Multiple studies have shown that the symptomatology of knee osteoarthritis weakly correlate to radiographic severity of disease. Current literature however does not have much in the way of comparing functional outcomes of those with OA knees with radiographic severity. Our objective was to compare radiographic measurements of OA knees with self-reported functional outcomes and determine if radiographic severity of OA knees correlated with loss of functional ability.

A retrospective review of prospectively collected registry data of 305 patients with osteoarthritis of the knee was collected. The patient's x-rays were reviewed, and radiographic measurements were taken to include medial, lateral and patellofemoral joint space distance measured in millimetres. The Kellgren and Lawrence, and Ahlback classifications of radiographic knee OA were computed. These were correlated with severity of functional limitations was measured using the SF36, Knee society score (KSS) and Oxford knee scores.

Statistical analysis were conducted with SPSS V22.0 statistical software. Demographic characteristics and functional assessments were analysed using one way ANOVA test. Post-hoc test using Tukey HSD and effect size (partial-eta squared η²) was performed if one-way ANOVA was found to be statistically significant. A p-value of 0.05 or less was considered statistically significant.

Pre-operative patient demographics are shown in table 1. Patients in with Grade 2 osteoarthritis were significantly younger than Grade 4 patients (post-hoc p=0.003). There were no statistically significant differences in age between the other Grades, and there were no differences in BMI or gender or operative site between all grades.

There were significant differences in KSS Function scores between Grade 2 and Grade 3 patients (post-hoc p=0.017) and Grade 2 and 4 patients (post-hoc p < 0 .001). Statistically significant differences were also found between Grade 1 and Grade 4 patients for the KSS Knee score (post-hoc p=0.016). There were significant differences in Oxford knee score (post-hoc p=0.026) and SF- Physical Function (post-hoc p < 0 .001) between Grade 2 and Grade 4 patients too.

The effect size η² for KSS Function, KSS Knee and Oxford knee score was 0.05, 0.06 and 0.33 respectively. When comparing the loss of joint space with the functional scores, there were no statistically significant correlations. Our study show that the radiological severity of knee osteoarthritis based on the two scoring methods was able to correlate with worsening functional scores. Most notably, the differences in KSS function scores correlated strongly between Grade 2 and Grade 3 patients. Of note, there was no correlation between the loss of joint space and the severity of functional limitations across any of the scoring systems. Our study showed that although both the Kellgren and Lawrence and Ahlback radiological grading of Osteoarthritis were able to correlate with worsening functional scores, this was not due to loss of joint space alone and further studies need to be conducted on the other contributors to the scoring system such as osteophytes and subchondral sclerosis.

Our study show that the radiological severity of knee osteoarthritis based on the two scoring methods was able to correlate with worsening functional scores. Most notably, the differences in KSS function scores correlated strongly between Grade 2 and Grade 3 patients. Of note, there was no correlation between the loss of joint space and the severity of functional limitations across any of the scoring systems. Our study showed that although both the Kellgren and Lawrence and Ahlback radiological grading of Osteoarthritis were able to correlate with worsening functional scores, this was not due to loss of joint space alone and further studies need to be conducted on the other contributors to the scoring system such as osteophytes and subchondral sclerosis.

For any figures or tables, please contact authors directly.