Data quality and completeness in a community based and participatory epidemiologic study

The adoption of community based participatory research (CBPR) principles represents movement away from traditional scientific foundations that guide epidemiologic inquiry. CBPR emphasizes complete collaboration between the affected population and the trained investigator and values resources that preexist within the community. In contrast, modern epidemiology is founded on standards that idealize objectivity and the idea of the expert researcher conducting a controlled experiment. Therefore, skepticism surrounds the scientific validity of data obtained by epidemiologic studies that undertake a CBPR approach. We addressed this issue by examining the quality and completeness of data that participants produced during the Community Health Effects of Industrial Hog Operations (CHEIHO) study, which undertook a CBPR approach. Approximately 100 residents of Eastern, NC who lived near at least 1 industrial hog farm participated in CHEIHO by collecting data on their own and in their own homes. We quantified the percentage of total records in which participants 1) left certain variables missing and 2) produced a sequence error, meaning that they recorded outcome information before they experienced the exposure. Of 2,932 journal records, 98% were free of a sequence error. Between 80% and 99% of the journal records had complete information for the 15 variables that we examined, and the median completeness rate was 97%. We plan to consider individual- and community-level predictors of data completeness and quality, including participants' time-in-study, presence or absence of a community based representative to offer assistance to participants, and participants' age and employment status.