PUTTING TOGETHER THE PULSENET PUZZLE · of lab reports, individual case interviews, ... l Report findings. Putting Together the PulseNet Puzzle lHow does PFGE and PulseNet help the
Post on 29-Mar-2020
0 Views
Preview:
Transcript
PUTTING TOGETHER PUTTING TOGETHER THE PULSENET THE PULSENET PUZZLEPUZZLE
Roshan Reporter MD, MPHAcute Communicable Disease
Control ProgramLos Angeles County Department
of Health Services
How are outbreaks How are outbreaks detected?detected?lBy collecting information and
analyzing it.– Recognized and reported by individuals,
doctors, emergency depts, media, ...– Detected by PH agency through review
of lab reports, individual case interviews, – Enhanced surveillance and notification
by federal, state and local health officials
Steps in outbreak Steps in outbreak investigationinvestigationl Detect problem by public health surveillance
l Verify diagnosis
l Confirm outbreak
l Identify / count cases
l Characterize data → time / place / person
l Take immediate control measures
l Formulate / test hypotheses
l Implement / evaluate additional control measures
l Report findings
Putting Together the PulseNet Putting Together the PulseNet PuzzlePuzzle
l How does PFGE and PulseNet help the epidemiologist?– Determine if an increase in cases by noted by
routine surveillance are related– Detect a cluster of related isolates not found
on routine surveillance– Help link an already known outbreak to other
outbreaks or a source– Help link sporadic cases from widespread area
as an outbreak
Long Local Report Local Immediate RangeSurveillance Evaluation Investigation Prevention PreventionPhase Phase Phase Phase Phase
HACCP
CentralReporting
ClinicalSpecimens
Maintainas appropriate
EPI outbreakInvestigation
Test
Ho
IDagent
ID vehicle
Developquestionnaire
ReportEvaluateAbstractPublish
TracebackFood item
Food Specimens
EHMaintainSurveillance
PMD
ER
Confirm detailof report
Evaluate by:--Previous Reports--Reportable diseases--Severity--PMD Report--Commercial food source--Incubation Symptom--High Risk Food
Victim
Case Investigation
MaintainSurveillanceEPI
Maintain
AssemblyTeam
communications
Analyze
ImmediateEH
PreventiveMeasures
ImmediateEPI
PreventiveMeasures
EPI-EH
COMMUNI-CATIONS
LabTesting
ExpandCommuni-
cations
NotifyOther JurisdictionsState & Federal,EPI, EH, LAB
EPI-EH-LABCommunications/
Data Sharing Ho GeneratingInterviews
Decide EpiMethodology
ObtainQuestionnaireData
DataManagement
EH DeterminedOutbreak
EPIDetermined
Outbreak
ACD / MT 3/00
ICP
Salmonella Thompson: Buns to Salmonella Thompson: Buns to RunsRuns
Los Angeles County Department of Health ServicesAcute Communicable Disease Control
CA Department of Health ServicesDivision of Communicable Disease Control
Southern California RegionalEpidemiologic Support Team
Tenant Appreciation DayTenant Appreciation Day
l July 14, 2000l Catered Luncheonl Office Building Complex in Whittierl All 33 Offices Invitedl Two Buffet Tablesl One Common Grill
Whittier
Los Angeles County
BackgroundBackground
l Initial Report received July 20, 2000
l 300 attended luncheon
l 27 illnesses
l Symptoms: diarrhea, vomiting
l Catered by Company C
l Some victims went to the ER/hospital
l One positive Salmonella case
MethodsMethods
l Questionnaires developed and distributed
l Site visit to office building complex
l Follow-up on those who did not return questionnaires
l Stool specimens collected on ill persons
– including all in sensitive occupations
l Data analyzed with Epi-Info and SAS
l Inspected caterer’s facility
Case DefinitionCase Definition
l A person associated with the luncheon -- and -- one of the following:
l Culture positive for Salmonella serotype Thompson -- or --
– Diarrhea with fever -- or --
– Diarrhea with 2 or more other symptoms (nausea, vomiting, abdominal cramps, fatigue, headaches, or body aches)
Results Results -- CasesCases
l Of 250, 202 questionnaires completed (81%)
l 78 attendees reported illness
