822
L. Chen et al.
bone and joint diseases, especially for elderly patients
or those with certain hepatorenal insufficiency who
are unable to undergo surgery or take medications (6).
Moreover, a large body of basic research has shown that
PEMF therapy can promote the proliferation of chondro-
cytes and the secretion of chondrocyte extracellular
matrix, which are beneficial to the repair of cartilage
damage caused by knee OA (7). However, a series of
randomized controlled trials (RCTs) and systematic re-
views on the therapeutic effects of PEMF therapy in the
clinical treatment of knee OA have yielded controversial
results about joint pain, stiffness and physical function
(8). In 2013, Ryang et al. (9) conducted a systematic
review of PEMF therapy in the treatment of knee OA,
which showed that studies with high-quality or low-
quality methodology may report different therapeutic
effects of PEMF therapy. Nevertheless, these 4 publis-
hed systematic reviews (9–12) contained either studies
with low-quality methodology or studies that reported
the results not of classical PEMF therapy but of pulsed
short-wave therapy.
Thus, based on controversial clinical trials and
systematic reviews, the therapeutic effects of classical
PEMF therapy in the management of knee OA remain
to be validated in this context. Several new RCTs on
this subject have been published recently. The aim of
this systematic review and meta-analysis of randomi-
zed placebo-controlled trials was therefore to assess
the efficacy of classical PEMF therapy on joint pain,
joint stiffness and physical function in patients with
knee OA.
MATERIAL AND METHODS
This systematic review and meta-analysis was based on pre-
viously published literature according to the Preferred Reporting
Items for Systematic Reviews and Meta-analyses (PRISMA)
statement, thus it was deemed exempt by the ethics committee of
West China Hospital, Sichuan University, Chengdu, China (13).
Search strategy
Electronic databases, PubMed, EMBASE, Web of Science, and
the Cochrane Library, were searched from inception to April
2018 to obtain relevant studies. The main search method was
to use both MeSH and the keyword “knee OA”, combined with
electromagnetic fields [MeSH] OR pulsed electromagnetic field
OR pulse electric-magnetic field OR PEMF OR magnetotherapy
OR magnetic therapy. The publication type was limited to clini-
cal trials or RCTs. In addition, a hand search was performed to
identify the relevant references included in articles.
Inclusion and exclusion criteria
Inclusion criteria for studies in this meta-analysis were:
• All patients had a clear diagnosis of chronic knee OA based
on the clinical or radiological criteria of the American College
of Rheumatology, with no restrictions on sex or race.
www.medicaljournals.se/jrm
• Treatment should be classic PEMF therapy, rather than short-
wave, electrical stimulation, magnetic resonance, or other
physical therapies.
• The control group should be a placebo group.
• Patients’ baseline and primary outcome both presented the
severity of joint pain, assessed by Western Ontario and
McMaster Universities Osteoarthritis Index (WOMAC)
questionnaire or 10-cm visual analogue scale (VAS) for pain
(where 0 means no pain, and 10 means pain as severe as the
subject can imagine). The WOMAC questionnaire contains
24 items divided into 3 subscales: pain (5 items), stiffness (2
items), and physical function (17 items). A higher WOMAC
score represents a worse outcome (14).
• Studies must be RCTs.
Exclusion criteria were:
• Animal or laboratory studies.
• Studies evaluating the severity of joint pain using the Knee
Injury and Osteoarthritis Outcome Score (KOOS).
• Insufficient research data, knee OA participants not reported
separately, data not extractable, or the corresponding authors
did not respond so that they could not be statistically analysed.
• The full text of the research was published in a language
other than English.
Study selection
First, 2 reviewers independently evaluated the eligibility of
the studies by reading the title and abstract according to the
aforementioned criteria. If the abstract information was insuf-
ficient to judge eligibility, further screening would be performed
by obtaining the full text. Disagreements were first resolved
through discussion; otherwise a third reviewer would conduct
an independent review if necessary.
Data extraction and quality assessment
A standard data extraction spreadsheet was used to collate
basic information from the eligible studies, including first
author, year of publication, number of patients, age, sex,
body mass index, disease duration, pulse frequency, magnetic
flux density, and treatment regimen. The primary outcomes
were then extracted, including joint pain, joint stiffness, and
physical function, as assessed by WOMAC scores or VAS-
related measurement. Data evaluated between 3 and 6 weeks
after the start of treatment were used to analyse the efficacy
of 1 month’s treatment. Data presented in other forms, such
as 95% confidence interval (95% CI) and standard error were
converted to a form of mean and standard deviation (SD) ac-
cording to the Cochrane Handbook for Systematic Reviews
of Interventions (available from http://handbook.cochrane.
org). When the raw data provided only baseline values and
values of changes in the form of mean and SD, the mean and
SD of the endpoint values were calculated using the formula
provided by the Cochrane Handbook for Systematic Reviews
of Interventions. All data extraction was carried out indepen-
dently by 2 reviewers and any disagreements were resolved
through discussion. However, when disagreements failed to
reach a consensus, the source data were re-examined and a
third reviewer was consulted.
As all included studies were RCTs, 2 reviewers used Review
Manager (RevMan) version 5.3 (The Cochrane Collabora-
tion, Oxford, UK) to independently assess the risk of bias in
accordance with the method recommended by the Cochrane
Handbook for Systematic Reviews of Interventions. The con-