Introduction to Epidemiology
Gastroenteritis in Texas: an outbreak case-study
October 26, 2009
Dr. N. Birkett,
Department of Epidemiology & Community
University of Ottawa
On the morning of March 11, the Texas
Department of Health (TDH) got a phone call from
a university student. He reported that he and his
roommate (who lived in a frat house) were
suffering from nausea, vomiting and diarrhea.
Both became ill during the night. Neither had seen
They believed they were ill due to food eaten at a
local pizzeria the previous night. They wanted
advice about whether to take a biology mid-term
Part 1 (Q’s)
• What questions would you ask the
• What advice would you provide?
• Do you think that this complaint should be
Part 2 (1)
Staff were skeptical, thinking students
wanted an excuse to miss the exam but
decided to make a few calls.
•Pizzeria was closed until 1100,
•No answer at the student health centre
Part 2 (2)
• ER at hospital A revealed 23 university students
had been seen for gastro in past 24 hours.
– Only 3 similar patients seen from March 5-9, none
• 10:30, MD from Student Health Center called.
20 cases of students with vomiting and diarrhea
seen in previous day.
– Normally, see only 1-2 similar patients per week.
Is this an outbreak at the university?
TDH visit hospital A that afternoon to review medical
Abdo cramps: 68%
Muscle aches: 49%
Bloody diarrhea: 5%
Temperature: 98.8 to 102.4 (median 100)
CBC: elevated white cell count
Stool specimens: pending
Part 3 (Q’s)
Broad categories of diseases to consider?
• Requires organism growth
– Incubation period days, not hours
• Symptoms: diarrhea, nausea, vomiting, cramps. Can get
fever and raised white count
• Toxin already in food disease in minutes to hours
• Symptoms: vomiting (**). Can include neurological
Pizza as source toxin.
• Local hospitals asked to report any cases of
vomiting/diarrhea seen in past week. Asked to
collect stool samples on new cases.
• Preliminary cultures on 17 ill students were
negative for 9 common bacterial causes of
– Salmonella, shigella, E. Coli and Listeria
What do the negative stool culture results mean?
Within 24 hrs, 75 persons with vomiting or
diarrhea had been reported to the TDH. All
were students from the university campus.
There were no cases among faculty, staff or
the local community. Median age was 19,
69% were freshmen and 62% were female.
Part 6 (Q’s)
• Is there anything else you would like to
see from this preliminary data?
– An epidemic curve!
Epidemic curve, based on 75 cases
16/10/08 Initial 11
Part 6 (Q’s)
• Is there anything else you would like to see from
this preliminary data?
– An epidemic curve!
• How do you interpret the curve?
• What about the case on March 6?
• Next steps?
– Contact university
• What would need to be discussed?
– Get more info
• Case series interviews to generate hypotheses
– Does the pizzeria hold up as a source?
Part 7 (1)
• University is in small town (pop=27,354).
Enrollment about 12,000. 2,386 live on
campus in 36 residences.
• University has municipal water and
– No breaks or work on water/sewage lines in
– No recent road work or digging near campus
Part 7 (2)
• Two cafeterias on campus (run by same
company) plus 6 fast-food places
– 2000 students on food plan
– Most on-campus students eat in main cafeteria
• Hot entrees
• Grill, deli bar and salad bar
– Second cafeteria used by off-campus students and
• Hot entrees, grill and salad bar
• No deli bar.
• Spring break starts in two days (March 13).
Cafeteria will be closed for 10 days.
Part 7 (3)
• Interview of 7 early cases revealed:
• 4 M; 3F. 6 freshmen;
– 2 psych major,1 English/Animal husbandry major; 3
• 5 different residence halls. All ate mainly in
main cafeteria. All had eaten from deli bar; 2
from salad bar and 3 from grill. No common
food items were noted
• Only psych students shared any classes. 5 were
in sorority/fraternity; 3 attended an ‘all school’
mixer on March 6; 2 went to a Sci Fi film festival
on March 7.
Part 7 (Q’s)
• State your leading hypothesis on pathogen,
mode of transmission, source and period of
– Viral infection
– Food or beverage spread
– Served at the main university cafeteria between
March 5 and 10
– Perhaps related to the Deli bar
• What actions would you take next?
– Epi study with controls
– Lab examination of foods (including ice)
– Inspect operational methods; interview staff
Part 8 (environmental - 1)
• 31 staff work in cafeteria, 24 are foodhandlers.
– 30 interviews. One person refused to participate (worked at the deli
• Asked about responsibilities, which meals they served and where,
use of gloves, hand washing practices, work schedule in key week
and illnesses. They provided stool samples.
• No illnesses in staff.
• Deli Bar
– Separate refrigerator and prep area.
– Sandwiches made to order by foodhandler
– New meat, cheese and condiments were added to partially depleted
stock daily (leftover food items were not discarded first)
– Food items were not refrigerated or kept on ice while deli was open
– Containers were not routinely cleaned
• Deli Bar was closed by the Health Department on March 12.
– Do you agree with this action?
Part 9 (epi)
• Need more information, including a control
• Matched case-control study (done on March 12)
– Case definition
• Vomiting or diarrhea (3+ loose stools in a 24 hr period) with
onset after March 5 in a university student seen at ER or the
health centre. Had to be in dormitory during survey
• Roommates of cases who did not become ill
– Telephone interviews (why?)
– 29 pairs of subjects
Part 9 (epi – results)
Epidemic curve, based on 75 cases
16/10/08 Exposure Initial 20
• By March 13, the outbreak had grown to include 125
patients. CDC was invited to assist. They performed
RT-PRC testing of stool samples for viruses.
• Unmatched case-control study done to explore source of
– Limited to students eating at main cafeteria
– Case definition modified
• Had to be member of university meal plan (need not be in
– Controls were in meal plan but had not got ill.
– 40 randomly selected cases
– 160 controls
– Telephone interviews
Final results (1)
• Water/ice negative for coliforms
• All stool cultures negative for bacteria
• 50% of fresh stool cultures on ill students were positive for Norwalk-
• Ham samples were positive for Norwalk-like virus (same DNA
sequence as stool samples)
• Staff member who refused interview agree to provide information
– She sliced ham for three meals on March 9/10
– Prepared sandwiches on those days
– Wore gloves. Therefore, believed that hand washing didn’t matter.
– She wasn’t ill
– BUT her infant had watery diarrhea from March 7 to 10.
– The same Norwalk-like virus was found in her child’s stool
– Transmitted by contamination, not through infection
• What needs to be done to re-open deli
– Throw away all leftover deli bar food
– Clean and disinfect all equipment and
surfaces in deli bar
– Require foodhandlers to submit stool sample
prior to return to work? NOT NEEDED
– Educate foodhandlers on proper procedures
– Develop a sick foodhandler policy