instances being the result of true outbreaks of disease. Periods of increased counts could have been the result of more rigorous surveillance efforts, increased veterinary visits in reaction to a public service campaign, or other cause not related to an outbreak of disease. Additional clusters may have gone unrecognized because of the limitations in the
65
Serfling method with very infrequent counts or because increases may have been species- specific. For diseases of very low prevalence, the presence of only one to a few cases might define an epidemic. Investigation of these data with other retrospective methods
for determining epidemic periods may better identify temporal clusters of less frequently
appearing pathogens or those affecting only a select species. Detection of such outbreaks
would most likely not occur from analysis of aggregated test orders since the number of
samples resulting from such an outbreak would be small in comparison to the number of
samples overall.
The difference in detection method performance between the two datasets
appeared to be attributable to some factor other than the size of the clusters. The pattern
of increased microbiology test orders that did not correspond to increases in specimen
isolates observed in the OSU dataset might have resulted in the decreased ability of the
detection methods to identify aberrations caused by increases due to other causes.
Additional study of the patterns in the time series and possible confounders may better
identify a detection method to indicate significant changes from normal.
Detection systems of this type can only be expected to identify outbreaks within
certain size parameters earlier than traditional methods (47, 48). Outbreaks with only a
relative small number of cases may be lost in the volume of data and better detected by
other methods (e.g., an astute clinician who happens to see a sufficient number of the
patients, or a reportable disease system). A large, rapidly progressing, geographically
concentrated outbreak will most likely not require any specialized surveillance for
discovery. There are also inherent tradeoffs of certain attributes in these systems (47) and
directly related to the users’ needs and desires (49). For example, the ability to identify
66
smaller outbreaks (i.e. increased sensitivity) is often achieved by sacrificing the
specificity of the system (48, 50), resulting in an increase in false alerts. If investigating
alerts that have a lower probability of being epidemiologically significant is not a
concern, then such a compromise may be
acceptable. The usefulness of the system determines the correct balance of these attributes (42).
Early outbreak detection systems using VDL data may provide solutions for
limitations encountered by other systems, especially those dependent on manual provider
entries. A review of these manual systems found decreasing motivation over time, heavy
reporting load, and reporter compliance as some of the challenges in maintaining reporter
participation (51-55). A surveillance system that uses existing electronic data,
automatically transferred from laboratories, would not be dependent on reports entered
manually. Hence, motivation and reporting load of people assigned to enter data would
not be factors created by the surveillance program. Another challenge of reporter driven
systems, where historic data is not available, is the ability to determine baseline or
reference levels (52). Electronic laboratory records provide a historical record that might
provide a means to determine these baselines of occurrence.
Results of this study indicate that data from these VDLs have potential value for
early outbreak detection in certain animal populations. Analyzing these data to detect
excessive counts of veterinary laboratory orders might provide for more timely detection
of disease outbreaks. Subsequent investigation of these data, including detection methods
and their relation to other data, is necessary to assert their true value for any specified
need. Further studies are justified by these findings and should continue to establish
67
where these and similar data sources fit into the overall biosurveillance efforts to detect outbreaks of emerging infectious disease.
68
Figure 4.1: Seven county region of central Ohio where microbiology specimens submitted to IDEXX and OSU laboratories originated.
69
120
100 OSU IDEXX
80
60 69
40
20
0 1 5 9 13 17 21 25 29 33 37 41 45 49 Week Specimen Received
Figure 4.2: Counts of IDEXX and OSU microbiology orders originating from central Ohio, 2003
IDEXX OSU specimen accessions 9,287 4,378 specimen accessions resulting in isolate/s 9,278 1,840
unique isolates* 10,761 3,101
unique microorganism
genera 79 48
turnaround (days) Range 0-42 0-67 Median 3 5 Mean 3.4 5.6
*multiple genera of microorganisms were sometimes isolated from a single specimen
Table 4.1: Select characteristics of microbiology datasets for specimens originating in central Ohio, January 2001 – December 2003.
70
IDEXX OSU Organism Count % of total Organism Count % of total Escherichia coli 2,891 26.9% Escherichia coli 614 19.8% Staphylococcus 2,305 21.4% Staphylococcus 569 18.3% Enterococcus 961 8.9% Streptococcus 355 11.4% Proteus 951 8.8% Pseudomonas 247 8.0% Pseudomonas 949 8.8% Enterococcus 199 6.4% Streptococcus 761 7.1% Proteus 134 4.3% Bacillus 375 3.5% Bacillus 98 3.2% Klebsiella 192 1.8% Pasteurella 94 3.0% Other 1376 12.8% Other 791 25.5%
Table 4.2: Microorganism genera isolated from specimens originating in central Ohio, January 2001 – December 2003.
71
IDEXX OSU Animal Species Count % of accessions Count % of accessions Avian 165 1.8 26 0.6 Bovine 9 0.1 81 1.9 Canine 5,480 59.0 2,477 56.6 Caprine 10 0.1 28 0.6 Equine 766 8.2 752 17.2 Exotic 29 0.3 107 2.4 Feline 2,008 21.6 735 16.8 Ovine 3 <0.1 8 0.2 Porcine 17 0.2 3 0.1 Reptile 234 2.5 15 0.3 Other 41 0.4 80 1.8 Unknown 525 5.7 66 1.5
Table 4.3: Frequency of species from central Ohio providing specimens for microbiology analysis, January 2001 – December 2003.
72
Cluster Period Animal % of % of total Cluster # VDL (weeks) Organism Species Count cases orders 1 IDEXX 9-10 Escherichia coli Canine 24 55.8 29.9 Feline 14 32.6 Equine 3 7.0 Unknown 2 4.7 2 IDEXX 13-14 Streptococcus Canine 9 50.0 13.0 Feline 3 16.7 Equine 3 16.7 Unknown 3 16.7 3 IDEXX 16-17 Staphylococcus Canine 26 68.4 21.7 Feline 10 26.3 Equine 1 2.6 Reptile 1 2.6 4 IDEXX 20-21 Staphylococcus Canine 24 51.1 25.5 Feline 21 44.7 Equine 1 2.1 Reptile 1 2.1 5 IDEXX 32-33 Escherichia coli Canine 37 44.0 38.0 Unknown 26 31.0 Feline 18 21.4 Equine 2 2.4 Other 1 1.2 6 IDEXX 32-34 Proteus Canine 24 66.7 11.7 Unknown 9 25.0 Feline 1 2.8 Equine 1 2.8 Reptile 1 2.8 7 IDEXX 41-47 Escherichia coli Canine 81 47.1 30.5 Unknown 40 23.3 Feline 37 21.5 Equine 7 4.1 Exotic 3 1.7 Reptile 3 1.7 Avian 1 0.6 8 IDEXX 49-52 Bacillus Feline 9 42.9 8.1 Canine 8 38.1 Equine 2 9.5 Reptile 1 4.8 Unknown 1 4.8
Table 4.4: Clusters determined from IDEXX microbiology test results of specimens originating from animals treated at clinics located in central Ohio, 2003.
Cluster Period Animal % of % of total Cluster # VDL (weeks) Organism Species Count cases orders 1 OSU 24-25 Escherichia coli Canine 12 80.0 41.7* Unknown 2 13.3 Feline 1 6.7 2 OSU 24-25 Pseudomonas Canine 8 80.0 Feline 1 10.0 Avian 1 10.0 3 OSU 28-29 Pseudomonas Canine 8 88.9 13.6 Exotic 1 11.1 *combined for clusters 1 & 2 since they occur during the same weeks
Table 4.5: Clusters determined from OSU microbiology test results of specimens originating from animals residing in central Ohio, 2003.
74
50
45
40
35
30
25
75 20
15
10 Isolates 5 Microbiology Orders 0 1 5913 17 21 25 29 33 37 41 45 49 Week Specimen Received
Figure 4.3: Counts of OSU microbiology orders (solid line) and isolates (dashed line) for specimens received from patients residing in central Ohio during 2003. Vertical dotted lines denote the weeks of senior student trimester rotation.
IDEXX OSU Baseline/ Buffer Method (weeks) Threshold Sn PVP PVN Sn PVP PVN Serfling NA 95% CI 0.50 0.67 0.63 0 0 0.92 EWMA 5/1 2σ 0.50 0.80 0.62 0 0 0.91 EWMA 5/0 2σ 0.50 0.80 0.62 0 0 0.92 EWMA 2/1 2σ 0.75 0.70 0.64 0 0 0.91 EWMA 2/0 2σ 0.75 0.67 0.65 0 0 0.92 EWMA 3/1 2σ 0.50 0.71 0.62 0 0 0.92 EWMA 3/0 2σ 0.63 0.71 0.62 0 0 0.91
CUSUM 5/1 St = 1 0.63 1.00 0.65 0 0 0.92
CUSUM 5/0 St = 1 0.50 1.00 0.61 0 0 0.92
CUSUM 2/1 St = 1 0.75 0.82 0.68 0.67 0.17 0.91
CUSUM 2/0 St = 1 1.00 0.64 0.66 0 0 0.91
CUSUM 3/1 St = 1 0.63 0.83 0.63 0.33 0.20 0.92
CUSUM 3/0 St = 1 0.63 0.83 0.63 0 0 0.91
Table 4.6: Performance indicators for detection algorithms in identifying microorganism clusters from time series of aggregate microbiology orders.
76
REFERENCES
1. Institute of Medicine. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, DC, National Academy Press.
2. Karesh WB et al. Wildlife Trade and Global Disease Emergence. Emerg Infect Dis 2005;11:1000-2.
3. Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:983-9.
4. Cleaveland S, Laurenson MK, Taylor LH. Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:991-9.
5. Meslin FX, Stohr K, Heymann D. Public health implications of emerging zoonoses. Rev Sci Tech 2000;19:310-7.
6. Woolhouse MEJ . Population biology of emerging and re-emerging pathogens. Trends Microbiol 2002;10:S3-S7.
7. Childs J et al. Emerging Zoonoses. Emerg Infect Dis 1998;4:453-4.
8. Kaufmann AF, Meltzer MI, Schmid GP. The Economic Impact of a Bioterrorist Attack: Are Prevention and Postattack Intervention Programs Justifiable? Emerg Infect Dis 1997;3:83-94.
9. Fouchier RAM, Osterhaus ADME, Brown IH. Animal influenza virus surveillance. Vaccine 2003;21:1754-7.
10. Kaye D, Pringle CR. Avian Influenza Viruses and their Implication for Human Health. Clin Infect Dis 2005;40:108-12.
11. U.S. Department of the Interior. An Early Detection System for Highly Pathogenic H5N1 Avian Influenza in Wild Migratory Birds U.S. Interagency Strategic Plan [Web Page]. 2006 Mar 14; Accessed 2006 Aug 24. Available at: www.doi.gov/issues/birdflu_strategicplan.pdf.
12. National Research Council. Making the Nation Safer: The role of science and technology in countering terrorism. 2002. Washington, DC, National Academy Press.
13. Conti L. Petborne Zoonoses: Detection and Surveillance Challenges. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press.
77
14. Ashford DA et al. Planning against Biological Terorism: Lessons from Outbreak Investigations. Emerg Infect Dis 2003;9:515-9.
15. Ashford DA et al. Biological terrorism and veterinary medicine in the United States. J Am Vet Med Assoc 2000;217:664-7.
16. Shephard R, Aryel RM, Shaffer L. Animal Health. In: Wagner MM, Moore AW, Aryel RM, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:111-27.
17. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality.
18. Doherr MG, Audige L. Monitoring and surveillance for rare health-related events: a review from the veterinary perspective. Philos Trans R Soc Lond B Biol Sci 2001;356:1097-106.
19. Wurtz RM, Popovich ML. Animal Disease Surveillance: A Framework for Supporting Disease Detection in Public Health. March 2002. Tuscon, AZ, Scientific Technologies Corporation.
20. Lynn T. Coordinating Zoonotic Disease Surveillance: Partnering Agriculture and Public Health. NAHSS Outlook [June]. 2005 . United States Department of Agriculture . Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook.
21. Kelly TK et al. The Office of Science and Technology Policy Blue Ribbon Panel on The Threat of Biological Terrorism Directed Against Livestock. March 2004. Arlington, VA, RAND Science and Technology.
22. Koo D, Parrish II RG. The Changing Health-Care Information Infrastructure in the United States: Opportunities for a New Approach to Public Health Surveillance. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford University Press, 2000:76-94.
23. Moodie M et al. Biological Terrorism in the United States: Threat, Preparedness, and Response. [200*1999*00132]. November 2000. Chemical and Biological Arms Control Institute.
24. Wagner MM, Shaffer L, Shephard R. Functional Requirements for Biosurveillance. In: Wagner MM, Moore A, Aryel R, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:51-64.
25. Fienberg SE, Shmueli G. Statistical issues and challenges associated with rapid detection of bio-terrorist attacks. Stat Med 2005;24:513-29.
78
26. Buehler JW et al. Framework for Evaluating Public Health Surveillance Systems for Early Detection of Outbreaks. MMWR Recomm Rep 2004;53.
27. Sosin DM. Draft Framework for Evaluating Syndromic Surveillance Systems. J Urban Health 2003;80:i8-i13.
28. Centers for Disease Control and Prevention. Addressing emerging infectious disease threats, a prevention strategy for the United States. 1994. Atlanta, GA, Centers for Disease Control and Prevention.
29. Reis BY, Mandl KD. Time series modeling for syndromic surveillance. Med Inform Decis Making 2003;3.
30. Wong W-K et al. Rule-Based Anomaly Pattern Detection for Detecting Disease Outbreaks. 2002. American Association for Artificial Intelligence.
31. Hutwagner LC et al. Using Laboratory-Based Surveillance Data for Prevention: An Algorithm for Detecting Salmonella Outbreaks. Emerg Infect Dis 1997;3:395- 400.
32. Focks D. Mathematical Models and Predictors of Disease Outbreaks. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washnigton, DC, National Academy Press.
33. Mandl KD et al. Implementing Syndromic Surveillance: A Practical Guide Informed by the Early Experience. J Am Med Inform Assoc 2004;11:141-50.
34. Sonesson C, Bock D. A review and discussion of prospective statistical surveillance in public health. J R Stat Soc A 2003;166:5-21.
35. Wagner MM et al. The Emerging Science of Very Early Detection of Disease Outbreaks. J Pub Health Manag Pract 2001;7:51-9.
36. Serfling RE. Methods for Current Statistical Analysis of Excess Pneumonia- Influenza Deaths. Public Health Rep 1963;78:494-506.
37. Tsui F-C et al. Value of ICD-9-Coded Chief Complaints for Detection of Epidemics. Proceedings of the AMIA Annual Symposium. 711-715. 2001.
38. Friedman CP, Wyatt JC. Evaluation Methods in Medical Informatics. New York, NY: Springer-Verlag, 1997.
39. Buckeridge DL et al. Algorithms for rapid outbreak detection: a research synthesis. J Biomed Inform 2005;38:99-113.
79
40. Burkom H. Manual for Accessible Alerting Algorithms. September 23, 2005. The Johns Hopkins University Applied Physics Laboratory.
41. Farrington P , Andrews N. Outbreak Detection: Application to Infectious Disease Surveillance. In: Brookmeyer R, Stroup DF, eds. Monitoring the Health of Populations: Statistical Principles and Methods for Public Health Surveillance. New York, NY: Oxford Univeristy Press, 2004:203-31.
42. Buehler JW. Surveillance. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins, 1998:435-57.
43. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board. Available at: www.porkboard.org/docs/mengle.pdf.
