Multiple tendon conditions · Review · 2023
Does adding shockwave to exercise help jumper's knee?
In this review, two trials added shockwave to a lengthening (eccentric) exercise programme for jumper's knee. In the short term, pain and knee function ended about the same as with exercise plus a placebo treatment. The review supports exercise as the main treatment.
Similar to exercise with placebo in the short term- trials added shockwave to exercise for jumper's knee
- 2
- trials in the whole review
- 68
- conditions: Achilles, kneecap tendon and outer hip pain
- 3
Good to know: Based on 2 short-term trials within a larger review of 68 exercise trials.
Challoumas D, Crosbie G, O'Neill S, et al. Sports Medicine - Open, 2023 · Read the paper ↗
How this study worked
What happened
- trials adding shockwave to exercise
- 2
- Compared with
- Eccentric (lengthening) exercise with a placebo treatment
Knee pain with shockwave added to exercise
Difference in pain score points; below zero favours shockwave
Show the numbers
| Difference (pain score points) | likely range | |
|---|---|---|
| Knee pain, short term | −0.09 | −1 to 0.84 |
What the trials used
- Intervention framework
- The review compared exercise loading (especially eccentric exercise) alone or with adjuncts, including ESWT, PRP, low-level laser, injections and other modalities. Many adjuncts were only studied on top of exercise; their monotherapy effects cannot be isolated. For patellar tendon pain/function, eccentric exercise plus ESWT was compared with eccentric exercise plus placebo ESWT.
- Analysis framework
- Pairwise meta-analysis used raw mean differences for the same measures at similar timepoints and pooled only within the same diagnosis. Network meta-analyses were performed for Achilles and patellar pain VAS and VISA outcomes when data allowed; no GTPS network was feasible.
Who took part
People in 68 trials: 1,792 with Achilles tendon pain, 1,109 with jumper's knee and 907 with outer hip pain. Average ages were 46, 28 and 55.
Side effects
In the outer hip pain trials, there were no tendon tears or other major complications with shockwave or steroid injections; skin irritation was shockwave's most common complication. No side-effect data were given for jumper's knee.
All the numbers and the full appraisal
The results, in numbers.
68 RCTs across non-insertional Achilles tendinopathy (31 studies, 1,792 participants), patellar tendinopathy (31, 1,109) and greater trochanteric pain syndrome (GTPS; 6, 907). Mean ages by condition were 46, 28 and 55 years. Follow-up was mostly short or intermediate term.
ESWT added to eccentric exercise: patellar pain VAS1 reported outcome
ESWT added to eccentric exercise: patellar pain VAS
Pain visual analogue scale mean difference, VAS points
Eccentric exercise plus ESWT versus eccentric exercise plus placebo ESWT (the Fig. 2 caption labels ESWT the intervention and placebo the control)
2 studies · I²=0%
No short-term difference was found for pain (MD −0.09, 95% CI −1.00 to 0.84; p=0.84; I²=0%; 2 studies); evidence strength was moderate.
Analysis details and source
The mean difference retains the source comparison direction and scale; a non-significant result does not establish equivalence. Strength of evidence and clinical importance are stated separately.
ESWT added to eccentric exercise: VISA-P1 reported outcome
ESWT added to eccentric exercise: VISA-P
Victorian Institute of Sport Assessment–Patella mean difference, VISA-P points
Eccentric exercise plus ESWT versus eccentric exercise plus placebo ESWT (the Fig. 2 caption labels ESWT the intervention and placebo the control)
2 studies · I²=0%
No short-term VISA-P difference was found (MD −1.8, 95% CI −8.0 to 4.4; p=0.57; I²=0%; 2 studies); evidence strength was moderate. Authors state this supports no added benefit of ESWT in this setting.
Analysis details and source
The mean difference retains the source comparison direction and scale; a non-significant result does not establish equivalence. Strength of evidence and clinical importance are stated separately.
PRP added to eccentric exercise: VISA-A3 reported outcomes
PRP added to eccentric exercise: VISA-A
Victorian Institute of Sport Assessment–Achilles mean difference, VISA-A points
Eccentric exercise plus PRP injection versus eccentric exercise plus placebo injection
3 studies · I²=5%
No short-term difference was found (MD 1.52, 95% CI −2.49 to 5.53; p=0.46; I²=5%; 3 studies); low-strength evidence.
Analysis details and source
The mean difference retains the source comparison direction and scale; a non-significant result does not establish equivalence. Strength of evidence and clinical importance are stated separately.
PRP added to eccentric exercise: VISA-A
Victorian Institute of Sport Assessment–Achilles mean difference, VISA-A points
Eccentric exercise plus PRP injection versus eccentric exercise plus placebo injection
3 studies · I²=0%
Mid-term results favored PRP adjunct (MD 5.05, 95% CI 1.45 to 8.64; p=0.006; I²=0%; 3 studies), but the authors judged the difference unlikely clinically significant and evidence low strength.
