Plantar fasciitis is one of the most common causes of adult heel pain, occurring in an estimated 10% of the population. Several treatment options are available to relieve plantar heel pain including rest, stretching, foot orthotics, night splinting, and invasive therapies. For the majority of patients, plantar fasciitis is a self-limiting condition and can be treated effectively by nonsurgical treatments.[4,5] Nevertheless, in recalcitrant cases, surgery is suggested.[6–8]
Focused extracorporeal shock wave therapy has become a popular alternative to traditional surgical approaches.A pneumatic generator produces focused extracorporeal shock waves that travel from the point of contact on the skin's surface to the affected area. Its exact treatment mechanism is unknown; however, this therapy likely promotes tissue healing and neovascularization. Notably, other therapies that apply different energy forms for treating plantar fasciitis are available and have demonstrated favorable results. These include ultrasound therapy[11,12] and the recently developed ultrasound-guided pulsed radiofrequency (UG-PRF) treatment for the gastrocnemius, low-level laser therapy (LLLT), radiotherapy (i.e., radiation therapy), noninvasive interactive neurostimulation (NIN) and intracorporeal pneumatic shock therapy (IPST). In contrast to surgical approaches or injection therapy, these interventions treat plantar fasciitis with different types of energy output that can propagate through a medium, such as sonic waves, mechanical oscillations, heat, or light. Their effects on tissue are dependent on stimulus intensity, stimulus frequency, pulse interval, treatment duration, and application method (e.g., minimally invasive).
A previous systematic review and network meta-analysis[18,19] revealed that extracorporeal shock wave therapy is clinically effective for relieving chronic recalcitrant plantar fasciitis. Additionally, radial extracorporeal shock wave therapy (RSW) is advantageous over focused extracorporeal shock wave therapy (FSW) in terms of pain relief. However, Chang et al only reported outcomes at 6 months or at the follow-up closest to 6 months for analysis because the follow-up duration varied among included studies. Furthermore, until now, no systematic reviews on LLLT, IPST, UG-PRF, and NIN for plantar fasciitis treatment have been reported. Thus, although all 7 treatment options appear to be more effective than placebo in alleviating plantar heel pain, the most effective treatment is currently unknown. Therefore, we conducted a Bayesian network meta-analysis to compare the therapeutic effectiveness among these 6 therapies for treating plantar fasciitis across different time intervals.
2.1 Protocol and registration
We prospectively registered this systematic review and network meta-analysis with PROSPERO (CRD42015017353) and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines for network meta-analyses.
2.2 Search strategy
We comprehensively searched Embase via embase.com, MEDLINE via PubMed, CENTRAL (Cochrane Central Register of Controlled Trials) and the Physiotherapy Evidence Database (PEDro) up to December 4, 2017, without language or date restrictions. Only randomized controlled trials (RCTs) were included. The search terms used in Embase are presented as an example in Appendix 1, http://links.lww.com/MD/C549. To avoid the potential omission of studies, we searched additional databases, such as opengrey.eu for gray literature. We also manually screened reference lists of previous systematic reviews and meta-analyses of plantar fasciitis. Two reviewers (XL and LZ) independently reviewed the titles and abstracts of all of the studies retrieved by the search. Duplicates were removed using Endnote X7 (Thomson Reuters Co, New York). Full-text articles were obtained and examined if necessary. Then, the reviewers selected potentially relevant studies according to the eligibility criteria. If a disagreement occurred regarding the inclusion or exclusion of a study, then a third reviewer (JY) was consulted.
2.3 Eligibility criteria
2.3.1 Types of studies
We included randomized controlled clinical trials for treating plantar fasciitis. Studies should evaluate at least 2 modalities, including sham therapy. Trials were excluded if they did not provide validated therapeutic protocols; did not report the results of pain relief; did not provide a specific range of treatment intensity of the intervention group; or compared different methods of focused extracorporeal shock wave therapy application within the same range of energy output.
2.3.2 Types of participants
Participants in the included studies were adults (older than 18 years) diagnosed with plantar fasciitis. Studies were excluded if they included participants belonging to specific populations such as athletes or those who had suffered from fracture or underwent surgery on the involved heels.
2.3.3 Types of interventions
Interventions included shock wave therapy (focused extracorporeal shock wave therapy and radial extracorporeal shock wave therapy). Additionally, low-level laser therapy, ultrasound therapy, intracorporeal pneumatic shock therapy, and ultrasound-guided pulsed radiofrequency treatment were included. The therapies were compared with each other or with a sham group. We also divided the treatment intensity of FSW into 3 levels according to Chang method: low intensity (energy flux density ≤0.08 mJ/mm2), medium intensity (energy flux density = 0.08–0.28 mJ/mm2), and high intensity (energy flux density ≥ 0.28 mJ/mm2).
