Text mining electronic health records to identify hospital adverse events

Stud Health Technol Inform. 2013:192:1145.

Abstract

Manual reviews of health records to identify possible adverse events are time consuming. We are developing a method based on natural language processing to quickly search electronic health records for common triggers and adverse events. Our results agree fairly well with those obtained using manual reviews, and we therefore believe that it is possible to develop automatic tools for monitoring aspects of patient safety.

MeSH terms

  • Algorithms*
  • Artificial Intelligence*
  • Data Mining / methods*
  • Data Mining / statistics & numerical data
  • Denmark
  • Humans
  • Medical Errors / classification*
  • Medical Errors / statistics & numerical data
  • Natural Language Processing*
  • Pressure Ulcer / classification
  • Pressure Ulcer / epidemiology*
  • Reproducibility of Results
  • Risk Management / methods*
  • Sensitivity and Specificity