From Absence to Leadership: The Social Work Informatics Equity Framework for Health AI

Background and Purpose

Health informatics is the interdisciplinary field that connects health information and data science to improve individual and population health outcomes. Despite early calls to establish social work informatics (SWI) as a professional specialty within this field, social work has remained absent from health informatics governance, research, and system design. In the intervening decades, AI systems have become embedded across child welfare, health, housing, and criminal justice, shaping high-stakes decisions including child maltreatment risk assessment, service access determinations, and clinical risk scoring. Social work-generated data represent a rich and underutilized source of health-relevant information. Whether structured as demographic fields, zip codes, and ICD-10 social condition codes, or unstructured as clinical narratives, social work assessments, and case notes, these data capture family relational dynamics, experiences of structural racism, community-specific risk and protective factors, and the conditions shaping whether care is accessible or trusted. They remain inadequately leveraged in current health AI systems. This paper advances the Social Work Informatics Equity Framework, arguing that health informatics cannot achieve health equity without the formal integration and leadership of SWI.

Methods

This paper draws on critical analysis across social work science, health informatics, and AI ethics to identify structural gaps in current informatics practice and to develop the Social Work Informatics Equity Framework. Key gaps include underintegration of social determinants of health data, bias detection, limited community participation in design and implementation, and absence of justice-grounded ethical frameworks in AI development. These gaps are analyzed against social work's disciplinary competencies, including reflexivity, community-based participatory research, and the ethical mandate of the Social Workers' Code of Ethics. Prior transdisciplinary work demonstrates feasibility, including applications of social work ethical frameworks to AI model development that surfaced racial bias and evaluation challenges not identified through standard informatics review, and uses of social work clinical documentation to reveal health-relevant social patterns across clinical and community contexts.

Results

The SWI Equity Framework identifies five domains where social work's competencies address core limitations in health informatics. Reflexive analysis enables systematic interrogation of structural power across the AI lifecycle. Communication and representation practice draws on social work's person-centered, culturally informed approach to clinical language to improve how social and health information is exchanged between practitioners and patients, enhancing data quality, interpretability, and trust. SDOH data integration leverages social work documentation to capture social determinants not available or interpretable without community context. Community co-design embeds participatory methods in the development and implementation of health data systems. Ethical accountability applies enforceable social work ethics to AI development and deployment, extending beyond technical fairness metrics.

Conclusions and Implications

Social work informatics is not peripheral to health informatics. It is foundational to achieving its mission. Social work education must integrate informatics training. Accreditation bodies must formalize competencies. Social work scientists must engage as co-investigators in informatics research. Health informatics institutions should formally recognize SWI and integrate social work perspectives into leadership, funding, and research. Health informatics cannot achieve health equity without the reflexive methods, community accountability, and justice-centered ethics that define social work.