Efficacy and tolerability of extracorporeal shock wave therapy in patients with plantar fasciopathy: a systematic review with meta-analysis and meta-regression

Abstract

INTRODUCTION

Plantar fasciopathy (PF) is a common musculoskeletal condition characterized by heel pain and functional impairment. Extracorporeal shock wave therapy (ESWT) has gained increasing interest in the treatment of PF, but the optimal ESWT program is still debated. Therefore, this systematic review with meta-analysis and meta-regression aimed at providing a comprehensive assessment of the efficacy and tolerability of ESWT in PF management.

EVIDENCE ACQUISITION

Randomized controlled trials (RCTs) published until February 2023 were systematically searched on PubMed/MEDLINE, Scopus, Web of Science, Cochrane Central Register of Controlled Trials (CENTRAL), and PEDro. Studies assessing adult patients with PF treated with ESWT were considered. The primary outcome was the tolerability of ESWT, measured by treatment adherence, dropouts, and safety. Secondary outcomes were pain intensity and functional outcomes. Meta-analysis and meta-regression were performed to examine the relationship between ESWT program characteristics and treatment outcomes. The quality of included studies was assessed using the Jadad scale and the Cochrane risk-of-bias tool.

EVIDENCE SYNTHESIS

Eleven studies met the inclusion criteria and were included in the analysis. Our findings showed that ESWT is effective in reducing pain intensity assessed by Visual Analogue Scale [focal-ESWT: -2.818 (SE 0.803, -4.393, -1.244; P< 0.0001; radial-ESWT: -3.038 (SE 0.428, -3.878, -2.199; P<0.001)]. Meta-regression analysis indicated a positive relationship between specific ESWT parameters (frequency, number of pulses, energy flux density and frequency, and number of pulses, pressure) and pain intensity (all P<0.05) and dropout (all P<0.05).

CONCLUSIONS

ESWT seems to be an effective and tolerable treatment for PF, albeit the peculiarity of parameters might affect both the efficacy in pain relief and the adherence to the treatment. Physicians should consider individual patient characteristics when selecting the ESWT parameters for PF treatment. Further high-quality studies are warranted to establish the optimal ESWT protocol to treat PF.

Introduction

Plantar fasciopathy (PF) is a chronic degenerative condition related to biomechanical overuse, characterized by plantar fascia pain, and impaired physical functioning, with crucial implications for patients’ quality of life (QoL).^{1, 2} Fasciopathy is the preffered term adopted in recent years rather than fasciitis to underline that the etiology of this condition probably involves a prolonged degenerative process rather than acute inflammation.^{1, 3} This painful disorder affects a significant proportion of the population, including athletes, individuals with excessive physical activity, and those with specific risk factors such as obesity or foot abnormalities.^{2, 4, 5}

Given the prevalence and impact of this disorder,⁶ the management of this condition is crucial for improving physical and psychosocial outcomes of people with PF. The primary goal of treatment is to alleviate pain, and promote healing of affected tissues,⁷ while conservative approaches such as physical therapy, orthotic devices, analgesics, and stretching exercises are commonly employed, they may not always provide satisfactory outcomes, especially in cases of chronic or recalcitrant PF.^{1, 8-10} Thus, more invasive interventions like corticosteroid injections or surgical procedures may be considered, although the recent systematic review by MacRae et al.¹¹ underlined low-grade evidence supporting surgical interventions in patients with refractory symptoms and no consensus is currently available on the best surgical approach.⁵

In the recent years, the extracorporeal shock wave therapy (ESWT) has been emerging as a promising non-invasive treatment modality for several musculoskeletal conditions, including tendinopathies, spasticity, and PF.^12-15 ESWT involves the application of high-energy shock waves to the affected area, targeting the site of pain and pathology.^12-14 These rapid pressure waves exceed the speed of sound in the medium they traverse. These waves comprise two distinct phases: an initial rapid rise phase, lasting less than 10 nanoseconds, characterized by focal pressures ranging from 50 to 80 megapascals (MPa), followed by a slower phase lasting milliseconds, during which negative pressures of up to 10 MPa occur.¹⁶ Despite their therapeutic potential, the exact mechanism by which shock waves exert their effects remains incompletely understood.¹⁶

In this context, two types of ESWT are currently available: focal ESWT (f-ESWT) and radial ESWT (r-ESWT).¹⁷

More in detail, the f-ESWT delivers shockwaves in a focused or convergent pattern, concentrating the energy on a specific area or target, and possess both positive and negative phases and can achieve high intensities deep within target tissues. This technique provides a deeper penetration, allowing for the targeting of deeper structures such as bones and joints.^{12, 16} The therapeutic effectiveness of f-ESWT depends significantly on the intensity of the shock waves at the focal point, which is quantified as energy flux density (EFD) per pulse, measured in mJ/mm². In clinical settings, EFD levels typically range from 0.001 to 0.5 mJ/mm².¹⁶