44. Amercian Veterinary Medical Association. U.S. Pet Ownership and Demographics Sourcebook. 2002. Schaumburg, IL, American Veterinary Medical Association.
45. Christensen J. Application of Surveillance and Monitoring Systems in Disease Control Programs. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:15-34.
46. Pfeiffer DU. Veterinary Epidemiology - An Introduction . London: The Royal Veterinary College, Univeristy of London, 2002.
47. Stoto MA, Schonlau M, Mariano LT. Syndromic Surveillance: Is It Worth the Effort? Chance 2004;17:19-24.
48. Reingold A. If Syndromic Surveillance Is the Answer, What Is the Question? Biosecur Bioterror 2003;1:77-81.
49. Pavlin JA et al. Innovative Surveillance Methods for Rapid Detection of Disease Outbreaks and Bioterrorism: Results of an Interagency Workshop on Health Indicator Surveillance. Am J Pub Health 2003;93:1230-5.
50. Pavlin JA. Investigation of Disease Outbreaks Detected by "Syndromic" Surveillance Systems. J Urban Health 2003;80:i107-i114.
51. Mauer WA, Kaneene JB. Integrated Human-Animal Disease Surveillance. Emerg Infect Dis 2005;11:1490-1.
52. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10.
80
53. DeGroot B. The Rapid Syndrome Validation Project for Animals - Augmenting Contact with the Network of Accredited Veterinarians. NAHSS Outlook [April]. 2005. United States Department of Agriculture. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook.
54. Gobar GM, Case JT, Kass PH. Program for surveillance of causes of death of dogs, using the internet to survey small animal veterinarians. J Am Vet Med Assoc 1998;213:251-6.
55. Robotham J, Green LE. Pilot study to investigate the feasibility of surveillance of small animals in the UK. J Small Anim Pract 2004;45:213-8.
81
CHAPTER 5
CLINICAL ROTATION OF SENIOR VETERINARY STUDENTS AS A
CONFOUNDER FOR OUTBREAK DETECTION USING MICROBIOLOGY
ORDERS IN A VETERINARY TEACHING HOSPITAL
5.1 Abstract:
We report on the clinical rotation of senior veterinary students as a confounder
that was identified during a retrospective study of microbiology orders from a veterinary teaching hospital. Laboratory orders might provide a useful source of data for surveillance systems utilizing prospective analysis to detect outbreaks of infectious disease, including those caused by emerging pathogens and potential bioweapon agents, earlier than possible through traditional reporting. This method, commonly referred to as syndromic surveillance, exploits the trends of indicators of health, in this case microbiology test orders, as a proxy for cases of disease. Sensitivity of detection is improved by establishing the baseline occurrence of these indicators in order to estimate expected counts. We investigated the proportion of cultures that resulted in isolates to the number of microbiology orders by week for a one-year period and found that order trends
82
may be affected by the clinical rotation of senior students. While increases observed in
the baseline may not have been the result of disease, they demonstrate a predictable
pattern. Awareness of this possible influence could be important to system designers
considering these and related data for outbreak detection.
5.2 Introduction:
A new requirement for biosurveillance systems now exists resulting in part from concerns over bioterrorism and emerging infectious disease; the very early detection of
disease outbreaks (1). One promising approach to achieve this goal is the application of
syndromic surveillance methods. Syndromic surveillance systems analyze the changes in
trends for health indicators as a proxy to disease events. The prospective analysis of
healthcare related data that is available before diagnosis is made, such as laboratory
orders (2, 3), may provide for improved timeliness of outbreak identification by providing alerts to increases in indicators that might signal the elevated incidence of disease (4).
Animal-based data might be a potentially rich source for these surveillance efforts
(5). Animals have repeatedly demonstrated their usefulness as sentinel indicators of disease (6-9). Many animals could show signs of certain infectious disease long before their human keepers are feeling ill (10, 11). Therefore, earlier detection of an outbreak in animals might provide for earlier warning of potential or impending disease in humans including pandemic influenza (12, 13) and those that might result from bioterrorist activities (14).
83
We had hypothesized that veterinary laboratory orders might be a valuable data
source for earlier detection of disease outbreaks in animals. During a retrospective
investigation of a dataset containing microbiology orders from the veterinary teaching
hospital at The Ohio State University for 2003 (15), we noticed what appeared to be
regular intervals when increases in the number of culture orders were not accompanied
by increases in the number of isolates (Figure 5.1). Decreased proportions of isolates to
orders appeared to coincide with the clinical rotation of students. Senior veterinary
students rotated through three different clinic groups in 2003. Clinics I was comprised of
small animal medicine, small animal surgery, dermatology, radiology, and ophthalmology. Clinics II, which coincided with Clinics I, included equine medicine, food animal, anesthesia, and general practice. Clinics III consisted of emergency, intensive care, necropsy, preventive medicine, and large animal and equine field service.
During each of the 16-week trimester rotations, students cycled through Clinics I & II every five weeks and Clinics III every 4. There were only four occasions when Clinics I
& II began during the same week as Clinics III. Students were in clinical rotation every week of 2003 except the first two weeks of January and the last two weeks of December.
5.3 Methods:
Records of all microbiology orders, and results, made in 2003 for specimens
provided by patients that lived in a seven county area of Central Ohio were obtained from
the Veterinary Teaching Hospital at The Ohio State University. The clinical laboratory at
the Veterinary Teaching Hospital services small and large animal clients treated on
84
outpatient, inpatient, or emergency basis by students, residents, and instructional staff of
the College of Veterinary Medicine. Specimens are not accepted from outside sources.
We determined the temporal distribution of specific microbiology orders by
considering the percentage of each test represented in the average weekly total. First
weeks of clinical rotation normally began on a Monday except for three occasions when
rotations for Clinics I and II started on a Thursday. For these occurrences, we defined the
first week of rotation as the subsequent week. We calculated the proportion of culture
isolates to microbiology test orders for each week that students were in clinical rotation
(weeks 3 through 50), excluding those that comprised the students’ winter break. We
investigated the association between these proportions and first weeks of rotation by
considering first weeks of rotations as exposure, compared to all other weeks as non-
exposure. We counted the number of times for both that the proportion of isolates to
orders was less than average and average or greater then analyzed these values using
Epitome Stats Calculator v1.3 (www.biostat.mcw.edu/phome/rh_files/epitome.html) to
determine the odds ratio, including the 95% confidence interval (CI), of a less than average proportion occurring during the first weeks of clinical rotation compared to all other weeks. This exercise was repeated three times. Once for the first weeks of any rotation, then for the first weeks of only Clinics I & II rotation and for the first weeks of only Clinics III.
5.4 Results:
There were 1,460 microbiology test orders submitted during the 48 weeks of
clinical rotation in 2003. Together, blood cultures, mycology cultures, and susceptibility
85
cultures comprised 97.1% of all microbiology orders (Table 5.1). The remaining 2.9%
consisted of Campylobacter cultures, Mycoplasma cultures, Trichomonas cultures, acid-
fast stains, and Gram stains. The average weekly proportion of isolates to test during the
48 weeks of clinical rotation was 0.28 (range, 0 to 1) for blood cultures, 0.23 (range, 0 to
1) for mycology cultures, and 0.71 (range, 0.29 to 1) for susceptibility cultures. The
overall average proportion of isolates to orders during clinical weeks was 0.66 (range,
0.29 to 1). The average proportion during the first weeks of Clinics III rotation was 0.63
(range, 0.36 to 1) and first weeks of Clinics I and II was 0.52 (range, 0.34 to 0.73).
A below average proportion of isolates to microbiology orders was five times
more likely to occur during the first weeks of clinical rotations than other weeks (Figure
5.2). The OR for a below average proportion during the first weeks of Clinics III rotation
compared to other weeks was 3.38 (95% CI, 0.90 to 12.68). For Clinics I & II, the OR
was 12.80 (95% CI, 2.05 to 79.84). When stratified by test, increased OR’s for below average proportions during the first weeks of Clinic I & II were also found for
susceptibility cultures (OR, 14.29; 95% CI, 2.33 to 87.49) and mycology cultures (OR,
3.53; 95% CI, 0.41 to 30.47). OR’s were also notable for susceptibility cultures (OR,
3.81; 95% CI, 1.02 to 14.28) and blood cultures (OR, 3.50; 95% CI, 0.55 to 22.11) for the first weeks of Clinics III rotation.
5.5 Discussion:
There are inherent biases involved when using data from veterinary laboratories
for disease surveillance. These include submissions predominantly originating from
individuals that are clinically ill and clinicians foregoing tests on additional individuals
86
once a diagnosis can be confirmed (16). Other factors that may influence these data
include the cost of testing, the value of the animal, the working relationship of the
veterinarian with the laboratory, and the integrity of the specimens (17). In addition to
timely availability, data should be reliable and stable. As a result of using data that are
proxy to an event, the specificity of these systems will be less than those systems based
on confirmed diagnoses and tests.
Identifying potential biases and confounders that may affect data quality is an
important consideration when evaluating surveillance systems (18, 19).Having the benefit
of predictable temporal trends is a key requirement to improve upon the specificity of
detecting outbreaks (20-22). Identification of factors that impact on the reliability of the temporal trends observed in the data may provide for the ability to improve the capability to identify aberrations in those trends.
These results suggest that the test ordering behavior of veterinary students may be
influenced by their introduction into a new clinical experience. This is especially true for
those rotations that include small animal medicine and dermatology. The goals and
methods of instructors might also mediate test ordering behavior. While clinical
instructors supervise the use of tests by students, such use may not be restricted, fostering
an environment where students learn from their individual choices made during the
diagnostic process. Test ordering from an ever and frequently changing group of
providers, a situation unique to a teaching environment, may influence the predictability
of microbiology order data. How student behavior in each clinical rotation contributes to
this effect remains to be determined.
87
Another possible influence suggested by the results may arise from the different clinics that are included in each rotation. The OR of a below average isolate to blood culture order proportion occurring during the first weeks of Clinics III was greater than that for the first weeks of Clinics I & II. We will recall that the experiences that make up
Clinics III included intensive care and necropsy where the potential for blood cultures to be more of a normally occurring test may be greater than for other experiences such as general practice or dermatology. Conversely, the OR for a below average proportion in mycology orders during the first weeks of Clinics I & II, which include such experiences as dermatology, were ten fold higher than during the first weeks of Clinics III.
The temporal pattern we observed requires further investigation to explore the potential to model the effect and accurately calculate expected values. Developers investigating data from veterinary teaching hospitals for use by syndromic surveillance systems should consider the unique biases that may exist and could influence the counts of certain values such as the number of laboratory test orders.
Preceptors of veterinary students may also find these results of interest. Compared to human medicine where many laboratory procedures are paid for through third party insurance carriers, laboratory testing in veterinary medicine may be inhibited by the financial abilities of the client (23, 24). In such an environment, accurate diagnosis may often need to be made without benefit of laboratory results. Veterinary students and their future clients could potentially benefit from teaching experiences where diagnoses do not rely substantially on laboratory testing.
88
50
45 Isolates Microbiology Orders
40
35
30
25 Count 89
20
15
10
5
0 1 5 9 13 17 21 25 29 33 37 41 45 49 Week
Figure 5.1: Counts of microbiology test orders and isolates originating from The Ohio State University Veterinary Teaching Hospital during 2003.
1.00
0.90
0.80
0.70
0.60
0.50
0.40 90 0.30
0.20 ratio of isolates of microorganismsof isolatesratio of microbiology to orders
0.10
0.00 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 Week
Figure 5.2: Proportion of isolates to microbiology orders by week originating from The Ohio State University Veterinary Teaching Hospital during 2003. Weeks 3, 9, 14, 19, 25, 30, 35, 41, and 46 were first weeks of rotation for Clinics I & II. Weeks 3, 6, 10, 14, 18, 22, 26, 30, 34, 38, 42, 46, and 50 were first weeks of rotation for Clinics III. Horizontal line denotes average proportion for all weeks of clinical rotation.
OR of Blood Culture 2.1 3.3 0.28 3.60 0.95 3.50 (0-8) (0-18.2) (0-1) (0.64,20.33) (0.13,7.23) (0.55,22.11) Mycology Culture 2.2 5.3 0.23 0.77 3.53 0.33 (0-8) (0-26.3) (0-1) (0.18,3.29) (0.41,30.47) (0.07,1.52) Susceptibility Culture 25.4 88.5 0.71 6.07 14.29 3.81 91 (12-39) (73.7-100) (0.29-1) (1.75,21.03) (2.33.87.49) (1.02,14.28) Table 5.1: Average weekly values for microbiology orders made during clinical rotations at The Ohio State University Veterinary Teaching Hospital in 2003 and the OR of the proportion of isolates to orders being less than average during the first weeks of rotation compared to any other weeks of rotation. Clinics I & II included small animal practice and surgery, dermatology, equine medicine, food animal, and general practice. Clinic III included emergency, intensive care, necropsy, preventive medicine, and large animal and equine field service. REFERENCES 1. Wagner MM, Shaffer L, Shephard R. Functional Requirements for Biosurveillance. In: Wagner MM, Moore A, Aryel R, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:51-64. 2. Buehler JW et al. Framework for Evaluating Public Health Surveillance Systems for Early Detection of Outbreaks. MMWR Recomm Rep 2004;53. 3. National Center for Infectious Diseases. Recognition of Illness Associated with the Intentional Release of a Biologic Agent. MMWR Morb Mortal Wkly Rep 2001;50. 4. Tsui F-C et al. Value of ICD-9-Coded Chief Complaints for Detection of Epidemics. Proceedings of the AMIA Annual Symposium. 711-715. 2001. 5. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality. 6. Centers for Disease Control and Prevention. Epidemic/Epizootic West Nile Virus in the United States: Guidelines for Surveillance, Prevention, and Control. 2003. Ft. Collins, CO. 7. Eidson M et al. Dead Bird Surveillance as an Early Warning System for West Nile Virus. Emerg Infect Dis 2001;7:631-5. 8. Gill JS et al. Serologic Surveillance for the Lyme Disease Spirochete, Borrelia burgdorferi, in Minnesota by Using White-Tailed Deer as Sentinel Animals. J Clin Microbiol 1994;32:444-51. 9. Morris CD et al. Comparison of chickens and pheasants as sentinels for eastern equine encephalitis and St. Louis encephalitis viruses in Florida. J Am Mosq Control Assoc 1994;10:545-8. 10. Babin SM et al. Early detection of possible bioterrorist events using sentinel animals. The 131st Annual Meeting of the American Public Health Association. 2003. 11. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category A agents. J Am Vet Med Assoc 2004;224:1084-95. 12. Fouchier RAM , Osterhaus ADME, Brown IH. Animal influenza virus surveillance. Vaccine 2003;21:1754-7. 92 13. U.S. Department of the Interior. An Early Detection System for Highly Pathogenic H5N1 Avian Influenza in Wild Migratory Birds U.S. Interagency Strategic Plan. March 14, 2006. August 24, 2006. 14. Conti L. Petborne Zoonoses: Detection and Surveillance Challenges. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press. 15. Shaffer LE et al. Evaluation of Microbiology Orders from two Veterinary Diagnostic Laboratories as Potential Data Sources for Early Outbreak Detection. Adv Disease Surveil. forthcoming. 16. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board. 17. Power C. Passive Animal Disease Surveillance in Canada: A Benchmark. Proceedings of a CAHNet Workshop. November 1999. Canadian Food Inspection Agency. 18. Centers for Disease Control and Prevention. Updated Guidelines for Evaluating Public Health Surveillance Systems: Recommendations from the Guidelines Working Group. MMWR Recomm Rep 2001;50. 19. Sosin DM. Draft Framework for Evaluating Syndromic Surveillance Systems. J Urban Health 2003;80:i8-i13. 20. Duchin JS. Epidemiological Response to Syndromic Surveillance Signals. J Urban Health 2003;80:i115-i116. 21. Reis BY, Mandl KD. Time series modeling for syndromic surveillance. Med Inform Decis Making 2003;3. 22. Wong W-K et al. Rule-Based Anomaly Pattern Detection for Detecting Disease Outbreaks. 2002. American Association for Artificial Intelligence. 23. Doherr MG, Audige L. Monitoring and surveillance for rare health-related events: a review from the veterinary perspective. Philos Trans R Soc Lond B Biol Sci 2001;356:1097-106. 24. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10. 93 CHAPTER 6 EARLY OUTBREAK DETECTION USING AN AUTOMATED DATA FEED OF TEST ORDERS FROM A VETERINARY DIAGNOSTIC LABORATORY 6.1 Abstract: Disease surveillance in animals remains inadequate to detect outbreaks resulting from novel pathogens and potential bioweapons. Mostly relying on confirmed diagnoses, another shortcoming of these systems is their ability to detect outbreaks in a timely manner. We investigated the feasibility of using veterinary laboratory test orders in a prospective system to detect disease outbreaks earlier compared to traditional reporting methods. IDEXX Laboratories, Inc. automatically transferred daily records of laboratory test orders submitted from veterinary providers in Ohio. Test products were classified to appropriate syndromic category using their unique identifying number. Counts of each category by county were analyzed to identify unexpected increases using a cumulative sums method. The results indicated that disease events can be detected through the prospective analysis of laboratory test orders and may provide indications of similar disease events in humans before traditional disease reporting. 94 6.2 Introduction: Prompt detection of outbreaks might provide for earlier intervention efforts that result in minimizing their overall impact (1). Some animals are susceptible to infection from many of the same pathogens as humans, sometimes showing signs of disease earlier (2). Therefore, animals might be used as sentinels and provide for earlier recognition of disease outbreaks that could affect humans (3). As pet animals share much of the same environment as their human owners, they especially might prove to be valuable outbreak sentinels (2). Most of the current disease surveillance systems used for animal populations are considered inadequate for detecting outbreaks of emerging disease, potential acts of bioterrorism, or outbreaks resulting from pathogens for which the system was not specifically designed for in a timely manner (4-7). Such functionality in animal-based systems has been considered important to our overall bioterrorism and disease outbreak preparedness capabilities (8-11). Syndromic surveillance methods utilize population health indicators to warn of potential outbreaks earlier than reports of confirmed diagnoses. Although many sources of data have been investigated for syndromic surveillance in humans, there is paucity in the literature describing similar studies in animals. Laboratories are recognized as important sources of data for disease surveillance in animals (12) as well as humans (13). Test orders for specimens submitted to commercial medical laboratories have been utilized as one of the data sources for syndromic surveillance in humans (14). Most of the private veterinary practitioners in the United States also submit specimens to commercial laboratories for diagnostic testing 95 (15). Through the utilization of data from these commercial laboratories, we might possibly achieve the benefit of the aggregation of many veterinary providers across a wide geographic area. Such centralized aggregation of data may be important in detecting certain outbreaks (1). In a previous study, we retrospectively investigated microbiology test orders made to a commercial veterinary diagnostics laboratory (VDL). The results of this study indicated that VDL test orders might be useful in identifying increased cases of disease potentially resulting from an outbreak. Although laboratory analyses are not as frequently a part of the veterinary care of pet animals compared to the medical care of humans (11), we hypothesize that the consistency of test orders over time is such that increases in cases of disease will result in detectable increases in the number of test orders submitted by veterinarians. 6.3 Methods: We conducted a prospective study of laboratory orders submitted to IDEXX Laboratories, Inc. (Westbrook, Maine) for specimens originating from veterinary clinics in Ohio between September 1, 2006 and November 30, 2006. IDEXX transferred once daily to a server located at the Real-time Outbreak and Disease Surveillance (RODS) Laboratory (University of Pittsburgh, Pennsylvania), via secure file transfer protocol, an automatically generated text file containing records of laboratory orders for specimens received within the previous 24-hour period. Each record included the accession number 96 assigned by IDEXX to the specimen, date and time that IDEXX received the specimen, 5-digit ZIP code of the clinic submitting the specimen, species of animal, and numerical code/s of the laboratory product/s ordered. 6.3.1 Mapping Laboratory Orders to Syndromic Category: We previously distributed a list of product descriptions ordered during a 2-week period to ten small and large animal veterinarians asking them to consider the diseases that they might use each product to confirm or rule out during the diagnostic process. The veterinarians would then assign each product to syndromic categories based on the expected presentation of these diseases. Eight categories were considered initially: respiratory, gastrointestinal, neurologic, behavioral, dermal, reproductive, non-specific, and sudden death (Table 6.1). Seven of the ten surveyed veterinarians returned the categorized lists. The behavioral and sudden death categories were subsequently removed based on zero responses from the surveyed veterinarians for these categories. In addition to the surveyed veterinarians, two IDEXX laboratorians also reviewed the list of products. Based on their input and advice, five categories were added to further describe many of those products that had been classified into the non-specific category. These additional categories were endocrine, hepatic, infectious, febrile, and renal. Records were mapped to syndromic category based on the identifying number for the laboratory product ordered (Appendix C) and appropriately classified as the RODS Laboratory server received them. 97 6.3.2 Descriptive Statistical Analyses: We used frequency analysis to describe the representation of species groups and distribution of accessions by day of the week. ArcView 8.1 (ESRI, Redlands, California) was used to generate a cartographical image to evaluate the geographical distribution of specimen origin. The percentage of the total daily records included in the dataset for each 24-hour period was used to describe the availability of records. 6.3.3 Detection Method: A cumulative sums (CuSum) method was used to analyze category counts, as records were received, for each Ohio County, as determined by the ZIP code. The value of the CuSum was calculated as St = { t−1 + XS − μtt + σ ))5.0((,0max σ tt }, Eq (6.1) where X t was the observed count at time t, μt the expected count (baseline), and σ t the standard deviation of the counts used to determine the baseline. Daily analysis was performed automatically using the count from the current and previous six days for the observed value. A moving 7-day period was chosen to reduce the anticipated day-of- week effect in the data. The expected value was calculated by averaging the weekly counts for the previous 4-week period. We defined alerts as instances when the CuSum value equaled or exceeded five. 98 An alert period was defined as at least two consecutive days where the CuSum value exceeded the threshold. By using this two-in-a-row rule we were able to somewhat reduce the impact of single-day increases on weekly counts. Using this rule has been shown to increase the “robustness” of CuSum methods (16). Alerts were considered for all syndromic categories except non-specific, which was mostly comprised of general screening tests such as blood chemistries. We investigated alerts by identifying the specific laboratory product or products involved and contacting select veterinarians located in the same area as the alert asking about their impressions of disease activity. Veterinarians may or may not have been IDEXX clients at the time of the alert. 6.4 Results: 6.4.1 Evaluation of Data Transfer: During the study, the daily transfer of data from IDEXX Laboratories was interrupted twice. The first interruption began on September 7 and continued through September 28. This interruption in data transfer occurred because the workflow involved in the transfer had been unscheduled and the job was mistakenly shut down. The second interruption occurred October 6 through October 9 for unknown reasons. The interruptions affected the transfer of 10,847 (22.6%) records. IDEXX forwarded records that were created during these times of interruption once the data feed was re-established. The study relied upon transfer of data from IDEXX that was being queued in a test environment. The reliability of this environment was known to be less stable than a production environment. The interruptions experienced during this study would not be expected in a more stable production platform. 99 6.4.2 Descriptive Statistics: During the study period, IDEXX transferred records for 48,086 accessions. Specimens originated throughout Ohio and appeared to correlate with the population of each area (Figure 6.1). Accessions displayed an obvious and predictable day-of-week effect (Figure 6.2) with Sunday, Monday, and days following holidays representing days with the lowest volume. Species represented by the accessions included canine (70.1%), feline (25.6%), and equine (2.1%) (Figure 6.3). An important consideration for the designers of any syndromic surveillance system is the timely availability of data. Earlier detection being the overall goal, the system must receive records, with the appropriate information for analysis, within a period that provides for improved timeliness of detection compared to traditional reporting systems. Excluding the accessions that occurred during the interruption periods (n=10,847), on average, 95% of daily records were received with the next day’s dataset (Figure 6.4). Almost all (99.4%) records were received by the fourth 24-hour period. 6.4.3 Aberration Detection: The system identified nine alert periods during the study period using the CuSum detection method as previously described (Table 6.2). All of the alerts involved canines and/or felines. The number of accessions generating the alerts ranged from eight to 43. No cause could be determined for three of the nine (33.3%) alert periods and two (22.2%) were possibly related to breeding operations that existed in the area (e.g. screening of litters for pathogens). Two (22.2%) others were potentially the result of provider interest. One veterinary practice located in an area where a gastrointestinal alert occurred reported 100 being especially interested in educating clients about the risks from parasite ova. Another provider in an area where an endocrine alert occurred had recently been ordering an increased number of thyroid tests that were unrelated in increases in clinical disease. The remaining two (22.2%) alert periods were linked to verified disease activity in the pet population during the time of the alert. 