Analysis details and source
The mean difference retains the source comparison direction and scale; a non-significant result does not establish equivalence. Strength of evidence and clinical importance are stated separately.
Low-level laser added to eccentric exercise: VISA-A
Victorian Institute of Sport Assessment–Achilles mean difference, VISA-A points
Eccentric exercise plus low-level laser therapy versus eccentric exercise plus placebo
2 studies · I²=85%
The pooled estimate favored laser adjunct (MD 6.29, 95% CI 1.72 to 10.85; p=0.007; I²=85%; 2 studies), but evidence was very low strength and the benefit was judged unlikely clinically significant.
Analysis details and source
The mean difference retains the source comparison direction and scale; a non-significant result does not establish equivalence. Strength of evidence and clinical importance are stated separately.
Pain VAS3 reported outcomes
Corticosteroid injection versus exercise: GTPS pain
Pain VAS mean difference, VAS points
The Results describe short-term pain as favoring corticosteroid injection over exercise and print MD 1.57, 95% CI “(1.1. 2.1)”, I²=94%, p=0.01, 2 studies. The interval punctuation is malformed/ambiguous; no corrected bounds are inferred. Direction coding is not explicit in this passage.
Analysis details and source
Source text contains two periods in its short-term CI string. The reported point estimate, p value, heterogeneity and study count are retained narratively; the malformed CI is not charted.
Corticosteroid injection versus exercise: GTPS pain
Pain VAS mean difference and 95% CI as printed
The source prints a long-term estimate of MD −0.7 with 95% CI −0.6 to 0.45 (p=0.78; I²=97%; 2 studies). The point estimate lies outside the stated interval, so it is preserved narratively and not structured as a valid effect; the authors report no significant difference.
Analysis details and source
Internal source inconsistency; do not repair or infer corrected interval without the figure/supplement.
Patellar tendinopathy: isometric vs isotonic exercise pain
Pain VAS mean difference in VAS points, as reported
For immediate post-intervention pain, the authors report no statistically significant difference between isometric and isotonic exercise (MD 1.03, p=0.19, I²=70%, 3 studies). The stated 95% CI is −2.57 to 0.5, which does not contain the point estimate; the inconsistency is preserved and the interval is not treated as a valid effect estimate.
Analysis details and source
Null pairwise finding retained from Results; source CI/estimate inconsistency prevents valid numeric charting. The review reports low-strength evidence.
Achilles NMA treatment rankings2 reported outcomes
Achilles NMA treatment rankings
Network treatment rank probability probability of rank as reported
Results prose reports eccentric exercise plus HVI with corticosteroid as having the highest probability of being most effective for short-term VISA-A and pain VAS (56% and 45%) and mid-term VISA-A (83%). For long-term VISA-A, eccentric exercise plus prolotherapy is reported at 55%, with one contributing study. However, Figure 3 captions describe its displayed panels as median treatment ranks; those graphics were not acquired, so the text-versus-caption metric discrepancy cannot be resolved. These ranking statements are not comparative effect estimates or proof of superiority.
Analysis details and source
Percentages are transcribed from the Results prose, not from the unavailable graphic. The caption for Figure 3 names median ranks, creating a source metric discrepancy; neither metric is converted or inferred. The Results text also cites Fig. 3a–b for short-term, 3c for mid-term and 3d for long-term results, whereas the Fig. 3 caption labels panel b as mid-term VISA-A, c as long-term VISA-A and d as short-term pain VAS; the panel letters in the locator follow the Results text.
Patellar NMA treatment rankings
Network treatment rank probability probability of rank as reported
Results prose reports eccentric exercise plus hyaluronic-acid injection as having the highest probability of being most effective for short-term VISA-P (75%); moderate-load slow resistance exercise is reported at 22% for pain VAS. Each of these top-ranked interventions was informed by a single study. Figure 4 captions describe the displayed panels as median treatment ranks; those graphics were not acquired, so the text-versus-caption metric discrepancy cannot be resolved. Rankings are not comparative effect estimates or proof of superiority.
Analysis details and source
Percentages are transcribed from Results prose, not from the unavailable graphic. Figure 4 caption names median ranks, creating a source metric discrepancy.
GTPS network feasibility1 reported outcome
GTPS network feasibility
Network evidence availability not estimable
The authors could not perform a network meta-analysis for GTPS because data were insufficient. Pairwise short-term corticosteroid-versus-exercise results were highly heterogeneous, and long-term source estimates contain an internal CI inconsistency.
Analysis details and source
No missing effects or ranks are inferred.
How treatment was delivered
The review compared exercise loading (especially eccentric exercise) alone or with adjuncts, including ESWT, PRP, low-level laser, injections and other modalities. Many adjuncts were only studied on top of exercise; their monotherapy effects cannot be isolated. For patellar tendon pain/function, eccentric exercise plus ESWT was compared with eccentric exercise plus placebo ESWT.