On the other hand, r-ESWT delivers lower-intensity pressure waves, which lack a negative phase, in a radial or divergent pattern, spreading the energy outwards as it reaches the target area. This technique is suitable for treating larger and superficial treatment areas and is commonly utilized for musculoskeletal disorders involving soft tissues such as tendinopathies or muscle injuries outcomes.^{9, 12, 16}

Both focal and radial shock waves might promote tissue regeneration, stimulate angiogenesis, and exert analgesic effects, offering long-lasting pain relief and improved function.^{9, 18, 19} In both scenarios, each session typically includes between 800 to 3000 pulses. Treatment protocols often involve three shock wave sessions with a one-week interval between sessions.¹⁶

Albeit no long-term complications are expected, patients typically experience a milder sensation of pressure, discomfort, and localized pain and high dosage protocols might increase the risk of treatment failure due to poor compliance.^{9, 20}

However, despite the growing interest on ESWT as an effective treatment option for PF, the evidence about the optimal ESWT program to treat this condition remains inconclusive. Moreover, several studies have been conducted to investigate the efficacy and tolerability of both f-ESWT and r-ESWT,^{21, 22} however the results were heterogeneous and several ESWT programs are currently available.

Thus, this systematic review of randomized controlled trials (RCTs) with meta-analysis aimed to provide a comprehensive assessment of the efficacy and tolerability of ESWT in the management of PF. Furthermore, with a meta-regression analysis, we aimed at exploring the relationship between ESWT parameters (e.g., energy intensity, frequency, number of sessions) and treatment outcomes to provide clinically meaningful indications to treat PF.

Evidence acquisition

Registration

This systematic review has been performed ethically in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement.²³ The international prospective register of systematic reviews (PROSPERO) was preliminarily searched for similar review protocols in progress. No similar review was identified; thus, the systematic review was submitted to PROSPERO and accepted on April 1^st, 2023 (available at https://www.crd.york.ac.uk/prospero; registration number CRD42023409664).

Search strategy

Five databases (PubMed/Medline, Scopus, Web of Science, Cochrane Central Register of Controlled Trials (CENTRAL), and PEDro) were systematically searched simultaneously by two different independent investigators for RCTs published until February 19th, 2023. The search strategies for each database are shown in Supplementary Digital Material 1, Supplementary Table I.

Selection criteria

In accordance with the PICO model,²⁴ we considered eligible randomized controlled trials (RCT) satisfying the following criteria:

1. Participants: patients over 18 years old with a diagnosis of PF.

   1. Intervention: ESWT.

      1. Comparator: any comparator.

         1. Outcome: the primary outcome was tolerability of ESWT, expressed as adherence to the treatment, adverse events reported, and number of drop-outs. The secondary outcomes were: 1) pain intensity; 2) functional outcomes.

Only RCTs published in peer-reviewed International Journal were included. The exclusion criteria were: 1) studies involving animals; 2) participants with pregnancy, clinical instability, or palliation; 3) masters or doctorate theses and conference proceedings; 4) language other than English; 5) studies associating any other intervention for treating PF. The database search was performed without applying any date restriction.

The retrieved records were assessed for duplication via automated tools, and the resulting papers were screened by two independent investigators (LL and AF), reviewing the titles and abstracts, excluding the papers that did not meet the inclusion criteria. Any disagreement was discussed involving a third (AdS) reviewer to reach consensus. Resulting studies were assessed full text for eligibility by two independent reviewers (LL and AF) and relevant data were extracted through Excel. Any disagreement was solved by discussion between the two reviewers or by consulting a third reviewer (MI).

Data extraction and synthesis

Lastly, a qualitative method has been used in data synthesis from the selected studies. More in detail, the following data have been presented: 1) title; 2) authors; 3) publication year; 4) nationality; 5) participants (number, mean age and age range, gender); 6) ESWT parameters (i.e., bar, mJ/mm², Hz); 7) comparator; 8) main findings (which presented the study results in terms of primary and secondary outcomes).

The data have been synthesized by two authors independently (LL and AF) from full-text documents. Any disagreement between the two reviewers has been solved by collegial discussion among the authors. In case of disagreement, a third author has been asked (AdS).

Text and tables have been used to provide a descriptive summary and explanation of study characteristics and findings. Subgroup analysis was performed to differentiate between r-ESWT and f-ESWT.

Then, meta-regression analysis was performed between the tolerability of ESWT and the specific ESWT parameters (e.g., energy intensity, frequency, number of sessions) for patients with PF. Similarly, ESWT efficacy in terms of pain intensity and functional outcome has been assessed using a meta-regression analysis for the same above-mentioned ESWT parameters.

