Herpes zoster (HZ) is caused by reactivation of the varicella-zoster virus, which lies dormant in the dorsal root ganglion and is reactivated frequently. In a healthy host, memory cells can help to avoid reactivation. However, when host immunity is lowered by aging, stress, and/or the administration of external immunosuppressants, reactivation cannot be suppressed, and the virus replicates and infects the body [1]. With the aging population, the number of patients with HZ is increasing. According to health insurance statistics in Korea, the number of patients treated for HZ steadily increased during 2010–2021, to 725,000 from 48,000 [2]. Also, patients with HZ showed an increase in coronavirus disease 2019 infection during the pandemic period [3].

The symptoms of HZ include a whole-body reaction: headache, fever, etc. The most characteristic symptom is skin lesions along the affected dermatome, accompanied by temporary and persistent pain that result in sleep disorder, anxiety, and lower quality of everyday life [4]. The pain can manifest as a variety of sensations such as burning, poking, and stabbing. The skin rashes become blisters within 1–2 days and last 3–4 days. The blisters turn into pustules within one week and subsequently into ulcerations and crusting [5].

Treatment in Western medicine is mainly antiviral medications and supplemental therapies such as steroids and painkillers. When those approaches are not sufficient to relieve the symptoms, antidepressants, anticonvulsants, and opioids are prescribed [6]. However, these drugs have side effects such as nausea, vomiting, and liver or kidney failure because of restricted drug metabolism [7]. With conventional drug treatment, the incidence rate of post-herpetic neuralgia (PHN) is 6.5%, which is still high [8].

With the increase of patients with HZ and the limitations of conventional treatment, an approach to HZ with Korean medicine is considered necessary [9,10]. Acupuncture in particular is known to be safe and to have fewer side effects [11]. Several studies have revealed benefits of acupuncture in HZ, using various acupuncture methods: acupuncture, fire needling, pharmaco-acupuncture, moxibustion, bloodletting, electro-acupuncture (EA), etc. However, no systematic reviews have been published that include only acupuncture and EA, which doctors of Korean medicine use most frequently in treatment and which are covered by national insurance. In this study, we therefore aimed to systematically review and investigate the efficacy and safety of acupuncture and EA for HZ.

1. Types of studies

This review included only randomized clinical trials. Case reports and protocols were excluded, as were nonrandomized and observational studies. No language restrictions were imposed.

2. Type of participants

The subjects of this study were patients with HZ. Patients with HZ occurring concurrently with other diseases and patients suffering from PHN were excluded.

3. Types of interventions

We included only acupuncture and EA as interventions. To see the effect of acupuncture alone, we excluded acupuncture techniques such as thread-embedding acupuncture, fire acupuncture, acupotomy, acupressure, and pharmaco-acupuncture. Studies of moxibustion, cupping, and their combinations with acupuncture were also excluded. Included studies were those that used acupuncture and EA as the sole treatment or in combination with drugs.

4. Types of outcome measures

The primary outcomes were the effective rate and visual analog scale (VAS) for pain. Secondary outcomes were time to pain relief, pain elimination, incrustation time, decrustation time, and incidence of PHN.

5. Data sources and search methods

The databases searched for relevant studies were PubMed, EMBASE, Cochrane Library, China National Knowledge Infrastructure, Citation Information by NII, Korean Studies Information Service System, Korean Medical Database, ScienceON, and Oriental Medicine Advanced Searching Integrated System up to October 2022. Terms associated with the diagnosis (HZ) and treatment (acupuncture, EA) were used. In addition, the gray literature, reports, and dissertations were searched.

6. Data extraction and quality assessment

Two researchers (SKC and JHM) independently screened titles and abstracts in the search results to eliminate duplicate and unsuitable studies. They then read the full-text articles to determine eligibility for inclusion based on the predefined criteria. Disagreements were resolved through discussion, and the final decision was made by a third reviewer (EJK). Data on authors, publication years, interventions, outcome measures, results of studies, and adverse events (AEs) were then extracted from the included studies. The data were analyzed using the Review Manager software application (version 5.4.1; Copenhagen, Denmark; The Nordic Cochrane Center, The Cochrane Collaboration, 2020).

