National Studies on Acute Gastrointestinal Illness:

1
National Studies on Acute Gastrointestinal
Illness Results to-date and upcoming research
Dr. Shannon Majowicz Public Health Agency of
Canada December 13, 2006
2
Background

• What is the picture of acute gastrointestinal
  illness in the population?
• Traditional surveillance and outbreak
  investigations are subject to bias and
  under-reporting
• In order to more accurately determine whats
  happening in the general community (sporadic
  illness, unreported illness), population-based
  studies are used
• In 2000, the Public Health Agency of Canada
  began the National Studies on Acute
  Gastrointestinal Illness (NSAGI) initiative, to
  determine the burden of acute GI in the
  population
• www.phac-aspc.gc.ca/nsagi-engma/index.html

3
Background
Component 1 BASELINE UNDER-REPORTING
Component 2 COMMUNITY ETIOLOGY ATTRIBUTION
Outcomes, Actions, Interventions, Policy
Component 4 RISK FACTORS HIGH RISK POPULATIONS
Component 3 CHRONIC SEQUALAE
4
NSAGI - Phase 1

• In 2000, four surveys aimed at addressing the
  burden and under-reporting were initiated
• Population survey (pilot)
• Conducted in Hamilton, ON, Feb. 2001 Feb. 2002
• Physician survey (pilot)
• Conducted in Hamilton, ON, Spring 2001
• Laboratory survey
• Conducted Canada-wide, covering the year 2000
• Public health survey
• ? Conducted in Ontario and British Columbia, 2001

5
Results Population Survey Pilot

• 1.3 episodes of acute GI occurred per person-year
  in Hamilton (95 C.I. 1.1 1.4)
• 10 monthly prevalence (95 C.I. 9.94 10.14)
• Prevalence of acute GI was higher in women, and
  those lt10 and 20-24 years of age
• When the data and analyses were standardized, the
  prevalence in Hamilton, Ontario was remarkably
  comparable to the US and Australia for the same
  year

Majowicz et al. 2004 Epidemiol Infect
132607-17 Scallan et al. 2005 Int J Epidemiol
34454-60
6
(No Transcript)
7
Results Population Survey Pilot

• Annual estimates per 100,000 population
• 126,320 cases of acute GI, with
• 37,000 physician visits
• 4,400 ER visits, 30 hospitalizations
• 25,000 cases taking anti-diarrheals
• 23,000 cases and 6,000 caregivers missing
  93,000 days of work
• The total cost of GI was
• 1,089 per case
• 115 per capita

Majowicz et al. 2006 J Food Prot 69651-9
8
Results Physician Survey Pilot

• In Hamilton, 3.4 of patients seen in last 30
  days were diagnosed with acute GI
• Of those, 22.3 were requested to submit a
  stool
• Factors that prompted a stool request
• bloody diarrhea
• immunocompromised status
• occupation
• recent travel overseas
• association with an outbreak
• Stool request practices
• Bacteria 80 always or often request
• Parasites 86 always or often request
• Viruses 18 always or often request

Online www.phac-aspc.gc.ca/nsagi-enmga/phys_e.htm
l
9
Results Laboratory Survey

• In 2000, approx. 3 of stools received across
  Canada were rejected without testing or referral
• Routine testing varied by laboratory
• Canadian laboratories processed 459,982 human
  stool specimens, of which
• 5 positive for bacteria (ex. C. difficile)
• 15 positive for C. difficile
• 8 positive for parasites
• 19 positive for viruses

Flint et al. 2004 Can J Public Health 95309-13
10
Results Public Health Survey

• In Ontario and B.C., local health units
  received reports from laboratories (91),
  physicians (2), or both (7)
• Nearly 80 of health units received additional
  information (e.g. serotype, phagetype) after the
  case had been reported to the province
• 29 sent this additional information on to the
  province
• Approx. 5 of lab-confirmed cases of GI
  reported to the health unit level were not
  reported to the province

Flint et al. 2004 Can J Public Health 95309-13
11
Majowicz et al. 2005 Can J Public Health 96178-81
12
Conclusions from Phase 1

• Epidemiology of GI observed in the pilot
  population appears similar to that observed in
  other developed countries
• GI represented a significant health burden in the
  pilot population, with costs high enough to
  justify investigating prevention and intervention
  efforts
• Notifiable enteric disease data appear to be
  highly under-reported and should be interpreted
  accordingly
• more research needed to have a Canadian picture

13
NSAGI - Phase 2

• In 2002, the population and physician surveys
  were replicated in three health unit areas in
  British Columbia
• 1. Population survey
• Conducted Jun. 2002 Jun. 2003
• Had extensive cost of illness/burden questions
• 2. Physician survey
• Conducted Oct. 2002 Jul. 2003
• Accounted for seasonality via four questionnaires

14
Results B.C. Population Survey

• 1.3 episodes of acute GI occurred per person-year
  in B.C. (95 C.I. 1.1 1.4)
• 9.2 monthly prevalence (95 C.I. 8.4 10.0)
• Adjusting for sex, those 0-9 years (OR1.8) and
  10-14 years (OR1.6) were more likely to
  experience GI than those 25-64 years
• Adjusting for age, females were 1.3 times more
  likely to experience GI than males
• Prevalence varied significantly by region
  (p0.03)
• Vancouver - 8.7
• East Kootenay - 9.6
• Northern Interior region - 11.1

Thomas et al. BMC Public Health (in press)
15
Results B.C. Population Survey

• Of 223 employed adults with GI, 14 (6) were food
  handlers, 2 (1) were day care workers, and 22
  (10) were health care workers
• Relative risk of discontinuing working when ill
  (compared to those working in low-risk employment
  settings)
• Health care workers 1.5
• Day care workers 2.0
• Food handlers 1.0
• Do we need to revisit the effectiveness of
  existing educational material?

