Segmented Linear Regression Models for Assessing Change in Retrospective Studies in Healthcare

<div>Time to surgery versus time <i>t</i> in days from the start of the study. The vertical dashed line marks the onset of the HFU. The black line shows line-plateau as best-fitting linear model for entire period: the line has equation <i>y</i> = −0.0414<i>t</i> + 40.1868; plateau at <i>y</i> = 24.7033 reached after 375 days. Dashed blue lines are best-fitting linear models for pre-HFU and post-HFU periods. For pre-HFU period, the best model in line-line model with first line <i>y</i> = −0.0172<i>t</i> + 35.8863 and second line <i>y</i> = 0.0120 (<i>t</i> − 894) + 20.5137 with delimiting time 894 days. For post-HFU period, the best model is the plateau at <i>y</i> = 24.7359.</div>

Computational and Mathematical Methods in Medicine

fig2

Figure 2

Figure 2: Segmented Linear Regression Models for Assessing Change in Retrospective Studies in Healthcare 