The use of low-intensity pulsed ultrasound for the treatment of established nonunion fractures: A systematic review and meta-analysis

BACKGROUND: Systematic reviews on low-intensity pulsed ultrasound (LIPUS) in nonunion fractures report a positive effect on healing. The quality of evidence is, however, limited as it is mostly based on observational studies. The literature on LIPUS in nonunion fractures has grown considerably over the past years. We aimed to conduct an up-to-date systematic review and meta-analysis of the evidence on the effect of LIPUS on healing rate and time in established nonunion fractures. METHODS: A literature search was conducted in Medline, CINAHL, Embase, and Web of Science, from their inception to May 6, 2025, using keywords and MeSH terms of LIPUS and nonunions. Case series, cohort studies, or randomized controlled trials (RCTs) investigating the effect of LIPUS on healing rate or healing time of nonunion fractures were included. Studies in a language other than English or French, on animal populations, or fresh fractures were excluded. Two independent reviewers performed study selection, data extraction, and quality assessment with the risk of bias tool for randomized controlled trial or the Risk of Bias In Non-randomized Studies of Interventions tools. Distinct meta-analyses were performed for observational studies and RCTs to estimate the pooled effects on healing rate and time. The certainty of evidence was assessed with the Grading of Recommendations Assessment, Development and Evaluation. RESULTS: Three RCTs (n = 264 participants) and 30 observational studies (n = 3437 participants) presenting low to serious risk of bias were included. The pooled healing rate following LIPUS in observational studies (n = 3439 nonunions) was 0.79 (95% confidence interval [CI]: 0.73-0.85; P < .01) with considerable heterogeneity (I2 = 92%) and a very low certainty of evidence. The meta-analysis of RCTs on healing rate (3 studies, n = 264 nonunions) reached a nonsignificant risk ratio of 1.13 (95% CI: 0.93-1.37, P = .23) in favor of LIPUS compared to sham, with substantial heterogeneity (I2 = 60%) and low certainty. The meta-analysis of RCTs on healing time (2 studies, n = 135 nonunions) reached a nonsignificant mean difference of 14.99 days (95% CI: -120.37 to 150.35; P = .83) between LIPUS and sham with substantial heterogeneity (I2 = 74%) and low certainty of evidence. CONCLUSION: Current evidence does not support the use of LIPUS as a first-line intervention in established nonunion fracture management, although uncertainty remains. High-quality placebo-controlled RCTs are warranted.