Issues in Brief Pulsenet: A Critical Food Safety Surveillance System Association of Public Health Laboratories APril 2010

PUlseNet A Critical Food Safety Surveillance System

Public health laboratorians are critical to the detection and prevention of foodborne illnesses. Through a national laboratory-based foodborne disease surveillance network known as PulseNet, public health and agricultural laboratories have detected high-profile outbreaks such as those from imported produce, peanut butter and peanut butter-containing products and raw cookie dough.1-3

In 2008 alone, PulseNet laboratorians DNA fingerprinting. Established detected more than 1,500 local clusters in 1996 by the Centers for Disease of foodborne illnesses and increased Control and Prevention (CDC), four the number of foodborne bacterial public health laboratories, the US isolates tested. Additional resources and Department of Agriculture (USDA) support are needed for public health and the Association of Public Health laboratories to test and investigate all Laboratories (APHL), PulseNet has since cases of . It is critical grown to more than 70 laboratories for the nation to recognize the impact nationwide, including state and that public health and agricultural local public health laboratories, state laboratories have on the overall agricultural laboratories and regulatory foodborne disease surveillance system laboratories within the US Food and to ensure a safer food supply. Drug Administration (FDA) and USDA. Molecular subtyping and computer The PulseNet network links public analysis are performed to generate health laboratories nationwide, and analyze the DNA fingerprint monitoring such as Shiga- patterns. These patterns generated toxin producing or from isolates of ill persons and/or food STEC (including E. coli O157:H7), and environmental samples are then Listeria monocytogenes, compared to national databases at CDC, species (sp.), Shigella species and allowing for the early identification Campylobacter species using a molecular of foodborne disease clusters. Such subtyping method called pulsed- information assists epidemiologists with field gel electrophoresis (PFGE) or their investigations and may potentially lead to the identification of the source conducted in 2006 to assess activities of an outbreak. for the 2005 calendar year. Some comparisons are included in this APHL, which represents state and report for questions that were largely local public health and agricultural unchanged between 2005 and 2008. laboratories, supports its members by providing various training opportunities, The survey was administered through disseminating information relevant to MR Interview, a web-based repository the public health laboratory community, and survey tool. Descriptive analyses facilitating the transfer of new were conducted, and responses were technology nationwide, and conducting grouped into three main categories: assessments of laboratory capacity and surveillance, communication, and capabilities. In 2009, APHL administered laboratory management. a survey to assess the capability and capacity of the PulseNet network and SURVEILLANCE to determine the challenges that face The PulseNet network has proven to be public health laboratories in achieving an essential component to the national the goals and objectives of the network. foodborne disease surveillance system. This issue brief presents the findings of Through active participation by local that survey. and state laboratories, this network has ensured improvements to the METHODS nation’s food safety system by detecting In April 2009, APHL conducted a survey numerous outbreaks and preventing to assess participating laboratories’ many illnesses. In recent years, capacities and capabilities for laboratories have continued to increase conducting molecular subtyping the number of isolates tested by this through the PulseNet network for the molecular subtyping method. 2008 calendar year. The survey was sent to 64 public In 2008, the PulseNet network In 2008, the PulseNet network health laboratories collectively subtyped a total of 48,194 and agricultural isolates of STEC, Salmonella, Shigella, L. collectively subtyped a laboratories— monocytogenes and Campylobacter. This total of 48,194 isolates of which included is a 47% increase from 2005 in which 50 state and 32,830 isolates of the same pathogens STEC, Salmonella, Shigella, territorial were subtyped. In 2008, PulseNet L. monocytogenes and public health laboratories were able to subtype 70% laboratories, 10 of the aforementioned foodborne Campylobacter. local public health pathogens received from clinical laboratories and 4 laboratory partners, as compared to 61% state agricultural in 2005. laboratories. Of those surveyed, APHL received 57 responses—comprised of Although PulseNet has increased 46 state and territorial public health the percentage of foodborne isolates laboratories, 9 local public health subtyped, all pathogens under PulseNet laboratories and 2 state agricultural surveillance that are submitted to laboratories—for an overall response public health laboratories should ideally rate of 89%. A similar survey was be subtyped. A significant delay or lack

2 APHL Public Health Laboratory Issues in Brief of outbreak recognition may occur very similar PFGE patterns.4-6 Currently, when laboratories do not subtype all E. coli O157:H7 isolates that are part of foodborne isolates. In 2008, only 61% a multi-state outbreak are sent to the of laboratories were able to subtype CDC for MLVA testing and analysis. In all STEC, Shigella, L. monocytogenes and 2008, only 12% of PulseNet laboratories Salmonella isolates received in their performed supplemental or additional laboratories. Laboratories did not subtyping methods (a 3% increase subtype all isolates due to inadequate funding for supplies, staff shortages and Figure 1. Comparison of Isolates PFGE Subtyped increased workload in all areas of public Comparison of Isolates PFGE Subtyped by PulseNet Laboratories in by2005 PulseNet and 2008 Laboratories in 2005 and 2008 health laboratory testing.