2.3.4 Types of outcomes
We chose pain relief as the outcome because pain is the predominant symptom in participants with plantar fasciitis. Validated measures of the change in pain relief included the visual analogue scale (VAS), the numerical rating scale pain score, the pain subscale of the validated Foot Function Index, or other indices of pain relief outcomes. All pain data were transformed to a range of 0 to 10. The postinterventional follow-up time points were defined as following: short-term (0–6 weeks), intermediate-term (2–4 months), and medium-term (6–12 months). If the data from more than 1 follow-up result were available for 1 period, then the time points closest to 4 weeks, 3 months, and 6 months were adopted. In addition, if several states of pain were present, then we applied the following priority levels: morning pain,[21,22] first-step pain,[11–14,23] daily activity pain,[24,25] overall pain,[26,27] night pain, rest pain, and pressure pain.
2.4 Data extraction and quality assessment
Two reviewers (XL and LZ) independently extracted the data using a predesigned extraction form. The extracted information included the participant characteristics, interventions, treatment method, dose/schedule, proportion of patients who were female, average age of the patients, duration of condition, outcome measures, funding, and conflicts of interest. Next, the data were integrated. Discrepancies within the data abstracted were typically resolved through discussions; however, a third reviewer was consulted when an agreement could not be reached. Two reviewers (XL and LZ) independently assessed the risk of bias. The Cochrane Collaboration Risk of Bias Tool (Review Manager, V.5.2; Revman, Copenhagen, Denmark: The Nordic Cochrane Centre, The Cochrane Collaboration) was used. This tool covers the following domains: random sequence generation, allocation concealment, blinding, incomplete outcome data, selective reporting, and other biases. We also assessed the risk of bias across trials. If more than 50% of the information was from trials at a low risk of bias, the domain was judged to be at a low risk of bias. Similarly, if most information was from RCTs had an unclear/high risk of bias, the domain was considered to be at an unclear/high risk of bias.
After all the outcomes were evaluated, a summary of findings table was created using the GRADE system, following the 4-step approach to rate the quality of evidence in each of the direct and mixed estimates based on methods developed by the GRADE working group.[30,31]
2.5 Statistical analysis
First, a pair-wise meta-analysis was performed using random-effects models. Every pair of studies with the same treatment was determined. Next, the results were reported as the mean difference (MD) with corresponding 95% confidence intervals (CIs), in addition to the number of pairs of studies. These statistical analyses were performed using STATA via the Metan package (Version 14.0; STATA Corporation, College Station, TX). Statistical heterogeneity across studies was quantified using the χ2 test and by calculating the inconsistency (I2).
Second, random-effects network models were developed within the Bayesian framework using the Markov Chain Monte Carlo algorithm in WinBUGS (Bayesian inference Using Gibbs Sampling for Windows, Version 1.4.3; Imperial College and MRC, UK). The model was based on 3 Markov chains for 100,000 iterations after a burn-in of 50,000. A thinning interval of 10 was applied, which collected 1 sample every 10 iterations. Consequently, 30,000 samples were gained for each parameter. Through this process, the Brooks–Gelman–Rubin method was used to assess the convergence between the direct and indirect variances. According to the theory of Brooks and Gelman, if the result of the potential scale reduction factor (PSRF) approximates or is equal to 1, then convergence has been reached. This result was also presented via the MD with 95% CrIs. If the null value was not included in the 95% CrIs of the MD, then a significant difference was indicated. The rank probability for each treatment was estimated using WinBUGS, and the data were then imported into STATA. Next, plots of the surface under the cumulative ranking curves (SUCRA) were generated. The value of SUCRA was presented as the percentage of the area under the curve, where 100% indicates the best treatment, and 0% indicates the worst treatment. Placebo compared with other treatments was always considered; therefore, the comparisons between placebo and other treatments are also presented. The network order in STATA was used to plot the MDs and 95% CrIs over these 3 follow-up durations.
2.6 Inconsistency analysis
If a “loop” (e.g., A-B-C) was present in the network, then each comparison in the loop (e.g., A-B) might be an indirect result of the other comparisons (e.g., A-C and C-B); consequently, the direct and indirect results might differ. Model inconsistency was assessed using the node-splitting method. If the P value was smaller than .05, then an inconsistency was considered as detected. The node-splitting models were generated via the Gemtc package (version 0.6–1, http://cran.r-project.org/package=gemtc) within R (version 3.2.3, http://www.r-project.org).
2.7 Sensitivity analysis and meta-regression
To test the influence of low-quality studies, a sensitivity analysis was conducted after excluding the low-quality studies. We recalculated the network result via the rank probability. If no significant difference occurred, then the results were considered credible.
A meta-regression was performed to assess the relationship between the sample size and the treatment effect (Table 1). As recommended by the UK's National Institute for Health and Care Excellence, a single interaction term was used as the covariate. No significant change in the deviance information criterion (DIC) was observed. The covariate did not include a null in the 2 to 4 months result, indicating that the sample size of the study at 2 to 4 months was associated with the treatment effectiveness.
2.8 Ethical approval
This systematic review does not require ethical assessment because only indirect literature will be included and evaluated.
3.1 Eligible studies
Our search strategy yielded 1699 potentially relevant articles. After carefully screening the titles and abstracts, 87 potentially eligible articles were obtained. After a careful full-text screening, 68 articles were discarded for the reasons listed in Fig. 1. Twenty-four RCTs met the inclusion criteria.[11,13–15,17,21–28,37–43]
However, 3 articles[15,42,43] on radiotherapy, 1 on radiofrequency and 1 on cryotherapy could not be used to construct a comprehensive network with other studies because of self-contrast and were therefore excluded. Finally, 19 RCTs that mentioned ultrasound therapy, UG-PRF, L-FSW, M-FSW, H-FSW, RSW, LLLT, IPST, and NIN were included in the quantitative analysis. The characteristics of the included trials are presented in Table 2.
Nineteen trials with 1676 patients were included. The network for the 0 to 6 week, 2 to 4 month, and 6 to 12 month follow-up durations included 14, 11, and 9 studies, respectively. The total numbers of participants in these studies were 1027, 1213, and 932 respectively.
3.2 Risk of bias and quality assessment
Quantification of the risk of bias assessment is presented in Fig. 2. A random sequence was adequately generated in 14 trials, and the risk of bias in the domain of randomization was judged to be low. We considered that the risk of bias was unclear for the domain of allocation concealment because more than half of the studies had unclear information about the methods used to conceal the allocation (n = 10, 52.6%). For the incomplete outcome data element, there was a low risk of bias because most information stemmed from studies with a low risk of bias (n = 16, 84.2%). There was also a low risk of selective outcome reporting because 17 of the studies had a low risk of bias in this domain (89.5%). In addition, the risk of other biases was also low (n = 14, 73.7%). The outcome assessors were successfully blinded in most trials (n = 16, 84.2%), and the risk of bias in these domains was judged to be low. The participants and personnel were unsuccessfully blinded in most trials (n = 13, 68.4%), and the risk of bias in this domain was judged to be high.
The results of the GRADE evaluation of interventions for plantar fasciitis are presented in Appendix 2, http://links.lww.com/MD/C550. All the reasons for downgrading are labeled. Because the design of inclusion criteria was rigid, there was no obvious nontransitivity. Due to the inconsistency and imprecision in the 2 to 4 month and 6 to 12 month results, the confidence of the evidence at the 2 follow-ups was low or very low.
3.3 Results of the pair-wise meta-analysis
All data were imported into STATA, and direct comparisons were made using metan with a random-effects model. The MDs and 95% CIs were calculated. The results are listed in the lower-left triangle of Table 3(A–C, respectively).
RSW, LLLT, and IPST therapy were associated with a significant pooled reduction in VAS compared with the placebo at 0 to 6 weeks [MD = 6.60, 95% CI: (6.04, 7.16); MD = 2.34, 95% CI: (1.60, 3.08,); MD = 2.24, 95% CI: (1.44, 3.04); respectively]. Similarly, at 2 to 4 months, the MD in the VAS score reduction compared with placebo was 0.82 [95% CI: (0.2, 1.45)] for H-FSW and 2.31 [95% CI: (1.26, 3.36)] for UG-PRF. At 6 to 12 months, the MD was 1.11 [95% CI: (0.07, 2.15)] for UG-PRF and 4.92 [95% CI: (4.11, 5.73)] for IPST.
We found a significant improvement for M-FSW treatment compared with L-FSW [MD = 1.03, 95% CI: (0.44, 1.62)]. No significant difference was found between RSW and ultrasound at 0 to 6 weeks. However, RSW conferred a greater pain reduction than ultrasound at 2 to 4 months [MD = 2.39, 95% CI: (1.58, 3.21)] and 6 to 12 months [MD = 3.06, 95% CI: (2.14, 3.98)]. In addition, NIN demonstrated better pain relief effectiveness than M-CSF at 0 to 6 weeks [MD = 3.00, 95% CI: (2.48 3.52)] and 2 to 4 months [MD = 3.00, 95% CI: (2.62, 3.38)].
There was significant heterogeneity across the groups when comparing RSW versus ULT at 0 to 6 weeks, PLA versus RSW at 2 to 4 months, and PLA versus L-FSW and PLA versus RSW at 6 to 12 months (I2 = 76.3%, P = .040; I2 = 98.9%, P = .000; I2 = 97.6%, P = .000; I2 = 98.6%, P = .000, respectively). All the heterogeneity results are presented in Appendix 3, http://links.lww.com/MD/C551.
3.4 Results of the network meta-analysis
Three comprehensive network graphs were built using STATA (Fig. 3; the size of the circle represents the number of participants, and the thickness of the edge represents the number of studies). All potential comparisons were calculated via WinBUGS and are presented as MDs and 95% CrIs. All of the PSRF parameters were approximately 1, indicating a strong convergence. The results are listed in the upper-right triangle of Table 3, and the significant differences are shaded.
RSW, ultrasound, LLLT, IPST, and NIN all corresponded to significant pain reduction compared with placebo at 0 to 6 weeks [MD = 6.61, 95% CrI: (5.05, 8.18) for RSW; MD = 5.8, 95% CrI: (3.86, 7.73) for ultrasound; MD = 2.33, 95% CrI: (1.10, 3.58) for LLLT; MD = 2.24, 95% CrI: (0.61, 3.84) for IPST; and MD = 3.00, 95% CrI: (2.48, 3.52) for NIN]. In contrast, statistical significance was observed only for RSW when compared with the placebo [MD = 3.67, 95% CrI: (0.31, 6.9)]. No significant differences were found for the 2- to 4-month and 6- to 12-month periods, as all the results had wide 95% CrIs.
Because placebo was the most commonly used control, comparisons with placebo across 3 different follow-up durations are shown in the plot. With regard to changes in pain scores, all of the treatments yielded better outcomes than placebo over the 3 time periods except for L-FSW at the short-term and M-FSW and H-FSW at the medium-term. However, only LLLT, ultrasound therapy, RSW, and IPST showed statistically significant superiority with regard to the short-term results. These results are presented in a forest plot (Fig. 4).
3.5 Rank probability based on SUCRA
The ranking probability of each treatment in terms of 3 follow-up durations is illustrated in Fig. 5. Larger areas under the SUCRA curve represent better effectiveness. NIN and IPST showed the highest probability of being the best treatment to relieve pain in the short (79.5%) and in the medium (81.1%) term, respectively, closely followed by RSW (79.4% and 78.6%respectively). RSW showed the highest probability of being the best treatment in the intermediate (83.9%) term.
3.6 Inconsistency analysis
Inconsistencies among the results are listed in Table 1. The P value for the inconsistency of PLA versus RSW at 0 to 6 weeks was .003475, indicating the detection of a significant inconsistency. The other P values were > .05, demonstrating that no inconsistencies were detected.
3.7 Sensitivity analysis and meta-regression
After excluding 1 low-quality study (Hawamdeh et al), the SUCRA changed slightly, although no change occurred in the rank probabilities, indicating that the results of the network meta-analysis are robust.
A meta-regression did not reveal a significant change in the DIC; however, we found that the covariate (sample size) did not include a null result over the 2 to 4-month period. This finding suggests that the covariate was associated with the treatment effectiveness at 2 to 4 months. No significant change in DIC was observed (Table 1).
The present meta-analysis compared the effectiveness of focused extracorporeal shock wave therapy, radial extracorporeal shock wave therapy, LLLT, NIN, pulsed radiofrequency treatment, and ultrasound therapy for the treatment of plantar fasciitis. The network meta-analysis indicated that RSW had the highest probability of providing the best outcome at 2 to 4 months (76.9%). NIN (79.5%) and IPST provided the best outcome (81.1%) at 0 to 6 weeks and 6 to 12 months, respectively, but were only slightly better than RSW (79.4% and 78.6%, respectively) at each time point.
We assessed 7 modalities in our network meta-analysis. These interventions treat plantar fasciitis with different types of energy output that can propagate through tissue, such as sonic waves, mechanical oscillations, heat, or light. These modalities are accompanied by processes involving energy generation and transformation, and effects on tissue are dependent on stimulus intensity, stimulus frequency, pulse interval, treatment duration, and application method (e.g., minimally invasive).
Extracorporeal shock wave continuously transmits sonic waves at a frequency of 16 Hz to 20 MHz and has been harnessed and used for various applications in medical science. RSW is a recently developed alternative to FSW for managing plantar fasciitis, and its major advantages over FSW include better patient tolerance and not requiring adjunct local anesthesia during treatment. In addition, RSW devices are cheaper, smaller, and easier to use. Furthermore, the present study showed that RSW corresponded to the second-best short- (0–6 weeks) and medium-term (6–12 months) outcomes and the best intermediate-term (6–12 months) outcome, that is, pain relief effectiveness, among all the modalities. This efficacy is satisfactory and stable. Thus, we strongly recommend RSW for adults with plantar fasciitis who have had symptoms for over 6 months and in whom conservative therapies have been unsuccessful but before surgical treatments are considered.
