Collecting data on key social determinants of health in primary care: a multi-site implementation evaluation

Context: Social data that is collected in a standardized fashion can be used to improve individual patient care by uncovering and then addressing social needs, to stimulate the development of new integrated health and social programs, and be used to identify inequities across an organization. This data can also support system- and policy-level planning to identify health inequities at the population-level and support policy change.

Objective: To report implementation outcomes using a standardized approach and tool in a multi-site study of routine and systematic demographic and social needs data collection in primary care settings.

Study Design and Analysis: Between Sept 2022-Oct 2023 we implemented the SPARK Tool at five primary care clinics. We collected implementation outcomes and used surveys to assess the perspectives of patients, clerical staff and providers.

Setting: We collected data at five primary care clinics in five Canadian provinces (Saskatchewan, Manitoba, Ontario, Nova Scotia, and Newfoundland and Labrador).

Instrument: SPARK Tool, which includes 18 demographic and social needs questions

Dataset: number of SPARK Tool surveys completed by patients (n=2063), patient feedback surveys (n=1368), clinic description and readiness checklists (n=5), provider and staff implementation surveys (n=36) and online training evaluation surveys (n=33)

Population Studied: patients, clerical staff, and providers who had used or completed the SPARK Tool

Outcome Measures: Implementation outcome measures including acceptability, adoption, feasibility, penetration, cost, appropriateness, fidelity, sustainability.

Results: SPARK Tool completion rates varied significantly (9.4%-48%), indicating moderate penetration. The SPARK Tool was highly acceptable, with 90.5% of patients agreeing that the tool was clear and easy to complete and 84.5% having a positive experience. 58.1% reported being comfortable answering the questions. Clerical staff and provider 96.7% finding the SPARK Tool to be useful and 81.8% reporting a positive experience with using it.

Conclusions: The results highlighted positive acceptability, feasibility, and the adoption of the SPARK Tool in diverse primary care clinics, as well as practical insight into implementing demographic and social needs data collection. Levels of penetration, patient comfort, and ease-of-use could be improved when using the tool at other sites.