Improved healing response in delayed unions of the tibia with low-intensity pulsed ultrasound: results of a randomized sham-controlled trial

Abstract

Background: We compared the healing response of tibial delayed unions between subjects treated with low-intensity pulsed ultrasound (LIPUS) (n = 51) and subjects treated with a sham device (n = 50). Fracture age was ≥ 4 months in all cases. Study personnel and participants were blinded to random treatment assignment throughout the study.

Methods: This multi-center randomized sham-controlled trial was undertaken at six hospitals in Germany. Adult patients who had sustained a tibial shaft fracture that subsequently showed inadequate progress toward healing (i.e., delayed union) were enrolled and randomized to receive either LIPUS (Exogen 2000/2000+, Smith & Nephew GmbH, Schenefeld, Germany) or an identical nonoperative sham device. The daily treatment duration was 20 minutes, for a period of 16 weeks. Subjects randomly assigned to active treatment had the ultrasound pressure wave signal set at the following parameters: 1.5 MHz frequency, 1 kHz repetition rate, 200 μs pulse duration, 30 mW/cm2 spatial intensity. Progress toward healing was estimated from changes in bone mineral density (BMD) and gap area as determined from computed tomography scans. Intention-to-treat analysis was conducted using a multiple imputation methodology.

Results: Based on log-transformed data, mean improvement in BMD was 1.34 (90% confidence interval (CI) 1.14 to 1.57) times greater for LIPUS-treated subjects compared to sham (p = 0.002). A mean reduction in bone gap area also favored LIPUS treatment (p = 0.014).

Conclusions: These findings demonstrate significantly greater progress toward bone healing after LIPUS treatment compared to no LIPUS treatment in subjects with established delayed unions of the tibia.

Figure 1

Line graph illustrating improvement in bone mineral density for each treatment group separately through 16 weeks of follow-up. Data based on 'completers' analysis of observed cases. The difference in mean improvements in log Hounsfield units was statistically significant (p = 0.014) with a corresponding effect size, 0.53, representing a medium degree of effectiveness.