Regarding FSW, L-FSW showed the most significant pooled reduction in VAS over the 6 to 12-month period (treatment effectiveness compared with placebo), followed by M- and H-FSW. This finding is consistent with Chang's result at 6 months. However, our data favor placebo treatment over M- or H-FSW, which may be in part due to the following reasons. We adopted 3 time points with different follow-up durations; however, the extracted data in Chang's meta-analysis corresponded to the follow-up assessment closest to 6 months (26 weeks), which is usually considered an intermediate-term assessment. Our work used different inclusion criteria than those in Chang's study. We included more recent RCTs and excluded RCTs of athletes. Moreover, we did not include Marks's work, which did not provide a baseline VAS score. (This author only provided changes in VAS scores at 1 and 6 months.) We included 6 other treatment types in our network meta-analysis (i.e., mixed treatment comparisons). Importantly, all of the 95% CIs regarding RSW in pain reduction included 0, and large 95% CIs including 0 were observed over the medium term. It should be noted that FSW was inferior to RSW at all time intervals, which may be because the shock wave energy of FSW may be weakened by disturbances (such as bone, calcifications, etc.) in the pathway to the target.Additionally, in terms of H-FSW, this method demonstrated unsatisfactory results and may cause more pain during treatment.[45,50] Thus, our data suggest that FSW is a suboptimal choice for treatment of plantar fasciitis compared with RSW.
Ultrasound therapy is another commonly used treatment for plantar fasciitis. It involves piezoelectric crystals that use high-frequency alternative current to transform electrical energy into mechanical oscillation energy. The pooled results of the network meta-analysis showed that ultrasound did not demonstrate significantly better outcomes than placebo over the short, intermediate, and medium terms due to the wide 95% CrIs. Thus, ultrasound therapy does not seem to be an optimal choice for reducing plantar fasciitis pain over the 3 time periods.
LLLT is a recently developed technology. It operates on the principles of photochemistry, using a discrete wavelength of light to initiate a signal transduction cascade by stimulating a protein capable of absorbing light energy, also known as a photoreceptor protein. Compared with placebo, LLLT showed the third-best short-term treatment effect [MD = 2.7, 95% CrI: (−0.42, 5.77)] in the network result, indicating its satisfactory short-term pain-relieving effects. Importantly, however, all studies included supported its short-term superiority; thus, long-term follow-up data are needed.
Only 1 study applied NIN. This method is a new form of electrotherapy that has already been shown to have positive results in treating myofascial syndrome and other musculoskeletal conditions. NIN might exert an anti-inflammatory effect in areas of local inflammation. Notably, this method demonstrated the best short-term [MD = 4.25, 95% CrI: (−1.4, 9.88)] and intermediate-term [MD = 3.84, 95% CrI: (−3.21, 10.94)] effectiveness. Thus, more relevant clinical RCTs of long-term results are required in the future.
Similarly, only 1 study applied IPST. This method showed promising results because it demonstrated the best medium-term effectiveness [MD = 4.92, 95% CrI: (−1.81, 11.78)] and better short-term effectiveness [MD = 2.22, 95% CrI: (−2.6, 7.01)] (ranked fifth) compared with placebo in the network result. This result may be due to the fact that it is typically used as a minimally invasive approach, which might engender superior effectiveness compared with noninvasive modalities. However, as an invasive approach, IPST may have a temporary lower efficacy compared with RSW because the VAS score is relatively higher at 0 to 6 weeks due to wound healing during that time. Although invasive, IPST is well tolerated under local anesthesia, imparts no weight bearing restrictions and allows participants to quickly return to work. Because a pneumatic lithotripter is much cheaper than commercially available shock wave machines and readily available in almost all hospitals, IPST may be a viable option to treat plantar fasciitis, especially in health centers where shock wave therapies are not available.
UG-PRF is a less neuro-destructive alternative than continuous radiofrequency heat lesions. It uses high-frequency current at 500 kHz, called radiofrequency, to generate heat at the electrode tip to cause coagulation. According to Ye et al, UG-PRF can inactivate the trigger points in the gastrocnemius, which can relieve muscle contracture and reduce plantar heel pain. During treatment, a needle is inserted into the gastrocnemius, rather than the heel, at trigger points. A traditional pair-wise meta-analysis associated UG-PRF with a significant pooled reduction in VAS compared with placebo over the intermediate and medium terms. The network result favored UG-PRF, although the trend was not significant.The potential failure of accurate needle placement and the risk of nerve injury should be carefully considered even under ultrasound guidance. Only 1 RCT included this intervention with unsatisfactory effectiveness; thus, more relevant studies of short-term results are expected in the future.
Collectively, regarding clinical implications, in patients who have had symptoms for over 6 months and in whom conservative therapy has been unsuccessful, we recommend RSW before surgical treatments are considered. ULT and FSW therapies, which are commonly used in current clinical practice, can also be considered treatment candidates. The benefits of IPST, NIN, and LLLT need to be carefully tested in additional clinical trials.
