Model 2 - Hierarchical logistic regression analysis of Persistent Post-surgical Pain 4–6 months (T3) following TKA and THA on demographic and clinical baseline measures (T1), and post-surgical pain and anxiety 48 h after surgery (T2).

Note. After removing 2 outliers, this final model correctly predicted 75% of all patients; χ2(5) = 27.935 pR2 = 0.372; OR = odds ratio; CI = confidence interval; bold = significant at p≤.05.1Continuous variable, NRS - Numerical Rating Scale (0–10) from BPI-SF: Brief Pain Inventory-Short Form;2Continuous variable;3Dichotomic variable: 0 = THA: Total Hip Arthroplasty; 1 = TKA: Total Knee Arthroplasty;aContinuous variable, HADS-A: Hospital Anxiety and Depression Scale - anxiety subscale.注：在剔除2个异常值后，本最终模型可正确预测全部患者中的75%；χ²(5)=27.935，pR²=0.372；OR为优势比（odds ratio），CI为置信区间（confidence interval）；结果以加粗格式标注表示在p≤0.05水平下具有统计学显著性。1 连续变量：源自简明疼痛量表简版（BPI-SF, Brief Pain Inventory-Short Form）的数字评定量表（NRS, Numerical Rating Scale，0~10分）；2 连续变量；3 二分类变量：0代表全髋关节置换术（THA, Total Hip Arthroplasty），1代表全膝关节置换术（TKA, Total Knee Arthroplasty）；a 连续变量：医院焦虑抑郁量表焦虑分量表（HADS-A, Hospital Anxiety and Depression Scale - Anxiety Subscale）。