Page 46 - The value of total hip and knee arthroplasties for patients

Chapter 3
was used to classify the severity of OA (grade 0: indicating no OA; grade 1: doubtful OA; grade 2: minimal OA; grade 3: moderate OA; and grade 4: severe OA).12 10% of the radiographs were scored twice: correlation between both readings was used to establish the intra-reader reliability (intra-class correlation, hip radiographs: 99% (95% CI: 85–93); intra-class correlation, knee radiographs: 95% (95% CI: 92–98)). The second reading was used for further statistical analyses.The KL grade in our study was classified as mild for KL 0–2 and severe for KL3–4.
Statistics
Comparisons between patient characteristics preoperatively that were included in the analysis and those that were excluded due to incomplete follow-up were performed with chi-squared tests (for categorical variables), or independent-samples t-tests or Mann-Whitney U-tests (for continuous variables).The choice between unpaired t-tests and Mann-Whitney U-tests was based on the distribution of the variables. For this purpose, we made normality plots of all continuous variables using the Kolgomorov-Smirnov test.
Comparisons of the preoperative characteristics and the change scores over time between the groups of patients with mild and severe OA were first done with ANOVA.Then multivariable linear regression analyses were performed, with the KL grade (mild/severe) as independent variable, for the following outcomes: the mean change in the HOOS/KOOS, OHS/OKS, SF subscales, SF summary scales, EQ score, and EQ-VAS scale. All analyses were adjusted for potential confounding factors (derived from the literature on determinants of outcome in THA and TKA and/or the above-mentioned crude analyses (p > 0.10, ANOVA)). Potential confounding factors considered were the KL grade (mild/severe); age (divided into the groups 0–65 years, 66–75 years, and ≥ 76 years); sex; BMI (categorized as 0–25, 26–30, and ≥ 31); and—only for the THA group—the Charnley classification (A, B, or C). Outcome variables (dependent) were the mean change scores of the HOOS/ KOOS, OHS/OKS, SF subscales, SF summary scales, EQ score, and EQ-VAS scale. Data were analyzed using the SPSS statistical package version 20.0. All statistical testing was performed with 2-tailed tests and the level of statistical significance was set at p ≤ 0.05 for all analyses.
Ethics
The study protocol was reviewed and approved by the local hospital review board (Rijnland Hospital, Leiderdorp; registration number 10/07), which is affiliated to the
44