6.4.4 Details about the Two Alerts Associated with Disease Activity: Alert #1: On September 11, 2006, the system generated an alert in Preble County located in western Ohio. Cases (20 cats and 2 dogs) were distributed equally between two ZIP codes. Follow-up with area veterinarians confirmed that many small animal practices were treating an increased number of animals that lived or spent a significant amount time out-of-doors for unspecified gastrointestinal distress. Following consultation with the Ohio Department of Natural Resources, veterinarians suspected that the cases may have resulted from corona virus infections acquired from rodents (Melissa Howell, Preble County Health Department, personal communication). An increased number of rodents were noted in the area, coinciding with the harvesting of local grain fields. Veterinarians speculated that pets may have captured and consumed some of the rodents, resulting in the self-limiting intestinal condition. Although health authorities received no reports of human cases, the Real-time Outbreak and Disease Surveillance System operated by the Ohio Department of Health indicated significant increases in both gastrointestinal-related chief complaints of emergency department patients and sales of anti-diarrheal medication in these areas during this time. 101 Alert #2: The pilot system generated a gastrointestinal alert for Lake County in northeastern Ohio on September 4, 2006. This alert included test orders for specimens originating from ten cats and three dogs submitted by clinics in two ZIP code areas. A local veterinarian from this county telephoned the State Public Health Veterinarian on September 26, 2006 to inquire about a number of clients that had brought their pets presenting with vomiting and diarrhea (Nancy Niehaus, Lake County Health Department, personal communication). These clients had shared with the local veterinarian that they also were experiencing diarrhea. The Lake County Health Department reported on October 4, 2006 that they were investigating “a cluster of diarrheal illness in humans and their associated pet dogs.” 6.5 Discussion: The primary purpose of this study was to explore the feasibility of using pre- diagnostic data from a VDL in a prospective manner to detect unexpected increases in the number of disease cases that might indicate an outbreak. We evaluated the feasibility by first determining the stability of electronic records and the success of automatically transferring them from the VDL for analysis, measured in terms of the percentage of complete records received in a timely manner. We then considered the representation of the records both by species of animal and geographic distribution. Finally, we investigated the alerts generated by the pilot system to validate if they might be associated with increases of disease. 102 While no single data source provides the capability to detect all outbreaks that may occur, veterinary providers may be desirable sources to include in surveillance activities for bettering our capabilities of detecting those outbreaks that result from emerging pathogens and potential bioweapon agents (3, 6, 17-20). The change in the number of laboratory orders submitted by veterinary providers may be a valuable proxy to measure the number of individual cases they are treating. An increase in the number of these individual cases may result from an outbreak, detection of which may be possible through the analysis of aggregated laboratory orders counts from several providers in the outbreak area. There are inherent biases to consider with using laboratory data. Laboratory testing in veterinary medicine is not as frequently used as in human medicine (11). Severity of clinical disease and cost benefit are two factors that influence the use of laboratory testing for animals (21). Severity of clinical disease as an influence on testing may provide for improved specificity since only animals with true disease/condition are included. As demonstrated in this study, the interests of the providers may also contribute to the potential biases encountered. The consistency of the veterinarians’ ordering behavior may help to control some bias by recognizing the effects in the counts over time and how they contribute to the normal baseline (i.e. expected number of test orders). The results of this study demonstrated the stability and timely availability of test order data for companion animals and how those data might be used in a prospective surveillance system to detect disease outbreaks. A significant number of daily records were received within the first 24-hour period following their creation. Using pre-existing data, generated by routine workflow, minimizes any additional burden for providers. 103 Employing an automated data transfer protocol further reduces burden and is an essential benefit to support a sustained surveillance effort (22). This system also achieved the benefit of obtaining provider-level data from a wide geographic area through a single source, creating no additional work for the veterinary providers and minimal work to establish and maintain the automated transfer mechanism for records from the VDL. The results from this study also indicated that increases in the number of test orders submitted by veterinarians might be detected in a timely manner using prospective analysis. The development of the syndrome categories and the detection method used most likely influenced the alerts generated by this pilot system. We described two alerts that linked unexpected increases in test orders to increased incidence of disease. One of these alerts may also have provided warning of human cases of disease. The number of true and verifiable outbreaks of disease that occur limits determining the performance of an outbreak detection system. Such a gold standard was lacking for this study. Therefore, we considered attempts to estimate sensitivity, specificity, or positive predictive value to be inappropriate. Additional investigation, following refinement of the syndrome categories, might be beneficial for better evaluating the potential of such a system to detect outbreaks of disease. The results support the continued consideration of VDL data by demonstrating the quality of data available, the ability to transfer and analyze the data in a timely manner, and the potential for detecting real disease events in the surveillance population. The true measure of a surveillance system lies in its usefulness (23). Additional benefits from this method of surveillance may exist that add intangible value to the system (22, 24, 25). Previous studies found that regular reports of conditions were beneficial when made 104 available to data providers (26, 27). While prospective analysis of orders includes methods designed to detect aberrant increases, reports of area syndromic trends may be valuable to veterinarians when treating individual animals as part of their practice. The addition of test results might also provide reports beneficial for veterinarians while potentially improving the specificity of outbreak detection. Input from all potential end users should be considered when further developing the utility of this type of surveillance system to ensure its maximum benefit. 105 Example Diseases Clinical Presentation Syndrome Category Glanders, Bordetella, Aspergillosis • Coughing Respiratory • Dyspnea • Nasal discharge Salmonellosis, Clostridia-associated • Diarrhea Gastrointestinal enterocolitis, Campylobacter • Vomiting • Colic Heartwater, plant poisoning, Botulism, • Convulsions Neurologic Tetanus • Paralysis • Staggering • Disturbed vision Poxvirus, allergies, Foot and Mouth • Abscesses Dermal Disease • Rash • Hair loss • Vesiculation Brucellosis, chronic Leptospirosis • Retained placenta Reproductive • Abortion • Orchitis Plague, Tularemia, Anemia, early • Lethargy Non-specific Leptospirosis • Malaise • Weakness • Fever without defining associated sign acute swine erysipelas, Anthrax, Red • Rapid onset of death Sudden Death Water Disease without apparent cause • Death occurring after brief display of illness Rabies, Listeriosis • Change in appetite Behavioral without defining associated signs • Unexplained aggression • Disorientation Table 6.1: Syndrome category descriptions distributed to veterinarian sample for grouping laboratory products 106 107 Figure 6.1: Distribution of specimen origin for accessions submitted from veterinary clinics in Ohio to IDEXX from September 1 through November 30, 2006 compared to ZIP code population. 1000 g y y 900 ivin g 800 Labor Da 700 Thanks Columbus Da 600 500 Count 400 108 300 200 100 0 6 6 6 6 6 6 0 0 0 0 0 006 0 0 0 /2 /20 /2 /2006 /2 /2 /2006 /1 /6/20 0 7 0 4 9 9/8/200 0 /2 /2 9/15 9/22/2006 9/29 1 0 11/3/2006 1 10/13/2006 1 10/2 11/1 11/17/2006 1 Date Figure 6.2: Counts of specimens received by IDEXX from veterinary clinics in Ohio from September 1 through November 30, 2006. Unknown, 1.4 Other, 0.8 Equine, 2.1 Feline, 25.6 Canine, 70.1 Figure 6.3: Representation of animal species in prospective accession datasets received from IDEXX September 1 through November 30, 2006. 109 100.0 80.0 60.0 110 40.0 % of daily total accessions total daily % of 20.0 0.0 012345671421 % 33.1 95.0 97.1 98.9 99.4 99.7 99.8 99.8 99.9 100.0 24-hour period Figure 6.4: Delay in receipt of daily records from IDEXX during prospective pilot. Date of first Number of alert County Syndrome category accessions Species (#) Findings 10/27/2006 Gallia Reproductive 9 Canine (9) Potentially related to breeding operation located in area 11/16/2006 Erie Gastrointestinal 11 Feline (7) No increase in disease noted by area veterinarians Canine (4) 10/11/2006 Preble Gastrointestinal 22 Feline (20) Increased gastrointestinal cases noted particularly in animals Canine (2) spending most of their time out-of-doors 11/15/2006 Preble Respiratory 43 Canine (43) Potentially related to breeding operation located in area 10/21/2006 Athens Endocrine 13 Canine (13) Possibly related to veterinarian ordering more T4 tests 111 recently 10/19/2006 Medina Gastrointestinal 8 Canine (8) No increase in disease noted by area veterinarians 9/1/2006 Tuscawarus Gastrointestinal 19 Feline (19) No increase in disease noted by area veterinarians 11/01/2006 Clermont Gastrointestinal 18 Canine (15) Possibly related to veterinarian with interest in educating Feline (3) clients about ova risks 9/4/2006 Lake Gastrointestinal 13 Canine (3) Increased gastrointestinal complaints noted in pets and their Feline (10) owners Table 6.2: Alerts generated by pilot prospective surveillance system using laboratory orders from select providers in Ohio from September 1 through November 30, 2006 and summary of follow up findings with area veterinary providers. REFERENCES 1. Dato V, Wagner MM, Fapohunda A. How Outbreaks of Infectious Disease are Detected: A Review of Surveillance Systems and Outbreaks. Public Health Rep 2004;119:464-71. 2. Backer L et al. Pet dogs as sentinels for environmental contamination. Sci Total Environ 2001;274:161-9. 3. Conti L. Petborne Zoonoses: Detection and Surveillance Challenges. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press. 4. Kearney B. Strengthening Safeguards Against Disease Outbreaks. In Focus 5[2]. 2005. Washington, D.C., The National Academy of Sciences. 5. Kelsey H. Improvements in methodologies for tracking infectious disease needed. The Newsbulletin . January 13, 2005. Los Alamos National Laboratory. 6. Green MS, Kaufman Z. Surveillance for Early Detection and Monioring of Infectious Disease Outbreaks Associated with Bioterrorism. Isr Med Assoc J 2002;4:503-6. 7. Moodie M et al. Biological Terrorism in the United States: Threat, Preparedness, and Response. November 2000. Chemical and Biological Arms Control Institute. 8. Federation of American Scientists. The Role of Disease Surveillance in the Watch for Agro-terrorism or Economic Sabotage. October 2001. 9. Doherr MG, Audige L. Monitoring and surveillance for rare health-related events: a review from the veterinary perspective. Philos Trans R Soc Lond B Biol Sci 2001;356:1097-106. 10. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board. 11. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10. 12. Conner CF. Review of efforts to protect the agricultural sector and food supply from a deliberate attack with a biological agent, a toxin or a disease directed at crops and livestock. July 20, 2005. Bio-security and Agro-terrorism. 2005. 112 13. Hutwagner LC et al. Using Laboratory-Based Surveillance Data for Prevention: An Algorithm for Detecting Salmonella Outbreaks. Emerg Infect Dis 1997;3:395-400. 14. Bradley CA et al. BioSense: Implementation of a National Early Event Detection and Situational Awareness System. MMWR Morb Mortal Wkly Rep 2005;54:11- 9. 15. Glickman LT et al. Purdue University-Banfield National Companion Animal Surveillance Program for Emerging and Zoonotic Diseases. Vector Borne Zoonotic Dis 2006;6:14-23. 16. Lucas JM. Counted Data CUSUM's. Technometrics 1985;27:129-44. 17. Shephard R, Aryel RM, Shaffer L. Animal Health. In: Wagner MM, Moore AW, Aryel RM, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:111-27. 18. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality. 19. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category A agents. J Am Vet Med Assoc 2004;224:1084-95. 20. National Research Council. Animal Health at the Crossroads: Preventing, Detecting, and Diagnosing Animal Diseases. July 2005. Washington, D.C., The National Academy of Sciences. 21. Power C. Passive Animal Disease Surveillance in Canada: A Benchmark. Proceedings of a CAHNet Workshop. November 1999. Canadian Food Inspection Agency. 22. Henning KJ. Syndromic Surveillance. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, D.C., National Academy Press. 23. Buehler JW. Surveillance. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins, 1998:435-57. 24. Begier EM et al. The National Capitol Region's Emergency Department Syndromic Surveillance System: Do Chief Complaint and Discharge Diagnosis Yield Different Results? Emerg Infect Dis 2003;9:393-6. 113 25. Buehler JW et al. Syndromic Surveillance and Bioterrorism-related Epidemics. Emerg Infect Dis 2003;9:1197-204. 26. Bartlett PC et al. Development of a computerized dairy herd health data base for epidemiologic research. Prev Vet Med 1986;4:3-14. 27. Mauer WA, Kaneene JB. Integrated Human-Animal Disease Surveillance. Emerg Infect Dis 2005;11:1490-1. 