Skip to main content

Table 2 Baseline Patient Characteristics

From: Development and internal validation of machine learning algorithms to predict patient satisfaction after total hip arthroplasty

Variables Training set
(n = 1206)
Test set
(n = 302)
P-value
Age 62.8 (12.1) 63.2 (11.8) 0.623
Sex (Female) 841 (69.7%) 211 (69.9%) 0.964
BMI 25.9 (4.8) 25.3 (4.3) 0.075
BMI (categorical)    0.488
  < 18.5 39 (3.2%) 14 (4.6%)
 18.5–29.9 982 (81.4%) 241 (79.8%)
  ≥ 30 185 (15.3%) 47 (15.6%)
Comorbidities
 Number of comorbidities 1.0 (1.1) 0.9 (1.1) 0.417
 Diabetes 133 (11.0%) 30 (9.9%) 0.584
 Hypertension 518 (43.0%) 115 (38.1%) 0.125
 High cholesterol 389 (32.3%) 93 (30.8%) 0.626
 IHD 52 (4.3%) 16 (5.3%) 0.460
 Stroke 17 (1.4%) 7 (2.3%) 0.259
 Renal disease 22 (1.8%) 7 (2.3%) 0.576
 Back pain 35 (2.9%) 7 (2.3%) 0.581
 Depression 7 (0.6%) 2 (0.6%) 0.999
Surgical History
 Previous knee surgery 133 (11.0%) 37 (12.3%) 0.548
 Previous hip surgery 236 (19.6%) 64 (21.2%) 0.527
 Previous lumbar spine surgery 69 (5.7%) 28 (9.3%) 0.025*
Preop PROMs
 SF-36 PCS 27.4 (8.9) 26.7 (9.8) 0.233
 SF-36 MCS 49.1 (12.1) 49.2 (12.2) 0.899
 WOMAC 49.5 (20.9) 48.5 (20.9) 0.473
 OHS 40.0 (9.2) 40.8 (9.3) 0.190
2-year PROM Improvement
 SF-36 PCS + 20.2 (12.0) + 19.8 (12.4) 0.601
 SF-36 MCS + 6.9 (12.1) + 6.7 (12.6) 0.804
 WOMAC + 41.4 (20.9) + 41.6 (22.1) 0.921
 OHS −24.1 (9.7) −24.4 (10.2) 0.658
2-year Satisfaction
 Satisfied 69 (5.7%) 17 (5.6%) 0.951
  1. Continuous outcomes are reported as mean (standard deviation) while categorical outcomes are presented as number (percentage)
  2. P-values are calculated using two-sample t-tests for continuous variables and Chi-squared test/Fisher’s exact test for categorical variables
  3. *: P-value < 0.05