Meta-analysis

The meta-analysis was performed by Comprehensive Meta-Analysis by Comprehensive Meta-Analysis V.3.7 (Biostat, NJ, USA) software, with significance set as <0.05. Changes in VAS were shown as mean difference (MD) and standard deviation (SD). The heterogeneity among comparisons was estimated by the Chi-squared and I2 statistic tests. An I²>75% determined significant heterogeneity across the articles. In the event of considerable heterogeneity, a random-effects model was adopted to determine the pooled estimates with the effect size (ES) and 95% confidence interval (CI). Missing means and SDs were estimated from medians, ranges, and interquartile ranges (IQRs) using the method introduced by Hozo et al.²⁵

Meta-regression

The meta-regressions were performed by Comprehensive Meta-Analysis by Comprehensive Meta-Analysis V.3.7 (Biostat, NJ, USA) software, with significance set as <0.05. We performed a random-effect meta-regression, to investigate the relationship between study-level characteristics (ESWT therapy variables, such as frequency, number of pulses, energy flux density, and pressure) and the effect sizes of individual studies of the logit event rate for ESWT tolerability (dropout and adherence) and pain intensity (in terms of VAS). Random-effect meta-regression accounts for both within-study variability and between-study variability. It assumes that the true effect sizes across studies follow a normal distribution with a mean representing the overall effect and a random component capturing the variability between studies.²⁶

The meta-regression was used to estimate the regression coefficients for each moderator (study characteristics), along with their standard errors and significance levels. Positive coefficients suggested a positive relationship between the moderator and the effect size, while negative coefficients indicate a negative relationship. The amount of heterogeneity (variation) between studies that was not explained by the moderators was quantified by Q statistic and I-squared. In order to assess the proportion of the total variation attributed to heterogeneity the I² (I-squared) statistic was used in meta-regression. The I² statistic ranged from 0% to 100%, with higher values indicating greater heterogeneity.

Quality assessment and risk of bias

Two of the authors (LL and AF) independently assessed the quality of the included studies using the Jadad scale. Disagreements were resolved by collegial discussion between the authors or by contacting a third reviewer (AdS). High quality was attributed with a Jadad Score of 3 to 5.²⁷

The Cochrane risk-of-bias tool for randomized trials (RoBv.2) was used to assess the risk of bias. Based on the RoBv.2 items, bias was graded as low, high, or uncertain. RoBv.2 specifically analyzed the following domains: 1) random process; 2) deviation from the intended interventions; 3) missing outcome data; 4) measurement of the outcome; 5) selection of the reported result.²⁸

Evidence synthesis

As shown in Figure 1, the database research identified 291 papers, which were screened for duplication removal via EndNote (EndNote 20; Bld 18004) automated tool. Subsequently, the duplication removal results were screened by two reviewers without finding issues. The resultant 251 papers were screened for title and abstract and 220 papers were excluded. As a result, 31 papers were selected and screened in full text. Of those, 20 were excluded for the following reasons: 11 studies included ESWT combined with other interventions, one study included patients with spondylarthritis, one study was in language other than English, one study was not an RCT, one study was duplicated, three studies had ongoing recruitment, and two studies did not assess primary outcome. Supplementary Digital Material 2, Supplementary Table II provides further details on the paper’s exclusion reasons. As a result, 11 papers^29-39 were included in this study.

Figure 1.

—PRISMA 2020 flow chart.

Main characteristics of the included studies

The 11 included studies were published between 2003³⁷ and 2022.^{29, 34, 39} Three studies (27.27%)^{29, 32, 33} were conducted in Türkiye, 2 (18.18%)^{34, 37} in the UK, while the remaining studies were conducted in Egypt (N.=1; 9.09%),³⁸ in Germany (N.=1; 9.09%),³¹ in China (N.=1; 9.09%),³⁵ in India (N.=1; 9.09%),³⁰ in Italy (N.=1; 9.09%),³⁹ and in the USA (N.=1; 9.09%).³⁶ The main characteristics of the included studies are summarized in Supplementary Digital Material 3, Supplementary Table III.^29-39

Study participants

The randomized population consisted of 1081 subjects, while the participant number at the end of the study was 997, measuring a cumulative dropout of 7.78%. The groups randomized to receive exclusively ESWT intervention added up to 555 participants, while at the end of follow-ups, the participants number was 516, measuring a cumulative drop out of 7.02%. The mean age range reported was from 50.1±6.7³⁰ to 59.3±12.2 years;³⁹ however, it should be noted that three studies^{31, 32, 37} reported the age only as mean years (minimum age – maximum age) per intervention group, ranging from 37.5 (25-62)³² to 52.5 (30-73) years.³⁷

The male/female ratio of the participants included in the analysis was 407/611. However, one study³⁰ did not report the sex distribution. Body Mass Index (BMI) information was reported by 6 studies.^{29-33, 38} Four studies^{29, 30, 33, 38} reported each group’s BMI in terms of mean BMI±SD ranging from 29.7±4.8³⁸ to 32.6±4.0 kg/m².³⁰ Two studies^{31, 32} reported each intervention group BMI in terms of mean BMI (minimum – maximum), ranging from 26.6 (19.6-29.1)³² to 29.3 (19-35) kg/m².³¹