l 47 attendees met case definition
– Attack rate = 19%
l 12 lab confirmed cases
l All 12 had indistinguishable PFGE pattern
l Age range 15 - 75 years
l 70% female
PFGE ResultsPFGE Results--ST OutbreakST Outbreak
Epidemic CurveEpidemic Curve
0
2
4
6
8
10
12
14
7/14am
7/14pm
7/15am
7/15pm
7/16am
7/16pm
7/17am
7/17pm
7/18am
7/18pm
7/19am
Onset by Half Day Intervals
Num
ber
Cas
es
Probable Cases(n=32)Lab ConfirmedCases (n=12)
Lunch
Suspect ItemsSuspect Items
RR=1.4, CI=0.8-2.362% cases
RR=1.8, CI=1.02-3.070% cases
RR=2.0, CI=1.2-3.662% cases
HamburgerTable A
Lemonade
Suspect ItemsSuspect Items
100% cases ate Hamburger or Chicken Burger, but 96% of controls did also
(RR = undefined; p = 0.5)
SummarySummary
l No single food implicated by analysis
l Suspect sources: – Hamburgers – Lemonade – Hamburgers and/or Chicken Burgers
l Suspect something about Table A
DiscussionDiscussion
l Which food item caused illness?
l Was this outbreak limited to Tenant Appreciation Day?– Caterer screened for illness at other
eventsl Part of bigger outbreak of Salmonella
Thompson in Southern California and Arizona investigated by State DHS
BackgroundBackgroundl 7/25/00: Los Angeles County (LAC)
reports ST outbreak l 8/00: ST reported from other counties
– Orange County (3)• 2 (67%) cases ate burgers at Chain A
– San Diego County (7)• 4 (57%) cases ate burgers at Chain A
– OC and SD had 1 case total in 7/9– LAC had “sporadic” cases as well (12)– Had same PFGE pattern as LAC outbreak
812
12
11 3
1 3
Arizona
California
S.S. Thompson “Sporadic” Case Thompson “Sporadic” Case Distribution by County/State (N = 50)Distribution by County/State (N = 50)
San Bern.
Riverside
Ventura
Orange
San Diego
Los Angeles
Confirmed Cases eligible for caseConfirmed Cases eligible for case--control control study, CA and AZ, July 2000 (N = 23)study, CA and AZ, July 2000 (N = 23)
0
1
2
3
4
5
6
7
13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30
onset date
# ca
ses
July
Chain A also located in No. CA, CO, NV, and TX, but Chain A associated cases only from So. CA and AZ
CaseCase--Control StudyControl Study
OR undefinedp < 0.00000001
No
Clinically ill ?Yes No
Ate Burger ? 26
Yes 23
0 26
4 27
23 30
The only food item at Chain A associated with illness wasa burger item; hamburger (19), turkey/chicken burger (3), and
unknown type of burger (1)
(cases) (controls)
# #
# #
# #
# #
The SuspectsThe Suspects
?
It was not clear what component of the burger caused illness.Based on prior experience with Salmonella outbreaks, it was thought to be either the beef or poultry patty.
Results: The Lucky Break Results: The Lucky Break
l Earliest potential case patient AF– onset of illness 7/13– did NOT eat at Chain A– employee of Bakery B
l Bakery B– largest customer is Chain A restaurants in
S. Cal and AZ; provides buns and rolls– also supplies burger buns to Company C– provides buns to few other restaurants
Case History: Patient AFCase History: Patient AFl AF responsibilities:
– remove baked burger buns from rack– feed buns through slicer– package buns for distribution
l Clinical course– worked while ill, no gloves– illness severe enough to require
hospitalization overnight on 7/17 (after working shift)
– resumed work until terminated on 7/23– brother also Bakery B employee, ill 7/17
S. Thompson by Date of Onset, CA and AZ, S. Thompson by Date of Onset, CA and AZ, July 2000 (n = 40)July 2000 (n = 40)
0
2
4
6
8
10
12
13 15 17 19 21 23 25 27 29
Onset
# o
f C
ases
Other rest (3)
Chain A (23)
Caterer C (12)
AF's brother
Case AF
July
Period AF worked at Bakery B while symptomatic
ResultsResults-- LaboratoryLaboratory
l Isolates from case AF, the catered lunch outbreak and from restaurant A patrons all were Salmonella serotype Thompson
l All had indistiguishable PFGE pattern
l In addition, other cases with the outbreak pattern had eaten sandwiches on a bun at other So Cal restaurants supplied by Bakery B.
