Skip to main content
. 2021 Aug 12;11:16459. doi: 10.1038/s41598-021-95963-y

Table 3.

Estimates obtained from least absolute shrinkage and selection operator (Lasso) regression analyses showing the prediction models of acute postsurgical pain. (N = 1,579) Independent variables are the preoperative predictors. The dependent variables are movement-evoked acute postsurgical pain (NRS-MEP) at day 1, 2, 3 and over the first three days after surgery.

Preoperative predictors Day 1 Day 2 Day 3 Day 123
Est RMSE R2 Est RMSE R2 Est RMSE R2 Est RMSE R2
2.52 0.162 2.44 0.163 2.18 0.176 2.11 0.207
Intercept 4.082 3.564 3.099 3.669
Hospital admittance 0.018 0.087 0.091
Female sex 0.101 0.019
Preoperative movement evoked pain 0.148 0.130 0.148 0.146
Age − 0.017 − 0.016 − 0.017 − 0.021
PCS total 0.010 0.019 0.014 0.015
APAIS anxiety 0.051 0.028 0.041 0.047
PDI total 0.012 0.011 0.008 0.012
Category expected pain
Category expected pain 1
Category expected pain 2
Category expected pain 3
Category expected pain 4 0.094
Category expected pain 5 0.138 0.286 0.161 0.272
Medical specialty
General surgery
Orthopedics
Urology 0.053
Gynecology
ENT − 0.157 − 0.087
Eye surgery
Plastic surgery
Neurosurgery
Oro-maxillofacial surgery
Other surgeries
Wound size unknown* 0.187
Wound size > 10 cm* 0.197 0.170 0.265

Est means estimate; RMSE means Root Mean Square Error; R2 means R-squared; PCS means Pain Catastrophizing Scale; APAIS means Amsterdam Pain Anxiety and Information Scale; PDI means Pain Disability Index. Meaningful predictors in one of the four prediction models are displayed in bold font.

* compared to wound size ≤ 10 cm.