This study has some advantages and strengths. It only included RCTs with a prospective design criterion. All potential treatments (e.g., rarely studied NIN and IPST) were included because the type of intervention was not limited in the database search. Furthermore, all P values regarding the inconsistencies were less than .05. The sensitivity analysis did not show a significant change in the cumulative probabilities rank, and no significant change in the DIC was found according to the meta-regression results. Therefore, the outcome of this meta-analysis is valid and reliable. We adopted 3 time points for the different follow-up durations. This strategy provided us certain insight with regard to time effectiveness. We also divided FSW therapies into 3 subgroups based on different energy intensity effluxes. In addition, unlike conventional meta-analyses techniques, Bayesian methodology advantageously enabled us to simultaneously compare every treatment. We calculated indirect comparisons via Bayesian statistics. Furthermore, this model was used to build inconsistency, sensitivity, and meta-regression tests.
We discuss 4 limitations here. First, some treatments were presented in only 1 study, including IPST and NIN, or without results for all 3 follow-up durations. The pooled results also showed that NIN and IPST had the highest probability of being the most effective treatment at 0 to 6 weeks and 6 to 12 months, respectively. Thus, we highlight the need for more high-quality RCTs in the future to corroborate these results. Second, this network meta-analysis did not include radiotherapy in its comparison because the 4 identified articles only used radiotherapy for self-comparison and, consequently, failed to form a network with another treatment. Thus, we suggest that more RCTs should focus on radiotherapy for comparison with other modalities to further elucidate its pain-relieving efficacy. Third, the qualities of the included studies varied. Some studies were better designed RCTs with high patient numbers and adequate randomization; however, other studies had few participants or weak blinding/allocation. This limitation can be addressed after more high-quality studies are conducted in the future. Fourth, significant heterogeneity existed within particular subgroups because treatment schedules and dosages varied across studies, and our review ignored these differences. This uncertainty is magnified when integrating these factors in mixed-treatment comparisons, as evidenced by the enlarged 95% CrIs. Our meta-regression indicated that the sample size was associated with the treatment effectiveness over the 2 to 4-monthperiod. This finding might explain why most of the indirect comparisons were insignificant in the intermediate and medium terms. There is inconsistency between the direct and indirect comparisons for PLA versus RSW (P = .003475). Thus, network comparisons (mixed evidence) of PLA versus RSW were rated down in GRADE due to incoherence.
The present meta-analysis compared the effectiveness of 8 modalities for treating plantar fasciitis. Regarding the 3 follow-up effectiveness time points, RSW provided relatively more effective and stable pain relief compared with other interventions and is therefore a promising candidate for clinical applications. Ultrasound therapy and FSW therapies can also be considered treatment candidates. However, H-FSW and UG-PRF are not recommended. Regarding the 0- to 6-week and 6- to 12-month periods, NIN and IPST provided the greatest pain relief, respectively, and, thus, have the potential to be more effective alternatives. Randomized trials comparing NIN, IPST, UG-PRF, and LLLT are needed to obtain more precise estimates of their relative efficacy.
The authors gratefully acknowledge the financial support from National Major Scientific and Technological Special Project for “National New Drug Innovation Program” during the Thirteenth Five-Year Plan Period (2017ZX09304003).
Conceptualization: Xian Li, Li Zhang, Shuming Gu, Rui Gao.
Data curation: Xian Li, Li Zhang, Shuming Gu.
Formal analysis: Xian Li, Zongshi Qin.
Investigation: Xian Li, Ning Ding, Rui Gao.
Methodology: Xian Li, Shuming Gu, Zongshi Qin, Yu Zhong.
Project administration: Xian Li, Li Zhang, Jianfeng Sun, Jiaji Yue.
Resources: Xian Li, Li Zhang, Zongshi Qin, Yu Zhong.
Software: Xian Li, Li Zhang, Jianfeng Sun, Yu Zhong, Ning Ding.
Supervision: Rui Gao.
Writing – original draft: Xian Li, Li Zhang.
Writing – review & editing: Xian Li, Li Zhang.
. Kamonseki DH, Gonçalves GA, Yi LC, et al. Effect of stretching with and without muscle strengthening exercises for the foot and hip in patients with plantar fasciitis
: a randomized controlled single-blind clinical trial. Man Ther 2016;23:76–82.
. James AM, Williams CM, Haines TP. Effectiveness of footwear and foot orthoses for calcaneal apophysitis: a 12-month factorial randomised trial. Br J Sports Med 2016;50:1268–75.
. Lee WC, Wong WY, Kung E, et al. Effectiveness of adjustable dorsiflexion night splint in combination with accommodative foot orthosis on plantar fasciitis
. J Rehabil Res Dev 2012;49:1557–64.
. Aqil A, Siddiqui MR, Solan M, et al. Extracorporeal shock wave therapy is effective in treating chronic plantar fasciitis
: a meta-analysis of RCTs. Clin Orthop Relat Res 2013;471:3645–52.
. Mahindra P, Yamin M, Selhi HS, et al. Chronic plantar fasciitis
: effect of platelet-rich plasma, corticosteroid, and placebo. Orthopedics 2016;39:e285–9.