7. Statistical analysis

Mean differences (MD) and odds ratios (OR) with 95% confidence intervals (CI) were calculated to evaluate the effectiveness of acupuncture for HZ as described by the data. Statistical heterogeneity between studies was assessed using the chi-squared test and I-squared (I²) statistic, as recommended in the Cochrane Handbook [12], with an I² statistic of 0–40% indicating substantial heterogeneity, 30–60% moderate heterogeneity, 50–90% substantial heterogeneity, and 75–100% significant heterogeneity. If the heterogeneity was >50%, a random effects model was applied. If the outcome measurement unit differed between studies, the standardized MD (SMD) was applied [13]. Subgroup analysis was done for the acupuncture types; acupuncture and EA. Sensitivity analysis was performed when heterogeneity was high.

8. Quality of evidence

Two researchers evaluated the risk of bias in three categories using the seven domains of the Cochrane Collaboration “risk of bias” criteria: random sequence generation, allocation concealment, blinding of participants, blinding of outcome assessors, incomplete outcomes data, selective outcome reporting, and other biases [12]. If there was disagreement between the researchers, a third was involved. Funnel plots were also constructed to examine publication bias.

1. Study selection

Of 728 studies identified from the search, 142 duplicates were removed, and 586 studies were screened by title and abstract. Based on the predefined criteria, 280 studies were excluded, and the remaining 306 studies were assessed for eligibility by full-text review. Of the 22 eligible studies included in this review and meta-analysis, 7 were dissertations, and the other 15 were journal articles. Fig. 1 presents the study selection process.

Figure 1. Flowchart of the study selection process. CNKI, China National Knowledge Infrastructure; KISS, Korean Studies Information Service System; KMBASE, Korean Medical Database; OASIS, Oriental Medicine Advanced Searching Integrated System; CiNii, Citation Information by NII.

2. Characteristics of included studies

Table 1 shows the studies’ associated characteristics. All studies were reported in China except for two from Italy and Germany [14,15]. Studies were reported for the years 2004–2020. The total number of patients was 1,963, and their mean ages ranged from 42 to 66 years. Seven studies had more than one experimental group, treated using different methods. Three had two experimental groups eligible for this review [14,16,17], and four had only one eligible experimental group among their interventions, except cupping or various acupuncture treatments in the intervention [18-21]. The number of treatments ranged from 5 to 28. All studies included acupuncture treatment as an intervention, and 12 of the studies used EA [14,17-20,22-28]. Most studies used the Ashi and Jiaji points; frequently used distal acupoints were TE6 and SI3. Antiviral drugs, vitamins, and painkillers were used as Western medicine treatments. In six studies of add-on acupuncture treatment, Western medicine and acupuncture were used together as an intervention in the experimental group [14,16,19-21,29].

Table 1 . Characteristics of included studies.