Thomas et al. 2006 CCDR 3216
16
Results B.C. Physician Survey

• Overall, 2.5 of patients seen in last 30 days
  were diagnosed with acute GI
• Highest in winter (3.5 95 C.I. 3.3, 3.7)
• Lowest in summer (2.0 95 C.I. 1.9, 2.2)
• Of those, 24.8 were requested to submit a
  stool
• Highest in the summer (34.9)
• Higher in East Kootenay than other regions
  (32.7)
• Factors that prompt a stool request were
  consistent with pilot results

Edge et al. Can J Public Health (in press)
17
NSAGI - Phase 3

• In 2005, the population survey was replicated in
  the province of Ontario
• 1. Population survey
• Conducted May 2005 May 2006
• Incorporated general risk perception questions

18
Other NSAGI Results
19
Estimating Community Rates

• Circa 2000, estimated Canadian annual rates per
  1,000 population
• 0.7 to 3.4 cases of VTEC infection
• (10 to 49 community cases per nationally reported
  case)
• 2.5 to 6.9 cases of salmonellosis
• (13 to 37 community cases per nationally reported
  case)
• 9.3 to 19.7 cases of campylobacteriosis
• (24 to 50 community cases per nationally reported
  case)

Thomas et al. 2006 Can J Infect Dis Med
Microbiol 17229-34
20
Food Consumption Study

• Food items consumed in the past week
• Unpasteurized milk ? 0.7
• Cheese made from unpasteurized milk ? 1.1
• Egg dishes with runny yolk ? 42.3
• Food items containing raw egg ? 5.9
• Meat pâté ? 5.6
• Raw fish (i.e. sushi) ? 7.9
• Chicken ? 91.7
• Chicken nuggets/strips 19.2
• Pork ? 60.2
• Pink/undercooked pork 3.3

Nesbitt 2006 M.Sc. Thesis (U. Guelph)
21
Food Consumption Study

• Food items consumed in the past week
• Raspberries ? 7.6
• Unpasteurized juices ? 6.2
• Bean sprouts ? 9.3
• Alfalfa sprouts ? 3.4
• Fresh basil ? 11.6
• Raw nuts ? 36.3

Nesbitt 2006 M.Sc. Thesis (U. Guelph)
22
Food Consumption Study

• Respondents ate an average of 3 meals outside the
  home a week
• Location of meal consumption
• Pizza or donut shop ? 43
• Sit down restaurant ? 36
• Fast food chain ? 34
• Salad bar ? 19
• Buffet or cafeteria ? 14
• Catered event ? 9

Nesbitt 2006 M.Sc. Thesis (U. Guelph)
23
Food Consumption Study

• Source of food safety knowledge
• Family/friend ? 75.1
• Television/news ? 18.6
• School/home economics ? 15.1
• Public health/primary health care ? 1.16
• Cooking practices
• How do you know when meat is cooked enough to
  eat?
• Visually ? 63.0
• Time ? 32.7
• Thermometer ? 13.6
• Taste ? 9.7

Nesbitt 2006 M.Sc. Thesis (U. Guelph)
24
Water Consumption Study

• Bottled water
• Primary drinking water source for 27 of
  respondents
• Not consumed by 60 of respondents
• Consumption was associated with age (increasing
  to
• age 31, then decreasing)
• Was more likely to be consumed on weekends than
  
• weekdays

Jones et al. 2006 J Water Health 4125-38
25
Upcoming Research
26
Upcoming Research

• Quebec Population Survey
• Modified population survey (n7,000) will be
  conducted
• in select Quebec communities in 2007
• Will evaluate the role of agricultural risk
  factors and GI,
• as well as the impact of climate variables

27
Upcoming Research

• Atlantic Canada Population Survey
• Pending funding by CIHR, population survey will
  be
• conducted in the four Atlantic provinces in
  2007/2008
• (n6,000)
• Survey will include questions on social
  determinants of
• acute GI

28
Upcoming Research

• Community Etiology Study
• Prospective, community-based study to measure
• pathogen-specific incidences in the community
• Still determining
• Study population (Canada-wide?)
• Pathogen list
• Anticipated start of pilot Fall 2007

29
Upcoming Research

• Burden of Foodborne Disease Study
• Objective is to determine the burden (in
  DALYs) of
• foodborne disease in Canada
• - still to define foodborne
• - will use incidence data from the Community
• Etiology Study
• Will also feed into the WHOs Global Burden of
  
• Foodborne Disease project

30
Questions?
NSAGI Contact for 2007 Kate Thomas kate_thomas_at_
phac-aspc.gc.ca