The symptom duration was from 12.6±9.3²⁹ to 33.4±36.3 months;³⁴ other studies^{31, 32, 37} reported the duration as mean months (minimum duration – maximum duration) per group, ranging from 8 (6-19)³¹ to 31.6 (6-240) months.³⁶ Only Lai et al.³⁵ included patients with less than 3 months of symptoms (from 7.94±2.92 weeks to 8.06±2.87 weeks), while Bahar-Ozdemir et al.³⁰ included patients from less than one month to up to six months. Lastly, Tognolo et al.³⁹ provided no information on the symptom duration. Further demographic characteristics are displayed in Supplementary Table III.

Interventions

Among the population assessed, 334 subjects underwent r-ESWT alone or in a protocol, in a total of 5 studies,^30-34 while 306 underwent f-ESWT alone or in a protocol, in a total of 5 studies.^35-39 Kesikburun et al.²⁹ administered both therapies to a total of 14 patients. The control groups included the following interventions: prolotherapy (PRL),^{29, 32} low-dye and sham kinesiotaping,³⁰ r-ESWT combined with a plantar-fascia-specific-stretching protocol (PFSS),³¹ platelet-rich plasma (PRP) injections,³² corticosteroid injections (CSI),^{32, 35} low-level laser therapy (LLLT),³³ ultrasound therapy,³³ photobiomodulation therapy (PBMT) and sham PBMT,³⁸ sham ESWT.^{36, 37} Lastly, two studies assessed different ESWT intensity protocols.^{34, 39}

Primary outcomes

The main goal of the present review was to assess the tolerability of ESWT treatment administered alone, evaluated as adherence to treatment, reported adverse events after ESWT administration, and number of dropouts. Further insights are presented below.

• Adherence to treatment: the RCTs administering r-ESWT reported between 82.19%⁴⁰ to 100%^{30, 32, 33} adherence to treatment, with a total adherence to treatment of 248/264 participants (93.94%). Interestingly, in the study by Wheeler et al.³⁴ the greatest adherence was found in the group that underwent higher intensity treatment (98.33% vs. 96.49%). On the other hand, the interventions with f-ESWT had between 84.45%³⁵ and 97.39%³⁶ adherence to treatment, with a total adherence to treatment of 250/272 participants (91.91%). Remarkably, Tognolo et al.³⁹ reported a higher adherence in the lower intensity f-ESWT group (100% vs. 73.33%). Lastly, the r- and f-ESWT subgroups had a total adherence of 93.33% (14/15).²⁹

• Dropouts: accordingly, the RCTs administering r-ESWT reported between 0^{30, 32, 33} to 13⁴⁰ dropouts, with total dropouts 16/264 participants (6.06%). On the other hand, the interventions with f-ESWT had between 3^{36, 38} and 8³⁵ dropouts, with total dropouts of 22/272 participants (8.09%). Lastly, the r- and f-ESWT subgroups registered 1 dropout (1/15).²⁹

• Adverse events: after r-ESWT, 3 studies reported 0 adverse events;^{30, 32, 33} minor adverse events were registered by Rompe et al.,³¹ which reported transient local reddening after ESWT administration, and pain (NRS>5) during treatment, but the exact number of participants reporting the symptoms was not specified. Wheeler et al.³⁴ reported minor bruising in one participant (1/60, 1.67%) undergoing high intensity r-ESWT vs. no adverse event in low intensity r-ESWT. After f-ESWT the minor adverse events registered were temporary local reddening and procedural VAS of 7 or higher reported by Takla et al.;³⁸ Malay et al.³⁶ reported bruising in 2 subjects (2/115, 1.74%), while local swelling was reported by one participant (1/115, 0.87%); Speed et al.³⁷ registered one syncope during f-ESWT administration (1/46, 2.17%); lastly, Lai et al.³⁵ reported local tenderness, but did not report the exact number of events. Tognolo et al.,³⁹ did not report adverse events.

