HOI METHODOLOGY.
The Virginia Health Opportunity Index (HOI) consists of 13 indicators that act as the building blocks. These indicators were chosen by our expert work group following an extensive review of the literature on the Social Determinants of Health (SDOH). Although there are innumerable variables and indicators that could be included, those chosen for the HOI were chosen based on the following criteria:
-
- Literature: Influence on health as expressed in literature.
- Expert Recommendations: Input from Local Health Districts and other stakeholders
- Available Data: Based on availability of data of consistent quality at the census tract level for all tracts in the state of Virginia.
It is important to note that each indicator, the profiles, and the HOI itself, are conceived as indications of the opportunity to live a long and healthy life in each area. For instance, the indicator “Access to Care” examines two variables: 1) the percent of residents without health insurance and 2) the number of full-time equivalent primary care physicians within 30 miles of the area. These two variables provide a good indication of access to care in each census tract. However, a more robust look requires examination of multiple variables, local dynamics, and the community.
The HOI provides a roadmap for the health opportunity landscape in an area, but does not tell the entire story.
Aggregation
To combine the multiple indicators, Z-scores were calculated for each indicator for each census tract in Virginia. Z-scores are a common standardization technique when using disparate data sources. These scores are weighted and combined using Principal Component Analysis into a single HOI score. The four factors identified using Principal Component Analysis, and scores are a weighted combination of the relevant Z-scores. The HOI and the Profiles are further aggregated into simple quintiles corresponding to “very low”, “low”, “average”, “high” and “very high” opportunity levels by census tract. County-level HOI and profile scores are population-weighted averages of each indicator, combined using Principal Component weights as described above. Note that while the HOI and Profiles draw from the same set of indicators, differences in weighting means that the HOI and each Profile are stand-alone measures.
- Principal Component Analysis (PCA) is a statistical procedure used to reduce the dimensionality of large datasets while retaining most of the original information. It does this by transforming a large set of variables into a smaller set of new variables, called principal components, which are uncorrelated and capture the maximum variance in the data. PCA is particularly useful for visualizing and exploring high-dimensional datasets, making it easier to identify trends, patterns, or outliers.
The Details
The variables and process used to develop each indicator are described in the following document, along with additional details on methodology and calibration.