Pulsed Electromagnetic Field Therapy in People with Knee Osteoarthritis: A Systematic Review and Meta-Analysis
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsSUMMARY
This systematic review and meta-analysis evaluates the efficacy of pulsed electromagnetic field (PEMF) therapy for knee osteoarthritis (KOA), synthesizing evidence from nine RCTs (n=457). The authors report time-dependent effects on WOMAC subscales and conduct subgroup analyses by device parameters. While the topic is clinically relevant and the analytical approach has merit, several methodological and reporting concerns limit confidence in the findings and must be addressed.
ABSTRACT
Systematic review abstracts should report I² values or other heterogeneity measures to allow readers to immediately assess the consistency of findings across studies (PRISMA 2020 guideline, Item #13). The authors present pooled effect estimates without indicating substantial heterogeneity was present (e.g., I²=94% for VAS pain at one month, as shown in Figure 5).
INTRO
Although the introduction states the aim of evaluating PEMF versus sham or other controls, it does not explicitly define the research question using a structured PICO format (Population, Intervention, Comparator, Outcomes). Specifically, it remains unclear:
Whether all KOA severities were eligible
Whether co-interventions were allowed (later mentioned ambiguously)
What follow-up durations were considered clinically relevant
Whether comparators were restricted to sham, active therapy, or botH
This is not a osteoarthritis specific journal, therefore the reader would appreciate being introduced to common conservative managements of knee
osteoarthritis in order to have a better understanding of the disease. Please acknowledge the role of conservative treatments first: "Established
conservative treatment for osteoarthritis include exercise (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3635671/ ), knee bracing
(https://pubmed.ncbi.nlm.nih.gov/29931372/ ), physical modalities ( https://pubmed.ncbi.nlm.nih.gov/25162407/ ), pharmacology
(https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6315310/ )
METHODS
Only three databases were searched (PubMed, Embase, Cochrane Library), with no reported searches of clinical trial registries (e.g., ClinicalTrials.gov), grey literature sources, or physiotherapy-specific databases (e.g., PEDro, CINAHL). The PRISMA 2020 statement recommends searching multiple databases and supplementary sources to minimize publication bias.
The inclusion criteria state participants "may have received other treatment with the PEMF" but do not clearly define acceptable comparators. The included studies compare PEMF to sham, other physical therapy modalities (laser, shockwave), and exercise combinations—representing substantial clinical heterogeneity. PRISMA and Cochrane guidance emphasize that PICO elements should be explicitly defined.
RESULTS
The authors present analyses at "18-21 days" and "one month" separately but do not explain in the Methods how these time points were determined a priori. This appears to be a post-hoc decision based on available data, raising concerns about selective reporting of time points that show favorable results.
The subgroup analyses by amplitude (>1mT vs. ≤1mT) and frequency (>100Hz vs. ≤100Hz) contain only 1-2 studies per subgroup (e.g., the "low amplitude with high frequency" subgroup contains only one study—Bagnato 2016). With such limited data, subgroup findings are highly unstable and should not be used to draw mechanistic conclusions.
DISCUSSION
Section 4.4 ("Role of amplitude and frequency") presents detailed biological mechanisms explaining why different PEMF parameters produce different effects (adenosine receptor modulation, chondrocyte proliferation). However, this review provides no direct mechanistic data—only associations from underpowered subgroup analyses. The Discussion inappropriately elevates these exploratory findings to causal claims (e.g., "High-amplitude fields provide the necessary energy to stimulate chondrocyte proliferation"). Such assertions require direct experimental evidence, not inference from clinical outcome patterns.
Author Response
Comments 1: Systematic review abstracts should report I² values or other heterogeneity measures to allow readers to immediately assess the consistency of findings across studies (PRISMA 2020 guideline, Item #13). The authors present pooled effect estimates without indicating substantial heterogeneity was present (e.g., I²=94% for VAS pain at one month, as shown in Figure 5).
Response 1: Thank you for pointing this out. We agree with this comment. Therefore, we have added the I² values to the pooled effect estimates in the Results section of the Abstract to accurately reflect the heterogeneity across the included studies. This change can be found on [Page 1, Line 28-30].
Comments 2: Although the introduction states the aim of evaluating PEMF versus sham or other controls, it does not explicitly define the research question using a structured PICO format (Population, Intervention, Comparator, Outcomes). Specifically, it remains unclear: Whether all KOA severities were eligible; Whether co-interventions were allowed (later mentioned ambiguously); What follow-up durations were considered clinically relevant; Whether comparators were restricted to sham, active therapy, or both.
Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have revised the final paragraph of the Introduction to explicitly state our research question using the PICO format. Furthermore, we have clarified the acceptable comparators and co-interventions in the Methods section to address the ambiguity regarding "other treatments". These changes can be found on [Page 2, Line 79-84] (Introduction) and [Page 3, Line 102-106] (Methods).
