Practical Applications: Analysis and Interpretation Considerations

Applying equity-centered approaches to analysis and interpretation

1. Disaggregate syndromic data to describe diverse and intersectional experiences.
Syndromic surveillance systems collect information on patient race and ethnicity, sex, age and geographic location. As discussed earlier, other important factors that do not yet correspond with widely implemented data standards and data fields (e.g., housing status, substance use, disability) can sometimes be identified through other data fields. Disaggregating syndromic surveillance data can help public health practitioners develop more pointed programmatic questions. In order to provide actionable insights to improve the health and well-being of all community members, however, data must be granular enough to detect differences in health events between groups. Throughout data work, consider why demographic data are being collected, what demographic categories are used and why, and which data analysis approaches may maximize benefits and reduce risks to populations and communities.

Analysts should disaggregate data across demographic factors and social determinants of health to understand patients’ unique health experiences and highlight disparities and trends. Depending on the quality of the data and sample size, analysts should examine intersectional experiences across multiple demographic variables and/or social determinants of health.

Data disaggregation is often complicated by small sample sizes, suppression rules, and low confidence in the estimate. Jurisdictions should consult their state and local guidance for suppression thresholds and strategies for communicating suppression decisions with data consumers. Not all data collection systems accommodate the demographic diversity represented across patients, which may limit the granularity of analysis. If granular demographic data are collected, then analyses should use those granular categories rather than collapsing people into less-specific groups.

Ample and transparent documentation should accompany any analyses that do remove or collapse categories to protect individual privacy or data stability. Analyses should communicate when the resulting sample is a count of cross-tabulations, for example, and specify when sample sizes were too small to disaggregate and report. Analysts should acknowledge the removal or collapsing of categories as a limitation of the analysis and not a reflection of the community itself.

2. Prioritize community engagement when analyzing and interpreting syndromic surveillance data.
Data interpretation is an inherently subjective process. There are layers of nuance behind syndromic surveillance data that are difficult to represent using quantitative data alone. Incorporate community input and needs when designing analyses and visualization products. Engaging communities in the data analysis and interpretation process provides a qualitative lens and includes community voice in the narrative.1

An example of effective community engagement is PRIDEnet, a national network of LGBTQ+ individuals and organizations. The organization actively engages the LGTBQ+ community in their long-term national health study by incorporating LGTBQ+ voices in the collection and interpretation of data.2 This is a powerful way to build trust and develop a deeper understanding of the data. Additional principles of community engagement include articulating the goals of your engagement, learning the communities by talking with partners and assessing capacity for engaging.3

3. Include community members in the development of syndrome definitions.
People and groups describe symptoms in similar and different ways. This is particularly important when thinking about cultural variation and stigma associated with certain conditions. For example, talking about mental health is stigmatized in some cultures and communities. Individuals may not describe a mental health symptom as part of their chief complaint or may describe it using indirect terms. Involving public health programs and partners in syndrome definition development can provide community perspectives helpful for contextualizing symptoms and capturing the variety of ways a health condition or health behavior may present.

4. Assess the extent to which principles of equitable data practice are incorporated in your work and opportunities to expand their practice.
Examine how data equity is applied to your syndromic surveillance data collection, analysis and use. Each approach should minimize harm and maximize benefits, uphold self-determination, promote transparency and accountability for surveillance decisions and limitations and apply tools and resources that expand opportunities for affected populations and communities. Partner with communities to inform data interpretation and analysis, share limitations and provide usable and actionable information to support community-engaged initiatives and equity work.