. Cutts S, Obi N, Pasapula C, et al. Plantar fasciitis
. Ann R Coll Surg Engl 2012;94:539–42.
. Othman AM, Hegazy IH. Endoscopic plantar fasciotomy versus injection of platelet-rich plasma for resistant plantar fasciopathy. J Orthop 2015;12(suppl 2):S176–81.
. Thomas JL, Christensen JC, Kravitz SR, et al. The diagnosis and treatment of heel pain: a clinical practice guideline-revision 2010. J Foot Ankle Surg 2010;49:S1–9.
. Wang CJ, Wang FS, Yang KD, et al. Shock wave therapy induces neovascularization at the tendon-bone junction. A study in rabbits. J Orthop Res 2003;21:984–9.
. Martin RL, Davenport TE, Reischl SF, et al. Heel pain-plantar fasciitis
: revision 2014. J Orthop Sports Phys Ther 2014;44:A1–33.
. Konjen N, Napnark T, Janchai S. A comparison of the effectiveness of radial extracorporeal shock wave therapy and ultrasound therapy in the treatment of chronic plantar fasciitis
: a randomized controlled trial. J Med Assoc Thai 2015;98:S49–56.
. Van Zundert J, de Louw AJA, Joosten EAJ, et al. Pulsed and continuous radiofrequency current adjacent to the cervical dorsal root ganglion of the rat induces late cellular activity in the dorsal horn. Anesthesiology 2005;102:125–31.
. Ye L, Mei Q, Li M, et al. A comparative efficacy evaluation of ultrasound-guided pulsed radiofrequency treatment in the gastrocnemius in managing plantar heel pain: a randomized and controlled trial. Pain Med 2015;16:782–90.
. Macias DM, Coughlin MJ, Zang K, et al. Low-level laser therapy at 635 nm for treatment of chronic plantar fasciitis
: a placebo-controlled, randomized study. J Foot Ankle Surg 2015;54:768–72.
. Niewald M, Holtmann H, Prokein B, et al. Randomized multicenter follow-up trial on the effect of radiotherapy on painful heel spur (plantar fasciitis
) comparing two fractionation schedules with uniform total dose: first results after three months’ follow-up. Radiat Oncol 2015;10:174.
. Razzano C, Carbone S, Mangone M, et al. Treatment of chronic plantar fasciitis
with noninvasive interactive neurostimulation: a prospective randomized controlled study. J Foot Ankle Surg 2017;56:768–72.
. Dogramaci Y, Kalaci A, Emir A, et al. Intracorporeal pneumatic shock application for the treatment of chronic plantar fasciitis
: a randomized, double blind prospective clinical trial. Arch Orthop Trauma Surg 2010;130:541–6.
. Chang KV, Chen SY, Chen WS, et al. Comparative effectiveness of focused shock wave therapy of different intensity levels and radial shock wave therapy for treating plantar fasciitis
: a systematic review and network meta-analysis. Arch Phys Med Rehabil 2012;93:1259–68.
. Yin MC, Ye J, Yao M, et al. Is extracorporeal shock wave therapy clinical efficacy for relief of chronic, recalcitrant plantar fasciitis
? A systematic review and meta-analysis of randomized placebo or active-treatment controlled trials. Arch Phys Med Rehabil 2014;95:1585–93.
. Hutton B, Salanti G, Caldwell DM, et al. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann Intern Med 2015;162:777–84.
. Haake M, Buch M, Schoellner C, et al. Extracorporeal shock wave therapy for plantar fasciitis
: randomised controlled multicentre trial. BMJ 2003;327:75.
. Theodore GH, Buch M, Amendola A, et al. Extracorporeal shock wave therapy for the treatment of plantar fasciitis
. Foot Ankle Int 2004;25:290–7.
. Speed CA, Nichols D, Wies J, et al. Extracorporeal shock wave therapy for plantar fasciitis
. A double blind randomised controlled trial. J Orthop Res 2003;21:937–40.
. Kiritsi O, Tsitas K, Malliaropoulos N, et al. Ultrasonographic evaluation of plantar fasciitis
after low-level laser therapy: results of a double-blind, randomized, placebo-controlled trial. Lasers Med Sci 2010;25:275–81.
. Lee SJ, Kang JH, Kim JY, et al. Dose-related effect of extracorporeal shock wave therapy for plantar fasciitis
. Ann Rehabil Med 2013;37:379–88.
. Gerdesmeyer L, Frey C, Vester J, et al. Radial extracorporeal shock wave therapy is safe and effective in the treatment of chronic recalcitrant plantar fasciitis
: results of a confirmatory randomized placebo-controlled multicenter study. Am J Sports Med 2008;36:2100–9.
. Kudo P, Dainty K, Clarfield M, et al. Randomized, placebo-controlled, double-blind clinical trial evaluating the treatment of plantar fasciitis
with an extracoporeal shockwave therapy (ESWT) device: a North American confirmatory study. J Orthop Res 2006;24:115–23.
