What parameters affect knee function in patients with untreated cartilage defects: baseline data from the German Cartilage Registry.

PURPOSE: To evaluate the factors influencing the baseline Knee Injury and Osteoarthritis Outcome Score (KOOS) in patients with knee cartilage defects and planned cartilage repair surgery and to provide baseline KOOS data from a large patient population.

MATERIAL AND

METHODS

Between October 2013 and April 2017, a total of 2815 patients assigned for cartilage repair surgery were included into the German Cartilage Registry (KnorpelRegister DGOU) and their data were analyzed for the present study. Multivariate regression model and ANOVA were used to detect patient- and defect-specific factors with an influence on baseline KOOS. In addition, KOOS baseline data was calculated and compared according to these parameters.

RESULTS

Sex, age, body mass index (BMI), and smoking status were revealed as patient-specific factors, and defect location and the number of previous knee and cartilage operations were revealed as defect-specific factors with a significant influence on baseline KOOS. Most subscores were affected in accordance with the total KOOS. Interestingly, defect ICRS grade, defect size, and symptom duration had no significant influence. The mean baseline KOOS was 56.7 (± 17.9). Men had significantly higher mean overall KOOS (60 ± 17.3 vs. 51.8 ± 17.6, p < 0.001) than women, and patients with a BMI over 30 and smokers scored significantly lower (58.07 ± 17.67 vs. 50.32 ± 17.29, p < 0.001; 57.64 ± 17.86 vs. 53.59 ± 18.06, p < 0.001). Patients with two or more previous knee operations as well as patients with more than one previous cartilage procedure also showed significantly lower overall KOOS (57.19 ± 17.89 vs. 54.56 ± 17.58, p < 0.001; 57.68 ± 18.01 vs. 52.72 ± 17.58, p < 0.001).

CONCLUSION

Several factors influencing baseline KOOS data in patients with knee cartilage defects assigned for cartilage repair surgery could be detected. Their individual influence in the multivariate linear regression model was not very strong. Baseline data according to these criteria is presented in this paper.