Fracture healing, delayed union and nonunion · Review · 2026
Can shockwave help broken bones heal?
In this review of 7 small studies, focused shockwave was used for fresh jaw fractures and for bones that were slow to heal, mostly in one session. Next to surgery, healing rates were often close (75% vs 71% at 6 months in one study), but one study found surgery with bone grafting did better.
- small studies
- 7
- people in each study
- 18–68
- session in most studies
- 1
Good to know: Seven small studies with different fractures, methods and comparisons.
Santos TIR, Souza E Silva LC, Bonifacio M, et al. Archives of Orthopaedic and Trauma Surgery, 2026 · Read the paper ↗
How this study worked
Conceptual animation: a focused front converges, then spreads. Not to scale. Tap the field to send a pulse.
What the trials used
- Focused ESWT dose and schedule
- Table 3 lists focused ESWT in all seven reports. Reported regimens: 2,000–6,000 pulses; energy flux density 0.21–0.62 mJ/mm² where stated; Yang reported 2–3 bar and 10 Hz. Most used one session; Ramon used three sessions, one week apart, and Yang 32 sessions over four cycles. Values are source-reported; bar is not converted.
- Guidance and anesthesia
- Application used radiography, fluoroscopy (including an image intensifier) or ultrasonography depending on study; anesthesia ranged from regional/general anesthesia or sedation to conscious treatment.
- Ahmed et al. (2022): ESWT protocol
- Focused ESWT; 0.35 mJ/mm²; 4,000 pulses; 1 session; next day after surgery; regional anesthesia; LIPUS and control comparison groups.
- Gökalp et al. (2023): ESWT protocol
- Focused ESWT; 0.6 mJ/mm²; 3,000 pulses; 1 session; 3–4 foci; general or regional anesthesia; radiographic localization/marking; cryotherapy within 24 h.
- Mittermayr et al. (2022): ESWT protocol
- Focused ESWT; 0.4 mJ/mm² at 4 Hz; 3,000 pulses; 1 session; fluoroscopic identification and targeting under anesthesia.
- Quadlbauer et al. (2019): ESWT protocol
- Focused ESWT; 0.41 mJ/mm²; 3,000 pulses; 1 session; general or regional anesthesia; image intensifier localization; targeted at three extensor-region points and proximal scaphoid pole.
- Ramon et al. (2023): ESWT protocol
- Focused ESWT; 0.21 mJ/mm² at 4 Hz; 2,000 pulses; 3 sessions, once per week for three weeks; sedation and linear-probe ultrasonography to locate fracture.
- Wang et al. (2007): ESWT protocol
- Focused ESWT; 0.62 mJ/mm² and 28 kV; 6,000 pulses; 1 session; traction-table positioning and radiographic confirmation of site/depth.
- Yang et al. (2025): ESWT protocol
- Focused ESWT; 2–3 bar at 10 Hz; 2,000–3,000 pulses; 32 sessions over four cycles (twice weekly, 3–4-day interval, eight sessions/cycle); conscious, comfortable positioning; radiographic targeting; ESWT plus kinesiotherapy (KT) arm. Tables 2–3 define KT as kinesiotherapy, as do the Results’ allocation and healing paragraphs and the title of reference 25; one Results sentence calls this arm "kinesiotaping".
Who took part
Seven small studies (18 to 68 people each) of fresh jaw fractures and of long or small bones that were slow to heal or had not healed.
Side effects
The review did not add up side effects. Several studies treated people under anaesthesia or sedation.
All the numbers and the full appraisal
Every reported result
What the full paper reports.
Seven clinical studies of fracture healing, with small samples (18–68 per study) across acute mandibular fractures, long-bone and small-bone delayed union/nonunion. Designs included randomized, prospective and retrospective reports.
Radiographic fracture healing1 reported outcome
Radiographic fracture healing
Narrative synthesis; study-specific units
Seven heterogeneous clinical studies evaluated fractures. Table 4 reports ESWT consolidation of 46% versus 43% surgery at 3 months and 75% versus 71% at 6 months in Mittermayr (no statistically significant difference at either time point); 81% with surgery plus ESWT versus 75% with surgery alone in Quadlbauer (not statistically significant); all Ramon participants healed by 33 months, with median healing time 3 months in each group; Gökalp et al. reported a significantly higher success rate with debridement and bone grafting than with ESWT; the review gives no numeric effect or denominator. Wang reported earlier/higher complete healing with ESWT without extractable numeric rates. Yang reported Lane–Sandhu radiographic scores rising from approximate mean values of 4–6 to about 6–8 after ESWT+kinesiotherapy, significantly versus control; these are approximate ranges, not per-arm estimates. Ahmed reported higher mean density with ESWT, but percentage change was not statistically different among three groups. These designs cannot be pooled as a common effect.
