Effects of therapeutic ultrasound in patients with knee osteoarthritis: A systematic review and meta-analysis

Luo Y, Rahmati M, Kazemi A, Liu W, Lee SW, Gyasi RM, López Sánchez GF, Koyanagi A, Smith L, Yon DK · Heliyon

Pain on the visual analogue scale improved with a standardised mean difference of -0.64 (95% CI -0.88 to -0.40, I2 71%) and the WOMAC total by -0.45 (95% CI -0.69 to -0.20, I2 67%). In the subgroups, pulsed ultrasound at or below 2.5 W/cm2 gave the pain effect, larger over 24 sessions (SMD -0.80, 95% CI -1.07 to -0.53, I2 0%) than over 10 (SMD -0.71, 95% CI -1.09 to -0.33, I2 68%). No treatment-related adverse event was reported. Every endpoint is symptomatic; the review measured no joint structure.

Population

Adults aged 40 to 72 with knee osteoarthritis in 21 randomised trials, ten of them from Turkey; unlike the 2021 pooling, trials that gave ultrasound on top of exercise or hot packs were included so long as both arms got the same co-treatment

Intervention

Therapeutic ultrasound, continuous or pulsed, mostly 1 MHz at 0.03 to 2.5 W/cm2

Comparator

Sham ultrasound, or usual care with no ultrasound

Limitations

All 21 trials carried an unclear risk of bias and only three had more than 40 people per arm; heterogeneity was high; Egger's test found publication bias on the WOMAC pain subscale, which the trim-and-fill adjustment removed. The abstract's Results and its Conclusion disagree about treatment duration: the Results sentence calls 4-8 weeks superior to 4 weeks or less while quoting a smaller effect for it (SMD -0.69, 95% CI -1.13 to -0.25, against -0.77, 95% CI -1.04 to -0.49), and the Conclusion says the opposite. The authors' own verdict is that the evidence is insufficient for firm clinical recommendations. Funding disagrees with itself: the Role of the funding source paragraph names only the National Research Foundation of Korea, and the acknowledgement adds Cosmax BTI, a health-supplement company.