. Rompe JD, Schoellner C, Nafe B. Evaluation of low-energy extracorporeal shock-wave application for treatment of chronic plantar fasciitis
. J Bone Joint Surg Am 2002;84-A:335–41.
. Guyatt GH, Oxman AD, Vist GE, et al. GRADE: an emerging consensus on rating quality of evidence and strength of recommendations. BMJ 2008;336:924–6.
. Salanti G, Del Giovane C, Chaimani A, et al. Evaluating the quality of evidence from a network meta- analysis. PLoS One 2014;9:e99682.
. Puhan MA, Schünemann HJ, Murad MH, et al. A GRADE Working Group approach for rating the quality of treatment effect estimates from network meta-analysis. BMJ 2014;349:g5630.
. Jonas DE, Wilkins TM, Bangdiwala S, et al. Findings of Bayesian Mixed Treatment Comparison Meta-Analyses. Rockville, MD: Agency for Healthcar Research and Quality; 2013.
. Brooks SP, Gelman A. General methods for monitoring convergence of iterative simulations. J Comp Graph Stat 1998;7:434–55.
. Chaimani A, Higgins JP, Mavridis D, et al. Graphical tools for network meta-analysis in Stata. PLoS One 2013;8:e76654.
. Dias S, Welton NJ, Caldwell DM, et al. Checking consistency in mixed treatment comparison meta-analysis. Stat Med 2010;29:932–44.
. Dias S, Sutton AJ, Welton NJ, et al. NICE DSU Technical Support Document 3:Heterogeneity: Subgroups, Meta-Regression, Bias and Bias-Adjustment. Sheffield: Decision Support Unit ScHARR; 2011; 1–24.
. Costantino C, Vulpiani MC, Romiti D, et al. Cryoultrasound therapy in the treatment of chronic plantar fasciitis
with heel spurs. A randomized controlled clinical study. Eur J Phys Rehabil Med 2014;50:39–47.
. Crawford F, Atkins D, Edwards J. Interventions for treating plantar heel pain. Foot 2001;11:228–50.
. Hawamdeh Z, Alghwiri AA, Nassar A. The short-term effect of extracorporeal shock wave in treating plantar fasciitis
: RCT. Jordan Med J 2016;50:1–1.
. Ibrahim MI, Donatelli RA, Schmitz C, et al. Chronic plantar fasciitis
treated with two sessions of radial extracorporeal shock wave therapy. Foot Ankle Int 2010;31:391–7.
. Liang HW, Wang TG, Chen WS, et al. Thinner plantar fascia predicts decreased pain after extracorporeal shock wave therapy. Clin Orthop Relat Res 2007;460:219–25.
. Ott OJ, Jeremias C, Gaipl US, et al. Radiotherapy for benign calcaneodynia: long-term results of the Erlangen Dose Optimization (EDO) trial. Strahlenther Onkol 2014;190:671–5.
. Ott OJ, Jeremias C, Gaipl US, et al. Radiotherapy for calcaneodynia. Results of a single center prospective randomized dose optimization trial. Strahlenther Onkol 2013;189:329–34.
. Osmon AM, El-Hamady DH, Kotb MM. Pulsed compared to thermal radiofrequency to the medial calcaneal nerve for management of chronic refractory plantar fasciitis
: a prospective comparative study. Pain Phys 2016;19:E1181–7.
. Schmitz C, Császár NB, Rompe JD, et al. Treatment of chronic plantar fasciopathy with extracorporeal shock waves (review). J Orthop Surg Res 2013;8:31.
. Lohrer H, Nauck T, Korakakis V, et al. Historical ESWT paradigms are overcome: a narrative review. BioMed Res Int 2016;2016:3850461.
. Rompe JD, Decking J, Schoellner C, et al. Shock wave application for chronic plantar fasciitis
in running athletes. A prospective, randomized, placebo-controlled trial. Am J Sports Med 2003;31:268–75.
. Marks W, Jackiewicz A, Witkowski Z, et al. Extracorporeal shock-wave therapy (ESWT) with a new-generation pneumatic device in the treatment of heel pain. A double blind randomised controlled trial. Acta Orthop Belg 2008;74:98–101.
. Schmitz C, Császár NBM, Milz S, et al. Efficacy and safety of extracorporeal shock wave therapy for orthopedic conditions: a systematic review on studies listed in the PEDro database. Br Med Bull 2015;116:115–38.
. Zhiyun L, Tao J, Zengwu S. Meta-analysis of high-energy extracorporeal shock wave therapy in recalcitrant plantar fasciitis
. Swiss Med Wkly 2013;143:w13825.
. Xia P, Wang X, Lin Q, et al. Effectiveness of ultrasound therapy for myofascial pain syndrome: a systematic review and meta-analysis. J Pain Res 2017;10:545–55.