Currently, the bulk of isolates 70,000 received and subtyped in PulseNet 69,074 laboratories are of clinical origin; 50,000 Isolates received in however, environmental samples 53,494 laboratories and food samples are also important Isolates subtyped in potentially identifying the source 30,000

Number of isolates 48,194 of foodborne outbreaks. Sixty- seven percent (67%) of PulseNet 32,830 laboratories subtyped isolates from environmental and food origins. This 2005 2008 is an increase from 2005, where 45% Source: APHL PulseNet Survey, 2008. Note: Comparison of of PulseNet laboratories subtyped isolates PFGE subtyped as share of the total amount food and environmental isolates. from 2005). With additional funding of isolates received in The importance of subtyping isolates and personnel, PulseNet laboratories PulseNet laboratories in of food and environmental origins would have the capability to perform 2005 and 2008. was demonstrated in the 2008-2009 these new subtyping techniques in their Salmonella Typhimurium multi-state own laboratories, thus eliminating the outbreak associated with peanut butter additional time needed for specimen and peanut butter-containing products. transport to CDC and potentially A match in DNA fingerprint profiles reducing the amount of time it takes to from the food source to the human identify an outbreak. outbreak strain prompted public health officials to issue consumer warnings Among the more than 1,500 foodborne and product recalls. disease clusters that were identified through PulseNet in 2008, three of PFGE has proven to be a powerful every four local clusters detected were tool for the detection of foodborne followed-up by an epidemiologist. The disease clusters. Additionally, newer detection of small clusters by local molecular subtyping methods, such public health laboratories can aid as multiple-locus variable number in the identification of larger multi- tandem repeats analysis (MLVA), can state outbreaks. This demonstrates greatly contribute to some foodborne the important role that state and outbreak investigations. In some cases, local agricultural and public health MLVA can further discriminate between laboratories contribute to the overall bacterial strains that have identical or foodborne investigation process.

Pulsenet: A Critical Food Safety Surveillance System April 2010 3 COMMUNICATION notification of new clusters or “matches” Communication and data sharing to an existing national cluster. In 2008, among PulseNet laboratories is critical available resources allowed for 72% of to detecting and identifying multi-state laboratories to respond to web-board postings within two working days. This Figure 2. Comparison of Isolates PFGE Subtyped is an increase from 2005, in which 49% by PulseNet Laboratories in 2008 of laboratories were able to respond Comparison of Isolates PFGE Subtyped by PulseNet Laboratories in 2008 within two working days. 50,000

47,030 Communication between laboratory and epidemiology partners is another key factor in promptly identifying foodborne outbreaks. Seventy-five percent (75%) of PulseNet laboratories in 2008 reported

Isolates received in communicating with their state or local laboratories 36,206 foodborne epidemiologists on a weekly Isolates subtyped basis or more. This is a significant solates I 30,000 improvement over 2005, in which 59% of laboratories reported communicating with their epidemiologists in the same

Number of timeframe.

Additional tools, such as an electronic interface linking laboratory and epidemiology data, could allow 12,219 outbreak investigations to be conducted 10,000 more efficiently. In 2008, only 11 of 57 laboratories (19%) reported 4,587 4,575 having a mechanism capable of data 6,321 exchange between the laboratory and 4,232 663 655 780 the epidemiology programs. If more L.monocytogenes Campylobacter STEC Shigella Salmonella

Note: Comparison of Source: APHL PulseNet Survey, 2008. In 2008, only 11 of 57 isolates PFGE subtyped by pathogens as share of total laboratories (19%) reported isolates received in 2008. outbreaks. One important component of the PulseNet surveillance system is having a mechanism capable the secure online databases that allow of data exchange between for the comparison of DNA fingerprints the laboratory and the from regulatory agencies, public health and agricultural laboratories across epidemiology programs. the nation. The rapid comparison of DNA fingerprints allows laboratories to determine whether a small, local laboratories can electronically exchange cluster is part of a larger multi-state information with epidemiologists, case outbreak. These databases, along with interviews could readily be linked to a secure online web-board available to isolate information allowing for rapid PulseNet participants, allow for timely response to outbreak investigations.

