POSA368 Real-World Dietary Data Collection: Current Challenges and Opportunities for Innovation

Abstract

Patient dietary intake is complex to obtain and assess, but the potential to help understand risk factors for chronic disease and impact on treatment outcomes is high. We describe current challenges in real-world dietary data collection and propose innovative approaches to address challenges. Common dietary data collection methods: are outlined and compared, including study design considerations. Technological innovations are highlighted, including how they help address current data collection challenges of patient burden, data accuracy, and extensive analysis time. There are many considerations in selecting a dietary collection method, such as study design/objectives, patient population, budget, and timelines. Although detailed data collection (24-hour recalls or multi-day food records) is generally preferred over other methods (short assessments: limited data collected; food frequency or history: average or previous intakes) due to accuracy and level of detail, it is challenged by patient burden, recall bias, cost, and long data analysis timelines. Key innovations include food image recognition platforms powered by artificial intelligence (AI), audio transcription software, and applications with guided questions/animations. These approaches may decrease patient burden, recall bias, data analysis time/cost, and increase data quality. Choosing to use an innovation should be partially informed by the target study population. App-based AI platforms and audio transcription may benefit populations whose dietary intakes are entered by a caregiver (e.g., children or cognitively impaired patients), or study populations with an already high patient burden who are technologically savvy. Populations who are less technologically savvy may benefit more from guided questions/animations. Technological solutions may address real-world dietary data collection challenges and allow more frequent dietary data collection in real-world studies, providing potential deeper insights into treatment outcomes. However, technological approaches are not always appropriate; therefore, decision-making to utilize innovations should be led by dietary data experts with experience in operationalizing technology solutions.