ConclusionsConclusions
l Outbreak of Salmonella Thompson due to an unexpected vehicle (burger buns)
l Illustrates the potential for ill food handlers to be the source of wide-spread outbreaks
l Underscores the need for appropriate education, hygiene, and need for sick leave policies that do not penalize food handlers who stay away from work when ill with contagious diseases
ConclusionsConclusions
l The use of PFGE helped to identify sporadic cases and a point source outbreak that were part of a larger outbreak
AN OUTBREAK OF AN OUTBREAK OF SALMONELLA POONA SALMONELLA POONA DUE TO CANTELOUPEDUE TO CANTELOUPE
AN OUTBREAK OF AN OUTBREAK OF SALMONELLA SALMONELLA POONAPOONA BACKGROUNDBACKGROUND
l S. Poona usually only about 1% of all Salmonella serotypes
l During March-May 2002, an increase occurred
lWidely distributed throughout US and Canada
AN OUTBREAK OF SALMONELLA AN OUTBREAK OF SALMONELLA POONA METHODS: SURVEILLANCEPOONA METHODS: SURVEILLANCE
l Surveillance for S. Poona cases in California and other states
l All cases of S. Poona were reviewed for demographic and medical factors
l Pulsed field gel electrophoresis (PFGE) was used to detect if isolates were related via PulseNet
AN OUTBREAK OF SALMONELLA POONA AN OUTBREAK OF SALMONELLA POONA METHODS: CASE CONTROL STUDYMETHODS: CASE CONTROL STUDY
l Cases were defined as persons > 2 years old with positive culture for S. Poona with the outbreak PFGE pattern and onset during March 15- May 3, 2002
l Controls matched by age and neighborhood
l In- depth interviews were done to generate hypotheses
l Cases and controls were interviewed about potential source items and food handling
AN OUTBREAK OF SALMONELLA AN OUTBREAK OF SALMONELLA POONA METHODS: TRACEBACKPOONA METHODS: TRACEBACK
l All patients with the outbreak strain asked where they purchased canteloupe in the week prior to onset
l Invoices were traced to distributorsl Distributors were traced to packers
AN OUTBREAK OF SALMONELLA AN OUTBREAK OF SALMONELLA POONA RESULTS: SURVEILLANCEPOONA RESULTS: SURVEILLANCE
l 58 S. Poona strains were identical by PFGE, all from March 30-May 31
l 36% of cases resided in California
lMedian age was 6 years(range 4 months-91 years)
l 10 cases were hospitalized, none died
AN OUTBREAK OF SALMONELLA POONA AN OUTBREAK OF SALMONELLA POONA RESULTS: CASE CONTROL STUDYRESULTS: CASE CONTROL STUDY
l There were 27cases and 54 controls lOnly significant food item was
canteloupe: 20/26 or 74% recalled eating; 11/54 (20%) of controls
MOR = 15.5 (95% CI 3.3, 125)
l Cases more likely to eat canteloupepurchased whole
MOR = 5.8 (95% CI 1.6, 23.3)
AN OUTBREAK OF SALMONELLA AN OUTBREAK OF SALMONELLA POONA RESULTS: TRACEBACKPOONA RESULTS: TRACEBACK
l 20/26 persons with S. outbreak strain recalled eating canteloupe
l Point of sale sources of canteloupewere traced back to shippers, then to farms in Mexico
lOutbreak ended with beginning of California canteloupe season
AN OUTBREAK OF AN OUTBREAK OF SALMONELLASALMONELLAPOONA CONCLUSIONS (1)POONA CONCLUSIONS (1)
lOutbreak detected because of an unusual Salmonella serotype with an unusual vehicle
