Abstract: BACKGROUND: There are individual differences in the effectiveness of arthroscopic surgery in improving knee osteoarthritis, and subjective scoring scales may be biased in evaluating the clinical efficacy of arthroscopic treatment for knee osteoarthritis.
OBJECTIVE: To explore the correlation between gait parameters and joint function recovery in patients with knee osteoarthritis after arthroscopic surgery.
METHODS: A total of 98 patients with knee osteoarthritis admitted to Chongqing Armed Police Corps Hospital from October 2023 to October 2024 were selected as the research subjects. According to Lysholm knee function score after 6 months of follow-up, they were divided into the excellent group (n=63) and the fair group (n=35). Clinical data including gender, age, body mass index, disease duration, respiration, heart rate, Kellgren-Lawrence grade, smoking history, drinking history, hypertension history, location of onset, and postoperative complications were collected. The intraoperative and postoperative indicators, as well as the knee joint function scores and gait parameters at different times before and after surgery were compared between the two groups. Multivariate Logistic regression was used to analyze the independent influencing factors of knee joint function recovery. Stratified regression analysis was conducted to explore the impact of different clinical and pathological characteristics after treatment on gait parameters. Generalized estimating equations were used to analyze the differences in gait parameters among patients with different knee joint functions after treatment. Generalized additive models were used to analyze the impact of gait parameters after treatment on the Lysholm score. Receiver operating characteristic curves were drawn to analyze the value of gait parameters in predicting the recovery of knee joint function after treatment.
RESULTS AND CONCLUSION: (1) There were statistically significant differences in age, disease duration, Kellgren-Lawrence grade, and postoperative complications between the excellent group and the fair group (P < 0.05). (2) The fair group had longer operation time, more intraoperative blood loss, longer time for swelling to subside, and longer rehabilitation time than the excellent group (P < 0.05). (3) The fair group had higher Western Ontario and McMaster Universities Osteoarthritis Index and Visual Analogue Scale scores, and lower Lysholm scores, step frequency, and step speed than the excellent group (P < 0.05). (4) Logistic regression analysis showed that age, Kellgren-Lawrence grade, postoperative complications, and time for swelling to subside were independent risk factors for knee joint function recovery (P < 0.05). (5) Stratified regression analysis showed that age, Kellgren-Lawrence grade, time for swelling to subside, and postoperative complications had negative effects on step frequency and step speed (β < 0, P < 0.05). (6) Generalized estimating equation analysis showed that there was a correlation between the degree of knee joint function recovery and gait characteristics (β > 0, P < 0.05). (7) Generalized additive model analysis showed that the impact of step frequency and step speed on the Lysholm score after treatment was linear. (8) Receiver operating characteristic curve analysis showed that the combined detection of step frequency and step speed had higher predictive efficacy (Area under the curve > 0.85, P < 0.05). (9) It suggests that arthroscopic surgery for knee osteoarthritis can improve knee joint function. By dynamically tracking changes in postoperative gait parameters and combining with functional scoring scales, the correlation between step frequency, step speed, and the outcome of knee joint function can be revealed, further clarifying its clinical predictive value. This provides a new quantitative tool for clinical functional assessment and enables more precise postoperative rehabilitation guidance.

Key words: knee osteoarthritis, arthroscopy, minimally invasive surgery, gait parameter, knee function, Logistic regression analysis