PH6998 Rapid Epidemiologic Assessment

1
PH6998Rapid EpidemiologicAssessment

• Stephen Waring, DVM, PhD
• Associate Director of Research
• Center for Biosecurity and Public Health
  Preparedness
• Assistant Professor of Epidemiology and
  Biological Sciences

2
Rapid Epidemiologic Assessment (REA)

• Origins from the Expanded Programme on
  Immunization (EPI)
• WHO methodology to assess immunization coverage
  for childhood vaccine-preventable diseases
• diphtheria, pertussis, tetanus, measles,
  poliomyelitis, and tuberculosis
• Effective, simple, and inexpensive

3
Rapid Epidemiologic Assessment (REA)

• Cluster sampling methodology
• Usually not possible to survey every child in
  target age range
• Simple random sampling not feasible since
  accurate lists of all individuals in target
  population not readily available or easily
  constructed
• Random selection of 30 groups of 7 children each
  (n210)

4
Rapid Epidemiologic Assessment (REA)

• Applications of REA methodology
• Cause-specific morbidity/mortality in selected
  populations
• Health service coverage
• Health service needs
• Rapid needs assessments following disasters
• hurricanes, tornadoes, blizzards, floods, ice
  storms, earthquakes, hazardous material spills

5
Rapid Epidemiologic Assessment (REA)

• References
• Henderson RH, Sundaresan T. Cluster sampling to
  assess immunization coverage a review of
  experience with a simplified sampling method.
  Bulletin of the World Health Organization.
  198260(2)253-260.
• International Journal of Epidemiology. 198918
  (Supplement 2)S1-S67.

6
Rapid needs assessment followingTropical Storm
Allison
7
Background

• Date of storm June 8-10, 2001
• Date of surveillance June 16, 2001
• Hardest hit areas were based on the information
  supplied by City of Houston Emergency Management
  Division

8
Background
9
Objective

• To obtain estimates of need for areas hardest hit
  by flooding from Tropical Storm Allison
• To provide a template for rapid needs assessment
  in future disasters

10
Sampling Methodology

• Modified cluster sampling
• n x k design
• of clusters (n) 30
• of respondents per cluster (k) 7
• Selection of clusters
• 168 tracts
• Sampling area A - 96 tracts in NE quadrant
• 148,654 Housing Units
• Sampling area B - 72 tracts in Western half
• 167,158 Housing Units
• 30 tracts from each area selected by random
  sampling procedure in SPSS (V. 10)

11
(No Transcript)
12
Demographics
13
Sampling Methodology

• Selection of area within cluster
• Household estimates for 1997 at the block group
  level were used to determine household density
  within a census tract
• Based on quintiles of household distribution,
  areas with the highest density of household units
  and streets were selected

14
Survey Methodology

• 10 teams of 7 volunteers each
• 1 team leader
• 3 interviewer pairs
• 60 sampling areas
• each team responsible for 6 tracts (2 per pair)
• Every 7th household
• If no response, consecutive selection until a
  completed response, then resume every
• Goal 7 completed responses per area

15
Survey Methodology

• Questionnaire
• One page instrument designed after samples used
  in other disaster settings
• Pilot tested among workers at HDHHS
• Interviewer training session held on Friday
  afternoon, and again Saturday morning

16
(No Transcript)
17
Survey Methodology

• Maps
• A block map of the sampling area with
  pre-determined starting point (NW corner) plus
  Key Map of area provided to team leaders and team
  members for the areas they were responsible for
• Team leaders were instructed on methodology
• Sampling areas assigned based on proximity

18
Results Household damage by area
One home was reported destroyed the rest were
reported as repairable
19
(No Transcript)
20
Results Utility disruption by area
Group A NE quadrant Group B Western half
21
Results Illness
Majority were heart-related or rash Group A
NE quadrant Group B Western half
22
Health effect of water in house among households
reporting injury or illness
Group A NE quadrant Group B Western half
23
Health effects of flooding among all households

24
Association of flooding (water in house) and
illness or injury

25
Percent of non-responders
Overall response rate 59.3
26
Immediate needs and actions

• Assessment was planned, conducted, and analyzed
  within 1 week
• Results were integral in
• assessing damage
• prioritizing service delivery
• directing assistance efforts

27
Immediate needs and actions

• Provided timely information to
• Houston City Mayors Office
• City Council
• City Office of Emergency Management
• American Red Cross
• Federal Emergency Management Agency
• CDCs Emergency Response team
• Texas Department of Health
• Other agencies assisting with flood recovery
  efforts

28
Immediate needs and actions

• numerous persons with special needs were
  identified and the appropriate assistance
  agencies quickly notified
• heavy trash pick-up crews were redirected to
  neighborhoods where most needed
• residents of an area with chronic drainage
  problems were put in contact with the city
  engineering department to work out plans for
  future improvements

29
Long range planning

• Results serve as a guide for
• future needs assessments
• disaster-response planning
• alleviation of disaster-effects

30
Acknowledgements

• HDHHS, TDH, and Harris County HD volunteers
• UT-SPH Center for Health Policy Studies
• Matthew Stone, MPH
• Paige Padgett, MA, MPH
• Anna Zakos-Feliberti, MA, MPH
• Robert Wood, MPH
• HDHHS Bureau of Epidemiology
• Scott Bishop, MS
• Sharon Marsh and the HDHHS Administrative and
  Support Division staff
• Dr. Randolph Daley, CDC

31
Reports

• June 19 Preliminary report to Mayors Office,
  City of Houston final report on June 23
• Presentations to HDHHS (Houston Health
  Department), TPHA
• Waring SC, et al. Tropical Storm Allison Rapid
  Needs Assessment - Houston, Texas, June 2001.
  MMWR, May 3, 2002 51(17)365-9.
• Waring SC, Reynolds KM, DSouza GA, Arafat RR.
  Rapid assessment of household needs in the
  Houston area following Tropical Storm Allison.
  Disaster Manag Response 2002 Sept3-9.