l PFGE used to identify casesl Probable low level contaminationl Higher percentage recalled eating
canteloupe (74%) than in other canteloupe outbreaks (44-62%)
AN OUTBREAK OF AN OUTBREAK OF SALMONELLASALMONELLAPOONA CONCLUSIONS (2)POONA CONCLUSIONS (2)
lOutbreak ended when distributors shifted to California grown canteloupes in May
l Canteloupes eaten by cases were traced back to a common source
l Canteloupe surface was found to have Salmonella contamination in 1990 outbreak associated with Mexican canteloupe
AN OUTBREAK OF AN OUTBREAK OF SALMONELLASALMONELLAPOONA RECOMMENDATIONS (1)POONA RECOMMENDATIONS (1)
l US melon industry should continue melon quality program, Mexico should initiate
l Need to examine HACCP program for fresh cut produce in US, Mexico
l Investigate farms implicated in outbreaks
AN OUTBREAK OF AN OUTBREAK OF SALMONELLASALMONELLAPOONA RECOMMENDATIONS (2)POONA RECOMMENDATIONS (2)
l Follow FDA recommendations for handling melons
l Consider irradiation or cold sterilization of fresh produce
l The consumer and retailer should be aware of potential risks and wash melons before cutting
Acute Communicable Disease ControlLos Angeles County
A foodborne outbreak of febrile gastroenteritis due to
Listeria monocytogenes
BackgroundBackground
l Sunday, 6/3/01 – Catered birthday party – Of approximately 60 individuals attending, 28
reported acute febrile gastroenteritis
l Acute Communicable Disease Control notified 6/6/01– No etiologic agent had been identified.– Stool cultures were performed in Public Health
Laboratory.
MethodsMethods
l Cohort study of party attendees– Interviewed by telephone with
standardized questionnaire• Food consumption• Symptom and illness history• Underlying conditions
l Employees– Interviewed for job duties and illness
l Caterer inspection
Case DefinitionsCase Definitions
Confirmed Case:A party attendee with culture-confirmed outbreak strain of Listeria monocytogenes
Probable Case:A party attendee with at least one symptom in each categoryvSystemic (fever, body ache or headache) vGastrointestinal (diarrhea, vomiting, cramps)
Methods: Case findingMethods: Case finding
l Letter mailed to caterer’s clients receiving risky food within 2 weeks of party– Follow-up telephone calls made to
ensure receipt of lettersl Letter faxed to hospital infection
control departmentsl Enhanced surveillance for listeriosis,
using supplemental reporting form
Methods: Lab/EnvironmentalMethods: Lab/Environmental
l Caterer inspectedl Environmental swabs takenl Food samples tested for enteric pathogensl Stool specimens tested for enteric pathogens
Party attendees ill, non-ill– Employees– Other symptomatic clients– PFGE performed on isolates
Results: Party attendeesResults: Party attendees
l 60 attendees, 44 interviewed – 73% Response Rate
l 16 met case definition– 36% Attack Rate
l Gender: 52% femalel Age: median 15.5 years (range 8 - 72)l Incubation: mean 21 hours (range 6 - 35)l Duration: mean 3 days (range 1 - 4 days)l Impact: none hospitalized, no deaths.
