AI-driven NLP models effectively extract social determinant data from clinical notes, as demonstrated by the Regenstrief Institute and Indiana University collaboration. Analyzing over six million clinical notes, their adaptable system identifies housing and financial keywords accurately. This advancement aids tailored patient care and referrals to relevant services. The broader trend emphasizes NLP’s potential in healthcare, necessitating provider engagement for optimal integration and equity-driven care improvements.
AI-powered natural language processing (NLP) models are successfully extracting social determinants of health (SDOH) data from clinical notes, according to a collaborative effort by the Regenstrief Institute and Indiana University. The research demonstrates that the NLP system can effectively identify keywords and phrases denoting housing or financial needs, providing accurate results.
The study, featured in the International Journal of Medical Informatics, examined over six million clinical notes from patients in Florida. The NLP system’s adaptability and accuracy were assessed, showcasing its capability to function effectively in different environments and adapt to varying data requirements.
The significance of this advancement lies in its potential to enable healthcare providers to extract valuable SDOH data from clinical notes, which often lack standardized terminology compared to electronic health records. This data can aid in tailoring medical care to patients’ social needs and facilitating referrals to appropriate services.
Dr. Chris Harle, a faculty member at Regenstrief and IU Fairbanks School and senior author of the study, emphasized the broader implications. He noted that this approach could be extended to extracting other social risk information from clinical text, such as transportation needs. Dr. Harle also highlighted the importance of evaluating and monitoring the effectiveness of NLP methods when applied in diverse healthcare systems.
This research aligns with a larger trend in which AI and NLP technologies are being leveraged to improve healthcare outcomes. Earlier, Regenstrief Institute researchers developed NLP algorithms to extract housing, financial, and employment data from electronic health records. Their efforts aimed to enhance risk models and provide clinicians with valuable factors for decision-making.
Moreover, the researchers developed an application called Uppstroms, which predicts patients requiring referrals to social services. The potential of NLP in identifying and addressing social risk factors across various healthcare domains, such as primary care, mental health, and public health, was also underscored.
While the applications of NLP are promising, widespread adoption requires the engagement of healthcare providers and the consideration of caregivers’ needs. The integration of NLP in healthcare holds the potential to lead to more personalized and proactive care interventions, ultimately contributing to achieving health equity.