Secondary outcomes

The secondary outcomes assessed were the following:

• pain intensity was measured through the visual analog scale (VAS) by 9 studies,^{29, 30, 32-38} albeit none of the included papers reported the ESWT procedural VAS. The r-ESWT protocols significantly reduced VAS in all the studies in which it was assessed (P<0.05).^{30, 32-34} Interestingly, in the study by Wheeler et al.³⁴ there was no significant difference between higher and lower intensity groups (P=NS). The r-ESWT compared to other interventions resulted in significantly better than US therapy (P=0.012).³³ The f-ESWT protocols achieved the following results. Lai et al.³⁵ reported a significantly higher efficacy in VAS reduction with respect to corticosteroid injection (P=0.001 at 4 weeks and P<0.001 at 12 weeks). Malay et al.³⁶ reported significant pain relief in the ESWT group compared to sham (P=0.045). On the other hand, Speed et al.³⁷ showed a significant improvement in both active treatment (0.12 mJ/mm²) and sham therapy (P<0.05), though no statistically significant differences were found between the groups (P=NS). Interestingly, Takla et al.³⁸ reported VAS results with statistically significant improvements in post-intervention and follow-up values in all treatment groups (all P<0.0001), while reporting none in the sham group (p>0.05); when comparing all four groups, the combination of PBMT and ESWT provided significant benefits compared to ESWT and PBMT alone (P<0.0001), and ESWT treatment alone had provided significant advantages compared to PBMT in terms of pain intensity (P<0.0001).

• Several functional outcomes were considered in the included papers and are furtherly discussed below.

• The Heel Tenderness Index (HTI) was assessed by two papers;^{30, 33} Bahar-Ozdemir et al.³⁰ reported significant improvement (P<0.05); however, no differences were reported compared to control group, while Ulusoy et al.³³ found that r-ESWT provided significant benefits compared to US therapy was (P=0.004).

• The Foot Function Index Scale (FFI) was assessed by 7 papers,^{29-32, 34, 38, 39} including the modified versions of the questionnaire, as following described. FFI in its classic version was assessed in two studies,^{29, 30} significance was reported one month after treatment (P≤0.001),³⁰ after 6 weeks (P<0.01),²⁹ after 3 months (P<0.01).²⁹ The revised version of the FFI (FFI-R) was found significantly improved at 6 months^{32, 34} with P<0.001. The pain subscale of the FFI (PS-FFI) was extracted in two studies;^{30, 31} in particular, Bahar-Ozdemir et al.³⁰ reported significant improvement (P<0.05), however, no differences with control group were underlined by the between-group analysis (P=NS). Rompe et al.³¹ reported significant differences between intervention and control groups, resulting in patients treated with ESWT plus PFSS (control group) having better results than those managed with ESWT alone (P<0.001). The disability subscale of the FFI (FFI-DS) was assessed in two studies;^{30, 38} in particular, Bahar-Ozdemir et al.³⁰ reported significant improvement in the intragroup analysis (P<0.05); however, no differences with control group were reported (P=NS). Takla et al.³⁸ reported statistically significant improvements in the intervention group (P<0.0001), while, significant differences were reported between the combination of PBMT and ESWT group, ESWT group and PBMT group (P<0.0001). In addition, ESWT showed significant advantages compared to PBMT in terms of physical function (P<0.0001). The activity subscale of the FFI (FFI-AS) was assessed by Bahar-Ozdemir et al.,³⁰ who reported significant improvement after the intervention (P<0.05), however no between group differences were reported (P=NS). Lastly, the Italian version of the FFI (17-iFFI) was assessed by Tognolo et al.,³⁹ reporting significant improvement in both groups (P<0.0001). In addition, significant advantages were reported in f-ESWT with low intensity in myofascial point group (P=0.0016).

• The Roles-Mosley score (RMS) was assessed by two studies.^{29, 33} Kesikburun et al.²⁹ reported significant improvement after the intervention (P<0.01); however, the between-groups analysis showed no differences between groups (P=NS). On the other hand, Ulusoy et al.³³ found significant differences between r-ESWT and US therapy groups (P=0.014).

• The plantar fascia thickness was assessed by 2 studies.^{33, 35} In more detail, Ulusoy et al.³³ studied magnetic resonance sagittal and coronal plane measurements, reporting significant thinning after r-ESWT intervention, though no differences were found compared to control group. Lai et al.³⁵ assessed the ultrasound plantar fascia thickness reporting greater increases in the ESWT group compared to CSI group at 4 weeks (P=0.048).

• The American Orthopedic Foot and Ankle Association score (AOFAS) was assessed by Ulusoy et al.,³³ reporting significant improvement in the intervention group after the treatment (P<0.001).

• The Manchester-Oxford Foot Questionnaire (MOXFQ) and the foot ankle ability measure in activities of daily living (FAAM-ADL) were assessed by Wheeler et al.,³⁴ who reported significant improvements in both groups (P<0.001), with no significant differences between higher and lower intensity groups (P=NS).

• The pain pressure threshold (PPT) was assessed by Takla et al.,³⁸ who reported statistically significant improvements in the intervention group (P<0.0001), while significant between groups differences were reported between the combination of PBMT and ESWT, ESWT and PBMT alone (P<0.0001).

• Lastly, the foot and ankle outcome score (FAOS) was assessed by Tognolo et al.,³⁹ showing significant improvement in both groups (P<0.0001), with significant advantages in f-ESWT on myofascial point group (lower intensity) (P=0.0072).

Meta-analysis

Three studies were included in the meta-analysis of f-ESWT mean VAS change^{35, 37, 38} resulting in an overall significant mean decrease of -2.818 (SE 0.803, -4.393, -1.244, P<0.0001).

On the other hand, four studies were included in the meta-analysis of r-ESWT mean VAS change,^{30, 32-34} resulting in an overall significant mean decrease of -3.038 (SE 0.428, -3.878, -2.199, P<0.0001).

Further details about the meta-analysis were shown in Figure 2.^{30, 32-35, 37, 38}

Figure 2.

—Meta-analysis of overall effect size of VAS change. A) Focal-ESWT;^{35, 37, 38} B) radial-ESWT.^{30, 32-34} ESWT: extracorporeal shockwave therapy; VAS; Visual Analogue Scale.

Meta-regression

Ten studies were included in meta-regressions,^30-39 while the study by Kesikburun et al.²⁹ was excluded due to the heterogeneity of the study intervention (f-ESWT combined with r-ESWT).

Adherence to f-ESTW was positively correlated with all the ESWT therapy variables considered (frequency, number of pulses, energy flux density) with statistical significance (all P<0.05). In accordance, adherence to r-ESTW was positively correlated with all the ESWT therapy variables considered (frequency, number of pulses, pressure) with statistical significance (all P<0.05). Random-effect meta-regression of moderators on adherence is shown in detail in Figure 3.

Figure 3.

—Random-effect meta-regression of moderators on adherence. A) Regression of Logit event rate on SW frequency (Hz) focal-ESWT; B) Regression of logit event rate on pulse-focal-ESWT; C) regression of logit event rate on energy flux density (mJ/mm²)- focal-ESWT; D) regression of logit event rate on SW frequency (Hz)-radial-ESWT; E) regression of logit event rate on pulse-radial-ESWT; F) regression of logit event rate on pressure (bar)-radial-ESWT. Coeff: coefficient; ESWT: extracorporeal shockwave therapy; SW: shockwave.

On the other hand, dropouts were negatively correlated with all the f-ESWT and r-ESWT therapy variables considered (frequency, number of pulses, energy flux density and frequency, number of pulses, pressure, respectively) with statistical significance (all P<0.05). Random-effect meta-regression of moderators on dropouts is shown in Figure 4.

Figure 4.

—Random-effect meta-regression of moderators on dropouts. A) regression of logit eventrate on SW frequency; (Hz)–focal-ESWT; B) regression of logit event rate on pulse-focal-ESWT; C) regression of logit event rate on energy flux density (mJ/mm²)-focal-ESWT; D) regression of logit event rate on SW frequency (Hz)-radial-ESWT; E) regression of logit event rate on pulse-radial-ESWT; F) regression of logit event rate on pressure(bar)-radial-ESWT. Coeff: coefficient; ESWT: extracorporeal shockwave therapy; SW: shockwave.

Eight studies^{30, 32-38} performed the analysis of pain outcomes and were subsequently included in this analysis. Overall, only three studies assessing f-ESWT were included,^{35, 37, 38} and VAS outcome was correlated to pulse and energy flux density, while correlation analysis was not possible in terms of f-ESWT frequency since it was reported only in one study.³⁸ In addition, the study by Malay et al.³⁶ has not been included because it did not report the SD of mean VAS. On the other hand, four studies using r-ESWT were included in the meta-regression.^{30, 32-34} Random-effect meta-regression of moderators on VAS is shown in detail in Figure 5.

Figure 5.

—Random-effect meta-regression of moderators on VAS. A) Regression of VAS mean difference on pulse-focal ESWT; B) regression of VAS mean difference on energy flux density (mJ/mm²)-focal-ESWT; C) regression of VAS mean difference on SW frequency (Hz)-radial-ESWT; D) regression of VAS mean difference on pulse-radial-ESWT; E) regression of VAS mean difference on pressure (bar)-radial-ESWT. Coeff: coefficient; ESWT: extracorporeal shockwave therapy; SW: shockwave.

Taken together, VAS changes were negatively correlated with all the f-ESWT and r-ESWT therapy variables considered (number of pulses, energy flux density and frequency, number of pulses, and pressure, respectively) with statistical significance (all P<0.05). Thus, these results showed that an increased number of pulses, energy flux density and frequency, number of pulses, and pressure were related to a greater decrease in VAS.

Lastly, all Q values were statistically significant (all P<0.05), with rather large I² statistic, suggesting presence of other factors affecting the effect size.

Study quality assessment and risk of bias

The quality assessment of the studies included in the present systematic review according to the Jadad scale²⁷ resulted in high quality for all the RCTs included (N.=11, 100%),^29-39 with a score between 3 and 5. More details are available in Table I.^29-39

Table I. —Quality assessment of the studies included in the present systematic review.^29-39.