Results: Symptoms (n = 15)Results: Symptoms (n = 15)
l Body ache 14 (93%)l Fever 13 (81%)
– (mean temp = 102.4º F)
l Headache 13 (81%)l Diarrhea 10 (63%)l Vomiting 9 (56%)
Epidemic Curve, Birthday PartyEpidemic Curve, Birthday PartyJune 3, 2001 (n=14)June 3, 2001 (n=14)
012345678
9a-4p 5p-12a 12a-8a 9a-4p 5p-12a 12a-8a
Party
6/3 6/4
Food Epidemiology ResultsFood Epidemiology Results
Food Items RR p-valueSandwich A 5.6 <0.002
Turkey * undef. 0.000004Jack cheese 6 0.0002
Sandwich B 0.1 <0.01Sandwich C 0.2 <0.01
* Significant after controlling for jack cheese
Deli KitchenDeli Kitchen
Market InspectionMarket Inspection
lGeneral lack of cleanlinessl Inadequate sanitizationl Hazardous storagel Repeated temperature violations
– walk-in refrigerator at 51ºF
l Upon re-inspection, all violations corrected
Lab: Stool Specimen ResultsLab: Stool Specimen Results
l 9 party attendee stool specimens– All negative for Salmonella, Shigella,
Campylobacter, Yersinia, and toxigenic E. coli
– 6 party attendee specimens positive for Listeria monocytogenes with indistinguishable PFGE patterns
Lab: Food Specimen ResultsLab: Food Specimen Results
l Leftover Sandwich A – Turkey– Pepper jack cheese– Different parts of Sandwich A
l All samples positive for Lml Leftover turkey sample yielded Lm at 109
organisms per graml All food samples from deli were negative
More Laboratory ResultsMore Laboratory Results
PFGEl All isolates from food and stool
specimens had indistinguishable PFGE patterns
l No similar outbreaks or clusters with the outbreak pattern
ENVIRONMENTALl All 30 swabs of caterer’s kitchen all
negative for Lm
Results: Employee informationResults: Employee information
l 9 of 13 employees completed questionnaires– No employees reported illness at time of
preparation– 7 stool specimens collected– Only 2 stool specimens tested
• Sandwich maker and one currently ill employee
• Both negative for Lm
LimitationsLimitations
l Delay in obtaining stool specimens
l Use of antibiotics prior to definitive diagnosis
l Mechanism of turkey contamination unknown
l No lot numbers to do traceback
ConclusionConclusion
l Turkey was most likely vehicle of transmission with high infective dose.
l This is only the third report of febrile gastroenteritis in immunocompetent individuals due to Lm in the U.S. and first to implicate a deli product.
l Public health action taken– EHS made sure violations corrected– ACDC sent out letters to those even
remotely exposed.
ConclusionsConclusions--Solving the PuzzleSolving the Puzzlel By strain typing with PFGE, and
sharing the data through CDC’s database, PulseNet, sporadic cases or local outbreaks can be linked to more widespread cases.
l The larger number of cases in different settings often makes it easier to find a common source.
MultiMulti--State State ShigellaShigellaoutbreak associated with a outbreak associated with a bean dipbean dip
BackgroundBackground
l January 18, 2000– Individual calls SD HD to report illness
after consumption of 5 Layer dip purchased at Trader Joe’s
l January 20,2000– Shigella isolated from patient who
reports eating dip prior to illness
l January 21, 2000– SDHD notified of 2 hospitalized patients
BackgroundBackground
l January 21, 2000– SCREST notified– Washington reports Shigella outbreak at
a baby shower implicating dip– CA DHS issues a press release– Voluntary recall begun– CDC and FDA notified
Outline of InvestigationOutline of Investigation
l Determine extent of outbreak– Counties notified through phone calls
and through the CD Brief– Other states through calls and MMWR
l Determine burden of illness– Cohort study
l Identify mechanism of contamination– Environmental investigation
MethodsMethods
l California counties asked to use a standardized interview form
l Interview form used on:– all persons with culture-confirmed
Shigella sonnei infection– anyone reporting diarrheal illness after
consuming a Brand X product– Susbmit isolates for PFGE
Results: Nationwide numbers Results: Nationwide numbers N= 409 confirmedN= 409 confirmedlOregon: 31 culture-confirmed caseslWashington: 132 culture-confirmed