Study	Control group (M/F)	Control group age (y)	Experimental group (M/F)	Experimental group age (y)	Add-on treatment	Experimental group treatment (points)	Control group treatment	Number of treatments	Outcome measures	AEs (C/E)
Bian 2010 [24]	30 (12/18)	45.70 ± 15	32 (13/19)	45.62 ± 12.52		EA (Ashi, TE6, SI3, EX-B2)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	ER, VAS, RP, EP, CB, In, De, PHN	None
Chen 2012 [25]	31 (17/14)	49.61 ± 16.34	27 (16/11)	42.39 ± 17.06		EA (Ashi, EX-B2, SJ6, SI3)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	ER, CB, In, De	Unknown
Ding 2006 [26]	42 (22/20)	18–70	59 (34/25)	24–69		EA (Ashi)	Acyclovir 0.2 g tid, polymyocytes 2 mg (IV, 2/wk), prednisone 10 mg tid, indomethacin 50 mg bid	14–28	ER	Unknown
Huang 2012 [27]	34 (14/20)	45.97 ± 14.66	35 (14/21)	44.08 ± 16.24		EA (circumference needle, Ashi)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	PHN	Unknown
Li 2006 [33]	148 (77/71)	≥40	43 (24/19)	≥40		Acupuncture (Jiao’s scalp acupuncture)	Vit B12 500 mg, anlagia (unknown), intramuscular inj. (unknown) qd	Unknown	ER	Unknown
Li et al. 2009 [28]	40 (19/21)	44.79 ± 9.43	20 (15/5)	46.36 ± 10.21		EA (Ashi, EX-B2, TE6, SI3)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	ER, VAS, In	Unknown
Li 2012 [30]	30 (14/16)	44.72 ± 12.88	30 (17/13)	45.25 ± 11.48		Acupuncture (Shu Zhua ci method; DU14, DU12, DU8, D66, DU2)	Acyclovir 0.3 g bid, Vit B1 10 mg tid	5	ER, VAS, RP, EP, CB, In, De, PHN	(5/0)
Li et al. 2012 [18]	98 (unknown/unknown)	43.76 ± 15.34	98 (unknown/unknown)	46.51 ± 15.30		EA (surrounding, EX-B2, TE6, SI3)	Valacyclovir 300 mg bid, Vit B1 10 mg tid	10	ER	Unknown
Liu et al. 2013 [19]	98 (unknown/unknown)	Unknown	98 (unknown/unknown)	Unknown		EA (circumference needle, Ashi, EX-B2)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	VAS, RP, EP, PHN	Unknown
Liu 2022 [35]	55 (30/25)	55.7 ± 4.11	55 (28/27)	54.24 ± 4.5	O	Acupuncture (Ashi and EX-B2)	Acyclovir 0.8 g tid, Vit B12 0.5 mg tid	14	ER, VAS, PHN	(3/1)
Mao et al. 2015 [20]	20 (12/8)	22–61	20 (9/11)	23–72	O	EA (Weici, LI11, TE3, GB34, LR3)	Ganciclovir 0.25 g inj., valacyclovir 0.3 g bid, methylcobalamin 0.5 mg tid	5	VAS	Unknown
Shi 2008 [22]	30 (10/20)	42.83 ± 16.80	30 (12/18)	44.91 ± 16.97		EA (Ashi, EX-B2, TE6, SI3)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	EP, CB, In, De, PHN	(7/1)
Song 2009 [23]	30 (15/15)	43.47 ± 13.57	30 (16/14)	42.23 ± 14.98		EA (Ashi, EX-B2, TE6, SI30)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	VAS, RP, EP, CB, In, De, PHN	(4/0)
Su 2017 [16]	32 (16/16)	42.50 ± 12.13	31 (18/13)	41.77 ± 12.08		Acupuncture (Gansu Zheng method, cold-reducing, Ashi, TE6, SI3, LR3, GV14, GB34, SP9, SP10, LR2, EX-B2)	Valacyclovir 0.3 g bid, methylcobalamin 0.5 mg tid	14	VAS, RP, EP, CB, In, De, PHN	(6/1)
	32 (16/16)	42.50 ± 12.13	30 (17/13)	42.97 ± 10.52		Acupuncture (Gansu Zheng method, twisting acupuncture, Ashi, TE6, SI3, LR3, GV14, GB34, SP9, SP10, LR2, EX-B2)	Valacyclovir 0.3 g bid, methylcobalamin 0.5 mg tid	14	VAS, RP, EP, CB, In, De, PHN	(6/1)
Sun et al. 2015 [17]	30 (15/15)	43 ± 12	30 (14/16)	43 ± 12		EA (Xie method, Ashi [in depth, 33 mm], EX-B2, BL60, GB43, LR2)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	ER, VAS, CB, In, De, PHN	None
	30 (15/15)	43 ± 12	30 (13/17)	42 ± 10		EA (Xie method, Ashi [16 mm], EX-B2, BL60, GB43, LR2)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid	10	ER, VAS, CB, In, De, PHN	None
Thoma 2012 [14]	16/6	53.5 ± 16.7	16/6	57.6 ± 17.6		EA (GB34, BI40, Ashi, EX-B2)	Gabapentin 300 mg tid	12	VAS	Unknown
Ursini et al. 2011 [15]	32 (12/20)	65.5 ± 12.8	34 (10/24)	67.1 ± 12.8	O	Acupuncture (CV12, CV4, LI11, LI4, ST44, SP10, LR2, PC6, Ashi)	Pregabalin bid, levobupivacaine (inj).)	8	PHN	Unknown
Zhang et al. 2009 [21]	25 (11/14)	20–60	25 (12/13)	19–67		Acupuncture (EX-B2, Ashi, TE6, SI3)	Valacyclovir 0.3 g bid, Vit B1 10 mg tid, Vit B12 0.25 mg tid	10	ER, VAS, RP, EP	Unknown
Zheng et al. 2011 [29]	30 (unknown/unknown)	Unknown	30 (unknown/unknown)	Unknown		Acupuncture (Jie-geng treatment, acupuncture syndrome differentiation)	Acyclovir 0.4 g tid, Vit B1 20 mg tid	5	ER	Unknown
Zhou 2018 [34]	25 (15/10)	51.8 ± 8.4	25 (13/12)	51.2 ± 8.6		Acupuncture (circumference needle, EX-B2, Ashi)	Acyclovir 200 mg × 5, Vit B12 0.5 mg tid	10	ER, VAS	Unknown
Zhou et al. 2020 [31]	64 (28/36)	62.47 ± 5.63	64 (27/37)	63.52 ± 5.95	O	Acupuncture (surrounding acupuncture method with thumbtack needle)	Famciclovir 0.25 g tid, Vit B12 0.5 mg tid, gabapentin 0.3 g tid, tetrandrine 40 mg tid	14	ER, VAS, RP, CB, In, PHN	Unknown
Zhu 2019 [32]	37 (12/25)	Unknown	43 (24/19)	Unknown	O	Acupuncture (press needle surround acupuncture, Ashi)	Valacyclovir 0.3 g bid, Vit B12 0.5 g tid	15	VAS, CB, In, De, PHN	None