Articles	Domain					Score
	Random sequence generation	Appropriate randomization	Blinding of participants or personnel	Blinding of outcome assessors	Withdrawals and dropouts
Kesikburun et al.29	1	1	1	1	1	5
Bahar-Ozdemir et al.30	1	1	1	1	1	5
Lai et al.35	1	1	0	1	1	4
Malay et al.36	1	1	1	1	1	5
Rompe et al.31	1	1	0	1	1	4
Speed et al.37	1	0	1	1	1	4
Takle et al.38	1	1	0	1	1	4
Tognolo et al.39	1	1	0	0	1	3
Uğurlar et al.32	1	1	0	0	1	3
Ulusoy et al.33	1	1	0	1	1	4
Wheeler et al.34	1	1	1	1	1	5

Points were awarded as follows: study described as randomized, 1 point; appropriate randomization, 1 point; subjects blinded to intervention, 1 point; evaluator blinded to intervention, 1 point; description of withdrawals and dropouts, 1 point.

Figure 6^29-39 shows the risk of bias assessed by RoB 2.²⁸ All studies (N.=11, 100%) ensured correct randomization, and no problems with measurement of the outcomes and selection of reported results.^29-39 Seven studies (63.64%)^{29, 30, 32, 33, 35, 38, 39} showed some concerns in the second domain due to the lack of details regarding the intention-to-treat analysis. Three studies^{29, 35, 37} presented some concerns in the third domain, related to missing outcome data. Overall, 8 studies (72.73%) were evaluated with some concerns,^{29, 30, 32, 33, 35, 37-39} while 3 (27.27%) were evaluated as low risk.^{31, 34, 36}

Figure 6.

—Risk of bias summary of the included studies.^29-39

Discussion

PF is a common and disabling condition affecting a growing number of people worldwide, leading to heel pain and significant reductions of QoL.^{1, 6} R-ESWT and f-ESWT are effective treatments to reduce pain and improve functional outcomes in patients affected by PF,^{41, 42} however, the optimal ESWT program to treat this specific condition has never been fully characterized. Thus, in this systematic review with meta-analysis and meta-regression, we aimed at assessing the effectiveness and tolerability of different ESWT program characteristics to treat PF and their relationship with treatment outcomes.

Our findings showed promising results in terms of tolerability and efficacy of both r-ESWT and f-ESWT. The overall adherence to treatment was high, with a range of 82.19%⁴⁰ to 100%^{30, 32, 33} in the r-ESWT group and 84.45%³⁵ and 97.39%³⁶ in the f-ESWT group. These results suggest that ESWT is well-tolerated by patients with PF. In this context, compliance rate to ESWT plays a crucial role in the success of this non-invasive treatment for several musculoskeletal conditions.^{43, 44} Besides this consideration, pain during ESWT treatment might have a crucial impact in treatment compliance and several studies are currently available combining ESWT with analgesic procedure in order to reduce pain during the treatment.^{45, 46} More in detail, Kudo et al.⁴⁵ assessed the effects of ESWT combined with calcaneal nerve block using xylocaine, while another potential strategy might include prilocaine hydrochloride injected in the most painful area.⁴⁶ On the other hand, ESWT parameters regulation might have a crucial impact in pain during the procedure and affect patients’ compliance to the treatment. Despite these considerations, no previous systematic review addressed the relationship between overall adherence to treatment and ESWT program characteristics.

Thus, we conducted a random-effect meta-regression to explore the association between ESWT therapy variables (such as frequency, number of pulses, energy flux density, and pressure), and ESWT adherence. The meta-regression analysis suggested several important findings. Firstly, we found a significant positive relationship between both f-ESWT and r-ESWT parameters (including the energy intensity and frequency of ESWT) and adherence, suggesting that treatment characteristics were positively associated with improved tolerability. Moreover, dropouts showed a statistically significant negative correlation with both f-ESWT and r-ESWT parameters (frequency, number of pulses, energy flux density and frequency, number of pulses, pressure). These findings are particularly important since an optimal ESWT regulation might be crucial for improving compliance rates and consequently improvement in functional outcomes in patients with PF. On the other hand, it should be noted that ESWT efficacy might be affected by the operator performing the treatment and potential differences in operator experience might contribute to the heterogeneity of data currently available in literature.^29-39

In recent years, there has been an increased interest on implementing effective strategies to engage patients and ensure their active participation in rehabilitation programs.^{47, 48} This is particularly important for achieving sustainable interventions and improving compliance rates. Tailoring rehabilitation programs have been proposed to enhance benefits and improving patient’s adherence, promoting patients actively participation in the therapeutic management.⁴⁹ Furthermore, incorporating technology-based interventions can facilitate remote monitoring of patients’ progresses and further individualized rehabilitation approach.^{50, 51} Despite these considerations, an evidence-based approach is needed to adapt rehabilitation strategies to patients’ needs, while precise indications on ESWT parameters variations in terms on adherence and drop out were lacking. The results of the present meta-regression analysis can provide valuable insights into the relationship between ESWT parameters and patients’ compliance, guiding clinicians in adopting an evidence-based approach to personalize ESWT treatment in patients with PF.

