Abstract
Background
To identify and rank the importance of key determinants of end-of-life (EOL) healthcare costs, and to understand how the key factors impact different percentiles of the distribution of healthcare costs.
Methods
We applied a principled, machine learning based variable selection algorithm, using Quantile Regression Forests, to identify key determinants for predicting the 10 th (low), 50 th (median) and 90 th (high) quantiles of EOL healthcare costs, including costs paid for by Medicare, Medicaid, Medicare Health Maintenance Organizations (HMO), private HMO, and patient’s out-of-pocket expenditures.
Results
Our sample included 7,539 Medicare beneficiaries who died between 2002 and 2017. The 10 th, 50 th and 90 th quantiles of EOL healthcare cost are $5,244, $35,466 and $87,241 respectively. Regional characteristics, specifically, the EOL-expenditure index, a measure for regional variation in Medicare spending driven by physician practice, and the number of total specialists in the hospital referral region, were the top two influential determinants for predicting the 50 th and 90 th quantiles of EOL costs, but were not determinants of the 10 th quantile. Black race and Hispanic ethnicity were associated with lower EOL healthcare costs among decedents with lower total EOL healthcare costs but were associated with higher costs among decedents with the highest total EOL healthcare costs.
Conclusions
Factors associated with EOL healthcare costs varied across different percentiles of the cost distribution. Regional characteristics and decedent race/ethnicity exemplified factors that did not impact EOL costs uniformly across its distribution, suggesting the need to use a “higher-resolution” analysis for examining the association between risk factors and healthcare costs.
A Method for Confidence Intervals of High Quantiles
