Efficacy and Safety of Acupuncture Therapy for Patients with Acute Ankle Sprain: A Systematic Review and Meta-Analysis of Randomized Controlled Trials

Background The efficacy of acupuncture for acute ankle sprain (AAS) is controversial. This study aimed to critically assess the efficacy and safety of acupuncture for AAS. Methods Parallel-group randomized controlled trials (RCTs) were included regardless of language or publication date. Participants with AAS were included regardless of age, sex, race, nationality, or diagnostic criteria for AAS. Experimental interventions included acupuncture alone or in combination with traditional therapies. Control interventions included no treatment, placebo, or traditional therapies. The primary outcome was the Kofoed ankle score. The secondary outcomes included visual analogue scale, duration of pain, use of painkiller, ankle circumference, effective rate, cure rate, and adverse events. PubMed, Embase, Cochrane Library, Web of Science, China National Knowledge Infrastructure, Wanfang Digital Periodicals, and Chinese Science and Technology Periodicals database were searched to identify potentially eligible studies from inception to September 10, 2020. World Health Organization International Clinical Trials Registry Platform (WHO ICTRP), ClinicalTrials.gov, Chinese Clinical Trial Registry (ChiCTR), and the reference list of eligible RCTs were checked to identify ongoing or unpublished studies. Risk of bias was assessed by the Cochrane Collaboration's tool. Statistical analyses were performed by RevMan 5.3 software. P < 0.05 indicated statistical significance. Results Seventeen eligible studies were included for the statistical analysis. There was no statistically significant difference of Kofoed ankle score between acupuncture and Rest, Ice, Compression, and Elevation (RICE) group (P=0.75). However, acupuncture could significantly relieve pain (P=0.02) and increase cure rate (P=0.004) compared with RICE. Moreover, acupuncture plus RICE could also significantly relieve pain (P < 0.00001) and increase cure rate (P=0.01) compared with RICE alone. Acupuncture combined with massage could significantly relieve pain (P=0.04) compared with massage alone. Acupuncture plus Chinese medicine might be more effective for relieving pain (P < 0.00001), reducing the duration of pain (P < 0.00001), and increasing cure rate (P=0.0002) compared with Chinese medicine alone. Two studies reported no adverse reactions. One study reported that a participant suffered from mild drug-related allergic reaction and was healed without treatment. Conclusions The findings of the present study suggest that acupuncture may be beneficial for AAS. However, more large-scale and well-designed RCTs are warranted.


Introduction
Acute ankle sprain (AAS) is defined as an acute injury of the ankle ligament [1]. Ankle sprain is one of the major injuries among the general population and athletes [2][3][4]. AAS may result in acute pain, swelling, high cost, chronic ankle instability, etc. [5,6]. ere are a variety of therapeutic interventions for AAS, involving pharmacological therapies (e.g., nonsteroidal anti-inflammatory drugs) and nonpharmacological therapies (e.g., functional support, exercise, and manual mobilization) [3,4]. However, no optimal therapies were recommended for treating AAS according to a latest evidence-based clinical guideline [4].
Acupuncture belongs to complementary and alternative medicine and is commonly used for relieving acute and chronic pain [7,8]. Two previous systematic reviews assessed the efficacy of acupuncture for ankle sprain [1,9]. However, the evidence on acupuncture for ankle sprain still remains inconclusive because of large heterogeneity [4].
ere are some methodological flaws in the two previous studies [1,9]. For example, high clinical heterogeneity might be caused by combining results from studies involving patients with acute and chronic ankle sprain. Different types of acupuncture resulted in the heterogeneity of interventions. Moreover, some new trials of acupuncture for AAS have been published in recent years [10][11][12][13][14][15][16][17]. However, the evidence has not been critically assessed. erefore, we conducted an updated systematic review to assess the efficacy and safety of acupuncture for AAS.

Methods
is systematic review was registered on PROSPERO (no. CRD42020156280). It was conducted in compliance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement [18].

Types of Studies.
Parallel-group randomized controlled trials (RCTs) were included regardless of language or publication date.

Types of Participants.
Participants with AAS were included regardless of age, sex, race, nationality, or diagnostic criteria for AAS.

Types of Interventions.
e experimental interventions included acupuncture alone or in combination with traditional therapies. e control interventions included no treatment, placebo, or traditional therapies. Traditional therapies for acute ankle sprain involve nonsteroidal antiinflammatory drugs, Rest, Ice, Compression, and Elevation (RICE), functional support, exercise, manual mobilization, etc. ere were no restrictions on frequency or duration of acupuncture. e following comparisons were considered if possible: (1) acupuncture alone versus no treatment/placebo/traditional therapies; (2) acupuncture plus traditional therapies versus traditional therapies alone; and (3) acupuncture plus traditional therapies versus traditional therapies plus placebo.

