Verikai Manual and Claims Risk Prediction Factors

How to Interpret Scores

 

Manual: Industry Loss Ratio Prediction Factor

This scalar score adjusts for age, gender, and geography in health insurance underwriting. It is used to modify base rates by applying demographic weights to determine partial manual rates for individuals and groups.

Formula Example:
Age * Gender * Geography = 1.03

Common Health Manuals:

  • HealthMAPS (Tillinghast)
  • Windsor
  • Custom Age/Gender/Zip5

Loss Ratio (Manual+ Adjustment)

The Loss Ratio score refines manual pricing by predicting loss ratios, helping adjust premiums to reach target profitability.

Definition:

  • Loss Ratio Score = (Verikai Predicted Claims) ÷ (Assumed Premiums from Manual)
  • Relative to 1.0:
    • 1.0 → Matches manual demographic expectations
    • <1.0 → Lower-than-expected risk
    • >1.0 → Higher-than-expected risk

Example Scenario

  • A 58-year-old man has a Loss Ratio Score of 1.65, meaning he is 65% riskier than manual assumptions.
  • Reason: Unmanaged diabetes likely to worsen.
  • Solution: Increase premiums by 65% to align with the target loss ratio.

Claims+ (Total Risk Score)

This score refines the manual base rate by incorporating demographic and morbidity adjustments, reflecting true expected claims relative to the manual base rate.

Relative to 1.0:

  • 1.0 → Matches the national average
  • <1.0 → Lower-than-average expected claims
  • >1.0 → Higher-than-average expected claims

Example Scenario

  • A 58-year-old man has a Claims+ Score of 2.79, meaning his expected claims are 179% higher than the manual base rate.
  • Reason: Unmanaged diabetes will likely cause increased medical costs.
  • Interpretation: If the base rate is $700 PMPM, the adjusted expectation is $1,953 PMPM.

 


 

CaptureLifestyle

Data Sources Used:
✔ 5,000+ behavioral & consumer attributes
✔ 2,000+ social determinants of health
✔ 250+ high-cost condition propensity models

 

CaptureHealth & Capture360: Advanced Risk Prediction Models

These models integrate clinical data, including ICD-10 codes, prescription data, and dependent-specific health insights, for a deeper risk analysis.

Additional Data Sources Used:

✔ 74,000+ high-cost condition ICD-10 codes
✔ 110,000+ high-cost prescription NDC codes
✔ Price & remittance data
✔ Dependent-specific health data

The Capture360 model expands further by integrating comprehensive behavioral, demographic, and clinical data for enhanced risk prediction and premium optimization.


Final Takeaways:

  • Manual+ adjusts premiums based on predicted loss ratios.
  • Claims+ estimates true expected claims relative to the base rate.
  • CaptureHealth and Capture360 refine risk assessment using deeper clinical and behavioral data.
  • Real-world application: Adjust premiums and underwriting strategies to improve accuracy and profitability.