Promenade Upper
Improving Response to the EMS for Children Program Survey: Applying Tailored Design Principles to a Challenging Establishment Survey (303588)
Michael Ely, University of UtahAndrea L Genovesi, University of Utah
Craig Hemingway, University of Utah
*Morgan M Millar, University of Utah
Lenora M. Olson, University of Utah
Kent Page, University of Utah
Patricia Schmuhl, University of Utah
Keywords: Response Rates, Web Survey, Mixed-Mode, Establishment Survey
A Tailored Design approach may be especially important for establishment surveys, because they pose more challenging circumstances for data collection. The Federal Emergency Medical Services for Children Program (EMSC), which supports pediatric emergency care improvement initiatives and projects across the U.S., faced considerable obstacles to obtaining state-level data to assess the readiness of EMS agencies to treat pediatric patients in the pre-hospital setting. During 2007--2009, the state EMSC program managers were tasked with administering a web survey of EMS agencies in their state to collect these data. Implementation approaches, and thus resulting response rates, to this initial survey varied dramatically across states, making it difficult to assess conclusions and make cross-state comparisons. Therefore, an organized effort was made to provide centralized, consistent implementation strategies and data quality goals to each state program manager for the next round of data collection in 2010--2011. In this paper, we compare the response rate results of these two rounds of data collection to assess the effectiveness of this improvement effort. The revised implementation strategy was informed by the Tailored Design Method and incorporated several techniques aimed at increasing response rates, including utilizing multiple, multi-mode, carefully-timed contacts and a designated time period for data collection. Additionally, state program managers were provided with extensively detailed standardized instructional materials and guided support to assist them through survey implementation. This approach resulted in a higher overall response rate (81% vs 69% in 2007--2009), which is notably high for a web survey, as well as greater consistency across states. The subsequent round of data collection (in 2013--2014) further demonstrated the usefulness of revised strategy by obtaining an overall response rate of 83%.