Types of Outcomes.
e primary outcome was the Kofoed ankle score. e secondary outcomes included visual analogue scale (VAS), duration of pain, use of painkiller, ankle circumference, effective rate, cure rate, and adverse events. Kofoed ankle score is comprised of pain, function, and mobility domain and ranges from 0 to 100 with higher score indicating less pain [19]. VAS ranges from 0 to 10 or 100 with higher score indicating more severe pain.  [20]. e detailed search strategy is available in Supplementary Material.

Selection of Studies and Data Extraction.
All studies identified from the electronic search were imported into EndNote software. Two reviewers (SWG and AFL) independently checked the title and abstract to remove duplicates and irrelevant studies. Full texts of the remaining studies were read to identify potentially eligible studies. e selection process was summarized using a PRISMA flow diagram. e following information was extracted independently by two reviewers (JXC and SWG). Disagreements were resolved by consensus or consultation with a third review author (JBZ).
(1) Study details: title, first author, country, year of publication, design, methods of randomization, allocation, and blinding (2) Patients: age, sample size (3) Interventions: type, frequency, and duration (4) Outcome measures: Kofoed ankle score, VAS, duration of pain, use of painkiller, ankle circumference, effective rate, cure rate, and adverse events 2.4. Assessment of Risk of Bias. Two reviewers (JXC and SWG) independently assessed the risk of bias in eligible studies using the Cochrane Collaboration's tool [21,22]. It includes seven important items: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective outcome reporting, and other potential sources of bias. e risk of bias for each item was classified as low, high, or unclear. e results were presented with risk of bias graph and summary figure.
2.5. Statistical Analysis. Mean difference (MD) with 95% confidence intervals (CIs) was calculated for continuous variables if the same tool was used to measure a certain outcome across different studies. Otherwise, standardized mean difference (SMD) was calculated. Risk ratio (RR) with 95% CIs was used for dichotomous variables. If clinical heterogeneity was low, meta-analysis was used to estimate the overall effect. Statistical heterogeneity was evaluated by chi-square test or I 2 statistics. If the P value of chi-square test was greater than 0.10 or I 2 was less than 50%, the fixed-effect model was used to estimate the effect size. Otherwise, the random-effect model was used. e funnel plot was used to assess publication bias when at least 10 studies were included in a meta-analysis. Subgroup analyses were performed based on intervention and comparison if possible. RevMan 5.3 software was used for the statistical analysis. P < 0.05 indicated statistical significance. If performing meta-analysis was infeasible, a narrative description was provided.

Characteristics of Included Studies.
e characteristics of the included studies are summarized in Table 1. Seventeen studies involving 1528 patients were published between 1999 and 2018 in China. Sample size ranged from 20 to 90 in the experimental group and 20 to 70 in the control group. Experimental interventions included acupuncture alone or in combination with RICE, dimethyl sulfoxide, Chinese medicine, or massage. Control interventions included RICE, dimethyl sulfoxide, Chinese medicine, massage, ice and hot pack, no treatment, or infrared radiation. [11,[14][15][16][17] used a random number table, one [25] used coin tossing, and one [28] used a computer random number generator to generate a random sequence. One study [30] used sealed envelopes to conceal allocation. Attrition bias was low because complete outcome data were reported in all included studies. Performance bias, detection bias, reporting, bias, and other bias were unclear because of insufficient information.

Kofoed Ankle
Score. One study [12] reported the response rate of Kofoed ankle score. Response was defined as having a Kofoed ankle score greater than or equal to 75. It indicated that no statistically significant difference was found between acupuncture and RICE group (n � 60, RR � 1.04, 95% CI: 0.80 to 1.36, P � 0.75).

Use of Painkillers.
Wu et al. [15] found no significant difference of use of painkillers between acupuncture plus massage and massage alone (n � 82, RR � 0.48, 95% CI: 0.16 to 1.46, P � 0.19).    [12] found that the difference of ankle circumference between the uninjured and injured ankle was not statistically different after acupuncture treatment compared with RICE (n � 60, MD � 0.24, 95% CI: −0.10 to 0.58, P � 0.17).

Effective Rate.
Fifteen studies reported the effective rate.
It is defined as a ratio of the number of patients labelled as cure, excellent, or effectivity divided by the number of patients in a certain group. Wu [28] found that the effective rate in the acupuncture group was statistically higher (P � 0.0002) than that in no treatment group ( Figure 4). However, the effective rate in the acupuncture group was not different statistically from that in massage (P � 1.00), "Ice and hot pack" plus Chinese medicine (P � 0.07), infrared radiation (P � 0.51), or RICE (P � 1.00) group. A metaanalysis showed that dimethyl sulfoxide could significantly increase the effective rate compared with acupuncture (P � 0.03). Yu [24] found that the effective rate in the acupuncture plus dimethyl sulfoxide group was statistically higher (P � 0.04) than that in the dimethyl sulfoxide group ( Figure 5). e results from meta-analyses showed that the effective rate in the acupuncture plus Chinese medicine or acupuncture plus RICE group was not different statistically from that in Chinese medicine (P � 0.14) or RICE group (P � 0.64). Moreover, Du et al. [11] reported that the effective rate in the acupuncture plus massage group was similar (P � 1.00) with that in the massage group.

