A Comparison of Population Characteristics by Assessment Approach (Paper vs. Web-Based Survey Methods) of the Together for Health = Virginia Study

Authors: Fuemmeler BF, Do EK, Ksinan AJ, Miller CA, Morris BB, Fugate-Laus K, Fallavollita WL, Kim SJ, Wheeler DC, Dahman B, Gal TS, Ginder GD

Category: Cancer Health Disparities
Conference Year: 2020

Abstract Body:
Purpose: To compare population characteristics by paper and web-based approaches within the on-going Together for Health = Virginia Study, a National Cancer Institute P30 Supplement to characterize the catchment area of Virginia Commonwealth University (VCU) Massey Cancer Center. Methods: Participants are recruited through a mail-based survey, with the option to complete the survey online via REDCap, and through a web-based and mobile application (app) platform used to form an eCohort. The mail-based survey was a probability-based sample of 6,000 community- dwelling adults, conducted within the VCU Massey Cancer Center catchment area, using a modified Dillman method. Mail-based surveys began in October 2019. Enrollment in the eCohort is expected to begin in January 2020. Survey information on sociodemographic characteristics, social determinants, health behaviors, healthcare use, health information utilization, health history, and cancer screening are collected. Response rates and demographic frequencies and percentages are provided by assessment approach. Results: Data collection is ongoing. Of the 559 participants who have responded, 327 (58.5%) elected to complete the survey by mail and 232 (41.5%) elected to complete the survey online using REDCap. The average age of participants using REDCap is younger (mean=49.3; SD=15.2 years) than paper-based surveys (mean=59.9; SD=14.8 years, t(422.4)=7.7, p<0.01). A higher percentage of REDCap vs. paper- based survey participants are White (78.1% vs. 68.9%, χ2(2)=9.4, p<0.01), have a bachelors' degree or higher (62.7% vs. 53.8%, t(493.9) = 2.4, p=0.02), and have an annual household income of $50,000 or higher (77.2% vs. 65.8%), χ2(1)= 6.7, p=0.01). No differences were found for marital status (χ2(1) = 1.4, p=0.24), self-identification as rural (χ2(1)=2.9, p=0.09), or being born in the US (χ2(1) =0.3, p=0.59). Conclusions: In addition to characterizing the population within the catchment area of VCU Massey Cancer Center, results of this study will aid in the development of future epidemiological research studies by providing practical guidance for public health and survey researchers regarding paper- based and web- based assessment approaches.

Keywords: health behaviors, cancer, population health