The SDOH Composite Stress Score is a proprietary Algorex Health model that measures compound stress caused by social influences on an individual. This model uses over 150 inputs (including age and variables such as household size and residence type) and has been trained against clinical and utilization outcomes on a Medicaid population across multiple states.

The score is particularly useful when assessing member or patient risk when the individual is lacking claims or clinical information, or when included as an additional factor on top of existing clinical risk scores.

The model output is a score represented as a decimal on a scale of 0 through 10, with higher scores suggesting increased costs and ED utilization in particular.

Member’s Composite SDOH quintile values include:

A combined score based on a range of factors such as income, criminal history, household composition, educational attainment and other indicators. Meant to provide an overall factor for social determinant risk stratification. It is similar to the SDOH Composite Stress Score, but is not trained against cost and utilization, so it is purely based off of member acuities.

Member’s Social Vulnerability quintile values include:

The Unstable Housing Score is calculated using a proprietary Algorex Health data science model designed to predict the likelihood this member will move within the upcoming 60 days. This is calculated primarily through attributes of affordability and recent length of residency as reported through client and external sources and is trained on a large set of patient data.

This score is useful for the identification of target members for outreach to request updated contact information and for identifying members who are likely to lose health plan coverage, as well as offering financial support for rent or utilities. These scores are a 0-10 probability value, with 10 being the highest likelihood of upcoming address change.

Member’s Unstable Housing tier values include:

Algorex Health maintains a national database of almost 10,000 homeless shelters, housing support agencies and state office buildings. These identified locations are compared to the members address:

The Neighborhood Stress Score builds on a model originally designed by the University of Massachusetts, which is used for risk adjustment in the state of MA. This score, known as NSS for short, uses seven census block group variables to establish an index with which neighborhoods can be compared to each other. The index is composed of the following seven values:

The score is a decimal calculated using a Z-Score based on the statewide population, with a score of 0 being the state mean. A negative score indicates below average stress, and a positive score indicates higher than average stress. It is calculated on the census block group level, so all individuals residing in the same census block group will have the same NSS.

Member’s Neighborhood Stress tier values include:

The Transportation Access Score estimates the level of private vehicle and public transit access and usage in the member’s respective geography (census block group). The score is derived from ZScore of Census Table S0801 ‘means of commuting to work’. The scores are represented as a decimal between 0 and 10, with a score of 10 indicating poorer access to transportation.

Member’s Transportation Access quintile values include:

The Food Access Score is a proprietary Algorex Health data science model that assesses physical access to food as well as the likelihood that the member will be able to reliably afford food. It is the result of a formula containing both accessibility and affordability (elements Q and S).

The resulting score is a decimal ranging from 0-16. A high score indicates worse overall access to high quality food. The model is used to target food-based interventions such as delivery, meal supplementation, and SNAP enrollment support. The model does not consider member choice (the third aspect of food security).

Member’s Food Access tier values include:

The Social Isolation Score is a proprietary Algorex Health model that is intended to identify members who are at risk of being isolated and lacking social supports. Inputs include access to public transportation, vehicle ownership, household composition, and neighborhood factors.

This model is often used to identify target members for a togetherness program or home visit. The model output is a score represented as a whole number on a scale of 0 through 4, with higher scores indicating greater likelihood of isolation.

Member’s Social Isolation tier values include: