Comparison of Efficacy of Acupuncture-Related Therapy in the Treatment of Postherpetic Neuralgia: A Network Meta-Analysis of Randomized Controlled Trials

Background Postherpetic neuralgia (PHN) is the most common sequela of herpes zoster, and the efficacy of the treatment regimens recommended in the guidelines is not entirely reliable. Acupuncture and moxibustion are widely used complementary alternative therapies that have a positive effect on the treatment of PHN. However, there are various forms of acupuncture and moxibustion, and there are differences in efficacy between the different forms. Methods The retrieval work of randomised controlled trials (RCTs) of acupuncture for PHN in English databases (including PubMed, Cochrane Library, Embase, Web of Science) and Chinese databases (including China National Knowledge Infrastructure (CNKI), WeiPu database, WanFang database, and China Biomedical Literature Database) were conducted from the time of database creation to June 2022. Literature screening, data extraction, and evaluation of risk of bias for the included studies were carried out independently by two researchers, and data analysis was performed using Stata 14.2 software. Results A total of 30 RCTs including 2138 patients with PHN were included. In terms of pain improvement, acupoint embedding + Western medicine group, bloodletting-cupping group, and bloodletting-cupping + Western medicine group ranked top. In terms of total efficiency, acupuncture + Western medicine group, bloodletting-cupping + Western medicine group, and acupoint embedding group ranked top. There were no statistically significant differences in the incidence of adverse events between treatment regimens. Conclusions In a comprehensive comparison of the outcome indicators of 14 different treatment regimens, we considered acupoint injection + Western medicine and bloodletting-cupping + Western medicine to be the best combinations for the treatment of PHN. Due to the limitations of the study, the above conclusions still need to be validated in further multi-centre, large-sample prospective randomised controlled clinical trials.


Introduction
PHN is a neuropathic pain that persists for 1 month or more after the healing of the herpes zoster (HZ) rash and is the most common complication of herpes zoster [1]. e nature of pain in PHN is varied and can be burning, electric shock, knife-like or hand-needle pain, intermittent or continuous [2]. Approximately 18∼41% of patients with HZ experience PHZ, a painful condition that can last for months or even years [3], during which physical, emotional, and social functioning are affected [4]. PHZ also adds to the medical burden at the individual and societal level [1].
Current treatment for PHN includes both pharmacological and interventional treatments, with pharmacological treatment being the most basic and commonly used method [5]. e drugs currently recommended for first-line treatment of PHN are calcium channel modulators (e. g. pregabalin, gabapentin, tricyclic antidepressants (TCAs) and lidocaine patch 5%) [6,7]. However, these drugs are not suitable for long-term use and their efficacy is not entirely reliable [8]. Other treatments include opioid analgesics, tramadol, topical capsaicin and botulinum toxin type A, but the long-term efficacy and safety of these drugs is uncertain [1,9]. erefore, how to optimise PHN treatment strategies is a matter of concern to clinicians.
Acupuncture-related therapy, based on meridian theory, is a traditional treatment modality of Chinese origin that plays an important role in the field of complementary and alternative therapies. A 2010 survey of global research trends in acupuncture [10] showed that pain management with acupuncture has been the most prevalent area of research and that acupuncture-related therapies have been widely used in the treatment of joint, muscle, and nerve-related pain and are safe and effective [11][12][13][14]. A large number of clinical studies on acupuncture-related therapies for the treatment of PHN have been conducted in China and abroad, and there are various forms of acupuncture, including acupuncture, moxibustion, electroacupuncture, fire acupuncture, and blood-letting [8]. Recent systematic evaluations have confirmed that acupuncture-related therapies (including fire acupuncture, acupuncture, electroacupuncture, and moxibustion) are effective for PHN and that they can reduce pain intensity, relieve anxiety, and improve quality of life in patients with PHN [15][16][17]. As these studies have mainly focused on the comparison of acupuncture therapies with traditional drug therapies, there is a lack of comparison of efficacy between different acupuncture therapies, and therefore, there is still controversy in clinical practice as to which acupuncture therapy is the best choice. Network meta-analysis (NMA) is a further development of traditional pairwise meta-analysis [18]. Based on current clinical research data, the NMA can complete both direct and indirect comparisons between different acupuncture therapies and further synthesise the results of direct and indirect comparisons to produce a ranking of the efficacy of different acupuncture treatments [11]. is study therefore used Network Meta-analysis to compare the efficacy of different acupuncture and moxibustion therapies in the treatment of PHN and to provide evidence for choosing the optimal combination for the clinical treatment of PHN.

Study Registration.
is network meta-analysis was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses for NMA guidelines [19].

Type of Study.
is study included parallel randomised controlled trials (RCTs) published in Chinese or English, which were subject to the limitations of the intervention. ere was no requirement for the study to be blinded.

Study Population.
Patients included in the study met the American Academy of Family Physicians' diagnostic criteria for PHN [2], or other accepted diagnostic guidelines for pain lasting 30 days to more than 6 months after healing of the HZ lesions. ere were no restrictions on the gender or age of the patients.

Interventions.
e treatment group was treated with different acupuncture therapies alone or in combination with Western medicine, which were defined as acupoint stimulation techniques guided by the meridian theory of Chinese medicine, including any of the following therapies: conventional acupuncture, warm acupuncture, electroacupuncture, fire acupuncture, bloodletting-cupping, moxibustion, acupoint burial, acupoint injection; the control group was Western medicine, which had to be first-line drugs recommended by guidelines for the treatment of PHN [6], including calcium ion channel modulators (e. g. pregabalin, gabapentin, tricyclic antidepressants (TCAs) and lidocaine patch 5%).

Outcome
Indicators. Primary outcome indicators: the primary objective of this study was to assess pain control, and pain measures included the Visual Analogue Scale (VAS) and Numerical Rating Scale.
Secondary outcome indicators. ① e efficiency rate, which was based on the evaluation of the overall efficacy and the evaluation criteria based on the Criteria of Diagnosis and erapeutic Effects of Diseases and Syndromes in Traditional Chinese Medicine [20] and the Guiding Principles for Clinical Research of New Chinese Medicines [21]; ② the occurrence of adverse effects.
Exclusion criteria. ① e presence of other painful conditions in the study population; ② lack of primary outcome indicators; ③ treatment protocols that included a combination of two or more acupuncture therapies, such as acupuncture combined with moxibustion, electroacupuncture combined with moxibustion, etc.; ④ duplicate published studies; and ⑤ studies for which complete data were not available in the article and relevant data were still not available after contacting the authors. neuralgia). English search terms include "acupuncture," "electroacupuncture," "warm needle," "fire needle," "bloodletting," "moxibustion," "acupoint embedding," "acupoint injection," "Postherpetic neuralgia," "PHN." e PubMed database search strategy is shown in Table 1.

Literature Selection.
e retrieved literature was imported into EndNote X9 software and checked for duplicates. After excluding duplicates, an initial screening was carried out by reading the abstracts to exclude irrelevant studies, and the remaining studies were further assessed by reading the full text, and finally data extraction was completed for those studies that met the inclusion criteria. All of the work above was done by two independent researchers, and where there was disagreement between the two researchers, a third researcher assisted in the judgement. Data extraction included title, author, date of publication, sample size, details of the intervention, details of the control measure, duration of treatment, and outcome indicators. For multiarm studies reporting different types of acupuncture interventions, data were extracted from all relevant arms. For pain scores, the mean and standard deviation of baseline and posttreatment change scores (defined as the baseline score minus the posttreatment score) were taken.

Risk of Bias Evaluation of Included Studies.
Two independent researchers evaluated the included studies backto-back by means of the Cochrane Systematic Evaluation Manual version 5.1.0 RCT Risk of Bias Assessment Tool [22]. e elements of the evaluation included random sequence generation, outcome allocation concealment, blinding of participants and personnel, blinding of assessment, incomplete outcome data, selective reporting bias and other bias. e final decision was "high risk," "low risk," or "unclear".
2.6. Statistical Analysis 2.6.1. Direct Comparison. Direct comparative meta-analysis was performed using Stata 14.2 software. For continuous variables, analysis was carried out using the standard mean difference (SMD), and for dichotomous variables, relative risk (RR) was used. Heterogeneity between the results of the included studies was analysed using the χ 2 test, while the magnitude of heterogeneity was determined quantitatively in conjunction with I 2 . If P ≥ 0.10, I 2 < 50%, there was no significant heterogeneity between studies and a fixed-effects model was used for meta-analysis; if P < 0.10, I 2 ≥ 50%, heterogeneity between studies was considered significant and a random-effects model was used for meta-analysis.

Reticulated Meta-Analysis.
e Stata 14.2 software was used to plot the net relationship between direct and indirect comparisons between the outcomes of the different treatment measures. For studies with three or more arms, they were first divided into pairwise contrast combinations.
e included studies were also tested for consistency and non-consistency, and local inconsistency was tested using the nodal split method; P > 0.05 was considered good for consistency. For each outcome indicator, the efficacy of the interventions was ranked using the surface under the cumulative ranking (SUCRA) values. e publication bias of included studies was tested by comparison-corrected funnel plots.

Literature Search Results.
A total of 5090 relevant literature was retrieved, and after initial screening and rescreening, 30 RCTs were finally included, comprising a total of 2138 patients. e literature screening process is shown in Figure 1.

Results of the Risk of Bias Evaluation of the Included
Studies. ① Random sequence generation: 17 studies [23, 25, 28, 29, 31, 32, 35, 36, 39-41, 44, 45, 48, 49, 51, 52] used random number tables, three studies [30,34,46] used computer-generated random numbers, two studies [37,42] used consultation sequences and the remaining eight studies [24,26,27,33,38,43,47,50] referred to "random" only. ② Allocation concealment: 4 studies [23,30,48,52] used sealed opaque envelopes; 2 studies [37,42] used order of attendance and the remaining 24 studies did not mention allocation concealment. ③ Blinding of patients, trialists: none of the studies were double-blinded due to the limitations of the intervention modality. ④Blinding of outcome assessors: none of the studies mentioned blinding of outcome assessors; ⑤Incomplete outcome data, selective reporting, other bias: all studies had complete outcome data, with no selective reporting or other bias. e results of the risk of bias evaluation are shown in Figure 2.   In a direct comparison regarding pain scores, meta-analysis showed that the pain score of warm acupuncture, electroacupuncture, electroacupuncture + Western medicine, acupuncture + Western medicine, bloodletting-cupping, bloodletting-cupping + Western medi cine, fire acupuncture, and acupoint embedding groups were superior to that of the Western medicine group (P < 0.05), while the scores in acupuncture and acupuncture point injection + Western medicine groups did not differ from the Western medicine group (P > 0.05). Descriptive analysis showed that both the acupoint injection and acupoint embedding groups were superior to the Western medicine group (P < 0.05). e acupuncture group was superior to the electroacupuncture group, the acupuncture + Western medicine group was superior to the acupuncture group (P < 0.05), and there was no difference between the fire acupuncture + Western medicine group and the Western medicine group, the fire acupuncture group and the bloodletting-cupping group compared to the acupuncture group (P > 0.05), see Table S2 in the supplementary material.

Total Efficiency.
In a direct comparison regarding pain scores, meta-analysis showed that the acupuncture, acupuncture + Western medicine, fire acupuncture, bloodletting-cupping, acupoint embedding groups were all superior to the Western medicine group (P < 0.05), with no difference in the electroacupuncture group compared to the Western medicine group. Descriptive analyses of acupoint embedding + Western medicine compared with Western medicine, fire acupuncture, bloodletting-cupping, and electroacupuncture compared with acupuncture showed no difference (P > 0.05), see Table S2 in the supplementary material.

Adverse Reactions.
In a direct comparison regarding adverse reaction rates involving 10 interventions (warm acupuncture, acupuncture, acupuncture + Western medicine, bloodletting-cupping, bloodletting-cupping + Western medicine, electroacupuncture, electroacupuncture + Western medicine, acupoint injection, acupoint injection + Western medicine, and acupoint embedding) versus Western medicine, none of their adverse reaction rates differed from those of Western medicine (P > 0.05), see Table S2 in the supplementary material.

Heterogeneity Analysis.
In the meta-analysis of direct comparisons, there was heterogeneity in some of the results. Analysis of the raw data revealed possible methodological heterogeneity due to the inclusion of studies with less description of blinding and allocation concealment, as well as possible clinical heterogeneity due to factors such as inclusion of populations, acupuncture points, and manipulation methods, but as the original studies did not specify these details and the small number of studies for some of the outcomes, further subgroup analysis could not be performed to explore sources of heterogeneity. However, we found by sensitivity analysis that the results were stable after excluding either study; see sensitivity analysis figures in the supplementary material ( Figure S1∼S19). We could therefore ignore this heterogeneity and used a random-effects model for Meta-analysis.

Results of Reticulated Meta-Analysis of Pain Scores.
irty studies reported pain scores , with inconsistency tests P � 0.992 > 0.05, and local inconsistency tests using the nodal split method with P > 0.05, showing good agreement between studies and reticulated metaanalysis under the consistency model. e results showed that acupoint injection + Western medicine was superior to acupuncture, acupuncture + Western medicine, electroacupuncture, bloodletting-cupping and Western medicine, and bloodletting-cupping + Western medicine was superior to Western medicine. e rest of the comparisons between the different treatments were not statistically different, see Table 4.

Results of the Reticulated Meta-Analysis of Adverse
Reactions. Adverse effects were reported in 14 studies [25, 26, 31, 33, 36, 39, 43-45, 47-49, 51, 52]. As they are all indirect comparisons and do not form a closed loop, no consistency test was required. A mesh meta-analysis was performed under the consistency model. e results showed that acupoint injection + Western medicine was superior to acupuncture, acupuncture + Western medicine, bloodletting -cupping, and Western medicine, and acupoint embedding was superior to Western medicine and bloodlettingcupping. e rest of the comparisons between the different treatments were not statistically different, see Table 6. e results of the ranking of the incidence of adverse     Figure 8 for the SUCRA ranking chart.

Small Sample Effect Estimation.
A comparativecorrected funnel plot of pain scores for the main outcome indicators was assessed by Stata 14.2 software, see Figure 9.
e results showed that the funnel plot was not fully symmetrical, suggesting that there may be some publication bias or small sample effect in the study network.