Analysis details and source
No pooled effect; fracture type, design, co-interventions and follow-up differ. The negative Gökalp comparison and approximate Yang radiographic ranges are source-reported but lack extractable effect sizes; no values were inferred.
Pain and functional outcomes1 reported outcome
Pain and functional outcomes
Narrative synthesis; study-specific units
Reports vary by fracture and instrument. Quadlbauer reported less VAS pain and higher Mayo scores in ESWT-treated cases, without a statistically significant increase in consolidation. Ramon reported complete pain reduction in both groups through about 3 months and return to sport median 13 weeks ESWT versus 16 weeks surgery, without significant differences. Wang reported significant pain reduction, improved limb function and faster return to activity with ESWT. Yang reported pain reduction in all groups without between-group difference and improved FAC in all groups, with greater gains in groups receiving ESWT. No common estimate or uniform follow-up is supplied.
Analysis details and source
No pooled effect; fracture type, design, co-interventions and follow-up differ.
Certainty and study limitations1 reported outcome
Certainty and study limitations
Narrative synthesis; study-specific units
The review rated GRADE low for union rate and time to union (seven and five studies, respectively); Table 6 labels all contributing designs nonrandomized, although the review describes Ahmed as randomized and Ramon as prospective randomized. ROBINS-I rated Gökalp, Quadlbauer and Wang at serious overall risk and the other four studies at moderate risk.
Analysis details and source
No pooled effect; fracture type, design, co-interventions and follow-up differ.
Study-specific fracture-healing findings: Gökalp and Yang1 reported outcome
Study-specific fracture-healing findings: Gökalp and Yang
Radiographic healing and Lane–Sandhu score Source-reported rates/approximate score ranges
Gökalp et al. (2023) reported significantly higher success with debridement and bone grafting than ESWT; no numerical effect or denominator was provided in this review table. Yang et al. reported Lane–Sandhu mean scores rising approximately from 4–6 to 6–8 after ESWT plus kinesiotherapy, significantly versus control. These approximate ranges are not arm-specific values and are not reconstructed.
Analysis details and source
This preserves the review’s direction and approximations without inventing effect sizes, denominators, or p values.
How treatment was delivered
Table 3 lists focused ESWT in all seven reports. Reported regimens: 2,000–6,000 pulses; energy flux density 0.21–0.62 mJ/mm² where stated; Yang reported 2–3 bar and 10 Hz. Most used one session; Ramon used three sessions, one week apart, and Yang 32 sessions over four cycles. Values are source-reported; bar is not converted.
Application used radiography, fluoroscopy (including an image intensifier) or ultrasonography depending on study; anesthesia ranged from regional/general anesthesia or sedation to conscious treatment.
Focused ESWT; 0.35 mJ/mm²; 4,000 pulses; 1 session; next day after surgery; regional anesthesia; LIPUS and control comparison groups.
Focused ESWT; 0.6 mJ/mm²; 3,000 pulses; 1 session; 3–4 foci; general or regional anesthesia; radiographic localization/marking; cryotherapy within 24 h.
Focused ESWT; 0.4 mJ/mm² at 4 Hz; 3,000 pulses; 1 session; fluoroscopic identification and targeting under anesthesia.
Focused ESWT; 0.41 mJ/mm²; 3,000 pulses; 1 session; general or regional anesthesia; image intensifier localization; targeted at three extensor-region points and proximal scaphoid pole.
Focused ESWT; 0.21 mJ/mm² at 4 Hz; 2,000 pulses; 3 sessions, once per week for three weeks; sedation and linear-probe ultrasonography to locate fracture.
Focused ESWT; 0.62 mJ/mm² and 28 kV; 6,000 pulses; 1 session; traction-table positioning and radiographic confirmation of site/depth.
Focused ESWT; 2–3 bar at 10 Hz; 2,000–3,000 pulses; 32 sessions over four cycles (twice weekly, 3–4-day interval, eight sessions/cycle); conscious, comfortable positioning; radiographic targeting; ESWT plus kinesiotherapy (KT) arm. Tables 2–3 define KT as kinesiotherapy, as do the Results’ allocation and healing paragraphs and the title of reference 25; one Results sentence calls this arm "kinesiotaping".
No structured harms summary is presented. One included study notes local anesthesia for pain associated with high-energy ESWT; no aggregate adverse-event counts are available.
Only seven small studies with diverse fracture sites, healing status, designs, protocols, comparators and outcome definitions were included.
Table 6 labels union and time-to-union evidence as nonrandomized, but Table 2 and Results describe Ahmed and Ramon as randomized; this internal design inconsistency is unresolved.
The review assigns low GRADE certainty to fracture union rate and time to fracture union.
Individual primary trial papers were not part of the retrieved source set, so this synthesis has not independently verified the review’s trial-level abstractions.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Only seven small studies with diverse fracture sites, healing status, designs, protocols, comparators and outcome definitions were included.
- Table 6 labels union and time-to-union evidence as nonrandomized, but Table 2 and Results describe Ahmed and Ramon as randomized; this internal design inconsistency is unresolved.
- The review assigns low GRADE certainty to fracture union rate and time to fracture union.
- Individual primary trial papers were not part of the retrieved source set, so this synthesis has not independently verified the review’s trial-level abstractions.
Source-linked critical reading
Systematic clinical review
Technical source record
Complete full-text XML read, including all six raw tables and two figure captions. Individual primary study articles were outside the acquired source set.
ESWT has been studied for several fracture types, but evidence comes from small and varied studies. Some report healing or symptom improvements, while comparisons are inconsistent and certainty is low; the review does not establish a standard protocol.
What clinical evidence exists for ESWT to support fracture healing, and how do protocols and outcomes vary?
Systematic review registered as CRD420251246364; seven studies (2007–2025) included one randomized controlled study, one prospective randomized clinical study, a prospective open clinical study and retrospective evaluations/follow-ups. ROBINS-I was used despite the review including randomized designs.
Small, diverse samples addressed fresh mandibular fractures, tibial/femoral nonunion, clavicle delayed union/nonunion, scaphoid nonunion, metatarsal nonunion, and tibia/fibula delayed union/nonunion. The article reports samples of 18–68 patients per study and mixes acute fractures with nonunion.
All Table 3 entries specify focused ESWT, but regimens vary: 2,000–6,000 pulses; energy flux density 0.21–0.62 mJ/mm² where reported, or 2–3 bar for Yang; one to three sessions in most studies, versus 32 sessions in Yang over four cycles. Imaging guidance, anesthesia and target localization varied. Do not convert bar to energy density.
Comparisons included surgery, decortication/bone grafting, control, low-intensity pulsed ultrasound, and ESWT plus kinesiotherapy. Several records are adjunctive comparisons (e.g., surgery+ESWT), not ESWT monotherapy.
Most reports describe radiographic consolidation and variable pain/function measures; no common outcome definition or pooled effect was available. Study-specific values and null findings are summarized in the extraction outcomes.
The review narratively reports favorable outcomes in six of seven studies, while Gökalp favored debridement/bone graft over ESWT. Several comparisons were null or lacked numeric effects; variation and nonrandomized designs preclude a general causal conclusion.
The review does not provide a structured adverse-event synthesis. Ahmed noted local anesthesia was required for pain associated with high-energy ESWT; this does not quantify overall harms.
Only seven heterogeneous, generally small studies; fracture type, design, protocols, imaging and follow-up varied. The review’s own design classification conflicts with Table 6 GRADE, which calls all studies nonrandomized despite descriptions of randomized designs. Low certainty was assigned for union rate and time.
These findings are hypothesis-generating across distinct fracture scenarios. They do not establish ESWT as a replacement for indicated fixation or grafting, nor an optimal dose or schedule.
Conduct adequately powered fracture-site-specific RCTs with standardized union definitions, blinded imaging assessment, comparator-matched co-care, prespecified follow-up, complete harms and functional outcomes, and reproducible focused-ESWT parameters.
Read the original article.
Extracorporeal shockwave therapy in bone fractures: a systematic review.
Santos TIR, Souza E Silva LC, Bonifacio M, Campos Melo CM, Sena BG, Simões MC, Risso Parisi J, da Silva Fernandes AC, Akashi HM, de Oliveira F, Renno ACM. · Archives of orthopaedic and trauma surgery · 2026
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →