Natural Language Processing System Improves Extraction of Social Risk Factors from Electronic Medical Records

Natural Language Processing System Improves Extraction of Social Risk Factors from Electronic Medical Records


Social risk factors, such as financial instability and housing insecurity, have a significant impact on an individual’s health. However, extracting this information from electronic medical records can be challenging due to the lack of standardized terminology.

A recent study conducted the Regenstrief Institute and Indiana University Richard M. Fairbanks School of Public Health has developed a natural language processing (NLP) system that showed excellent performance in identifying social risk factors. The system was tested on over six million clinical notes from patients in Florida and evaluated for its generalizability and portability.

The NLP system, which searched through physicians’ clinical notes, successfully identified key phrases and words that indicated housing difficulties or financial needs of patients. Despite challenges such as regional language variations, the NLP models provided highly accurate performance compared to human review.

Identifying social risk factors is crucial for clinicians and healthcare systems to provide better care to patients. By incorporating this information into their decision-making process, clinicians can better treat the whole person and connect them to services that address their needs. For example, knowing a patient’s housing instability can help healthcare systems provide proactive interventions and connect them to housing services or financial resources.

The study highlights the importance of natural language processing and other artificial intelligence methods in understanding patients’ needs. The more efficiently social information can be extracted, the more cost-effective and comprehensive population health perspectives can be developed.

This research has significant implications for patient care as it emphasizes the relationship between social factors and health outcomes. By incorporating social risk factors into electronic health records, clinicians can gain a better understanding of the individual and provide tailored care that accounts for their unique circumstances.