Developing a survey instrument to assess flood health beliefs among residents of river flood zones in the Heartland.

INTRODUCTION: Flooding represents a significant health risk to residents of states bordering the Missouri and Missis-sippi Rivers. In Nebraska and Iowa, nearly 280,000 individuals live in floodplains that put them at higher risk for expe-riencing a flood. In Nebraska, 8.60 percent of residents live within floodplains, while in Iowa, 3.83 percent of residents live in floodplains.1 Flooding poses immediate risks including injury and loss of life, but it additionally threatens long-er-term impacts due to increased exposure to environmental health risks and damage to physical and social infrastruc-ture.2,3 Health education interventions depend on risk communication based on accurate knowledge of the health beliefs of at-risk individuals. This paper outlines the development of a survey instrument designed to solicit flooding health beliefs of individuals living in river flood zones.

METHODS: This study developed the survey instrument through a series of revisions designed to increase the con-struct validity of the data collected. These steps included item sourcing and modification, expert review, and cognitive interviewing. Content for the initial survey draft was sourced from existing flood risk survey instruments and other standardized demographic surveys.4-8 This study then validated the instrument through sequential expert review and cognitive interviewing, collecting and conducting qualitative analysis of data collected through each phase.

RESULTS: Themes from the expert review include (1) improvement of the user experience through revision of re-sponse categories and removal of unnecessary structural elements, (2) improvement of construct representativity through attentiveness to existing flood preparedness policies in the Heartland, and (3) streamlining of data analysis through revision and clarification of item prompts and responses. Themes from the cognitive interviewing include (1) the addition of items soliciting respondents' awareness of their flood risk as well as open-ended items soliciting respond-ents' existing personal flood plan components, (2) revision of existing demographic items to solicit respondents' esti-mated-rather than exact-household income, and (3) clarification of our definition of a flooding event.

DISCUSSION: This study adapted existing survey items into an instrument designed to gather data on respondents' flood health beliefs and relevant demographic characteristics, producing a developed instrument that can be used to measure a population's flood health beliefs with reasonable evidence of validity. This instrument may be further tested by gathering baseline data and conducting psychometric testing to determine which items wield significant influence over respondents' decision to develop a personal or household flood plan.9 Additionally, this study provides a model for risk communication professionals seeking to adapt existing survey items to develop instruments designed to gather data on locally relevant health outcomes.10 .