M, males; F, females; AEs, adverse events; C/E, control group/experimental group; EA, electro-acupuncture; bid, two times daily; Vit, vitamin; tid, three times daily; ER, effective rate; VAS, visual analog scale for pain; RP, time to pain relief; EP, time to pain elimination; CB, time to new blister cessation; In, time to incrustation; De, time to decrustation; PHN, incidence of post-herpetic neuralgia; IV, intravenous; qd, four times daily..

3. Risk of bias assessment

In 13 studies, the randomization principle was used to choose random numbers, which resulted in a low risk of bias in random sequence generation [14-17,19,24,25,27, 28,30-32]. Allocation concealment was reported in only six studies [14-16,18,24,27]. In terms of blinding of participants, all studies had a high risk of bias. Considering the characteristics of acupuncture, a completely blind study cannot be guaranteed unless sham acupuncture is used, which would cost more. One study [14] used laser-acupuncture in the control groups, but that technique is not enough to measure against sham needles. In terms of detection bias, only two studies had a low risk of bias [16,24], mentioning that a third party collected the results; the others had no reports on it. Outcomes data were incomplete in three studies. Liu et al. (2013) [19] found that the total number of people differed for each outcome. Shi (2008) [22] missed reporting the VAS after treatment, and Zheng et al. (2011) [29] gave insufficient information on dropouts. Selection bias was mostly unclear because the protocol could not be found. Regarding other bias, no study ranked highly because there could be room for further bias; however, we found no information to evaluate it (Fig. 2).

