Predictive modeling of health care costs: do cardiovascular risk markers improve prediction?

Area of Science:

• Cardiovascular Disease Research
• Health Economics
• Predictive Modeling

Background:

• Cardiovascular disease (CVD) poses a significant burden on healthcare systems.
• Identifying individuals at risk is crucial for proactive management and cost containment.
• Existing models for predicting healthcare costs often lack comprehensive CVD risk assessment.

Purpose of the Study:

• To evaluate the predictive capability of multiple cardiovascular disease (CVD) markers for future healthcare costs.
• To determine if combining traditional and novel CVD markers enhances cost prediction accuracy.
• To assess the incremental value of CVD markers over established socioeconomic predictors.

Main Methods:

• Utilized data from 2233 participants in the Study of Health in Pomerania, Germany, with no prior history of major cardiovascular events.
• Included traditional CVD risk factors (e.g., BMI, hypertension, cholesterol) and newer markers (e.g., hs-CRP, eGFR, Lp(a)).
• Employed predictive modeling to assess total cost variation at baseline and after 5-year follow-up.

Main Results:

• A basic model (sex, age, education, income) explained only 0.9% of baseline and 1.5% of 5-year cost variation.
• Incorporating a combination of significant CVD markers increased R-squared for total costs by 70% at baseline and 69% at 5-year follow-up.
• The final model achieved an R-squared of 0.030 at baseline and 0.048 at 5-year follow-up.

Conclusions:

• The simultaneous assessment of multiple CVD risk markers significantly enhances the prediction of future healthcare costs.
• This improved predictive power is observed even in individuals without a history of cardiovascular disease.
• CVD markers offer valuable additions to existing models for predicting healthcare expenditure.