Cure Rate.
irteen studies reported the cure rate. It is defined as a ratio of the number of patients labelled as cure divided by the number of patients in a certain group. Figure 6 shows that the cure rate in the acupuncture group is statistically higher than that in the no treatment (P � 0.08), infrared radiation (P � 0.01), or RICE (P � 0.004) group. However, the cure rate in the acupuncture group was not statistically different from that in the massage (P � 0.52), or "Ice and hot pack" plus Chinese medicine (P � 0.10) group.
A meta-analysis showed that acupuncture plus Chinese medicine could significantly increase the cure rate (P � 0.0002) compared with Chinese medicine alone (Figure 7). Other meta-analysis showed that acupuncture plus RICE could significantly increase the cure rate (P � 0.01) compared with RICE alone (Figure 8). Du et al. [11] found that the cure rate in the acupuncture plus massage group was higher (n � 40, RR � 1.55, 95% CI: 1.00 to 2.39, P � 0.05) than that in the massage group.

Adverse Events.
ree included studies reported the information on adverse events. Yu [24] and Wu [28] reported no adverse reactions. Another study [23] reported that a participant suffered from mild drug-related allergic reaction and was healed without treatment.

Assessment of Publication Bias.
No funnel plots were provided to assess publication bias because no meta-analyses involving more than ten studies were performed.

Main Findings and Interpretation.
e present study critically assessed the efficacy and safety of acupuncture for AAS. Overall, risk of bias assessment was limited because of incomplete reporting on risk of bias items. In view of the heterogeneity of interventions, main findings were interpreted based on comparisons between experimental and control groups.

Evidence-Based Complementary and Alternative Medicine
Rest, ice, compression, and elevation (RICE) are generally used to treat acute ankle sprain in clinical practice [32]. However, a systematic review found insufficient evidence on RICE for AAS in adults [33]. RICE was also not recommended for the management of lateral ankle sprain according to a recent clinical guideline [4]. e present study found that acupuncture alone or in combination with RICE could significantly relieve pain and increase cure rate in patients with AAS compared with RICE alone. It suggests that acupuncture may be considered as a complementary and alternative therapy for treating AAS. However, more large-scale RCTs with objective outcomes are warranted to confirm these findings.
Massage belongs to nonpharmacological therapies and usually is used for the management of musculoskeletal disorders. A study found that massage might improve the flexibility and balance function of the ankle joint [34]. An evidence map showed that massage was used for treating a     Evidence-Based Complementary and Alternative Medicine variety of pain conditions [35]. e present study found that acupuncture combined with massage could significantly relieve pain in patients with AAS compared with massage alone. However, the underlying mechanism of the combination for AAS is still poorly investigated. Patients with AAS often experience acute pain and swelling associated with inflammatory reactions [36]. Previous studies showed that some Chinese medicines had antiinflammatory properties and were used for treating musculoskeletal disorders [37][38][39]. e present study found that acupuncture plus Chinese medicine might be more effective for relieving pain, reducing the duration of pain, and increasing cure rate than Chinese medicine alone in patients with AAS. Nonsteroidal anti-inflammatory drugs were recommended for reducing pain and controlling swelling in patients with acute lateral ankle sprain [4]. However, the efficacy of acupuncture versus nonsteroidal anti-inflammatory drugs for AAS was not assessed because no eligible trials were identified. e present study provided some insights into the management of AAS. Combined-modality therapy may provide additional benefits in patients with AAS. ese findings may be useful for updating clinical practice guidelines. Moreover, a review reported that acupuncture could relieve pain by activating acupoints and transmitting signals to the spinal cord and brain associated with the regulation of inflammatory factors [40]. An experiment showed that acupuncture might relieve neuropathic pain in rats by inhibiting microglial activation and inflammatory responses [41]. Chen et al. [42] found that acupuncture might relieve inflammatory symptoms at the ankle joint by reducing serum TNF-α and anti-cyclic citrullinated peptide antibody levels. e evidence may partly explain why acupuncture may be effective for AAS.

Limitations.
is systematic review had several limitations. Firstly, effect size might be overestimated because of         Evidence-Based Complementary and Alternative Medicine small sample size. Secondly, performing the meta-analysis was limited because of the heterogeneity of interventions. irdly, the results should be interpreted cautiously because of methodological flaws and rarely reported objective outcomes in included studies.

Conclusions
e findings of the present study suggest that acupuncture may be beneficial for AAS. However, more large-scale and well-designed RCTs are warranted.

Data Availability
All datasets presented in this study are included in the article or supplementary material.

Disclosure
AFL and SWG are considered as co-first authors. AFL and JBZ are considered as corresponding authors.