Pairwise meta-analysis used raw mean differences for the same measures at similar timepoints and pooled only within the same diagnosis. Network meta-analyses were performed for Achilles and patellar pain VAS and VISA outcomes when data allowed; no GTPS network was feasible.
The review summarizes selected intervention harms narratively (for example, no complications reported for ultrasound-guided hyaluronic-acid injection in included patellar trials; topical GTN may cause headache; corticosteroid injection has a theoretical tendon-rupture risk). No overall pooled safety rate is provided.
Most findings were low or very low strength; Achilles and patellar network analyses were downgraded for risk of bias and intransitivity.
Placebo was treated as no treatment despite possible independent effects; many interventions were disconnected from networks.
GTPS intervention groups mixed loading types; disease duration, age and activity could not be analyzed as subgroups.
Supplementary effect tables/figures were unavailable; one GTPS long-term contrast has an impossible printed CI relative to its point estimate.
Findings checked against the full source by a separate automated reviewer.
Limitations noted by the reviewers
- Most findings were low or very low strength; Achilles and patellar network analyses were downgraded for risk of bias and intransitivity.
- Placebo was treated as no treatment despite possible independent effects; many interventions were disconnected from networks.
- GTPS intervention groups mixed loading types; disease duration, age and activity could not be analyzed as subgroups.
- Supplementary effect tables/figures were unavailable; one GTPS long-term contrast has an impossible printed CI relative to its point estimate.
Source-linked critical reading
Systematic review with network meta-analysis
Technical source record
Complete full-text XML reviewed including all pairwise comparisons, condition-specific NMA findings/rank probabilities, evidence-strength conclusions, limitations and disclosures. A malformed long-term GTPS interval is preserved as a source inconsistency and not charted.
Exercise is the main treatment supported by this review. Adding shockwave therapy to eccentric exercise did not improve patellar outcomes in the available trials; evidence for other add-ons is less certain.
Which exercise programmes and adjuncts improve pain and function for Achilles, patellar and greater-trochanteric tendinopathy? ESWT was specifically assessed as an adjunct to eccentric loading in patellar disease.
Systematic review and pairwise/network meta-analysis of 68 RCTs. The review included 31 Achilles studies, 31 patellar studies and 6 GTPS studies, analysed separately by condition.
Included 3,808 participants overall: 1,792 with non-insertional Achilles tendinopathy, 1,109 with patellar tendinopathy and 907 with GTPS. Follow-up was mostly short or mid-term.
Exercise, often eccentric, was evaluated alone and with adjuncts. The ESWT comparison added it to eccentric exercise; results do not estimate ESWT monotherapy.
Controls included exercise plus placebo adjunct, other loading types, or alternative conservative care. Pairwise analyses were condition- and timepoint-specific.
In patellar tendinopathy, ESWT plus eccentric exercise did not improve short-term pain VAS (MD −0.09, 95% CI −1.00 to 0.84) or VISA-P (−1.8, −8.0 to 4.4); evidence strength was moderate. The immediate patellar isometric-versus-isotonic pain result was nonsignificant (MD 1.03; p=0.19; 3 studies), but the source CI (−2.57 to 0.5) excludes the point estimate and is not charted as valid. GTPS short-term CI is printed “(1.1. 2.1)” and remains ambiguous. NMA probabilities come from Results prose; Figures 3–4 captions instead identify displayed ranks as medians, and the graphic files were unavailable.
The review identifies no benefit of adding ESWT to eccentric exercise for patellar pain/function as its strongest practice-relevant conclusion. Most other pairwise/network findings are low or very low strength. GTPS short-term comparison has I²=94%.
Harms were discussed for selected treatments, but no overall pooled safety rate is reported. GTN may cause headache; local corticosteroid has theoretical tendon risks.
Low/very-low certainty for most analyses; network intransitivity/disconnection; placebo modeled as no treatment; mixed GTPS exercise loading; patient subgroups unavailable; supplements not acquired.
Exercise should remain first-line. Evidence does not support ESWT as an added component for patellar tendinopathy; other rankings are exploratory, low-certainty and condition-specific.
Conduct robust trials of loading monotherapies and adjuncts with common comparators, prespecified follow-up, clinically meaningful outcome thresholds, and patient-level modifiers.
Read the original article.
Effectiveness of Exercise Treatments with or without Adjuncts for Common Lower Limb Tendinopathies: A Living Systematic Review and Network Meta-analysis.
Challoumas D, Crosbie G, O'Neill S, Pedret C, Millar NL. · Sports medicine - open · 2023
Open the publication record or available full-text files.
Checked against the full paper · Clinical review: Dr. Marcus Yu Bin Pai, · How we review →