Figure 2. Distribution of Clinicians Affiliated vs Not Affiliated With Health Systems by Their 2019 Merit-based Incentive Payment System (MIPS) Final Performance Score.

Density approximates the number of clinicians at each MIPS data point. The calculation is smoothed because MIPS score is continuous. Epanechnikov kernel with a bandwidth of 2 was used to minimize the mean integrated squared error of the kernel density estimate. A larger bandwidth would result in a smoother plot and a narrower bandwidth would result in a more jagged plot.