Comments 3: This is not a osteoarthritis specific journal, therefore the reader would appreciate being introduced to common conservative managements of knee osteoarthritis in order to have a better understanding of the disease. Please acknowledge the role of conservative treatments first: "Established conservative treatment for osteoarthritis include exercise (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3635671/ ), knee bracing (https://pubmed.ncbi.nlm.nih.gov/29931372/ ), physical modalities ( https://pubmed.ncbi.nlm.nih.gov/25162407/ ), pharmacology (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6315310/ )"
Response 3: Thank you for pointing this out. We agree with this comment. Therefore, we have incorporated a new paragraph in the Introduction detailing established conservative treatments (exercise, knee bracing, physical modalities, and pharmacology) and cited the recommended references to provide a better understanding of the disease management. This change can be found on [Page 2, Line 58-62].
Comments 4: Only three databases were searched (PubMed, Embase, Cochrane Library), with no reported searches of clinical trial registries (e.g., ClinicalTrials.gov), grey literature sources, or physiotherapy-specific databases (e.g., PEDro, CINAHL). The PRISMA 2020 statement recommends searching multiple databases and supplementary sources to minimize publication bias. The inclusion criteria state participants "may have received other treatment with the PEMF" but do not clearly define acceptable comparators. The included studies compare PEMF to sham, other physical therapy modalities (laser, shockwave), and exercise combinations—representing substantial clinical heterogeneity. PRISMA and Cochrane guidance emphasize that PICO elements should be explicitly defined.
Response 4: Thank you for pointing this out. We agree with this comment. Therefore, we have explicitly added the constraint regarding the limited number of searched databases to the Limitations section to ensure readers are aware of potential publication bias. (Note: The concern regarding PICO elements and comparators has been addressed in our response and revisions to Comment 2). This change can be found on [Page 14, Line 410-414].
Comments 5: The authors present analyses at "18-21 days" and "one month" separately but do not explain in the Methods how these time points were determined a priori. This appears to be a post-hoc decision based on available data, raising concerns about selective reporting of time points that show favorable results.
Response 5: Thank you for pointing this out. We agree with this comment. Therefore, we have clarified in the Methods section that these specific time points were not defined a priori but were categorized post-hoc based on the natural clustering of follow-up periods reported in the included trials, rather than selective reporting. This change can be found on [Page 4, Line 144-147].
Comments 6: The subgroup analyses by amplitude (>1mT vs. ≤1mT) and frequency (>100Hz vs. ≤100Hz) contain only 1-2 studies per subgroup (e.g., the "low amplitude with high frequency" subgroup contains only one study—Bagnato 2016). With such limited data, subgroup findings are highly unstable and should not be used to draw mechanistic conclusions. Section 4.4 ("Role of amplitude and frequency") presents detailed biological mechanisms explaining why different PEMF parameters produce different effects (adenosine receptor modulation, chondrocyte proliferation). However, this review provides no direct mechanistic data—only associations from underpowered subgroup analyses. The Discussion inappropriately elevates these exploratory findings to causal claims (e.g., "High-amplitude fields provide the necessary energy to stimulate chondrocyte proliferation"). Such assertions require direct experimental evidence, not inference from clinical outcome patterns.
Response 6: Thank you for pointing this out. We agree with this comment. Therefore, we have extensively revised Section 4.4 to tone down the language, removing causal assertions and reframing these points as potential mechanistic hypotheses. We also explicitly acknowledged the limitation of the small number of studies in these subgroups. This change can be found on [Page 13, Line 377-381, 388-391, 394-398].
Reviewer 2 Report
Comments and Suggestions for AuthorsThis article addresses the current and clinically relevant issue of electromagnetic field therapy (PEMF) in the treatment of knee osteoarthritis. The authors conducted a systematic review and meta-analysis of randomized clinical trials, analyzing the impact of device parameters (frequency and amplitude) on clinical outcomes. The topic of this work is important from the perspective of rehabilitation, physical therapy, and conservative treatment of osteoarthritis. The work has scientific value and publication potential, but requires clarification of the methodology, expanded discussion, and improvement of certain aspects of the interpretation of the results and study limitations. Detailed comments and my own comments on the work are presented below.
Major comments:
- The introduction of the work requires expansion to include a more detailed discussion of the biomechanics of knee osteoarthritis and the mechanisms of pain and functional limitation. I also recommend supplementing the introduction with the latest publications from the last 2-3 years on conservative treatment of OA and physical therapy. I recommend adding: DOI: 10.3390/app15126896 ; DOI: 10.3390/ijms26157493
- The authors limited their literature search to English-language publications only, which may lead to publication bias. I recommend briefly discussing this limitation in the limitations section.
- The number of included studies is relatively small (9 RCTs, 457 patients), which limits the statistical power of the meta-analysis. I recommend more explicitly emphasizing the pilot nature of the analysis and more cautiously interpreting the results.
- A significant problem is the significant heterogeneity of PEMF protocols (different amplitudes, frequencies, and treatment durations), which complicates the interpretation of the meta-analysis results. I recommend expanding the discussion regarding the impact of intervention heterogeneity on the analysis results.
- The study analyzed results primarily from short-term follow-up (18–21 days and 1 month), which prevents assessment of the long-term effectiveness of PEMF therapy. I recommend more explicitly emphasizing this limitation in the limitations section.