114 CHAPTER 7 SUMMARY OF DISSERTATION 7.1 Summary of Conclusions: A review of the literature reveals the concern that exists for the early detection of disease outbreaks, especially those that arise because of emerging infectious pathogens and those that may be the result of an act of bioterrorism. The hope is that earlier detection will lead to earlier intervention and reduce the morbidity, mortality, and economic loss. Surveillance of animal populations may be very helpful in detecting outbreaks early as the majority pathogens associated with emerging infectious diseases and potential bioweapons are zoonotic. Most existing disease surveillance systems used in animal populations are inadequate to detect outbreaks of emerging infectious diseases or potential bioweapons. These systems tend to be very disease specific, driven more by regulatory requirements than disease risk to humans, and not very timely, requiring laboratory verification of the pathogen. Syndromic surveillance is a rather new method that uses indicators of health to measure disease trends in a population. Specificity of these systems is dependent on the data source and how closely the information correlates to true health events. Potential benefits from this method of surveillance include the capability to detect signature 115 changes that occur earlier in the timeline of an outbreak and capturing data from a large number of providers over a wide geographic area. Veterinary data sources for use in syndromic surveillance systems have not been as well investigated compared to human sources. Orders submitted to medical laboratories are a source used by some human- based syndromic surveillance systems. We investigated the potential benefit of using orders submitted to veterinary diagnostic laboratories to detect outbreaks of infectious disease that could affect human populations as well as animals. We calculated an estimate of the number of Ohio households that might be reached by a surveillance system using data from a single VDL in Chapter 3. Using data regarding animal ownership from a national survey by the American Veterinary Medical Association, we determined that between 4.8% and 8.8% of all Ohio households could be included using just canine laboratory orders from this single source. This represents almost a quarter of the dog-owning households in Ohio. In Chapter 4, we discuss the retrospective study of a sample of microbiology test orders and results from two VDLs with clients in Ohio. Analysis of these datasets showed that companion animals were well represented by these potential sources. Comparing the date each laboratory received the specimens to the date of results indicated that, on average, detection of outbreaks might be made three to five days earlier using test orders for surveillance. We determined the expected baseline for specific pathogens, used this baseline to identify significant increases, then demonstrated how select detection methods might be used to identify this clusters from the aggregate of orders. 116 The retrospective study revealed an unexpected confounder in the data set of one of these laboratories. We explored this confounder in more detail in Chapter 5, associating it with the rotation schedule of senior veterinary students through various clinical experiences. Finally, in Chapter 6, we prospectively studied the order data from one VDL by establishing a pilot surveillance system. This system received records of laboratory orders, classified them into syndrome categories, and analyzed the counts of each category to detect unexpected increases. Results from this pilot system study demonstrated that the automated transfer was feasible and that records were available in a timely manner for analysis. They also demonstrated the ability to automatically classify the orders and analyze them for aberrations. Descriptions of select alerts generated by the system indicate that the detection of true disease events in the population is possible including those events that include humans. 7.2 Future Studies: We recognized the need for improved outbreak detection efforts and discussed how worthwhile it is to include animal populations. The representation of human households by pet animals was estimated and the representation of these animals by the proposed data source was established. Further study of this data source demonstrated the ability to establish historic baselines of disease and indicators that might be useful for surveillance. The quality of data was further established by demonstrating the availability of records within a reasonably short period. Results of investigations have also indicated that disease events in the population are detectable in these data and may provide earlier 117 indications of an outbreak compared to traditional reporting methods. However, further investigation might better determine the true value of these data and further develop the analytical methods to analyze them and the utility platform to display those results. The next step in achieving these goals is to refine the syndrome categories and expand the geographic coverage and/or increase the study period of the pilot system used to evaluate the availability of the data. Detection of aberrations is limited by the syndrome categories and detection methods utilized by the system. The definition of syndrome categories will define the specificity of the system. In spite of our initial efforts, many of the test products used during the pilot study were non-specific. Veterinary providers used them to diagnose more than one category of disease. Many could probably be considered as screening tests; tests used to gain more knowledge about the general condition of a patient. Refinement of the syndrome categories through reclassification of these non-specific and screening tests could help to improve the detection capabilities of the system. A shortcoming in the ability to evaluate the performance of detection in this and similarly based systems is the infrequency of outbreaks. Many measures of performance are calculated through proportions based on the number of true outbreaks. Without a sufficient size denominator (number of true outbreaks), the power of these calculations is diminished. During our pilot study, lasting 90 days, two outbreaks of gastrointestinal illness in pet animals were indicated. The lack of any reporting registry for this condition in animals limits our determination to the two events that the system alerted to that could be verified through provider interviews. A power calculation for a one-proportion study indicates that eight such events are needed to achieve a power of 0.85 (Figure 7.1). Nine 118 outbreaks would provide a power of 0.90. If we err on the side of conservancy and assume that one outbreak occurs in this size of pet population every 90 days the next pilot should operate between 720 and 810 days to achieve sufficient power of analysis. An increase in the geographic area to expand the number of individuals included in a 90-day pilot by a factor of eight to nine should achieve similar benefit. As investigation continues, it will be important to determine the characteristics of diseases that are most important for surveillance by the end users of the system. Since there is no one-size-fits-all detection method it will be difficult to include the proper one without some consideration of the signature pattern that analysis is expected to detect. This consideration should include the incubation period of diseases, susceptible population, potential rate of infectivity, and cost versus benefit associated with detection. This point actually alludes to a broader requirement for successful system development; identification of the end users and consideration of their needs and desires. The ultimate measure of usefulness and value of any system will be reflected in acceptance by the end user. Input regarding information and analysis important to them should be an integral procedure in system development. System developers should anticipate many levels of end user and items of importance may differ with each. Information considered critical to population health users may be of little or no use to providers treating individuals. 119 The use of syndromic surveillance for outbreak detection in pet animals is but one step toward a one-medicine approach to disease surveillance. This process will need to continue with an air of awareness to the needs of each discipline and a spirit of cooperation if we are to achieve a true integrated approach to protecting humans and animals together from infectious disease. 120 1 0.9 0.85 0.8 0.7 0.6 0.5 Power 0.4 0.3 0.2 0.1 0 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Number of Outbreaks Included Figure 7.1: Determining the number of outbreaks required to occur during pilot surveillance system to achieve the desired level of power in evaluation measures. 121 BIBLIOGRAPHY Ackelsberg J et al. Syndromic Surveillance for Bioterrorism Following the Attacks on the World Trade Center -- New York City, 2001. MMWR Morb Mortal Wkly Rep 2002;51. Altekruse SF, Cohen ML, Swerdlow DL. Emerging Foodborne Diseases. Emerg Infect Dis 1997;3:285-93. Amercian Veterinary Medical Association. U.S. Pet Ownership and Demographics Sourcebook. 2002. Schaumburg, IL, American Veterinary Medical Association. Ashford DA et al. Biological terrorism and veterinary medicine in the United States. J Am Vet Med Assoc 2000;217:664-7. Ashford DA et al. Planning against Biological Terorism: Lessons from Outbreak Investigations. Emerg Infect Dis 2003;9:515-9. Babin SM et al. Early detection of possible bioterrorist events using sentinel animals. The 131st Annual Meeting of the American Public Health Association. 2003. Baca R, Cassidy K. NAHLN: A Standards-based Approach for Laboratory Diagnostic Data. NAHSS Outlook . October 2004. Fort Collins, CO, United States Department of Agriculture. Backer L et al. Pet dogs as sentinels for environmental contamination. Sci Total Environ 2001;274:161-9. Bartlett PC et al. Development of a computerized dairy herd health data base for epidemiologic research. Prev Vet Med 1986;4:3-14. Barton LL, Villar RG, Connick M. Pet-Associated Zoonoses. Emerg Infect Dis 1999;5. Begier EM et al. The National Capitol Region's Emergency Department Syndromic Surveillance System: Do Chief Complaint and Discharge Diagnosis Yield Different Results? Emerg Infect Dis 2003;9:393-6. Bender JB, Shulman SA. Reports of zoonotic disease outbreaks associated with animal exhibits and availability of recommendations for preventing zoonotic disease transmission from animals to people in such settings. J Am Vet Med Assoc 2004;224:1105-9. 122 Besag J, Newell J. The Detection of Clusters in Rare Diseases. J R Stat Soc A 1991;154:143-55. Birmingham J. Zoonotic concerns put veterinarians on the front lines: Leaders urge heightened vigilance. Veterinary Forum 23[7], 25-33. 2006. Bradley CA et al. BioSense: Implementation of a National Early Event Detection and Situational Awareness System. MMWR Morb Mortal Wkly Rep 2005;54:11-9. Buckeridge DL et al. Algorithms for rapid outbreak detection: a research synthesis. J Biomed Inform 2005;38:99-113. Buehler JW et al. Framework for Evaluating Public Health Surveillance Systems for Early Detection of Outbreaks. MMWR Recomm Rep 2004;53. Buehler JW. Surveillance. In: Rothman KJ, Greenland S, eds. Modern Epidemiology. Philadelphia, PA: Lippincott Williams & Wilkins, 1998:435-57. Buehler JW et al. Syndromic Surveillance and Bioterrorism-related Epidemics. Emerg Infect Dis 2003;9:1197-204. Burkom H. Manual for Accessible Alerting Algorithms. September 23, 2005. The Johns Hopkins University Applied Physics Laboratory. Center for Infectious Disease Research and Policy. Agencies launch warning system for animal diseases [Web Page]. 2006 Jul 25; Accessed 2006 Jul 27. Available at: www.cidrap.umn.edu/cidrap/content/influenza/avianflu/news/jul2506warning.html. Centers for Disease Control and Prevention. Addressing emerging infectious disease threats, a prevention strategy for the United States. 1994. Centers for Disease Control and Prevention. Centers for Disease Control and Prevention. Epidemic/Epizootic West Nile Virus in the United States: Guidelines for Surveillance, Prevention, and Control. 2003. Ft. Collins, CO. Centers for Disease Control and Prevention. Syndrome Definitions for Diseases Associated with Critical Bioterrorism-associated Agents. October 23, 2003. Centers for Disease Control and Prevention. Updated Guidelines for Evaluating Public Health Surveillance Systems: Recommendations from the Guidelines Working Group. MMWR Recomm Rep 2001;50. Chapman WW et al. Classifying free-text triage chief complaints into syndromic categories with natural language processing. Artif Intell Med 2005;33:31-40. 123 Cherry B et al. Salmonella Typhimurium Outbreak Associated with Veterinary Clinic. Emerg Infect Dis 2004;10:2249-51. Childs J et al. Emerging Zoonoses. Emerg Infect Dis 1998;4:453-4. Chin N. Dr. James Buehler On the Promise of Syndromic Surveillance. Homeland First Response [May/June]. 2004. Jems Communications. Chretien JH, Garagusi VF. Infections associated with pets. Am Fam Physician 1990;41:831(15). Christensen J. Application of Surveillance and Monitoring Systems in Disease Control Programs. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:15-34. Cleaveland S, Laurenson MK, Taylor LH. Diseases of humans and their domestic mammals: pathogen characteristics, host range and the risk of emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:991-9. Committee on Science and Technology for Countering Terrorism. Making the Nation Safer: The Role of Science and Technology in Countering Terrorism. 2002. Washington, D.C., National Research Council. Conner CF. Review of efforts to protect the agricultural sector and food supply from a deliberate attack with a biological agent, a toxin or a disease directed at crops and livestock. July 20, 2005. Bio-security and Agro-terrorism. 2005. Conti L. Petborne Zoonoses: Detection and Surveillance Challenges. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press. Council of State and Territorial Epidemiologists. National Assessment of Epidemiologic Capacity in Public Health: Findings and Recommendations. March 2003. Crawford PC et al. Transmission of Equine Influenza Virus to Dogs. 10.1126/science.1117950. September 29, 2005. Available at: www.sciencexpress.org. Crom RL. Veterinary Surveillance for Zoonotic Diseases in the United States. Burroughs, Tom, Knobler, Stacey, and Lederberg Joshua . The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washington, DC, National Academy Press. Dafni UG et al. Algorithm for Statistical Detection of Peaks--Syndromic Surveillance System for the Athens 2004 Olympic Games. MMWR Morb Mortal Wkly Rep 2004;53:86-94. 124 Das D et al. Monitoring Over-The-Counter Medication Sales Early Detection of Disease Outbreaks--New York City. MMWR Morb Mortal Wkly Rep 2005;54:41-6. Das D et al. Enhanced Drop-in Syndromic Surveillance in New York City Following September 11, 2001. J Urban Health 2003;80:i76-i88. Dato V, Wagner MM, Fapohunda A. How Outbreaks of Infectious Disease are Detected: A Review of Surveillance Systems and Outbreaks. Public Health Rep 2004;119:464-71. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category A agents. J Am Vet Med Assoc 2004;224:1084-95. Davis RG. The ABCs of bioterrorism for veterinarians, focusing on Category B and C agents. J Am Vet Med Assoc 2004;224:1096-104. DeGroot B. The Rapid Syndrome Validation Project for Animals - Augmenting Contact with the Network of Accredited Veterinarians. NAHSS Outlook [April]. 2005. United States Department of Agriculture. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook. Doherr MG, Audige L. Monitoring and surveillance for rare health-related events: a review from the veterinary perspective. Philos Trans R Soc Lond B Biol Sci 2001;356:1097-106. Duchin JS. Epidemiological Response to Syndromic Surveillance Signals. J Urban Health 2003;80:i115-i116. Dufour B, LaVielle S. Epidemiological surveillance of infectious diseases in France. Vet Res 2000;31:169-85. Eidson M et al. Dead Bird Surveillance as an Early Warning System for West Nile Virus. Emerg Infect Dis 2001;7:631-5. Eidson M et al. Dead Crow Densities and Human Cases of West Nile Virus, New York State, 2000. Emerg Infect Dis 2001;7:662-4. Eitzen EM, Takafuji ET. Historical Overview of Biological Warfare. In: Sidell FR, Takafuji ET, Franz DR, eds. Medical Aspects of Chemical and Biological Warfare. Bethesda, MD: Office of The Surgeon General, Department of the Army, United States of America, 1997:415-23. Engle MJ. The Value of an "Early Warning" Surveillance System for Emerging Diseases. The Value of an "Early Warning" Surveillance System for Emerging Diseases. 