Table 4.
Estimates obtained from least absolute shrinkage and selection operator (Lasso) regression analyses showing the prediction model of postsurgical pain at three months after surgery. (N = 1,049).
Intercept | Individual intercept | (CI) | P-value | Slope | (CI) | P value | R2 | RMSE | |
---|---|---|---|---|---|---|---|---|---|
NRS-MEP on day 1,2,3 | 0.125 | 0.325 | 0.285–0.365 | <0.001 | 0.836 | 0.710–0.961 | <0.001 | 0.142 | 1.85 |
NRS-MEP on day 1,2,3,7 | 0.086 | 0.413 | 0.371–0.455 | <0.001 | 1.476 | 1.304–1.647 | <0.001 | 0.198 | 1.79 |
NRS-MEP on day 1,2,3,7, week 6 | 0.034 | 0.669 | 0.631–0.708 | <0.001 | 3.364 | 3.178–3.551 | <0.001 | 0.460 | 1.47 |
Independent variables are pain trajectories over different time points. The dependent variable is movement-evoked pain (NRS-MEP) at three months after surgery.
NRS-MEP movement-evoked numerical rating scale, CI confidence interval, R2 means R-squared; RMSE means Root Mean Square Error.