4 APHL Public Health Laboratory Issues in Brief LABORATORY MANAGEMENT state outbreaks—such as the Salmonella Many laboratories are struggling Saintpaul outbreak associated with to keep up with the demand of fresh produce and the E. coli O157:H7 foodborne testing due to outbreak associated with raw cookie budget decreases at the state and dough products—turnaround time local government levels. PulseNet laboratories need the capabilities, FTurn-Around-Timeigure 3. Turnaround (in Median T imeWorking (in Days)Median of Foodborne Working Pathogens under capacity and resources to promptly Days)PulseNet of Surveillance Foodborne in 2005 Pathogens and 2008 under subtype all foodborne pathogens PulseNet Surveillance in 2005 and 2008 received in their laboratories. Findings from our survey suggest that lack of Pathogen 2005 2008 funding in 2008 has impacted laboratory STEC 5 4 capacity at the local and state levels. 5  46% encountered personnel L. monocytogenes 4 shortages Salmonella 8 5  41% reported inadequate funding for supplies Shigella 8 5  27% reported a lack of certified Campylobacter 13 5 laboratory staff to perform DNA subtyping Source: APHL PulseNet Survey, 2008.

TurnAround Time affects the timeliness of identifying related ill cases and the source of Since 2005, PulseNet laboratories have the outbreak. Approximately 33% of decreased the amount of time it takes laboratories subtyped Salmonella and to subtype an isolate. Turnaround Shigella isolates within CDC’s suggested time is defined as the time (in working turnaround time of four days, as days) from the receipt of an isolate in compared to approximately 75% of the PulseNet laboratory to the time laboratories that subtyped STEC and the isolate’s image is uploaded to L. monocytogenes. The large amount of the national PulseNet database. The Salmonella and Shigella isolates received PulseNet surveillance system uses in laboratories may account for the five- turnaround time as a measure to day turnaround time. In 2008, PulseNet determine its efficiency. CDC guidelines laboratories received 59,249 isolates suggest a turnaround time of four of Salmonella and Shigella compared working days for all PulseNet organisms. to 5,238 isolates of STEC and L. In 2008, the median turnaround time monocytogenes. Due to the sheer number for STEC and L. monocytogenes was four of Salmonella and Shigella isolates days, a decrease from 2005 by one day. received in PulseNet laboratories, some The median turnaround time in 2008 laboratories were unable to perform for Salmonella and Shigella was five testing within the four-day period. days, a significant decrease of three days from 2005. Another factor that affects turnaround time is the point at which molecular Although turnaround time has subtyping is performed. Some decreased over the years, there is pathogens require an additional always room for improvement. As testing method known as serotyping, demonstrated by recent large, multi-

Pulsenet: A Critical Food Safety Surveillance System April 2010 5 which is a traditional method to completed serotyping first then characterize bacterial isolates. In continued with molecular subtyping. general, once serotyping is completed, the laboratory then proceeds with While it appears that the best practice molecular subtyping. Serotyping may to decrease turnaround time in PulseNet take a few days to perform and, thus, laboratories is to simultaneously increases the overall turnaround time. perform serotyping and molecular On average, 42% of PulseNet laboratories subtyping on all isolates, some simultaneously performed serotyping laboratories do not find this application and molecular subtyping of STEC, L. feasible. With workforce shortages, monocytogenes, Salmonella sp., Shigella decreasing funds and lack of trained sp. and Campylobacter sp. in 2008. This is staff, some laboratories do not have the an improvement from 2005, where 30% capability to simultaneously serotype of laboratories reported simultaneous and subtype PulseNet pathogens, nor do testing. Thus, as expected, laboratories they have the resources to subtype all that reported simultaneous serotyping these pathogens once they have been and molecular subtyping had a lower serotyped. turnaround time than those that Staff Shortage Figure 4. General PulseNet Algorithm for The impact of inadequately staffed Laboratory Testing of Foodborne Pathogens: laboratories on the foodborne Serotyping and Subtyping surveillance system is tremendous. In 2005, one in three PulseNet laboratories had a vacancy rate of 25% or greater. In Confirm identification of foodborne pathogen 2008, one in two PulseNet laboratories isolated from ill patient, food or environmental had a vacancy rate of 25% or greater. sample As funds continue to decrease, and layoffs of public health personnel Serotyping performed to characterize and work furloughs continue, many the pathogen public health laboratory staff are faced with multiple responsibilities and Simultaneous or non-simultaneous testing priorities that negatively affect their ability to perform molecular subtyping Molecular subtyping and analysis on foodborne pathogens in a timely performed fashion. Additionally, with emerging diseases such as the novel 2009 strain Upload the DNA fingerprint to of Influenza A/H1N1, it is very possible Check for other isolates that match national PulseNet databases that laboratories will encounter staff this DNA fingerprint (clusters) in the local databases shortages due to the need to redirect Check the national databases for efforts to new public health threats. multi-state clusters CONCLUSION PulseNet laboratories at the state and local levels have been and will continue Notify epidemiologists of DNA fingerprint matches or clusters to be an integral part of the nation’s foodborne disease surveillance system. The 2009 APHL survey demonstrated