casesl California:
– 217 culture-confirmed cases– 123 probable cases
l 29 culture-confirmed cases from other states
ResultsResults-- CaliforniaCalifornia
l All reports of illness following consumption of bean dip
l 217 culture-confirmed– Onset from 1/8-2/2/2000– Majority of cases from LA (69) and SD
(52), and Orange Counties (34) – 65% female– Median age is 35 years– 87%Caucasian
Description of Description of Shigella Shigella casescases
l Symptoms:– Vomiting: 51%– Cramps: 96%– Fever: 92% (102F)– Diarrhea 100%– Illness Severity:
• Seen by MD: 97%• Antibiotics: 87%• Hospitalized 6%
Environmental investigationEnvironmental investigation
l Parent company is Brand X– Los Angeles, CA
l 5 Layer Bean Dip– black beans, fresh salsa, fresh
guacamole, nacho cheese, and sour cream, with cheese garnish
l Distributed nationwide– Trader Joe’s– Costco
Plant inspectionPlant inspection
lOld plant, in disrepairl Poor record-keeping practicesl Salsa and guacamole used in the
five-layered dip is also packaged and sold separately
l Employees responsible for the dip do not have duties elsewhere
Environmental investigationEnvironmental investigation
l Cultures taken: – of finished product, and raw ingredients
(FDA/FDB lab)– leftover products (FDA/FDB)– all members of the Fiesta Line (LAC)– all SF employees (OH)
l Problem:– infectious dose is very low, but need
very heavy contamination to isolate from food
Environmental investigationEnvironmental investigation
lOne employee reported diarrhealillness on 1/3– illness lasted one day– no antibiotics given– prepared the cheese layer for the dip
l Stool samples from ill employee and other members of the dip line negative
Results Results --EnvironmentalEnvironmental
l Product distributed nationwidel Plant had numerous violations, cultures
taken at the plant grew coliforms, but no pathogens
l Shigella sonnei isolated from open and unopen container of dip by Seattle FDA lab– cheese garnish layer?– Nacho cheese layer?– fresh cilantro?
Number of cases by date of Number of cases by date of exposure, CA 2000exposure, CA 2000
0123456789
10
1/7
1/8
1/9
1/10
1/11
1/12
1/13
1/14
1/15
1/16
1/17
1/18
1/19
1/20
1/21
1/22
ResultsResults-- PFGEPFGE
l Isolates from cases nationwide had indistinguishable PFGE “pattern A”
l The isolates from the bean dip also had the outbreak “pattern A”
l This same pattern was identified in a 1998 outbreak of shigellosis due to contaminated parsley which was traced back to a farm in Mexico
l Common pattern in Southern Californial Resistant to Ampicillin and TMP-SMX
ConclusionsConclusions
l A large multi-state outbreak of shigellosis occurred due to contaminated 5-layer bean dip
l Source of contamination may have been an ill foodhandler or a contaminated produce item
l PFGE helped define the extent of the outbreak and confirm the 5-layer dip as the vehicle for the outbreak
ConclusionsConclusions--Solving the PuzzleSolving the Puzzlel By strain typing with PFGE, and
sharing the data through CDC’s database, PulseNet, sporadic cases or local outbreaks can be linked to more widespread cases.
l The larger number of cases in different settings often makes it easier to find a common source.
AcknowledgementsAcknowledgements--ListeriosisListeriosis OBOB
l ACDC: Irene Lee, Rachael Zweig, Douglas Frye, Michael Tormey, Laurene Mascola
l LAC Public Health Laboratory: Joan Sturgeon, Leonard Lawani, Sydney Harvey
l Environmental Health Services: Michelle LeCavalier
AcknowledgementsAcknowledgements--Buns to RunsBuns to Runs
•L.A. County ACDC: Rachelle Velasco, Trina Pate, Michael Tormey, Rachael Zweig, David Cardenas, and Doug Frye
•Whittier District Public Health Nurses
•CDHS: Akiko Kimura and Jeff Higa
•Orange County Communicable Disease Control and Epidemiology: Mike Carson and Steve Klish
AcknowledgmentsAcknowledgments--Buns to RunsBuns to Runs
l City of Pasadena HDl Riverside County HD l San Bernardino County HDl San Diego County HD l Ventura County HDl Arizona HDl Food and Drug Branch Emergency Response
Teaml Microbial Disease Laboratoryl Centers for Disease Control and Prevention
top related