2000. National Pork Board. Available at: www.porkboard.org/docs/mengle.pdf. Engle M. Improving Emerging Disease Surveillance in the Pork Industry. Swine Health Report [Fall], 4-5. 2004. Bowling Green, KY, National Institute for Animal Agriculture. 125 Ezell H et al. Outbreaks of Multidrug-Resistant Salmonella Typhimurium Associated With Veterinary Facilities -- Idaho, Minnisota, and Washington, 1999. MMWR Morb Mortal Wkly Rep 2001;50. Fairchild A, Bayer R. Ethics and the Conduct of Public Health Surveillance. Science 2004;303:631-2. Farrington P, Andrews N. Outbreak Detection: Application to Infectious Disease Surveillance. In: Brookmeyer R, Stroup DF, eds. Monitoring the Health of Populations: Statistical Principles and Methods for Public Health Surveillance. New York, NY: Oxford Univeristy Press, 2004:203-31. Fayer R, Morgan U, Upton SJ. Epidemiology of Cryptosporidium: transmission, detection and identification. Int J Parasitol 2000;30:1305-22. Federation of American Scientists. The Role of Disease Surveillance in the Watch for Agro-terrorism or Economic Sabotage. October 2001. Fiala J. Veterinary roles must expand to protect public health, reports say. DVM: The Newsmagazine of Veterinary Medicine [Sept], 21-24. 2005. Fienberg SE, Shmueli G. Statistical issues and challenges associated with rapid detection of bio-terrorist attacks. Stat Med 2005;24:513-29. Focks D. Mathematical Models and Predictors of Disease Outbreaks. Burroughs, T., Knobler, S., and Lederberg, J. The Emergence of Zoonotic Diseases: Understanding the Impact on Animal and Human Health. 2002. Washnigton, DC, National Academy Press. Food and Agriculture Organization of the United Nations. Manual on Livestock Disease Surveillance and Information Systems. Rome: 1999. Fouchier RAM, Osterhaus ADME, Brown IH. Animal influenza virus surveillance. Vaccine 2003;21:1754-7. Friedman CP, Wyatt JC. Evaluation Methods in Medical Informatics. New York, NY: Springer-Verlag, 1997. Fèvre EM et al. Animal movements and the spread of infectious diseases. Trends Microbiol 2006;14:125-31. Gill JS et al. Serologic Surveillance for the Lyme Disease Spirochete, Borrelia burgdorferi, in Minnesota by Using White-Tailed Deer as Sentinel Animals. J Clin Microbiol 1994;32:444-51. Glickman LT et al. Purdue University-Banfield National Companion Animal Surveillance Program for Emerging and Zoonotic Diseases. Vector Borne Zoonotic Dis 2006;6:14-23. 126 Gobar GM, Case JT, Kass PH. Program for surveillance of causes of death of dogs, using the internet to survey small animal veterinarians. J Am Vet Med Assoc 1998;213:251-6. Green MS, Kaufman Z. Surveillance for Early Detection and Monioring of Infectious Disease Outbreaks Associated with Bioterrorism. Isr Med Assoc J 2002;4:503-6. Greenland S, Morgenstern H. Confounding in Health Research. Annu Rev Public Health 2001;22:189-212. Greiner M, Dekker A. On the surveillance for animal disease in small herds. Prev Vet Med 2005;70:223-34. Grimes DA, Schulz KF. Bias and causal associations in observational research. Lancet 2002;359:248-52. Guardabassi L, Schwarz S, Lloyd DH. Pet animals as reservoirs of antimicrobial-resistant bacteria. J Antimicrob Chemother 2004;54:321-32. Guarnera EA et al. Canine echinococcosis: an alternative for surveillance epidemiology . Vet Parasitol 2000;88:131-4. Guptill S et al. Early-Season Avian Deaths from West Nile Virus as Warnings of Human Infection. Emerg Infect Dis 2003;9:483-4. Hartl, Gregory and Meslin, François. Launch of global early warning system for animal diseases transmissable to humans [Web Page]. 2006 Jul 24; Accessed 2006 Jul 27. Available at: www.who.int/mediacentre/news/new/2006/nw02/en/print.html. Hashimoto S et al. Detection of epidemics in their early stage through infectious disease surveillance. Intl J Epidemiol 2000;29:905-10. Haydon DT et al. Identifying Reservoirs of Infection: A Conceptual and Practical Challenge. Emerg Infect Dis 2002;8:1468-73. Heffernan R et al. Syndromic Surveillance in Public Health Practice, New York City. Emerg Infect Dis 2004;10:858-64. Heinrich J. Infectious Diseases: Gaps Remain in Surveillance Capabilities of State and Local Agencies. [GAO-03-1176T]. September 24, 2003. Select Committee on Homeland Security, Subcommittee on Emergency Preparedness and Response. 2003. Heinrich J. SARS Outbreak: Improvements to Public Health Capacity Are Needed for Responding to Bioterrorism and Emerging Infectious Diseases. [GAO-03-769T]. May 7, 2003 . Committee on Energy and Commerce, House of Representatives, Subcommittee on Oversight and Investigations. 2003. 127 Henderson DA. Bioterrorism as a Public Health Threat. Emerg Infect Dis 1998;4:488-92. Henning KJ. Syndromic Surveillance. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, D.C., National Academy Press. Henning KJ. What is Syndromic Surveillance? MMWR Morb Mortal Wkly Rep 2004;53:7-11. Hill SL et al. Prevalence of enteric zoonotic organisms in cats. J Am Vet Med Assoc 2000;216:687-92. Hlavsa MC, Watson JC, Beach MJ. Cryptosporisiosis Surveillance--United States, 1999- 2002. MMWR Surveill Summ 2005;54. Hlavsa MC, Watson JC, Beach MJ. Giardiasis Surveillance--United States, 1998-2002. MMWR Surveill Summ 2005;54. Huerta M, Leventhal A. The epidemiologic pyramid of bioterrorism. Isr Med Assoc J 2002;4:498-502. Hung J et al. Electronic Surveillance of Disease States: A preliminary study in electronic detection of respiratory diseases in a primary setting. Chute, Christopher. Proc AMIA Ann Symp. 1998. Hutwagner LC et al. Using Laboratory-Based Surveillance Data for Prevention: An Algorithm for Detecting Salmonella Outbreaks . Emerg Infect Dis 1997;3:395-400. Hutwagner L et al. Comparing Aberration Detection Methods with Simulated Data. Emerg Infect Dis 2005;11:314-6. Hutwagner L et al. The Bioterrorism Preparedness and Response Early Aberration Reporting System (EARS). J Urban Health 2003;80:i89- i96. Inglesby TV, Grossman R, O'Toole T. A Plague on Your City: Observations from TOPOFF. Clin Infect Dis 2001;32. Institute of Medicine. Smolinski, Mark S., Hamburg, Margaret A., and Lederberg, Joshua. Microbial Threats to Health: Emergence, Detection, and Response. 2003. Washington, DC, National Academy Press. Janes GR et al. Descriptive Epidemiology: Analyzing and Interpreting Surveillance Data. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford Univeristy Press, 2000:112-67. 128 Johnson C. The Role of the Center for Emerging Issues in Coordinating International Safeguarding Information via the Offshore Pest Information System. NAHSS Outlook [Aug]. 2005. Fort Collins, CO, United States Department of Agriculture. Johnson CL. Electronic Surveillance for Emerging Diseases. National Veterinary Electronic Surveillance Meeting. 2004. Johnson HA, Wagner MM, Saladino RA. A New Method for Investigating Non- traditional Biosurveillance Data: Studying Behaviors Prior to Emergency Department Visits. 2005. Available at: www.rods.pitt.edu. Johnson JL et al. Passive Tick Surveillance, Dog Seropositivity, and Incidence of Human Lyme Disease. Vector BorneZoonotic Dis 2004;4:137-42. Kanof M. Emerging Infectious Diseases: Review of State and Federal Disease Surveillance Efforts. [GAO-04-877]. September 30, 2004. Committee on Governmental Affairs, Permanent Subcommittee on Investigations. 2004. Available at: www.gao.gov/cgi-bin/getrpt?GAO-04-877. Karesh WB et al. Wildlife Trade and Global Disease Emergence. Emerg Infect Dis 2005;11:1000-2. Karlen A. Man and Microbes: Disease and Plagues in History and Modern Times. New York, NY: Simon & Schuster Paperbacks, 1995. Kassenborg H et al. Human Ingestion of Bacillus Anthracis-Contaminated Meat -- Minnesota, August 2000. MMWR Morb Mortal Wkly Rep 2000;49. Kaufmann AF, Meltzer MI, Schmid GP. The Economic Impact of a Bioterrorist Attack: Are Prevention and Postattack Intervention Programs Justifiable? Emerg Infect Dis 1997;3:83-94. Kaye D, Pringle CR. Avian Influenza Viruses and their Implication for Human Health. Clin Infect Dis 2005;40:108-12. Kearney B. Strengthening Safeguards Against Disease Outbreaks. In Focus 5[2]. 2005. Washington, D.C., The National Academy of Sciences. Available at: http://infocusmagazine.org/5.2/env_animal_health.html. Kelly TK et al. The Office of Science and Technology Policy Blue Ribbon Panel on The Threat of Biological Terrorism Directed Against Livestock. March 2004. Arlington, VA, RAND Science and Technology. Kelsey H. Improvements in methodologies for tracking infectious disease needed. The Newsbulletin . January 13, 2005. Los Alamos National Laboratory. Available at: www.lanl.gov/orgs/pa/newsbulletin/2005/01/13/includes/text02.inc. 129 King L, Khabbaz R. Converging Issues in Veterinary and Public Health. Emerg Infect Dis 2003;9:510-1. Kleinman KP et al. A model-adjusted space-time scan statistic with an application to syndromic surveillance. Epidemiol Infect 2005;133:409-19. Kleinman K, Lazarus R, Platt R. A Generalized Linear Mixed Models Approach for Detecting Incident Clusters of Disease in Small Areas, with an Application to Biological Terrorism. Am J Epidemiol 2004;159:217-24. Koo D, Parrish II RG. The Changing Health-Care Information Infrastructure in the United States: Opportunities for a New Approach to Public Health Surveillance. In: Teutsch SM, Churchill RE, eds. Principles and Practice of Public Health Surveillance. New York: Oxford University Press, 2000:76-94. Kortepeter M. USAMRIID's Medical Management of Biological Casualties Handbook. Fort Detrick, Frederick, MD: U.S. Army Medical Research Institute of Infectious Diseases, 2001. Kruse H, Kirkemo A-M, Handeland K. Wildlife as Source of Zoonotic Infections. Emerg Infect Dis 2004;10:2067-72. Lazarus R et al. Use of Automated Ambulatory-Care Encounter Records for Detection of Acute Illness, Including Potential Bioterrorism Events. Emerg Infect Dis 2002;8:753-60. Lazarus R et al. Using automated medical records for rapid identification of illness syndromes (syndromic surveillance): the example of lower respiratory infection. Public Health 2001;1. Lober WB et al. Roundtable on Bioterrorism Detection: Infomration System-based Surveillance. J Am Med Inform Assoc 2002;9:105-15. Lucas JM. Counted Data CUSUM's. Technometrics 1985;27:129-44. Lurie N, Wasserman J, Nelson CD. Public Health Preparedness: Evolution Or Revolution? Health Aff 2006; 25:935-45. Lynn T. Coordinating Zoonotic Disease Surveillance: Partnering Agriculture and Public Health. NAHSS Outlook [June]. 2005 . United States Department of Agriculture . Available at: www.aphis.usda.gov/vs/ceah/ncahs/nsu/outlook. Ma H et al. Implementation of laboratory order data in BioSense Early Event Detection and Situation Awareness System. MMWR Morb Mortal Wkly rep 2005;54:27-30. 130 Magruder SF. Evaluation of Over-the-Counter Pharmaceutical Sales A a Possible Early Warning Indicator of Human Disease. Johns Hopkins APL Technical Digest 24[4], 349- 353. 2003. Mandl KD et al. Implementing Syndromic Surveillance: A Practical Guide Informed by the Early Experience. J Am Med Inform Assoc 2004;11:141-50. Mark L. Strengthening Veterinary Diagnostic Lab Network is Critical. February 24, 2006. Richmond, VA, United States Animal Health Association. Mauer WA, Kaneene JB. Integrated Human-Animal Disease Surveillance. Emerg Infect Dis 2005;11:1490-1. Meeks, Brock N. Animal docs fill homeland security slots. MSNBC. Washington, D.C.; 2005 Mar 18. Meslin FX, Stohr K, Heymann D. Public health implications of emerging zoonoses. Rev Sci Tech 2000;19:310-7. Mitchell J. State Strategies for Fully Integrating Public Health into Homeland Security. November 23, 2005. Washington, D.C., NGA Center for Best Practices. Moodie M et al. Biological Terrorism in the United States: Threat, Preparedness, and Response. [200*1999*00132]. November 2000. Chemical and Biological Arms Control Institute. Morris CD et al. Comparison of chickens and pheasants as sentinels for eastern equine encephalitis and St. Louis encephalitis viruses in Florida. J Am Mosq Control Assoc 1994;10:545-8. Murrell KD, Pozio E. Trichinellosis: the zoonosis that won't go quietly. Int J Parasitol 2000;30:1339-49. National Association of State Public Health Veterinarians I. Compendium of Measures To Prevent Disease Associated with Animals in Public Settings, 2005. MMWR Recomm Rep 2005;54. National Center for Animal Health Surveillance. National Animal Health Reporting System Operational Manual and Uniform Methods and Rules. March 2006. Fort Collins, CO, United States Department of Agriculture. National Center for Infectious Diseases. Recognition of Illness Associated with the Intentional Release of a Biologic Agent. MMWR Morb Mortal Wkly Rep 2001;50. National Institute for Animal Agriculture. Emerging Diseases Resolution and Position Statement. 2005. 131 National Research Council. Animal Health at the Crossroads: Preventing, Detecting, and Diagnosing Animal Diseases. July 2005. Washington, D.C., The National Academy of Sciences. National Research Council. Making the Nation Safer: The role of science and technology in countering terrorism. 2002. Washington, DC, National Academy Press. Nichols GL. Fly Transmission of Campylobacter. Emerg Infect Dis 2005;11:361-4. Nowotny N et al. Prevalence of Swine Influenza and Other Viral, Bacterial, and Parasitic Zoonoses in Veterinarians. J Infect Dis 1997;176:1414-5. O'Brien SJ, Christie P. Do CuSums have a role in routine communicable disease surveillance. Public Health 1997;111:255-8. Olsen CW et al. Serologic Evidence of H1 Swine Influenza Virus Infection in Swine Farm Residents and Employees. Emerg Infect Dis 2002;8:814-9. Olszewski RT. Bayesian Classification of Triage Diagnoses for the Early Detection of Epidemics. 2003. American Association for Artificial Intelligence. Pavlin JA. Epidemiology of Bioterrorism. Emerg Infect Dis 1999;5:528-30. Pavlin JA. Investigation of Disease Outbreaks Detected by "Syndromic" Surveillance Systems. J Urban Health 2003;80:i107-i114. Pavlin JA. Medical Surveillance for Biological Terrorism Agents. Hum Ecol Risk Assess 2005;11:525-37. Pavlin JA et al. Innovative Surveillance Methods for Rapid Detection of Disease Outbreaks and Bioterrorism: Results of an Interagency Workshop on Health Indicator Surveillance. Am J Pub Health 2003;93:1230-5. Pfeiffer DU. Veterinary Epidemiology - An Introduction. London: The Royal Veterinary College, Univeristy of London, 2002. Platt R. Homeland Security: Disease Surveillance Systems. September 24, 2003. Select Committee on Homeland Security, Subcommittee on Emergency Preparedness and Response. 2003. Platt R et al. Syndromic Surveillance Using Minimum Transfer of Identifiable Data: the Example of the National Bioterrorism Syndromic Surveillance Demonstration Program. J Urban Health 2003;80:i25-i31. 