Background: Addressing health inequity requires innovative analytics and dynamic application of data to accurately identify disparities in a timely manner. To date, proxy indicators (e.g. quality of life and mortality), are limited and rarely applied at a population-level within a health system. Objective: We sought to design and implement a Health Equity Index (HEI), a novel empirical approach to identify and prioritize health outcome inequalities for ambulatory care sensitive conditions (ACSC) in healthcare systems, calculated from widely available data. Methods: The proposed HEI, an overall summary statistic, is also available by race category. It represents the average ratio of observed-to-expected hospital encounters for a given ACSC in a set timeframe at a given hospital in Sutter Health, a 24-hospital health system in Northern California. Expected encounters for a condition are adjusted by age/sex/race for underlying population distribution of hospital catchment area, condition prevalence, propensity to utilize the given hospital, and average frequency of that utilization. Underlying population estimates are from the American Community Survey; prevalence estimates from the Centers for Disease Control; and hospital utilization from electronic health records. HEI’s for each condition and the underlying raw numbers are published monthly on a leadership dashboard, and a team of health equity analysts is available to help interpret. Results: In 2017, 3027 patients utilized the Sutter system of hospitals 3918 times for a primary diagnosis of diabetes, yielding a summary HEI=1.8. This HEI indicates 80% more encounters than expected given the underlying population characteristics. This was primarily driven by the youngest age group (20-44) in general, and by African-American (AA) patients, in particular (AA males 20-44 - HEI=10.4: AA females 20-44 - HEI=3.8). All other subgroups utilized less than expected. Conclusions: In this system, utilization for diabetes by Asians and all older individuals, regardless of race, was less than expected. A possible next step for hospitals with this profile is to identify interventions targeting young people at risk for diabetes. Implications: The HEI can be adapted for any health system with an accessible EHR, and it provides information to identify population subgroups that might benefit from patient-centered ACSC interventions to reduce health inequities.

background The County Health Rankings (CHR) is an excellent model to assess health equity at the county level. However, heterogeneity in these large areas masks health differences at the community level, which often vary considerably from county-level estimates. We developed the Chicago Community Health Rankings (CCHR), based on the CHR model, to assess and rank the health status of Chicago’s communities. By using the CCHR tool, we can increase awareness of factors that influence health, stimulate policy change and development, and promote community engagement. objectives 1.) Develop a community friendly, replicable tool to mobilize action in urban areas. 2.) Incorporate the social determinants of health in a comprehensive, multi-sector approach. 3.) Visually map community area profiles using a geographic information system (GIS). methods We compiled publicly available data to create health rankings for Chicago’s 77 community areas. Health measures adopted from the CHR model provided a rubric, along with urban-specific measures such as gun violence and drug possession. We calculated composite health rankings at the community level using the CHR framework of social and economic factors, health behaviors, clinical care, and the physical environment. Each community area profile was mapped using GIS. results Upon incorporation of 40 distinct health measures across multiple sectors, we replicated the CHR to create the CCHR. The variability across communities observed enabled us to identify the leading determinants of health within each community, such as violent crime, children living in poverty, and lack of health insurance among adults. conclusion The CCHR can be employed to draw comparisons across communities and to identify and visualize strengths and weaknesses of community areas within a city. A bound set of 77 user-friendly profiles are disseminated to communities, partners, and stakeholders. public health implications The CCHR is a beneficial tool for government agencies, local organizations, planners and developers, and community stakeholders to identify and develop working strategies to properly allocate resources based on the health of a community. The rankings can also be utilized to motivate community members to take part in the improvement of their own communities.

Background: For many Americans the community in which they live has a significant impact on their opportunity for optimal health. Traditional county level analysis often hides vulnerable populations, diluting their impact in analyses. Fairfax County, for instance, has one of the lowest poverty rates in Virginia but is home to more people living in poverty (68,000) than any other Virginia county or independent city, most of them clustered into a few small areas. Objective: The Virginia Department of Health developed the Virginia Health Opportunity Index (HOI) to identify vulnerable populations at the Census Tract level and to help communities, providers and policymakers understand how neighborhood level factors influence the opportunity residents have to live long and healthy lives. Methods: The HOI uses social, economic, educational, demographic, and environmental variables related to the well-being of a community and related directly or indirectly to population health. The HOI is a composite measure comprising 13 indicators that reflect a broad array of social determinants of health: (1) Affordability, (2) Income Inequality, (3) Material Dependency, (4) Job Participation, (5) Employment Access, (6) Education, (7) Air Quality, (8) Population Density, (9) Population Churning (10) Segregation, (11) Food Accessibility, (12) Walk-ability, and, (13) Access to Care. Indicators were suggested by stakeholders and selected based on their influence on health as expressed in the literature and the availability of quality data for all Virginia Census Tracts. Indicators were weighted using Principal Component Analysis to develop the final HOI. Results: Predictor models that take spatial, place-based, considerations into account (geospatial-weighted regression) were utilized. The HOI variables predict over 60 percent of the variation in Life Expectancy and Disability Free Life Expectancy when space is considered. Indicators contributed differentially to the overall coefficient weights depending on their location within the state. Conclusion: Place is significant to understanding the distribution of life chances. Public health implications: The HOI enables public health practitioners to examine how place-based indicators relate to broader community conditions to develop public policies and programs that promote health and health equity. The Health Opportunity Index has been replicated in Ohio and is being considered in other states.

The National Association of Community Health Centers, the Association of Asian Pacific Community Health Organizations, the Oregon Primary Care Association, and the Institute for Alternative Futures have collaborated with community health center partners to implement PRAPARE (Protocol for Responding to and Assessing Patient Assets, Risks, and Experiences), a nationally recognized and tested standardized patient risk assessment protocol that goes beyond medical acuity to account for patients’ social determinants of health. PRAPARE (www.nachc.org/prapare) is both evidence-based and stakeholder-driven, containing measures on 21 social determinants of health that align with national initiatives, such as the National Academy of Medicine’s recommendations for social demographic factors. PRAPARE partners worked with teams of health centers and their networks to create the social determinants of health data collection tool; mapped PRAPARE response choices with relevant codes (e.g., ICD-10 z codes, LOINC codes, SNOMED codes); developed templates in four different Electronic Health Record systems including eClinicalWorks, Epic, GE Centricity, and NextGen that can add relevant codes to PRAPARE responses; and tested the tool in different health center workflows. The project team has completed pilot implementations with 18 health centers nationally and has collected health-center level PRAPARE social determinants of health data from nearly 20,000 unique patients using a standardized reporting template. Preliminary data revealed that patient social determinant risks (e.g., lack of housing, food insecurity) ranged from 4.8 to 8.6 with an overall mean of six risks per patient. However, chronically ill patients tended to face up to eleven social determinant risks. A significant correlation was found between the number of social determinant risks a patient faces and having hypertension. High stress, less than a high school education, unemployment, and lack of insurance were among the top five risks for many health centers. The presentation will discuss: 1) the impact social determinants have on health; 2) the utility of having this information in the EHR; 3) the variation of different workflow models that can be used to collect PRAPARE data; and 4) the different ways PRAPARE data can be used to affect change at the patient, organizational, and population levels through interventions, partnerships, and advocacy.