Link → Use predictive modeling in Medicare populations
Predictive modeling has many implications for compliance enhancment.
Excerpts:
Most predictive modelers are heavily dependent on medical and pharmacy claims data. It is well known that not all claims are created equal; for example incomplete or inaccurate coding, bundling or up-coding, and lags in submission may occur. Relying on information from claims data alone for validation of risk can lead to inappropriate allocation of services and missed opportunities for impact. Integration of key non-claims driven information is necessary to adjust the predicted risk.
Considerations include:
- Beneficiary-provided information can be invaluable in identifying risks. Behavioral aspects of a beneficiary, such as self-confidence, perceived barriers and readiness for change, can have significant impact on a person’s ability to actively participate in management of their health.
- Chronically ill and elderly individuals are at a greater risk for isolation and depression. Offering free access to a Health Risk Assessment (HRA) for beneficiaries, either on a Web portal or via a mailed paper form, can assist in identifying individuals with potential socialization or depression identified risks.
- The literacy level of an individual can influence their health outcomes and risk. If a beneficiary cannot understand instructions or read educational materials given to them on managing their chronic medical conditions, the chance for compliance with treatment regimes is greatly reduced. Non-compliance increases overall risk of disease progression and less than optimal outcomes.
… Integration of key information from disparate sources provides a more comprehensive view of a beneficiary across the healthcare continuum.
System driven processes and integrated quality of care alerts/reminders facilitate efficient, consistent workflows designed to meet the complex needs of Medicare care management programs.
Ongoing monitoring of targeted interventions for at-risk and high-risk beneficiaries through analysis and reporting tools enables efficient, comprehensive identification of program trends, risks and opportunities.
… Predictive modeling can help programs working with Medicare beneficiaries focus on identification and stratification of individuals at highest risk for costs and adverse outcomes. Combining this information with non-claims based beneficiary information can facilitate successful care management strategies and interventions resulting in positive clinical and financial outcomes.
