Assessment of structured data elements for social risk factors

Area of Science:

• Health Informatics
• Social Determinants of Health
• Data Science

Background:

• Routinely collected structured data in electronic health records (EHR) and health information exchanges (HIE) offer a potential avenue for measuring social risk factors.
• Developing computable phenotypes from these data sources could enhance understanding and address of social determinants of health.

Purpose of the Study:

• To identify and assess the quality of structured data elements within EHR and HIE systems suitable for creating computable social risk factor phenotypes.
• To evaluate the suitability of existing data elements for measuring socioeconomic status, cultural context, social relationships, and community context.

Main Methods:

• A two-round Delphi technique involving 17 experts with knowledge of EHR and HIE data.
• Round 1: Nominal group approach to generate candidate data elements related to social factors.
• Round 2: Panelists rated data elements on quality and potential for systematic bias.

Main Results:

• A total of 89 structured data elements were identified, with 45 related to socioeconomic characteristics.
• Data elements used in reimbursement processes generally received higher quality ratings.
• Concerns were raised regarding potential implicit bias and system-level biases within certain data elements.

Conclusions:

• Structured data within EHR and HIE systems hold promise for reflecting patient social risk factors.
• This foundational work in identifying and assessing data elements is crucial for developing future computable social factor phenotypes.