Interestingly, ESWT demonstrated significant improvements across the included studies in terms of pain intensity.^{29, 30, 32-38} More in detail, the results of the meta-analysis for f-ESWT showed a significant mean decrease in VAS scores of -2.818 (standard error [SE] 0.803, -4.393, -1.244, P<0.0001). In accordance, meta-analysis for r-ESWT underlined a significant mean decrease in VAS scores of -3.038 (SE 0.428, -3.878, -2.199, P<0.0001). These findings indicate that both f-ESWT and r-ESWT are effective in reducing pain intensity as measured by the VAS in patients with PF. Interestingly, the meta-regression analysis revealed a significant relationship with energy intensity flux of ESWT and pain relief, suggesting that higher energy intensities were associated with greater reductions in pain intensity for both f-ESWT and r-ESWT.

In this context, pain reduction is a crucial aspect of PF rehabilitation management since it directly affects patients’ daily activities and overall function.^{52, 53} The mechanisms of action underlying ESWT’s pain-relieving effects are far from being fully understood but might be related to substance P and prostaglandin E2 release.^{54, 55} On the other hand, previous systematic review assessed the effects of ESWT in pain reduction. More in detail, the meta-analysis by Sun et al.⁴² showed significant effectiveness of both r-ESWT and f-ESWT in terms of improvement rates compared to placebo. However, due to the heterogeneity of ESWT protocols, it is particularly difficult to draw strong conclusions and few studies were included in the analysis.⁴² More recently, the systematic review by Li et al.⁵⁶ assessed the effects of ESWT compared to corticosteroid injections. The authors reported that high-intensity ESWT might promote superior pain relief compared to corticosteroid injections. However, ESWT and CSI were related to similar recurrent rates and functional improvements.⁵⁶ Despite these considerations, the authors did not characterize the differences between f-ESWT and r-ESWT and due to their different mechanisms of action, it is particularly difficult to provide clinical implications of the study results.

To the best of our knowledge, this is the first systematic review underlining the effectiveness of both f-ESWT and r-ESWT, supporting the effectiveness of this promising therapeutic approach in the tailored rehabilitation management of patients with chronic PF. In addition, our quantitative synthesis underlined that more common adverse events were generally minor, with temporary local reddening being the most reported.^{31, 34-39} Thus, ESWT should be considered a safe and effective intervention for people with PF and might be integrated in therapeutic path as a first-line approach.

Limitations of the study

Despite the promising results, different limitations of this study should be acknowledged. First, the number of included studies was relatively small, with negative implications for the generalizability of the results. Second, it is important to note that there was considerable heterogeneity among the studies, as indicated by the I-squared statistic. This indicates that a substantial portion of the variation in effect sizes across studies could not be explained by the ESWT parameters included in the meta-regression model. These results underline that the compliance rate to ESWT treatments might be affected by several variables that might be undetected by clinical studies currently available in literature. However, it is important to notice that the quantitative synthesis of the present meta-regression significantly showed that ESWT program characteristics should be considered in a tailored treatment plan aiming at improving both compliance and pain intensity in patients with PF. Further research is needed to explore other potential moderators and sources of heterogeneity in ESWT compliance and treatment outcomes. Lastly, our review focused on ESWT as a conservative treatment for chronic PF, without providing data about any differences from other therapeutic interventions. On the other hand, it is important to note that the aim of this study was to characterize the multilevel interactions between ESWT parameters and treatment outcomes, providing clinically meaningful data about the effects of different ESWT parameters on ESWT effectiveness and tolerability. To our knowledge, this is the first meta-regression analysis on ESWT treatment providing specific evidence-based data on ESWT program characteristics to guide clinicians in optimizing adherence and effectiveness of this intervention in chronic PF patients.

Conclusions

Taken together, findings of this systematic review with metanalysis and meta-regression highlighted a significant correlation between ESWT parameters (including frequency, number of pulses, energy flux density, and pressure) and the tolerability and the efficacy in pain relief. However, it is important to interpret these results with caution due to the heterogeneity observed across the included studies. To date, r-ESWT and f-ESWT are widely used, considering their evidence as non-pharmacological interventions to treat chronic PF. In this context, future research should focus on standardizing treatment protocols and evaluating the long-term effects of different f-ESWT and r-ESWT parameters in order to optimize their therapeutic benefits in patients affected by chronic PF.

Supplementary Digital Material 1

Supplementary Table I

Search strategy

Supplementary Digital Material 2

Supplementary Table II

Characteristics of excluded studies assessed in full text.

Supplementary Digital Material 3

Supplementary Table III

Main characteristics of the studies included.^29-39