JMIR Publications

ABSTRACT

Background:

Data on the social determinants of health could be used to improve care, support quality improvement initiatives, and track progress towards health equity. However, sociodemographic data collection is not widespread across Canada. Artificial intelligence (AI), specifically natural language processing and machine learning, could be used to derive social determinants data from electronic medical records. This could reduce the time and resources required to obtain social determinants data, which otherwise requires surveys of each individual patient.

Objective:

The aim was to understand perspectives on the use of AI to derive social determinants of health information from electronic medical record data.

Methods:

Using a qualitative description approach, in-depth interviews occurred with 195 participants purposefully recruited from Ontario, Newfoundland and Labrador, Manitoba, and Saskatchewan. Transcripts underwent inductive and deductive coding and we conducted a content analysis.

Results:

Four themes were identified: 1) AI as the inevitable future: facilitating more efficient, accessible SDoH information and use; 2) Healthcare harms: distrust in AI and public systems; 3) Loss of the human touch: preference for provider relationship and individualized care; 4) Consent is critical: strong safeguards are needed to protect patients’ data and trust.

Conclusions:

These findings provide important considerations for the use of AI in healthcare, and particularly when healthcare administrators and decision-makers seek to derive social determinants data.