JA:2021-25. Increasing Efficiency and Reliability of Agricultural Injury and Fatality Coding with Standard Operating Procedures: Lessons Learned from AgInjuryNews.org

Abstract

ABSTRACT Purpose: News media reports of agricultural injuries can be useful to better inform safety and health stakeholders and prevention efforts. These data are particularly useful for youth incidents, since the National Institute of Occupational Safety and Health discontinued their survey-based agriculture surveillance program in 2015, leaving a gap in reliable, consistent data on both youth and adult agricultural injuries and fatalities. In this study we describe how coding data derived from news reports, with a detailed Standard Operating Procedure (SOP) and refined inclusion/exclusion criteria, can lead to more efficient workflows to capture and code agricultural injury data. Methods: To capture quality data and improve inter-rater reliability we employ a multi-coder process. The first coder enters primary data into the Case Report Form (CRF), and a second coder reviews these data points for agreement. While reviewing for agreement, the second coder diverges into two other occupational/agricultural injury coding schemas (Occupational Illness and Injury Classification System [OIICS] and Farm and Agricultural Injury Classification [FAIC]) and employs an additional multi-coder and interrater reliability process, adding to the multifaceted workflow and detailed data output. This secondary coding schema and reliability ratings are discussed in a separate manuscript. Any discrepancies from the initial review are then discussed and settled between first and second coders. All CRFs are sent to another team member who reviews >10% and publishes the reports for public use on AgInjuryNews.org. International cases endure a similar process, but without OIICS or FAIC coding. In partnership with the Canadian Agricultural Safety Association, our team began Canadian injury report capture and coding in the fall of 2019. Findings: We developed a structured and detailed SOP for coding agricultural data obtained from news reports. Use of the SOP and refined inclusion/exclusion criteria were crucial when: 1) onboarding additional coders; 2) employing multiple coders; 3) refining search terms in data collection; 4) coding of emerging technologies and trends; and 5) improving the efficiency and quality of the data. Translation: Coders’ ability to reference a thoroughly vetted SOP and inclusion/exclusion criteria may lead to improved data coding for agricultural injuries and fatalities.