Discussion
Patients with PHN are often accompanied by prolonged and persistent pain, which triggers other symptoms such as anxiety, depression, and sleep disturbances [53]. Although a variety of treatment options have been proposed, first-line treatment options are still predominant in clinical practice [54], but their efficacy is not entirely reliable [8]. Acupuncture therapy based on meridian theory is widely regarded as potentially beneficial and safe for the treatment of neuropathic pain [55,56]. e efficacy and safety of acupuncture alone or in combination with Western medicine in the treatment of PHN has been clinically proven, and the selection of the optimal combination has become the focus of current research.
is study evaluated the effects of acupuncture-related therapies alone or in combination with Western medicine on pain scores, overall effectiveness, and adverse reaction rates in patients with PHN. e results of the study showed that acupoint injection + Western medicine was superior to acupuncture, acupuncture + Western medicine, electroacupuncture, bloodletting-cupping, and Western medicine, and bloodletting-cupping + Western medicine was superior to Western medicine in terms of improving pain scores. e results of the probability ranking showed that acupoint injection + Western medicine > bloodletting-cupping >bloodletting-cupping + Western medicine > warm acupuncture > acupuncture + Western medicine > fire acupun cture > electroacupuncture + Western medicine > acupuncture > electroacupuncture > acupoint embedding + Western medicine > fire acupuncture + Western medicine > acupoint embedding > Western medicine. In terms of total efficiency, acupuncture + Western medicine was superior to bloodletting-cupping, bloodletting-cupping + Western medicine and Western medicine, acupuncture was superior to Western medicine and fire acupuncture was superior to Western medicine. e results of the probability ranking showed that acupuncture + Western medicine > bloodletting-cupping + Western medicine > acupoint embedding > fire acupu ncture > acupuncture > bloodletting-cupping > electroacupu ncture > acupoint embedding + Western medicine > Western medicine. In terms of adverse reaction rates, acupoint injection + Western medicine was superior to acupuncture, acupuncture + Western medicine, bloodletting-cupping and Western medicine, and acupoint embedding was superior to Western medicine and bloodletting-cupping. e results of the probability ranking showed that electroacupuncture > acupoint injection + Western medicine > acupoint injection > acupoint embedding > warm acupuncture > acupuncture > bloodle tting-cupping + Western medicine > electroacupuncture + Western medicine > Western medicine > acupun cture + Western medicine > bloodletting-cupping. e analysis of the above indicators showed that Western medicine ranked low in terms of improvement in pain, overall effectiveness, and adverse reaction rate, which could be seen as a complementary and alternative option to firstline treatment options. Although the ranking of efficacy varied between indicators, the rankings of acupoint injection + Western medicine, bloodletting-cupping, bloodl etting-cupping + Western medicine, and acupuncture + Western medicine were ranked at the top, and the combined results of the pain score and the ranking of the total effective rate showed that acupoint injection + Western medicine and bloodletting-cupping + Western medicine were outstanding in the treatment of PHN. Combining the results of the direct meta-analysis followed by the reticulated meta-analysis, we found no significant differences in adverse reaction rates between the treatment regimens. Considering the moderate quality of the included studies, the selection   Cumulative Probabilities Rank Graphs by Treatment Figure 6: SUCRA ranking of pain scores for different interventions. Notes. In the figure, RANK is the horizontal coordinate, indicating the possible ranks, cumulative probabilities is the vertical coordinate, indicating the probability of being at this rank, and the area under the curve is used to represent the overall probability of the treatment measure.  Cumulative Probabilities Rank Graphs by Treatment Figure 7: SUCRA ranking of the total effectiveness of different interventions. Notes. In the figure, RANK is the horizontal coordinate, indicating the possible ranks; cumulative probabilities is the vertical coordinate, indicating the probability of being at this rank, and the area under the curve is used to represent the overall probability of the treatment measure.  (iv) ere was some publication bias and a small sample effect in the included studies, which affects the reliability of the findings.

Conclusion
In summary, after a comprehensive comparison of the outcome indicators of 14 different treatment regimens, acupoint injection + Western medicine and bloodlettingcupping + Western medicine were considered to be the best combination regimens for the treatment of PHN. e appropriate treatment modality should be selected in clinical practice in the context of the actual situation. Due to the limitations of the study, the above conclusions still need to be validated by further multicentre, large-sample prospective randomised controlled clinical trials.

PHN:
Postherpetic neuralgia RCTs: Randomised controlled trials HZ: Herpes zoster TCAs: Tricyclic antidepressants NMA: Network meta-analysis VAS: Visual analogue scale CNKI: China knowledge network SMD: Standard mean difference RR: Relative risk SUCRA: Surface under the cumulative ranking.

Data Availability
e data used to support the findings of this study are included with in the article and the supplementary information files.

Disclosure
Haiyan Wang and Renhong Wan are the co-first authors.

Conflicts of Interest
e authors declare that there are no conflicts of interest.