Predictive models in the Usual-Support Group were aimed at identifying individuals through medical claims and administrative data (such as hospitalization notification). The output of the predictive models is a rank-ordered, or stratified, list of individuals who have support needs. These lists were then used to generate outbound mail, interactive voice response (IVR) calls or calls by health coaches.

The Enhanced-Support Group intervention used more sophisticated predictive models, more extensive outreach to engage individuals, and provided tighter feedback loops to inform the care support process.

The Enhanced-Support Group program used lower risk thresholds for targeted outreach of individuals with chronic illness and for those likely to face discrete surgical interventions. Individuals at high financial risk but without one of the five chronic conditions were included in the Enhanced-Support Group.

The Enhanced-Support Group received up to five initial reach attempts while the Usual-Support Group received up to three attempts.

The Enhanced-Support Group used a 'tight loop' management structure including detailed, timely coach level feedback reports on efficiency and effectiveness.