© 2019 by Curalogix LLC

Episodic Risk Stratification

In addition to our customized healthcare data solutions, we have also developed an episodic risk stratification model to assist organizations that are participating in Medicare's Bundled Payment for Care Improvement (BPCI) or Medicare's Comprehensive Care for Joint Replacement (CJR) programs. 


Unlike other risk stratification methodologies that are based on insurance risk for populations, the Curalogix episodic risk stratification is built specifically for the inpatient clinical episodes of the BPCI Advanced or CJR program.  To effectively manage patients from the onset of a clinical episode, it is necessary for organizations to take into account the clinical complexity of each patient.  Our proprietary predictive model provides BPCI and CJR participants with the ability to understand the clinical and financial risks of each patient at the onset of the episode.

Model Inputs:

  1. Patient's Clinical Episode

  2. Patient Age

  3. Patient Diagnosis (all primary and secondary diagnosis codes)

  4. Is Patient Dual Eligible (Medicare & Medicaid)?

  5. Is Patient Clinically Managed?

Model Output:

In the above example, the patient is predicted to have an Episode Risk Factor (ERF) of 1.06, indicating that they are expected to be 6% more costly than average.  For patients with the same risk profile, 45.2% were discharged to home whereas 43.8% were discharged to a skilled nursing facility.  The additional services indicates the incidence rates of the different post-acute settings for each discharge disposition.  In the above example, 2% of the patients discharged to home also utilized IRF, 3% of the patients utilized SNF, 77% of the patients utilized HH, 88% of the patients utilized OP and 11% of the patients were readmitted. 

In addition to providing an Episode Risk Factor, we are also able to assist organizations in calculating an estimated real dollar amount per episode.  We would simply need a couple of data elements from the Target Price calculation provided by Medicare.

To learn more about the predictive capability of the model, please download our white paper by entering your email below: