Clinical Effects and Safety of Auricular Acupressure as an Adjunct Therapy on Postoperative Pain among Patients with Hip Fracture: A Meta-Analysis

Objectives To evaluate the short-term outcome of treatment by auricular acupressure (AA) on postoperative pain among hip fracture (HF) patients. Methods A systematic search for randomized controlled trials on this topic was conducted through May 2022 by searching multiple English and Chinese databases. The methodological quality of the included trails was assessed by the Cochrane Handbook tool, and relevant data were extracted and statistically analyzed by RevMan 5.4.1 software. The quality of the evidence supporting each outcome was evaluated by GRADEpro GDT. Results Fourteen trials with a total of 1390 participants were included in this study. Compared with conventional treatment (CT) alone, the combination of AA and CT had a significantly greater effect on the visual analog scale at 12 h (MD −0.53, 95% CI −0.77 to −0.30), 24 h (MD −0.59, 95% CI −0.92 to −0.25), 36 h (MD −0.07, 95% CI −0.13 to −0.02), 48 h (MD −0.52, 95% CI −0.97 to −0.08), and 72 h (MD −0.72, 95% CI −1.02 to −0.42), amount of analgesics used (MD −12.35, 95% CI −14.21 to −10.48), Harris Hip Score (MD 6.58, 95% CI 3.60 to 9.56), effective rate (OR 6.37, 95% CI 2.68 to 15.15), and adverse events (OR 0.35, 95% CI 0.17 to 0.71). Conclusions Compared with CT alone, the combination of AA and CRT had a significantly greater effect on postoperative pain in HF patients. However, trails with a rigorous methodology, including standard protocols for AA and multiethnic subjects, are still needed.


Introduction
As the aging process of the population continues to accelerate, the proportion of the elderly (>60 years) will continue to increase [1]. It is estimated that by 2050, the proportion of the elderly population will reach 21.1% worldwide [2,3]. Hip fracture is a common type of fracture in the elderly and ranks among the top 10 of disability [4]. It is estimated that the absolute number of hip fractures is expected to increase from 1.6 million in 2000 to 6.3 million by the year 2050 [5]. Hip fracture (HF) has become a worldwide health problem, it is estimated that the annual cost of HF treatment has increased from approximately 10.3 to 15.2 billion dollars in 1990 to 17 billion in 2002 [6].
Timely surgery for hip fractures remains the mainstay of treatment, including internal fxation, total hip arthroplasty, and hemiarthroplasty [7]. Many ofcial clinical societies recommend postoperative multi-modal analgesia [8] because elderly patients with inadequate postoperative pain control are reluctant to mobilize, thus increasing the potential risk of complications and slowing recovery [9]. Adequate analgesia is of great signifcance. Current strategies for pain management include oral and parenteral systemic analgesia, and systematic administration of opioids remains the most commonly used analgesia protocol [10]. While opioids are efective in relieving static pain, they may not be sufcient for dynamic pain [11]. Furthermore, the use of opioids may bring side efects, such as delirium, drowsiness, and even respiratory depression, which may afect the prognosis of patients [12]. Tus, to lower the risk of adverse events and also guarantee treatment efcacy, complementary and alternative therapies have been investigated and compared.
Acupuncture is a traditional nonpharmacological treatment in China and has been widely recognized worldwide [13]. Available evidence suggests that acupuncture is efective for pain relief, thus the World Health Organization recommends the use of acupuncture for a variety of pains, including postoperative pain [13]. As an important component of acupuncture [14], auricular acupressure (AA) has also been deemed efective for pain management by the National Institutes of Health [15]. A literature search yielded many published clinical randomized controlled trials (RCTs) of AA for postoperative pain among HF patients. Consequently, the aim of this study was to evaluate the short-term outcome of treatment by AA on postoperative pain among HF patients by conducting a systematic literature review and meta-analysis of RCTs.

Methods
Te protocol of this study was registered in PROSPERO (https://www.crd.york.ac.uk/prospero). Tis work was performed following criteria in the Cochrane Handbook [16] and reported in line with preferred reporting items for systematic reviews and meta-analyses (PRISMA) [

Criteria for Considering
Studies. Included criteria were as follows: (i) type of study: published randomized controlled trial (RCT) in English and Chinese; (ii) intervention: AA with conventional treatment (CT); (iii) comparison: CT; (iv) population: diagnosed as having a hip fracture confrmed by imaging, regardless of race, sex, or age; and (v) outcome: pain intensity (visual analog scale (VAS)) at 12, 24, 36, 48, and 72 h after surgery the amount of analgesics used, the Harris Hip Score (HPS), the efective rate (ER), and adverse events (AE). In general, ER is defned by the formula: ER � ("total number of patients" − "number of patients without response")/total number of patients; "no response" is defned as no signifcant change in VAS score after treatment. Trails of AA with more than one Traditional Chinese medicine treatment technique as an intervention were also excluded.

Study Identifcation.
Search results were imported into Endnote and duplicates were removed. Two research studies independently screen the titles and abstracts of the retrieved articles, evaluate the potential full texts, and determine the eligibility of the reviews. Any discrepancies were solved by introducing a third researcher for judgment.
Data were extracted by two independent research studies using a predefned form, including: frst author, year, country, simple size, characteristics of patients, course of disease, treatment protocol, outcome indicators, and consequences of and outcomes. Resolve any discrepancies through consultation.

Quality Assessment.
Two research studies independently assessed the risk of bias by using the Cochrane Collaboration tool [18]. Each trail could be judged to be at "low," "high," or "unclear" risk of bias according to the domains of random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete dataset, selective reporting, and other bias. If more than half of the domains were assessed as having a low risk of bias, the trial was assessed as having a low overall risk of bias; if more than half of the domains were assessed as having a high risk of bias or an unclear risk of bias, the trial was assessed as having a high overall risk of bias. Resolve any discrepancies through consultation.

Data Synthesis.
Te odds ratio (OR) with 95% of CIs for dichotomous outcomes and the mean diference (MD) for continuous variables. Te heterogeneity of the studies was assessed using the I 2 . If I 2 < 50%, there was no signifcant heterogeneity among studies, and a fxed efect model was used to analyze the data. If I 2 ≥ 50% of the heterogeneity among the studies was signifcant and a random efect model was used to provide the evaluations of the intervention. Subgroup analyses were determined by whether the obtained data were sufcient. A funnel plot was used to assess publication bias.

Level of Evidence.
Two research studies independently assessed the certainty of evidence for each outcome using a grading of recommendations assessment, development, and evaluation (GRADE) system [19]. Te GRADE guideline consists of seven domains, namely, risk of bias, inconsistency, indirectness, imprecision, and publication bias. Te certainty of the evidence was rated as high, moderate, low, and very low. Te summary of fndings table was created by the GRADEpro GDT (https://gradepro.org/). Resolve any discrepancies through consultation.

Literature Screening.
Literature screening fowchart is shown in Figure 1. A total of 460 trails were identifed. After removing duplications, the titles and abstracts of 267 trails were further evaluated. Next, the full texts of the remaining 21 records were assessed, and 14 trails were fnalized for inclusion in our meta-analysis [20][21][22][23][24][25][26][27][28][29][30][31][32][33]. Additional records identified through other sources (n = 0)  2022. All of them were conducted in China and reported nonsignifcant diferences in their patient baseline characteristics. Te main assessment tools were the VAS and the amount of analgesics used, HPS: Harris Hip Score, ER, and AE. Figures 2 and 3, the risk of bias in the included studies was mainly derived from random sequence generation, allocation concealment, blinding of participants and personnel, and blinding of outcome assessment. Finally, 8 of the 14 RCTs were assessed as having a low overall risk of bias, and the rest had a high overall risk of bias.

Visual Analog Scale
(1) Visual Analog Scale at 12 h. VAS was used to score the pain degree of patients, which the score was in direct proportion to the pain degree. Eight trials with a total of 886 participants recorded VAS at 12 h of the intervention. Te random efect model was applied, and pool results showed that AA could reduce VAS signifcantly in the experimental group than control group (MD −0.53, 95% CI −0.77 to −0.30) as shown in Figure 4. In addition, the data provided by the included studies were insufcient to support subgroup analyses, and therefore no further subgroup analyses were performed in subsequent analyses. Te funnel plot was given in supplementary fle B.

Harris Hip Score.
Five trials with a total of 688 participants recorded HHS at the end of the intervention. Te random efect model was applied and the pool results showed that AA could improve HHS signifcantly in the experimental group than control group (MD 6.58, 95% CI 3.60 to 9.56), as shown in Figure 9. Te funnel plot was given in supplementary fle B.

Efective
Rate. Two trials, with a total of 176 participants, recorded an efective rate at the end of the intervention. Te fxed efect model was applied, and pool results showed that AA could increase the efective rate signifcantly in the experimental group than control group (OR 6.37, 95% CI 2.68 to 15.15), as shown in Figure 11. Te funnel plot was given in supplementary fle B.   Finally, one of the outcomes was assessed as low moderate quality, and the rest were of low or very low quality. Funnel graphs for these outcomes are given in Appendix B. Details are outlined in Table 2.

Discussion
In recent years, there have been more publications on AA for postoperative pain among patients with HF. However, interpretation of this evidence from these studies is difcult. We conducted this study to systematically evaluate the clinical efect of AA on postoperative pain in HF patients.

Summary of Main Results.
In this study, 14 trails involving 1390 patients were included for meta-analysis. First, the pooled results suggested that AA combined with CT was signifcantly superior to CT alone in terms of VAS at 12, 24, 48, 36 h, and 72 h, HHS, amount of analgesics used, ER, and AE. Tese results indicated that AA can help reduce postoperative pain degree, reduce the amount of analgesics, improve hip function, and reduce the incidence of AE among HF patients. Second, it should be emphasized that methodological faws are prevalent in existing RCTs. Most RCTs did not report proper allocation concealment or blinding procedures for outcome assessments. Furthermore, no study was blinded to participants and personnel, suggesting potential performance bias. Of the 7 bias items, only one study met the requirement for low risk of bias. Tird, the evidence quality was generally evaluated as "moderate" "low" or "very low" by the GRADE system. No study was rated as having a high level of evidence. Although AA has been widely used in China, and the pooled analysis of this study has yielded that AA combined with CT was likely to have potential therapeutic benefts in postoperative pain among HF patients, but the level of evidence was not high. Tere may still be gaps between the evidence supporting the efcacy of AA and its clinical implementation. Further trials with rigorous methodology, including standard protocols for AA and multiethnic subjects, are still warranted to provide stronger evidence. Fourth, limited by insufcient data, this study was unable to evaluate the long-term efects of AA. Te included studies only evaluated the use of AA for several days and assessed outcomes before and immediately after treatment, therefore, the long-term efects of AA on postoperative pain could not be revealed. -20 -10 0 10 20 Figure 10: Comparison of the amount of analgesics used between the AA group and the control group.  Figure 12: Comparison of the adverse events between the AA group and the control group.

Agreements and Disagreements with Other Studies or
Reviews. Tis review agrees with the results of the other study [34] in the aspect that AA as a complement to conventional drugs reduces postoperative pain, though with uncertainty. A systematic review [34] reported that in postoperative patients with fractures treated with AA, the degree of pain signifcantly improved 24 hours after surgery and was also signifcantly lower than in the control group in terms of ER. Tese results agree with our study. However, this review [34] focused only on pain at one time point and did not address the question of duration of action, and this study included patients with all types of fractures. Te diference in our review is that we focused on patients with HF and observed pain at multiple time points after surgery and also on HHS, amount of analgesics used, and AE, making the assessment more comprehensive. In addition, as in the previous review, this study was also limited by the high risk of bias in the included trials, so the level of evidence obtained was not high. More welldesigned, rigorous, and large trials are needed in this feld.

Implications for
Practice. In addition to shorter hospital stays and reduced morbidity and mortality, efective relief of acute postoperative pain is associated with increased patient satisfaction [33]. AA is an ancient Chinese nonpharmacological treatment that has been reported to be efective and safe in improving multiple factors in fracture patients and has the potential to promote postoperative recovery in combination with CT. Although the current review did not provide the best evidence, the results suggested that the combination of AA with CT postoperative pain degree, reduce the amount of analgesics, improve hip function, and reduce the incidence of AE among HF patients. AA has been used to treat various types of pain, including postoperative, musculoskeletal pain, and pain associated with anesthesia [35]. Te analgesic efect of AA has been preliminarily revealed. Research studies have found that AA stimulation can activate the descending pain inhibitory pathway in the brainstem-spinal cord and inhibit the ascending pain pathway, which in turn exerts analgesic efects [36]. It has been found that acupoint stimulation of one or both ears can increase the pain threshold [37], and this efect peaks 5-10 minutes after stimulation and lasts for hours to days [38]. Furthermore, it has been suggested that the analgesic efect of AA is also associated with the endogenous opioid system [39].

Limitations.
Considering that all included trails were conducted in China, it should be noted that publication bias was also observed in this meta-analysis, which indicates that the results of this review may be challenging to generalize, especially in countries other than China. As a country that has practiced AA for a long time, Chinese attitudes towards AA may be more favorable than those of other ethnic groups, which may contribute to the placebo efect. Terefore, further studies in countries other than China are still needed. Furthermore, the diversity of the AA protocol used in the included trails may contribute to the heterogeneity of the fndings [40]. While complementary and alternative therapies, such as AA, may emphasize tailoring treatments to individual patient characteristics, developing basic treatment standards that allow for some modifcations can improve the quality of clinical evidence in this feld.

Conclusion
Compared with CT alone, the combination of AA and CRT had a signifcantly greater efect on postoperative pain in HF patients. However, trails with a rigorous methodology, including standard protocols for AA and multiethnic subjects, are still needed.