Skip to main content
. 2013 Mar 1;3(3):e002453. doi: 10.1136/bmjopen-2012-002453

Table 2.

Repeated measures analysis of covariance models to identify predictors of the average Oxford Hip Score (OHS) between 1-year and 5-year follow-up

Variable Univariable
Δ coefficient (95% CI)
p Value Multivariable
Δ coefficient (95% CI)
p Value
Patient variables
 Baseline total Oxford Hip Score (10 units) 3.68 (3.16 to 4.20) <0.001 2.68 (2.16 to 3.21) <0.001
 Year 0.02 (−0.10 to 0.13) 0.77 0.01 (−0.11 to 0.13) 0.88
 Age
  <50 −1.44 (−3.92 to 1.03) 0.25
  50–60 −0.96 (−2.27 to 0.35) 0.15 −1.87 (−3.22 to −0.53) 0.006
  60–70 0.00 (0.00 to 0.00) 0.00 (0.00 to 0.00)
  70–80 −0.37 (−1.20 to 0.46) 0.38 −1.49 (−2.37 to −0.61) 0.001
  80+ −2.29 (−3.69 to −0.88) 0.001 −3.81 (−5.29 to −2.33) <0.001
 BMI (10 units) −1.14 (−2.05 to −0.22) 0.02 −1.54 (−2.45 to −0.64) 0.001
 Number of coexisting diseases −1.06 (−1.43 to −0.69) <0.001 −0.90 (−1.27 to −0.54) <0.001
 SF-36 mental health score (10 units) 0.80 (0.50 to 1.11) <0.001 0.76 (0.46 to 1.07) <0.001
Surgical variables
 Femoral component offset size (mm offset) 0.15 (0.04 to 0.26) 0.01 0.17 (0.06 to 0.28) 0.002
 R2 17.4%
 Optimism 0.8%
 Bias-corrected R2 16.6%

Δ: represents the average follow-up OHS between 1, 2, 3, 4 and 5 years follow-up.

Variables included in the final regression model are those that are retained in at least 70% of the 200 bootstrap backward selection regression models.

Univariable—each predictor in the model is adjusted for baseline OHS only.