132 Power C. Passive Animal Disease Surveillance in Canada: A Benchmark. Proceedings of a CAHNet Workshop. November 1999. Canadian Food Inspection Agency. Reingold A. If Syndromic Surveillance Is the Answer, What Is the Question? Biosecur Bioterror 2003;1:77-81. Reis BY, Mandl KD. Time series modeling for syndromic surveillance. Med Inform Decis Making 2003;3. Riley PY, Chomel BB. Hedgehog Zoonoses. Emerg Infect Dis 2005;11:1-5. Rizzo SL et al. Using Case Studies to Evaluate Syndromic Surveillance. 2005. The RODS Laboratory, Center for Biomedical Informatics, University of Pittsburgh. Robertson ID et al. The role of companion animals in the emergence of parasitic zoonoses. Intl J Parasitol 2000;30:1369-77. Robotham J, Green LE. Pilot study to investigate the feasibility of surveillance of small animals in the UK. J Small Anim Pract 2004;45:213-8. Rogerson PA, Yamada I. Approaches to Syndromic Surveillance When Data Consist of Small Regional Counts. MMWR Morb Mortal Wkly Rep 2004;53:79-85. Rosen T, Jablon J. Infectious threats from exotic pets: dermatological implications. Dermatol Clin 2003;21:229-36. Rossi G, Lampugnani L, Marchi M. An Approximate CuSum Procedure for Surveillance of Health Events. Stat Med 1999;18:2111-22. Saini PK, Ransom G, McNamara AM. Emerging public health concerns regarding cryptosporidiosis. J Am Vet Med Assoc 2000;217:658-63. Salman MD. Surveillance and Monitoring Systems for Animal Health Programs and Disease Surveys. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:3-13. Salman MD, Stark KDC, Zepeda C. Quality assurance applied to animal disease surveillance systems. Rev Sci Tech 2003;22:689-96. Schwabe CW. Veterinary Medicine and Human Health. Baltimore, MD: Williams and Wilkins, 1984. Serfling RE. Methods for Current Statistical Analysis of Excess Pneumonia-Influenza Deaths. Public Health Rep 1963;78:494-506. 133 Shaffer LE et al. Evaluation of Microbiology Orders from two Veterinary Diagnostic Laboratories as Potential Data Sources for Early Outbreak Detection. Adv Disease Surveil. forthcoming. Shapiro AR. Taming Variability in Free Text: Application to Health Surveillance. MMWR Morb Mortal Wkly Rep 2004;53:95-100. Shears P. Communicable disease surveillance with limited resources: the scope to link human and veterinary programmes. Acta Trop 2000;76:3-7. Shephard R, Aryel RM, Shaffer L. Animal Health. In: Wagner MM, Moore AW, Aryel RM, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:111-27. Siegrist D et al. Evaluation of Algorithms for Outbreak Detection Using Clinical Data from Five U.S. Cities. October 15, 2004. Defense Advanced Research Projects Agency. Skerget M et al. Cat or Dog Ownership and Seroprevalence of Ehrlichiosis, Q Fever, and Cat-Scratch Disease. Emerg Infect Dis 2003;9:1337-40. Smith K et al. Outbreak of Multidrug-Resistant Salmonella Typhimurium associated with Rodents Purchased at Retail Pet Stores -- United States, December 2003-October 2004. MMWR Morb Mortal Wkly Rep 2005;54. Smith KE et al. Outbreaks of Enteric Infections Caused by Mulitple Pathogens Associated With Calves at a Farm Day Camp. Pediatr Infect Dis J 2004;23:1098-104. Sonesson C, Bock D. A review and discussion of prospective statistical surveillance in public health. J R Stat Soc A 2003;166:5-21. Sosin DM. Draft Framework for Evaluating Syndromic Surveillance Systems. J Urban Health 2003;80:i8-i13. Sosin DM. Syndromic Surveillance: The Case for Skillful Investment View. Biosecur Bioterr 2003;1:247-53. Spain CV et al. Prevalence of Enteric Zoonotic Agents in Cats less than 1 Year Old in Central New York State. J Vet Intern Med 2001;15:33-8. Stoto MA, Schonlau M, Mariano LT. Syndromic Surveillance: Is It Worth the Effort? Chance 2004;17:19-24. Stärk KDC et al. Concepts for risk-based surveillance in the field of veterinary medicine and veterinary public health: Review of current approaches. Health Serv Res 2006;6. 134 Tansey B. Debate over pets' role in spread of avian flu: No cases passed yet by dogs or cats, but experts want study. June 12, 2006. San Francisco, CA. Tauni MA, Österlund A. Outbreak of Salmonella typhimurium in cats and humans associated with infection in wild birds. J Small Anim Pract 2000;41:339-41. Taylor LH, Latham SM, Woolhouse MEJ. Risk factors for human disease emergence. Philos Trans R Soc Lond B Biol Sci 2001;356:983-9. Tenter AM, Heckeroth AR, Weiss LM. Toxoplasma gondii: from animals to humans. Int J Parasitol 2000;30:1217-58. Thompson RCA. Giardiasis as a re-emerging infectious disease and its zoonotic potential. Int J Parasitol 2000;30:1259-67. Tissot-Dupont H et al. Wind in November, Q fever in December. Emerg Infect Dis 2004;10:1264-9. Torok TJ et al. A Large Community Outbreak of Salmonellosis Caused by Intentional Contamination of Restaurant Salad Bars. J Am Med Assoc 1997;278:389-95. Trevors JT. Emerging Infectious Diseases: The Link Between Environment and Human Health. Water Air Soil Poll 2004;157:vii-viii. Tsao JI et al. An ecological approach to preventing human infection: Vaccinating wild mouse reservoirs intervenes in the Lyme disease cycle. Proc Natl Acad Sci U S A 2004;101:18159-64. Tsui F-C et al. Technical Description of RODS: A Real-time Public Health Surveillance System. J Am Med Inform Assoc 2003;10:399-408. Tsui F-C et al. Data, Network, and Application: Technical Description of the Utah RODS Winter Olympic Biosurveillance System. Proc AMIA Ann Symp. 2002. Tsui F-C et al. Value of ICD-9-Coded Chief Complaints for Detection of Epidemics. Proc AMIA Ann Symp. 711-715. 2001. U.S. Census Bureau. American Factfinder [Web Page]. Accessed 2006 Aug 29. Available at: http://factfinder.census.gov. U.S. Department of the Interior. An Early Detection System for Highly Pathogenic H5N1 Avian Influenza in Wild Migratory Birds U.S. Interagency Strategic Plan [Web Page]. 2006 Mar 14; Accessed 2006 Aug 24. Available at: www.doi.gov/issues/birdflu_strategicplan.pdf. 135 United States Department of Agriculture Animal and Plant Health Inspection Service. National Animal Health Reporting System [Web Page]. Accessed 2006 Jun 22. Available at: www.aphis.usda.gov/vs/ceah/ncahs/nahrs/. United States Department of Agriculture Animal and Plant Health Inspection Service. U.S. Animal Health and Productivity Surveillance Inventory [Web Page]. Accessed 2006 Jun 22. Available at: http://nsu.aphis.usda.gov/inventory/faces/index.jsp. United States Department of Agriculture National Veterinary Services Laboratories. The National Animal Health Laboratory Network: A State and Federal Partnership to Safeguard Animal Health [Web Page]. Accessed 2006 Apr 22. Available at: http://nahln.vs/. Van Immerseel F et al. Cats as a Risk for Transmission of Antimicrobial Drug-resistant Salmonella. Emerg Infect Dis 2004;10:2169-74. Venkatesh S, Memish ZA. Bioterrorism -- a new challenge for public health. Int J Antimicrob Agents 2003;21:200-6. Vourc'h G et al. Detecting Emerging Diseases in Farm Animals through Clinical Observations. Emerg Infect Dis 2006;12:204-10. Wagner MM et al. Design of a national retail data monitor for public health surveillance. J Am Med Inform Assoc 2003;10:409-18. Wagner MM et al. The Emerging Science of Very Early Detection of Disease Outbreaks. J Pub Health Manag Pract 2001;7:51-9. Wagner MM, Aryel R, Dato V. Availability and Comparative Value of Data Elements Required for an Effective Bioterrorism Detection System. November 28, 2001. Agency for Healthcare Research and Quality. Wagner MM, Shaffer L, Shephard R. Functional Requirements for Biosurveillance. In: Wagner MM, Moore A, Aryel R, eds. Handbook of Biosurveillance. New York, NY: Elsevier Inc., 2006:51-64. Wall PG et al. Multiresistant Salmonella typhimurium DT104 in cats: a public health risk. Lancet 1996;348:471. Waller LA. Invited Commentary: Surveilling Surveillance--Some Statistical Comments. Am J Epidemiol 2004;159:225-7. Webster RG. The importance of animal influenza for human disease. Vaccine 2002;20:S16-S20. Welliver RC et al. Sales of Nonprescription Cold Remedies: A Unique Method of Influenza Surveillance. Pediatr Res 1979;13:1015-7. 136 Wethington H, Bartlett P. The RUsick2 Foodborne Disease Forum for Syndromic Surveillance. Emerg Infect Dis 2004;10:401-5. Wilson M. Travel and the emergence of infectious diseases. Emerg Infect Dis 1995;1:39- 46. Witt CJ. Electronic Surveillance System for the Early Notification of Community-based Epidemics (ESSENCE). December 11, 2003. Department of Defense Global Emerging Infections System. Wong W-K et al. Rule-Based Anomaly Pattern Detection for Detecting Disease Outbreaks. 2002. American Association for Artificial Intelligence. Woolhouse MEJ. Population biology of emerging and re-emerging pathogens. Trends Microbiol 2002;10:S3-S7. Wurtz RM, Popovich ML. Animal Disease Surveillance: A Framework for Supporting Disease Detection in Public Health. March 2002. Tuscon, AZ, Scientific Technologies Corporation. Yingst SL, Saad MD, Felt S. Qinghai-like H5N1 from Domestic Cats, Northern Iraq. Emerg Infect Dis 2006;12:1295-7. Yudkowsky, Eliezer. An Intuitive Explanation of Bayesian Reasoning [Web Page]. Accessed 2006 Apr 17. Available at: http://yudkowsky.net/bayes/bayes.html. Zeng X, Wagner M. Modeling the Effects of Epidemics on Routinely Collected Data. J Am Med Inform Assoc 2002;9:S17-S22. Zepeda C, Salman MD. Planning Survey, Surveillance, and Monitoring Systems -- Roles and Requirements. In: Salman MD, ed. Animal Disease Surveillance and Survey Systems: Methods and Applications. Ames, Iowa: Iowa State Press, 2003:35-43. 137 APPENDIX A DISEASES NOTIFIABLE TO THE OIE 138 Updated: 23/01/2006 Diseases Notifiable to the OIE (http://www.oie.int/eng/maladies/en_classification.htm) Multiple species diseases • Anthrax • Aujeszky's disease • Bluetongue • Brucellosis (Brucella abortus) • Brucellosis (Brucella melitensis) • Brucellosis (Brucella suis) • Crimean Congo haemorrhagic fever • Echinococcosis/hydatidosis • Foot and mouth disease • Heartwater • Japanese encephalitis • Leptospirosis • New world screwworm (Cohliomyia hominivorax) • Old world screwworm (Chrysomya bezziana) • Paratuberculosis • Q fever • Rabies • Rift Valley fever • Rinderpest • Trichinellosis • Tularemia • Vesicular stomatitis • West Nile fever Cattle diseases • Bovine anaplasmosis • Bovine babesiosis • Bovine genital campylobacteriosis • Bovine spongiform encephalopathy 139 • Bovine tuberculosis • Bovine viral diarrhoea • Contagious bovine pleuropneumonia • Enzootic bovine leukosis • Haemorrhagic septicaemia • Infectious bovine rhinotracheitis/infectious pustular vulvovaginitis • Lumpky skin disease • Malignant catarrhal fever • Theileriosis • Trichomonosis • Trypanosomosis (tsetse-transmitted) Sheep and goat diseases • Caprine arthritis/encephalitis • Contagious agalactia • Contagious caprine pleuropneumonia • Enzootic abortion of ewes (ovine chlamydiosis) • Maedi-visna • Nairobi sheep disease • Ovine epididymitis (Brucella ovis) • Peste des petits ruminants • Salmonellosis (S. abortus ovis) • Scrapie • Sheep pox and goat pox Equine diseases • African horse sickness • Contagious equine metritis • Dourine • Equine encephalomyelitis (Eastern) • Equine encephalomyelitis (Western) • Equine infectious anaemia • Equine influenza • Equine piroplasmosis • Equine rhinopneumonitis • Equine viral arteritis • Glanders 140 • Surra (Trypanosoma evansi) • Venezuelan equine encephalomyelitis Swine diseases • African swine fever • Classical swine fever • Nipah virus encephalitis • Porcine cysticercosis • Porcine reproductive and respiratory syndrome • Swine vesicular disease • Transmissible gastroenteritis Avian diseases • Avian chlamydiosis • Avian infectious bronchitis • Avian infectious laryngotracheitis • Avian mycoplasmosis (M. gallisepticum) • Avian mycoplasmosis (M. synoviae) • Duck virus hepatitis • Fowl cholera • Fowl typhoid • Highly pathogenic avian influenza • Infectious bursal disease (Gumboro disease) • Marek's disease • Newcastle disease • Pullorum disease • Turkey rhinotracheitis Lagomorph diseases • Myxomatosis • Rabbit haemorrhagic disease 141 Bee diseases • Acarapisosis of honey bees • American foulbrood of honey bees • European foulbrood of honey bees • Small hive beetle infestation (Aethina tumida) • Tropilaelaps infestation of honey bees • Varroosis of honey bees Fish diseases • Epizootic haematopoietic necrosis • Infectious haematopoietic necrosis • Spring viraemia of carp • Viral haemorrhagic septicaemia • Infectious pancreatic necrosis • Infectious salmon anaemia • Epizootic ulcerative syndrome • Bacterial kidney disease (Renibacterium salmoninarum) • Gyrodactylosis(Gyrodactylussalaris) • Red sea bream iridoviral disease Mollusc diseases • Infection with Bonamia ostreae • Infection with Bonamia exitiosa • Infection with Marteilia refringens • Infection with Mikrocytos mackini • Infection with Perkinsus marinus • Infection with Perkinsus olseni • Infection with Xenohaliotis californiensis Crustacean diseases • Taura syndrome • White spot disease • Yellowhead disease 142 • Tetrahedral baculovirosis (Baculovirus penaei) • Spherical baculovirosis (Penaeus monodon-type baculovirus) • Infectious hypodermal and haematopoietic necrosis • Crayfish plague (Aphanomyces astaci) 143 APPENDIX B NATIONALLY NOTIFIABLE INFECTIOUS DISEASES 144 Nationally Notifiable Infectious Diseases United States 2006 (http://www.cdc.gov/epo/dphsi/phs/infdis2006.htm) • Acquired Immunodeficiency Syndrome (AIDS) • Anthrax • Arboviral neuroinvasive and non-neuroinvasive diseases o California serogroup virus disease o Eastern equine encephalitis virus disease o Powassan virus disease o St. Louis encephalitis virus disease o West Nile virus disease o Western equine encephalitis virus disease • Botulism o Botulism, foodborne o Botulism, infant o Botulism, other (wound & unspecified) • Brucellosis • Chancroid • Chlamydia trachomatis, genital infections • Cholera • Coccidioidomycosis • Cryptosporidiosis • Cyclosporiasis • Diphtheria • Ehrlichiosis o Ehrlichiosis, human granulocytic o Ehrlichiosis, human monocytic o Ehrlichiosis, human, other or unspecified agent • Giardiasis • Gonorrhea • Haemophilus influenzae, invasive disease • Hansen disease (leprosy) • Hantavirus pulmonary syndrome • Hemolytic uremic syndrome, post-diarrheal 145 • Hepatitis, viral, acute o Hepatitis A, acute o Hepatitis B, acute o Hepatitis B virus, perinatal infection o Hepatitis, C, acute • Hepatitis, viral, chronic o Chronic Hepatitis B o Hepatitis C Virus Infection (past or present) • HIV infection o HIV infection, adult(> =13 years) o HIV infection, pediatric (<13 years) • Influenza-associated pediatric mortality • Legionellosis • Listeriosis • Lyme disease • Malaria • Measles • Meningococcal disease • Mumps • Pertussis • Plague • Poliomyelitis, paralytic • Psittacosis • Q Fever • Rabies o Rabies, animal o Rabies, human • Rocky Mountain spotted fever • Rubella • Rubella, congenital syndrome • Salmonellosis • Severe Acute Respiratory Syndrome-associated Coronavirus (SARS-CoV) disease • Shiga toxin-producing Escherichia coli (STEC) • Shigellosis • Smallpox • Streptococcal disease, invasive, Group A • Streptococcal toxic-shock syndrome • Streptococcus pneumoniae, drug resistant, invasive disease • Streptococcus pneumoniae, invasive in children <5 years 146 • Syphilis o Syphilis, primary o Syphilis, secondary o Syphilis, latent o Syphilis, early latent o Syphilis, late latent o Syphilis, latent, unknown duration o Neurosyphilis o Syphilis, late, non-neurological o Syphilitic Stillbirth • Syphilis, congenital • Tetanus • Toxic-shock syndrome (other than Streptococcal) • Trichinellosis (Trichinosis) • Tuberculosis • Tularemia • Typhoid fever • Vancomycin - intermediate Staphylococcus aureus (VISA) • Vancomycin - resistant Staphylococcus aureus (VRSA) • Varicella (morbidity) • Varicella (deaths only) • Yellow fever 147 APPENDIX C IDEXX LABORATORIES PRODUCT CODE TO SYNDROME CATEGORY MAPPING 148 149 174 155 148 145 136 134 132 130 128 125 124 119 114 113 112 111 109 106 23 22 18 Code Product IDEXX AMYLASE +LIPASE PRE-OPERATIVE PANEL IDEXX ProductName AVIAN PANEL#3 AVIAN PANEL#2 AVIAN PANEL#1 K VALUE RENAL PANEL AVIAN CHEMISTRY PANEL AVIAN CHEMISTRY PANEL SMALL MAMMAL PANEL#2 SMALL MAMMAL PANEL#1 ADD-ON FeLVANTIGEN-ELISA ACTH RESPONSE -3POST ACTH RESPONSE -2POST ACTH STIMULATION REPTILIAN PANEL#2 REPTILIAN PANEL#1 DEXAMETHASONE SUPPRESSION ADD-ON T4 CHEM 25 LARGE ANIMAL PANEL Respiratory X X Gastrointestinal Neurologic Dermal Reproductive X X X X X Endocrine Hepatic Infection Febrile X X X X X X X X X X X X X Non-specific X Renal IDEXX Product Code IDEXX Product Name Respiratory Respiratory Gastrointestinal Neurologic Dermal Reproductive Endocrine Hepatic Infection Febrile Non-specific Renal 181 FREE T4 X 192 LIVER PANEL X 200 ALBUMIN X 203 AMYLASE X 204 BILIRUBIN X 205 BILIRUBIN X 207 BUN X 150 208 CALCIUM X 210 CHLORIDE X 211 CHOLESTEROL X 212 CPK X 213 CREATININE X 214 GGT X 215 GLUCOSE X 219 LIPASE X 220 MAGNESIUM X 221 PHOSPHORUS X 222 TOTAL PROTEIN X 223 PROTEIN ELECTROPHORESIS X 228 URIC ACID X 229 POTASSIUM X 151 318 314 313 309 303 302 301 300 275 274 273 267 266 261 258 257 242 234 232 231 230 Code Product IDEXX DEXAMETHASONE DEXAMETHASONE SUPPRESSION SUPPRESSION BILE BILE ACIDS ACIDS CHOLINESTERASE IDEXX ProductName TP ALBGLOB RETICULOCYTE PLATELET PANEL COUNT CBC COMPREHENSIVE GLUCOSE CURVE PLASMA PROTEIN ADD-ON DIFFERENTIAL CHEMISTRY PANEL REPTILIAN PANEL#3 REFLEX RETIC COOMBS (37 DEGREE) URINE CORTISOL SODIUM TRIGLYCERIDE BLOOD COUNT ETHYLENE GLYCOL Respiratory X XX X X Gastrointestinal X Neurologic Dermal Reproductive X X X Endocrine X X X Hepatic X Infection Febrile X X X X X X X X X Non-specific Renal 152 427 418 414 411 409 408 407 406 405 402 401 400 399 375 355 350 337 336 335 326 322 Code Product IDEXX FUNGAL CULTURE IDEXX ProductName DIRECT GRAMSTAIN WBC RETICULOCYTE ESTIMATE/HEMATOCRIT PANEL ACID FASTSTAIN CLIN PATH REVIEWBY REQUEST MYCOPLASMA CULTURE COMPLETE EXOTIC CBC ADD-ON AVIANEXOTICCBC(+PP) SPECIAL EYE CULTURE (CIP/TOB) GRAM STAIN AEROBIC CULTURE AEROBIC CULTURE -ID ONLY FUNGAL WET MOUNT CBC BASIC BLOOD CULTURE FOLLOW-UP CBC FOLLOW-UP URINE CULTURE ANAEROBIC &AEROBIC CULTURE PROTEIN AVIAN/EXOTIC CBCAND PLASMA WBC X X Respiratory Gastrointestinal Neurologic Dermal Reproductive Endocrine Hepatic X X X X X X X X X X X X X X X Infection Febrile X X X X Non-specific Renal 153 702 700 648 647 628 612 517 514 513 511 510 509 508 507 503 502 501 497 445 444 437 Code Product IDEXX COMP CULTURE INPROGRESS CULTURE INPROGRESS OVA &PARASITES AVIAN PARASITESCREEN GIARDIA ELISA BLOOD PARASITES SPECIAL STAINS IDEXX ProductName FECAL EXAMDIRECT BAERMANN TEST FORLUNGWORMS ANTINUCLEAR ANTIBODY(ANA) MICROFILARIA FIA CRYPTOSPORIDIUM ELISA RABBIT/RODENT AEROBIC CULTURE BRUCELLA CANIS (IFA) PARASITE IDENTIFICATION RAPID-GR MYCOBACTERIA(AFB) DERMATOLOGY CONSULTATION CYTOLOGY with microscopic (3 sites) CYTOLOGY with microscopic (2 sites) GIARDIA/CRYPTOSPORIDIUM X X XX X Respiratory X X X X X X X Gastrointestinal Neurologic X Dermal X Reproductive Endocrine Hepatic X X X X Infection X Febrile X X X Non-specific Renal 154 733 729 727 725 724 723 721 719 718 717 716 715 714 713 712 710 709 707 706 705 704 Code Product IDEXX CHLAMYDIA ELISA TOXOPLASMA IGGIFA HERPES IFA PARVOVIRUS FECALANTIGEN-ELISA LEPTOSPIROSIS PANEL(MAT) IDEXX ProductName EHRLICHIA CANIS ABIFA DISTEMPER AB(IgG)IFA LYME (BORRELI CORONAVIRUS AB DISTEMPER (CANINE) AB -IgM(IFA) CRYPTOCOCCUS ANTIGEN TITER DISTEMPER (CANINE) FA FeLV ANTIGEN (ELISA) EQUINE IgG SCREEN HEARTWORM ANTIGEN TEST FeCoV (FIP) ANTIBODY EHRLICHIA EQUI -CANINE FeLV ANTIGEN -IFA COGGINS (EIA) AGID COCCI-IMMUNODIFFUSION PARVOVIRUS ANTIBODY TITER(IgG)-IFA OSIS) TITER X X X X X X X X Respiratory X X X X X X Gastrointestinal X Neurologic Dermal Reproductive Endocrine Hepatic X Infection X X X X X X X Febrile Non-specific Renal 155 813 812 811 810 808 807 806 805 804 803 802 800 778 759 758 757 756 748 743 737 736 Code Product IDEXX PARATHYROID HORMONE ASSAY IDEXX ProductName DISTEMPER (CANINE) AB -IgG/M(IFA) QUANTITATIVE IGA-CANINE TESTOSTERONE HISTOPLASMOSIS ANTIBODY PARVOVIRUS AB IgG & IgM (IFA) PARVOVIRUS ABIgG& QUANTITATIVE IGM-CANINE PROGESTERONE ESTRONE SULFATE BRUCELLA (AGID2) HERPES/CALICI TITERFELINE ACTH ENDOGENOUS POST-T4 ESTROGEN -CORNELL ROCKY MOUNTAIN SPOTTEDFVR PRE-T4 INSULIN CORTISOL T4 T3 CHEMISTRY PANEL X X X Respiratory X X Gastrointestinal X Neurologic Dermal X X X X X Reproductive X X X X X X X Endocrine Hepatic X X Infection Febrile X Non-specific Renal 156 908 907 903 902 901 896 875 856 855 853 850 849 848 843 839 832 830 826 825 816 814 Code Product IDEXX T4 BYRIA IDEXX ProductName LIVER PANEL SPERM COUNT &MORPHOLOGY EHRLICHIA PLATYS OCCULT BLOOD GERIATRIC FELINEPANEL EHRLICHIA RISTICII ADD-ON FREE T4 (BYED) BABESIA CANISTITER PRE-ACTH CORTISOL COPPER BROMIDE (LIVER) FREE T4(EQUIL. DIALYSIS) FIV WESTERNBLOT THYROID PANEL#5 FECAL FECAL STARCH FAT FECAL TRYPSIN POST-ACTH CORTISOL cTSH ACETYLCHOLINE RECP AB TITER Respiratory X X X X X Gastrointestinal X X X X Neurologic Dermal X Reproductive X X X X X X Endocrine Hepatic Infection X X X Febrile X X Non-specific Renal 157 1026 1025 1024 1021 1020 1016 1002 1001 1000 994 993 962 957 948 945 943 941 940 939 935 916 Code Product IDEXX ASPERGILLUS ANTIBODY EQUINE VIRAL ARTERITIS BOVINE LEUKEMIAVIRUS EQUINE INFLUENZA IDEXX ProductName URINE, ROUTINE MICROSCOPICEXAM BLASTOMYCOSIS ANTIBODY RABBIT RABBIT PASTEURELLA SYPHILIS CHEMISTRY PANEL PHENOBARBITAL SERUM ELECTROLYTE IRON PANEL BAL CELLCOUNT&CYTOLOGY URINE PROTEIN/CREATININE RATIO ARSENIC DIGOXIN CULTURE IF (4035 - WITH MIC) RHINOPNEUMONITIS TITER BLOOD LEAD ENCEPHALITOZOAN URINE CREATININE X X X X X Respiratory X Gastrointestinal X X X X X Neurologic X Dermal Reproductive Endocrine Hepatic X X Infection X Febrile X X X X X Non-specific X Renal 158 1237 1233 1232 1231 1195 1150 1119 1113 1079 1071 1069 1068 1065 1047 1046 1043 1039 1036 1034 1031 1029 Code Product IDEXX ADD-ON FIVANTIBODY-ELISA MUSCLE HISTOCHEMISTRY FIV ANTIBODY-IFA ANTITHROMBIN III GLYCOSYLATED HEMOGLOBIN ADD-ON FIVANTIBODY-IFA FIV ANTIBODY BOVINE IGG PARVOVIRUS ANTIBODY (HI) RABIES SEROLOGY RFFIT IDEXX ProductName CHEMISTRY PANEL CHEMISTRY PANEL DISTEMPER (CANINE) AB (SVN) FELINE HEARTWORM ANTIGEN PANLEUKOPENIA (HAI) VIRAL SELENIUM ISOLATION OSMOLALITY -SERUM EQ GRANULOSA CELLTUMOR PNL CHEM 27 LLAMA IGG X X Respiratory X X X X X X Gastrointestinal X Neurologic Dermal X Reproductive Endocrine Hepatic X X X X Infection Febrile X X X X X X Non-specific X Renal 159 1367 1366 1365 1358 1357 1356 1336 1327 1326 1322 1319 1316 1315 1314 1313 1293 1271 1256 1255 1250 1238 Code Product IDEXX ASPERGILLUS AB/AGAVIAN (+EPH) RABIES SEROLOGYOIE FAVN FELINE HERPES&CHLAMYDIA PCR URINE PROTEIN ELECTROPHORESIS T4AA andT3AA FOLLOW-UP CHEMISTRYPANEL FELINE HEARTWORM ANTIBODY IDEXX ProductName LEAD/ZINC EXOTIC CHEMISTRY PANEL CHEMISTRY PANEL PARATHYROID RELATED PROTEIN TSH STIMULATION(EAST) CYCLOSPORINE CANINE THYROGLOBULIN AB ALDOSTERONE INSULIN PANEL FERRET ANDROGEN PANEL CHEM CHEM 11 21 ZINC NEOSPORA IFA X X X Respiratory X X Gastrointestinal X X X Neurologic Dermal Reproductive X X X X X X Endocrine Hepatic Infection Febrile X X X X X X X Non-specific Renal 160 1557 1554 1553 1545 1543 1497 1478 1474 1460 1449 1444 1439 1434 1429 1419 1404 1398 1396 1384 1370 1369 Code Product IDEXX ROTAVIRUS -ELISA HEAVY METALSCREEN CAPRINE ARTHRITISENCEPH MERCURY ZINC/MAMMALS IDEXX ProductName CHEMISTRY CHEMISTRY PANEL PANEL CHEMISTRY CHEMISTRY PANEL CHEMISTRY PANEL CHEMISTRY PANEL CHEMISTRY PANEL PANEL TRIGLYERIDES FLUID BABESIA GIBSONI CHOLESTEROL +TRIGLYCERIDE URINE CHEMISTRY PANEL CHLAMYDIA DNA(PROBE) CORTISOL X3 COPPER (SERUM) LIVER CHEMISTRIES LYTES YOUNG PRE-ANESTHETIC PANEL+ X X Respiratory Gastrointestinal X X X X Neurologic Dermal Reproductive X Endocrine X Hepatic Infection X Febrile X X X X X X X X X Non-specific X X X Renal 161 1803 1777 1766 1751 1746 1726 1725 1724 1723 1720 1717 1714 1705 1704 1702 1697 1680 1666 1646 1600 1565 Code Product IDEXX LEISHMANIA IFA FECAL PANEL VITAMIN E HEMOBARTONELLA PCR THYROID OFA TESTING CANINE PREGNANCY TEST IDEXX ProductName CHEMISTRY CHEMISTRY PANEL PANEL CHEMISTRY CHEMISTRY PANEL PANEL MULTI TICK PCR PANEL TRACE ELEMENT SCREEN EPM WESTERNBLOT -SERUM BLADDER TUMOR ANALYTES (K9) FIP PCR LYME +RMSF VITAMIN D EGG EHRLICHIA COUNT PCR EPM PCRADDONTO WESTERN BLOT COMBINATION ACTH/DEX SUPPRESSION X Respiratory X X Gastrointestinal X X Neurologic X Dermal X X Reproductive X X Endocrine Hepatic Infection X X X Febrile X X X X X Non-specific X X X Renal 162 2137 2121 2111 2091 2088 2084 2076 2031 2025 2023 2014 2011 2002 1991 1975 1963 1941 1911 1881 1857 1849 Code Product IDEXX KIDNEY PANEL ADD-ON AMYLASE WEST NILEVIRUSIgMC-ELISA TRYPSIN-LIKE IMMUNOREACT. EPM WESTERN BLOT -CSF DEXAMETHASONE SUPPRESSION WEST NILEVIRUSPLQ REDUCT IDEXX ProductName CHEMISTRY CHEMISTRY PANEL PANEL CHEMISTRY PANEL CHEMISTRY PANEL CHEMISTRY CHEMISTRY PANEL PANEL CHEMISTRY PANEL SPEC cPL BABESIA PANEL ELECTROLYES +CALCIUM EQUINE ENDOGENOUS ACTH ESTROGEN (DAVIS) PNL CANINE ADRENAL& FOLATE &VITAMINB12(COBALAMIN) FELINE ADRENAL XX Respiratory X X X Gastrointestinal X X X Neurologic Dermal Reproductive X X X X Endocrine Hepatic Infection X Febrile X X X X X X X X X Non-specific Renal 163 2251 2244 2243 2242 2241 2240 2231 2225 2224 2223 2220 2214 2212 2201 2199 2198 2197 2192 2191 2160 2138 Code Product IDEXX ADD-ON LIPASE STREPTOCOCCUS EQUI -ELISA ALDOSTERONE -SINGLE SAMPLE CANINE VACCINATION PANEL CANINE VACCINATION PANEL ADD-ON PROT ELECTROPHORESIS FELINE CANINE PROFILE PROFILE LIVER PANEL8 LIVER PANEL7 TOXOPLASMOSIS IgG/IgM (HA) RENAL PANEL3 ADD-ON MAGNESIUM HISTOPLASMA ANTIGEN IDEXX ProductName CHEMISTRY CHEMISTRY PANEL CHEMISTRY PANEL PANEL FELINE CALICI PCR LUTEINIZING HORMONE CHEM PROFILE-AVIAN PCR FELINE HERPES/CHLAMYDIA/CALICI XX XX X X X X Respiratory X X Gastrointestinal X Neurologic Dermal X Reproductive X Endocrine Hepatic Infection Febrile X X X X X X X X X X Non-specific Renal 164 4021 4010 4009 3141 3131 3091 3032 3013 3009 2571 2421 2401 2400 2362 2310 2301 2286 2285 2284 2272 2265 Code Product IDEXX ADD-ON BILEACIDS WEST NILEVIRUSABTITER SUSCEPTIBILITY ONLY ADD-ON PLATELET FECAL PANEL#6 ANAEROBIC CULTURE &ID PLASMA PROTEIN IDEXX ProductName WBC-EST CHEMISTRY PANEL STOOL CULTURE SALMONELLA ADD-ON RETICULOCYTE PANEL ADD ON FREET4byRIA PARATHYROID HORMONE DISTEMPER PCR TRITRICHOMONAS FOETUS ADD-ON TRIGLYCERIDE IONIZED CALCIUM COOMBS (37 +4) FREE T4byRIA CANINE SCREEN EARLY RENALDISEASE- HCT X Respiratory X X X Gastrointestinal Neurologic Dermal X Reproductive X X X Endocrine X Hepatic X X X X Infection X Febrile X X X X X X Non-specific X Renal 165 6313 6222 6137 6130 6128 6121 5141 5131 5033 5010 4183 4111 4091 4063 4062 4035 4031 4030 4027 4026 4023 Code Product IDEXX ADD-ON CRYPTOSPORIDIUM ELISA ADD-ON GIARDIA ELISA URINE CULT &SUSCEPTIBILITY C. DIFFICILE TOXINAELISA EXOTIC FECALSCREEN AVIAN FECALCULTURE MICROFILARIA OVA ANDPARASITES3OR MORE RETICULOCYTE COUNT STOOL CULTURE CAMPYLOBACTER IDEXX ProductName AVIAN CHEMISTRY C.PERFRINGENS ENTEROTOXIN AVIAN CBC GRAM STAIN-ADD-ON ADD MYCOPLASMA CULTURE TOTAL PROTEIN BLOOD CULTURE - 3SETS BLOOD CULTURE - 2SETS ADD-ON DERMATOLOGY CONSULT ADD-ON DERMATOLOGY CONSULT FOLLOW-UP URINE CULTURE (5 DAYS) X X Respiratory X X X X X X X X X Gastrointestinal Neurologic X X Dermal Reproductive Endocrine Hepatic X X X X X X Infection Febrile X X Non-specific Renal 166 7246 7244 7233 7231 7191 7171 7161 7159 7157 7151 7131 7101 7093 7077 7042 7041 7022 7001 6485 6484 6480 Code Product IDEXX TOXOPLASMA IGG & IGM ELISA TOXOPLASMA IGG&IGM ADD-ON ANA TOXOPLASMA IgMIFA LYME C6QUANT ABELISA (K9) ADD-ON DISTEMPER INCLUSION -FA FELV AG(ELISA) ADD-ON FIP BRUCELLA CANIS -RSAT ADD-ON DISTEMPER ANTIBODY(IgG) ADD-ON FeLV ANTIGEN-IFA ADD-ON TOXOPLASMOSIS TITER(IFA) DISTEMPER (K9)AB-IgG &IgM-CSF IDEXX ProductName SNAP* 3Dx* ADD-ON HEARTWORM ANTIGEN -ELISA HEARTWORM ANTIGEN TEST COGGINS (EIA) -ELISA CYTOLOGY with microscopic (5 sites) CYTOLOGY with microscopic (4 sites) CYTOLOGY no microscopic (3 sites) IFA ADD-ON EHRLICHIACANIS ANTIBODY- ADD-ON PARVOVIRU S FECALANTIGEN X X X X X X Respiratory X X X X X Gastrointestinal X X X Neurologic Dermal X Reproductive Endocrine Hepatic X Infection X X X X Febrile X X X Non-specific Renal 167 8230 8220 8063 8062 8060 8059 8041 8033 8032 8031 8030 8003 8002 8001 8000 7724 7723 7721 7434 7271 7249 Code Product IDEXX PROGESTERONE byRIA ADD-ON LYMES(BORRELLIOSIS)TITER T3 byRIA ADD-ON DISTEMPER ANTIBODY(IgM) PROGESTERONE IDEXX ProductName EHRLICHIA EQUI -EQUINE PROGESTERONE Preand Post INSULIN (EXOTIC SPECIES) CORTISOL -PRE CORTISOL by RIA PROMOTIONAL HEARTWORM ANTIGEN ADD-ON T3 HERPES TITERCANINE POST-T4 POST-T3 ADD-ON CORTISOL PRE-T4 PRE-T3 FOLLOW-UP T4 CORTISOL -POST AB PROMOTIONAL LYMEC6QUANTITATIVE X X X Respiratory X Gastrointestinal Neurologic Dermal X X X Reproductive X X X X X X X X X X X Endocrine Hepatic Infection X X X Febrile Non-specific Renal 168 72441 72440 64850 64840 20111 18492 18491 18490 12381 12371 10341 10081 10010 9362 9361 9258 9101 8961 8531 8491 8251 Code Product IDEXX ADD-ON URINALYSIS ADD-ON FIVANTIBODY WESTERN BLOT LAB 3Dx FELINE TLI ADD ON TOXOPLASMOSIS TITER IDEXX ProductName LYME WESTERNBLOT EQUINE URINE BILE ACIDS PLI -FELINE ADD-ON FELINE HEARTWORM ANTIGEN LYME WESTERNBLOT PHENOBARBITAL Trough &Peak FOLLOW UP FREET4(BYED) ADD ON LAB 3Dx URINE OSMOLALITY ADD-ON cTSH CYTOLOGY no microscopic (5 sites) CYTOLOGY no microscopic (4 sites) ADD-ON SPECcPL ANTIBODY ADD-ON FELINE HEARTWORM EQUINE IgG - QUANTITATIVE SPEC cPLNOCHARGE X X X X Respiratory X X X X X X Gastrointestinal X X Neurologic Dermal Reproductive X X Endocrine Hepatic X Infection X X X Febrile X X X X Non-specific X X Renal 169 772319 772318 772317 772316 772315 772314 772313 772312 772311 772310 80630 77239 77238 77237 77236 77235 77234 77233 77232 72466 72461 Code Product IDEXX LYME C6ABELISA-CONVALESCENT ADD-ON LYMEC6QUANTITATIVE AB IDEXX ProductName BATCH HEARTWORM BATCH X19 HEARTWORM BATCH X18 HEARTWORM BATCH X17 HEARTWORM BATCH X16 HEARTWORM BATCH X15 HEARTWORM BATCH X14 HEARTWORM BATCH X13 HEARTWORM BATCH X12 HEARTWORM BATCH X11 HEARTWORM X10 BATCH BATCH HEARTWORM BATCH HEARTWORM X9 BATCH HEARTWORM X8 BATCH HEARTWORM X7 BATCH HEARTWORM X6 BATCH HEARTWORM X5 BATCH HEARTWORM X4 HEARTWORM X3 X2 PROGESTERONE byRIA X X X X X X X X X X X X X X X X X X Respiratory Gastrointestinal Neurologic Dermal X Reproductive Endocrine Hepatic Infection X X X X Febrile Non-specific Renal IDEXX Product Code IDEXX Product Name Respiratory Gastrointestinal Neurologic Dermal Reproductive Endocrine Hepatic Infection Febrile Non-specific Renal 772320 BATCH HEARTWORM X20 X 772321 BATCH HEARTWORM X21 X 772322 BATCH HEARTWORM X22 X 772323 BATCH HEARTWORM X23 X 772324 BATCH HEARTWORM X24 X 772325 BATCH HEARTWORM X25 X 772326 BATCH HEARTWORM X26 X 170 772327 BATCH HEARTWORM X27 X 772328 BATCH HEARTWORM X28 X 772329 BATCH HEARTWORM X29 X 772330 BATCH HEARTWORM X30 X 772331 BATCH HEARTWORM X31 X 772346 BATCH HEARTWORM X46 X