Public Health CBRN course Rapid Community Needs Assessment Bonnie Henry, MD, FRCPC Goals of Session To learn the principles of rapid community risk assessment using modified cluster sampling methods To understand when this technique may be useful in response or recovery from a disaster Goals of Rapid Needs Assessment To rapidly obtain population-based estimates of need in the early aftermath of a disaster Should also be a way of reaching the affected areas with information about relief efforts and services History Based on methods developed by the WHO for estimating immunization coverage in the EPI program Used to estimate community impact of Hurricane Andrew in South Florida in 1992 Also used after the Sept 11, 2001 attacks in NYC to assess needs in residents of Lower Manhattan Who would use the data Government organizations Disaster relief organizations (Red Cross etc) Provincial, regional and local public health and healthcare organizations Law enforcement Utilities, (especially electricity, water) Media, etc. Cluster Sampling Methods Systematic sampling of 30 ‘clusters’ Create a grid over the area to be sampled – ¼ mile squares on street map Assign each square or ‘cluster’ a number Cluster Sampling Methods Determine the interval you need to end up with 30 clusters to sample – For example if you have 120 squares in your grid you will have an interval of 4 (120/30=4) Randomly determine where to start – In this case pick a number between 1 and 4 – You can use a random number generator or the serial number on a $5 bill Select your 30 clusters Cluster Sampling Methods Go to the centre of the first cluster (i.e. square or cluster 3 on your map) Determine your sampling unit – Usually people or households Determine the number of units to be sampled in each cluster (i.e. 10 households) Proceed in a randomly picked direction to the first occupied household Cluster Sampling Methods Interview the first adult at the occupied residence Then go consecutively to the next occupied household until you have reached 10 in that cluster If a multi-unit dwelling or apartment building pick first occupied unit for interview Move on to next selected cluster Example: Systematic sampling Source: D. Coulombier, Epicentre Source: MSF Sampling methods – organization of sampling units; Cegrane camp, Cegrane, Macedonia - 1999 Source; B. Woodruff, US-CDC Sampling interval = 12; Starting number = 3; Cycle = 3 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 .... Source: EPIET Interview Methods Interview team usually 3-4 people including an epidemiologist and volunteer interviewers – from medical schools, public health staff etc. Use a simple questionnaire Ensure team able to provide information on: – Location of medical treatment, supply distribution sites – Provide preventive health messages on food, water safety, handwashing, injury prevention etc. Sample Questions How many and what percentage of surveyed households reported the following? – Not enough food – No running water – No electricity – No heat/air conditioning (in temperature extremes) – No functioning toilet – No telephone – No car or truck – Injured resident(s) – Ill resident(s) – Resident(s) in need of medical care – Resident(s) unable to obtain needed medications – Resident(s) in need of counseling – Resident(s) with special needs – Displaced from home Time frame for initiating study After ground transportation is restored Approximately 3 and 10 days after the event for severely affected areas Assessments of less severely affected areas phased in as possible Benefits Practical and standardized methods allow for rational assessment (not based on rumours only) Gives a picture of what needs are and helps direct relief operations – E.g. switch from mass casualty trauma services to primary care and preventive services Can monitor over time “Being roughly right is generally more useful than being precisely wrong” Guha-Sapir, D. Rapid assessment of health needs in mass emergencies: Review of current concepts and methods. World Health Stat Q. 1991;44:171-181.