6 APHL Public Health Laboratory Issues in Brief In 2005, one in three PulseNet and provide tools that will enhance communication between laboratorians laboratories had a vacancy and epidemiologists. rate of 25% or greater. In 2008, one in two PulseNet The challenges faced by PulseNet laboratories can be addressed with laboratories had a vacancy additional resources and commitment rate of 25% or greater. from local, state and federal governments. In 2009, the Council to Improve Foodborne Outbreak Response that public health and agricultural (CIFOR) released guidelines aimed laboratories overall increased the at standardizing practices among all percentage of isolates subtyped, food safety disciplines responsible for identified an increasing amount of foodborne disease surveillance and foodborne disease clusters and have outbreak response.7 Hopefully, with decreased the turnaround time for these guidelines, PulseNet laboratories PFGE testing. However, challenges will continue to improve surveillance to the network include the ability and response to foodborne outbreaks at to timely subtype all foodborne all levels of government. pathogens, hire additional certified staff

References

1. Outbreak of Salmonella Serotype Saintpaul 5. Noller AC, McEllistrem MC, Pacheco AG, Boxrud DJ, Associated with Multiple Raw Produce Items—United States, Harrison LH. Multilocus variable-number tandem repeat 2008. MMWR 2008 Aug 29;57(34):929-934. analysis distinguishes outbreak and sporadic Escherichia coli O157:H7 isolates. Journal of Clinical Microbiology, 2004 2. Multistate Outbreak of Salmonella Infections Associated Feb;42(2):940. with Peanut Butter and Peanut Butter-Containing Products— United States, 2008-2009. MMWR 2009 Feb 6;58(04):85-90. 6. Multistate Outbreak of E. coli O157:H7 Infections Associated with Beef from Fairbanks Farms [Internet]. 2009 3. Multistate Outbreak of E. coli O157:H7 Infections Linked Nov 24 [cited 2009 Dec 29]. Available at www.cdc.gov/ to Eating Raw Refrigerated, Prepackaged Cookie Dough ecoli/2009/index.html. [Internet]. 2009 June 30 [cited 2009 Nov 3]. Available from: http://www.cdc.gov/ecoli/2009/0630.html. 7. Council to Improve Foodborne Outbreak Response (CIFOR). Guidelines for Foodborne Disease Outbreak Response. 4.Hyytia-Trees E, Smole SC, Fields PA, Swaminathan B, Ribot Atlanta: Council of State and Territorial Epidemiologists, 2009. EM. Second generation subtyping: a proposed PulseNet protocol for multiple-locus variable number of tandem repeat analysis of Shiga-toxin producing Escherichia coli O157 (STEC O157). Foodborne Pathogens and Disease, 2006 Spring; 3(1):118-131.

Pulsenet: A Critical Food Safety Surveillance System April 2010 7 Association of Public Health Laboratories Analysis. Answers. Action.

The Association of Public Health Laboratories is a national non-profit located in Silver Spring, MD, that is dedicated to working with members to strengthen governmental laboratories with a public health mandate. By promoting effective programs and public policy, APHL strives to provide public health laboratories with the resources and infrastructure needed to protect the health of US residents and to prevent and control disease globally.

Funders This issue brief was supported in part by Cooperative Agree- ment Number U60/CCU33019 from CDC. Its contents are solely the responsibility of the authors and do not necessarily represent the official views of CDC.

8515 Georgia Avenue, Suite 700 Silver Spring, MD 20910 Phone: 240.485.2745

Fax: 240.485.2700 Cert no. SGSNA-COC-003887 Web: www.aphl.org