Patient-level data, including claims and encounter data, provide important insight into health care utilization, spending, and outcomes. Yet, these data often tell only part of the story. Health care needs and outcomes are also shaped by the communities in which people live, the resources available to their communities, and the social, economic, environmental, and health system factors that influence access to care.

Dobson | DaVanzo conducts population health and community-level analyses to help clients better understand how health care utilization, spending, and outcomes vary across populations, providers, and geographies, and how those differences may be shaped by Social Drivers of Health (SDoH), community context, and local health system factors. Depending on the research question, these analyses can be conducted as standalone assessments of community health needs, disease burden, SDoH factors, or geographic disparities. In other cases, we link patient-level claims data or claims data aggregated at the provider-level to population health and community-level measures to provide additional context.

These analyses typically draw on population health measures such as chronic disease prevalence, provider supply, poverty levels, rurality, transportation, healthcare access and quality, age, race and ethnicity, language, household composition, and other SDoH characteristics. While these measures can be looked at individually or via existing composite measures¹, our team has also developed in-house composite measures that summarize broader patterns of community need and social vulnerability into a combined score. For example, the firm developed a county-level Community Driver Index (CDI) using socioeconomic and health outcomes measures. By linking the CDI with Medicare claims data, we calculated provider-level, weighted-average “scores” based on the counties of residence of patients served. This analysis showed that safety net providers tended to care for patients from communities with higher levels of need.

Dobson | DaVanzo applies these measures at the geographic level most appropriate for the analysis, such as census block group, census tract, ZIP Code or ZIP Code Tabulation Area, county, hospital service area, or health referral region. Analyses conducted at more granular geographic levels can help identify variation at the neighborhood-level, while analyses conducted at broader geographic levels may be better suited for examining rural health, provider supply, market-level variation, and policy impacts.

Areas of Analysis

Dobson | DaVanzo’s population health analytics work may include:

      -  Assessing how health needs, disease burden, social risk, and barriers to care vary across communities and geographies;

      -  Linking claims data with demographic, provider, geographic, and Social Drivers of Health measures;

      -  Evaluating variation in healthcare utilization, access, spending, payment impacts and outcomes, across populations, providers, and geographies;

      -  Developing indices or scoring methodologies to measure community or patient need; and

      -  Translating analytic findings into policy, payment, and strategic considerations.

This work provides our client with a nuanced understanding of how health care needs and outcomes differ across populations and how those differences may affect policy and delivery system decisions.

 ¹ Charles N. Kahn III, Kimberly Rhodes, Sarmistha Pal, Tilithia J. McBride, Donald May, Joan E. DaVanzo, and Allen Dobson. (2023). CMS Hospital Value-Based Programs: Refinements Are Needed To Reduce Health Disparities And Improve Outcomes. Health Affairs. 42 (7) https://doi.org/10.1377/hlthaff.2022.00844