- The authors analyzed both the 0–10 and 0–100 VAS scales, which may influence the results of the meta-analysis. I recommend a more detailed description of the data standardization method and the potential impact of this procedure on the results, along with current literature.
- The study demonstrated statistically significant differences in some WOMAC parameters, but these did not reach the MCID. I recommend that the distinction between statistical significance and clinical significance be more clearly emphasized in the discussion.
- The subgroup analysis of PEMF amplitude and frequency is interesting and a strength of the study, but the number of studies in the subgroups is very small. I recommend emphasizing the exploratory nature of this analysis.
Author Response
Comments 1: The introduction of the work requires expansion to include a more detailed discussion of the biomechanics of knee osteoarthritis and the mechanisms of pain and functional limitation. I also recommend supplementing the introduction with the latest publications from the last 2-3 years on conservative treatment of OA and physical therapy. I recommend adding: DOI: 10.3390/app15126896 ; DOI: 10.3390/ijms26157493
Response 1: Thank you for pointing this out. We completely agree with this comment. Therefore, we have expanded the Introduction to discuss the biomechanics of KOA and the mechanisms underlying pain and functional limitations. Additionally, we have incorporated the recommended recent publications to provide a more comprehensive and up-to-date overview of conservative treatments. These changes can be found on [Page 2, Line 48-53].
Comments 2: The authors limited their literature search to English-language publications only, which may lead to publication bias. I recommend briefly discussing this limitation in the limitations section.
Response 2: Thank you for pointing this out. We agree with this comment. Therefore, we have added a clear statement in the Limitations section acknowledging that restricting our search to English-language publications may have introduced publication bias. This can be found on [Page 14, Line 410-414].
Comments 3: The number of included studies is relatively small (9 RCTs, 457 patients), which limits the statistical power of the meta-analysis. I recommend more explicitly emphasizing the pilot nature of the analysis and more cautiously interpreting the results.
Response 3: Thank you for pointing this out. We agree with this comment. Therefore, we have revised the Limitations section to explicitly emphasize the relatively small number of included studies and patients. We noted that this constrains statistical power and that the findings should be interpreted with caution, highlighting the pilot nature of the synthesis. This change can be found on [Page 13-14, Line 406-410].
Comments 4: A significant problem is the significant heterogeneity of PEMF protocols (different amplitudes, frequencies, and treatment durations), which complicates the interpretation of the meta-analysis results. I recommend expanding the discussion regarding the impact of intervention heterogeneity on the analysis results.
Response 4: Thank you for pointing this out. We agree with this comment. Therefore, we have expanded Section 4.2 in the Discussion to address how the substantial heterogeneity in PEMF parameters (amplitudes, frequencies, and durations) complicates the interpretation of our pooled results. This change can be found on [Page 12, Line 335-340].
Comments 5: The study analyzed results primarily from short-term follow-up (18–21 days and 1 month), which prevents assessment of the long-term effectiveness of PEMF therapy. I recommend more explicitly emphasizing this limitation in the limitations section.
Response 5: Thank you for pointing this out. We agree with this comment. Therefore, we have more explicitly stated in the Limitations section that focusing on short-term follow-ups prevents any conclusions regarding the long-term effectiveness or durability of PEMF therapy. This change can be found on [Page 14, Line 414-416].
Comments 6: The authors analyzed both the 0–10 and 0–100 VAS scales, which may influence the results of the meta-analysis. I recommend a more detailed description of the data standardization method and the potential impact of this procedure on the results, along with current literature.
Response 6: Thank you for pointing this out. We agree that mixing different scales can influence results. Therefore, to preserve clinical interpretability and avoid the potential distortion of data standardization (such as converting to a Standardized Mean Difference), we deliberately chose not to pool the 0-10 and 0-100 VAS scales into a single analysis. Instead, we analyzed them separately as distinct subgroups using the Mean Difference (MD), as illustrated in Figures 4 and 5 . We have added a clearer description of this methodological choice to the Methods section to prevent any confusion. This change can be found on [Page 4, Line 135-138].
Comments 7: The study demonstrated statistically significant differences in some WOMAC parameters, but these did not reach the MCID. I recommend that the distinction between statistical significance and clinical significance be more clearly emphasized in the discussion.
Response 7: Thank you for pointing this out. We agree with this comment. Therefore, we have strengthened the Discussion (Section 4.3) to more clearly distinguish between statistical significance and clinical relevance, emphasizing that the observed statistical improvements did not reach the recognized Minimal Clinically Important Difference (MCID) thresholds. This change can be found on [Page 12, Line 355-359].
Comments 8: The subgroup analysis of PEMF amplitude and frequency is interesting and a strength of the study, but the number of studies in the subgroups is very small. I recommend emphasizing the exploratory nature of this analysis.
Response 8: Thank you for pointing this out. We agree with this comment. Therefore, we have added text to Section 4.4 to explicitly emphasize the exploratory nature of the subgroup analysis and the limitation of the small number of included studies. This change can be found on [Page 13, Line 394-398].
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsAuthors fully addressed my comments
