Sign in Subscribe
Medical records

Developing a Pressure Injury Predictive Indicator System for Data Mining

Strategic Insights

— 1 min read

Researchers at Second Xiangya Hospital in Changsha, Hunan, China, have developed a predictive indicator system for hospital-acquired pressure injuries (PIs), a critical measurement of medical care quality. The study aimed to identify and automatically mine predictive indicators from electronic medical record systems. The research was funded by the National Natural Science Foundation of China and has been peer-reviewed. The predictive indicator system consists of 3 categories and 14 indicators, which were extracted from the health care information system (HIS) using structured query language and the Random Forest technique. The system has been shown to be accurate in predicting PIs, with an accuracy rate of 95.26%. However, the researchers suggest that further revisions should be conducted in real-life medical environments.

Key Takeaways:

  • The prevalence of hospital-acquired pressure injuries has shown an upward trend, making it a critical measurement of medical care quality.
  • The predictive indicator system was developed using a modified Delphi method, including clinical healthcare provider interviews, literature review, research group meetings, and Delphi survey.
  • The system consists of 3 categories and 14 indicators, which were extracted from the health care information system (HIS) using structured query language and the Random Forest technique.
  • The agreement between manual extraction and the computer's automatic extraction was good, with a Cohen kappa score of 0.64 to 1.00.
  • The accuracy of the predictive model was 95.26%, indicating its potential in predicting PIs.
  • The researchers suggest that further revisions should be conducted in real-life medical environments to improve the system's performance.

Statistics:

  • Accuracy of the predictive model: 95.26%
  • Cohen kappa score: 0.64 to 1.00
  • Number of indicators in the predictive indicator system: 14
  • Number of categories in the predictive indicator system: 3
  • Accuracy rate of the system's automatic extraction: Not specified
  • Number of healthcare providers involved in the Delphi survey: Not specified

Sources:

  • Developing a Pressure Injury Predictive Indicator System for Data Mining In Health Care Information Systems: a Sequential Mixed-methods Study. Advances In Skin & Wound Care, 2025;38(9).
  • National Natural Science Foundation of China (http://dx.doi.org/10.13039/501100001809)
  • NewsRx. New Data Systems Findings Has Been Reported by Investigators at Second Xiangya Hospital (Developing a Pressure Injury Predictive Indicator System for Data Mining In Health Care Information Systems: a Sequential Mixed-methods Study). Information Technology Newsweekly. October 21, 2025; p 434.

Read more