Patient survey data boosts opioid use disorder prediction accuracy

A new study published on arXiv demonstrates that incorporating patient-reported survey data significantly enhances the prediction of opioid use disorder (OUD) when combined with electronic health records (EHRs). Researchers found that augmenting EHR data with survey responses improved prediction accuracy across various timeframes and machine learning models, with LightGBM showing a notable improvement from 0.6219 to 0.6603 for a 24-month prediction window. The study highlights that patient-reported information offers valuable predictive signals beyond structured EHRs, underscoring the importance of survey availability in clinical settings. AI

IMPACT Enhances predictive modeling for health outcomes by integrating diverse data sources.

RANK_REASON Academic paper detailing a new methodology for improving predictive models. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• All of Us
• arXiv
• Electronic health records
• gated recurrent unit
• Hugging Face
• LightGBM
• long short-term memory
• Opioid Use Disorder
• Transformer++
• XGBoost

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →