Development and Implementation Methods of an Offline Survey in Low- and Middle-Income Countries (303168)

Esteban Avendaño, Universidad de Ciencias Médicas 
Margaret M. Demment, University of Rochester Medical Center 
Timothy De Ver Dye, University of Rochester Medical Center 
*Scott McIntosh, University of Rochester 
Deborah J Ossip, University of Rochester Medical Center 
José G. G. Pérez-Ramos, University of Rochester Medical Center 
Carmen Milagros Vélez Vega, Departamento de Ciencias Sociales 

Keywords: Offline data collection, REDCap, Survey Development and Implementation

In related research projects in Pacific US Territories, and in Low and Middle Income Countries (LMIC) we are studying 1) perceptions of genetic testing (Puerto Rico), Diabetes prevention (Marshall Islands, Micronesia), and Maternal Health (Costa Rica, Dominican Republic, and Honduras). Internet-based survey instruments were developed, beginning with qualitative research and community-engagement, to include offline functionality and sustainable standardization. Fieldwork in Puerto Rico began with the development of a survey initially available at Mechanical Turk (mTurk), Amazon.com's crowdsourcing website, greatly expanding the ability for low cost timely survey development. To create an appropriate Spanish language version, it was translated then back translated by six bilingual investigators from the US and Puerto Rico for readability, skip patterns, formatting, and content. This survey was pretested using the mobile (offline) application of REDCap (Research Electronic Data Capture). REDCap, a HIPAA-compliant software for electronic collection and management of data, provides an both an online and offline interface, which eliminates the need for an Internet connection; allows direct data entry (IOS and Android); and allows for tracking of data entry staff by registering individual usernames. Six local data collectors were trained in a 3-hour workshop. The instrument was then piloted with iPads at health centers in Puerto Rico, resulting in 31 completed surveys. Feedback from data collectors included concerns that the survey was too long, and that some common survey aspects (e.g., Likert Scales) were difficult for respondents. Further adaptation for cultural competence and training on the use of the offline instrument will occur prior to a larger scale implementation in the Summer of 2016, and implementation of surveys in the other locales. Preliminary results, and the methodological strengths and limitations of using the offline survey will be discussed.