Figure 2. Risk of bias summary and risk of bias graph: review authors’ judgments about each risk of bias item for each included study presented as percentages across all included studies.

4. Effects of interventions

We classified interventions as acupuncture alone and acupuncture with Western medicine. Subgroups were divided by acupuncture type: acupuncture and EA. The primary outcomes were the effective rate and VAS. Secondary outcomes were time to pain relief, pain elimination, incrustation time, decrustation time, and incidence of PHN.

5. Effective rate

Regarding effective rate, 12 studies compared acupuncture with Western medicine [16-18,21,24-26,28-30,33,34], and three studies compared acupuncture to Western medicine [20,31,35].

In treatment with acupuncture alone, six results showed a significant difference (OR 5.63; 95% CI 2.45–12.92; p < 0.0001; I² = 0%). Treatment with EA showed a significant difference (OR 2.95; 95% CI 1.79–4.84; p < 0.0001; I² = 5%). The overall pooled result showed a statistically significant difference (OR 3.54; 95% CI 2.32–5.40; p < 0.00001; I² = 0%) (Fig. 3A).

Figure 3. Treatment effective rate. (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. M–H, Mantel–Haenszel; CI, confidence interval.

In add-on acupuncture treatment (two studies), the results showed a significant difference (OR 4.26; 95% CI 1.35–13.38; p = 0.01; I² = 0%). Only one result with EA showed no significant difference (OR 11.18; 95% CI 0.56–222.98; p = 0.11). The overall pooled result showed a statistically significant difference (OR 4.98; 95% CI 1.72–14.38; p = 0.003; I² = 0%) (Fig. 3B).

6. VAS

The 15 studies that evaluated VAS as the outcome calculated the SMD, because two versions of the VAS were used: 0–10 and 0–100. Nine studies evaluated acupuncture alone [16,17,19,21,23,24,28,30,34], and six evaluated acupuncture added to Western medicine [14,15,20,31,32,35].

In treatment with acupuncture alone, four results with acupuncture showed a significant difference (SMD 0.82; 95% CI 0.39–1.25; p = 0.0002; I² = 61%). The six results with EA treatment showed a significant difference in the pooled result (SMD 0.71; 95% CI 0.39–1.03; p < 0.0001; I² = 60%). The overall pooled result showed a statistically significant difference (OR 0.76; 95% CI 0.51–1.00; p < 0.00001; I² = 58%) (Fig. 4A).

Figure 4. Treatment visual analog scale for pain. (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. SD, standard deviation; IV, weighted mean difference; CI, confidence interval.

In add-on acupuncture treatment, four studies with acupuncture showed a significant difference (SMD 0.81; 95% CI 0.47–1.14; p < 0.00001; I² = 59%). Two results with EA showed no significant difference (SMD −0.46; 95% CI −0.99 to 0.06; p = 0.08). The overall pooled result showed no statistically significant difference (SMD 0.46; 95% CI −0.01 to 0.94; p = 0.06; I² = 81%) (Fig. 4B).

7. Time to pain relief

Six studies recorded the time (days) to pain relief as an outcome [16,19,21,23,24,30].

In treatment with acupuncture alone, four results with acupuncture showed a significant difference (MD −3.53; 95% CI −4.69 to −2.36; p < 0.00001; I² = 72%). Three results with EA treatment showed a significant difference in the pooled result (MD −2.62; 95% CI −4.25 to −0.98; p = 0.002; I² = 58%). The overall pooled result showed a statistically significant difference (MD −3.20; 95% CI −4.11 to −2.29; p < 0.00001; I² = 65%) (Fig. 5A).

Figure 5. Treatment time to pain relief (days). (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. SD, standard deviation; IV, weighted mean difference; CI, confidence interval.

In add-on acupuncture treatment, only one study with acupuncture showed a significant difference (MD −1.56; 95% CI −1.90 to −1.22; p < 0.00001) (Fig. 5B).

8. Time to pain elimination

Seven studies with acupuncture treatment alone measured pain elimination as the outcome [16,19,21-24,30]. Three studies with four results for acupuncture showed a significant difference (MD −7.59; 95% CI −10.95 to −4.23; p < 0.00001; I² = 93%). Four studies of EA showed a significant difference (MD −9.75; 95% CI −15.55 to −3.95; p = 0.001; I² = 89%). The pooled result showed a significant difference (MD −8.47; 95% CI −11.23 to −5.71; p < 0.00001; I² = 91%) (Appendix 1).

9. Time to new blister cessation

Nine studies recorded the time (days) to new blister cessation as an outcome [16,17,22-25,30].

In treatment with acupuncture alone, three results with acupuncture showed no significant difference (MD −0.69; 95% CI −1.45 to 0.07; p = 0.07; I² = 55%). Results with EA showed a significant difference (MD −0.97; 95% CI −1.88 to −0.07; p = 0.04; I² = 77%). The overall pooled result showed a statistically significant difference (MD −0.86; 95% CI −1.47 to −0.26; p = 0.005; I² = 71%) (Fig. 6A).

Figure 6. Treatment time to new blister cessation (days). (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. SD, standard deviation; IV, weighted mean difference; CI, confidence interval.

In add-on acupuncture treatment, two studies with acupuncture showed a significant difference (MD −0.61; 95% CI −1.17 to −0.05; p = 0.03; I² = 86%) (Fig. 6B).

10. Time to incrustation

Nine studies recorded the time (days) to incrustation as an outcome [16,17,22-24,28,30].

In treatment with acupuncture alone, three results with acupuncture showed a significant difference (MD −2.68; 95% CI −3.31 to −2.04; p < 0.00001; I² = 0%). Six results with EA showed a significant difference (MD −2.39; 95% CI −3.64 to −1.15; p = 0.0002; I² = 85%). The overall pooled result showed a statistically significant difference (MD −2.50; 95% CI −3.32 to −1.67; p < 0.00001; I² = 79%) (Fig. 7A).

Figure 7. Treatment time to incrustation (days). (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. SD, standard deviation; IV, weighted mean difference; CI, confidence interval.

In add-on acupuncture treatment, two studies with acupuncture showed a significant difference (MD −3.24; 95% CI −4.57 to −1.91; p < 0.00001; I² = 95%) (Fig. 7B).

11. Time to decrustation

Eight studies recorded the time (days) to decrustation as an outcome [16,17,22-25,30].

In treatment with acupuncture alone, three results with acupuncture showed a significant difference (MD −3.83; 95% CI −5.67 to −2.00; p < 0.0001; I² = 66%). For EA treatment, the pooled result showed a significant difference (MD −4.15; 95% CI −7.26 to −1.04; p = 0.009; I² = 85%). The overall pooled result showed a statistically significant difference (MD −3.96; 95% CI −5.79 to −2.13; p < 0.0001; I² = 83%) (Fig. 8A).

Figure 8. Treatment time to decrustation (days). (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. SD, standard deviation; IV, mean weighted difference; CI, confidence interval.

In add-on acupuncture treatment, only one result with acupuncture showed a significant difference (MD −3.16; 95% CI −3.61 to −2.71; p < 0.00001) (Fig. 8B).

12. Incidence of PHN

The PHN incidence rate was reported in 14 studies [15-17,19,22-24,27,30-32,35].

In treatment with acupuncture alone, three results with acupuncture showed a significant difference (OR 0.40; 95% CI 0.18–0.89; p = 0.03; I² = 0%). In EA, the pooled result showed a significant difference (OR 0.15; 95% CI 0.08–0.27; p < 0.00001; I² = 0%). The overall pooled result showed a statistically significant difference (OR 0.20; 95% CI 0.13–0.32; p < 0.00001; I² = 0%) (Fig. 9A).

Figure 9. Incidence of post-herpetic neuralgia. (A) Comparison of acupuncture treatment alone and Western medicine treatment. (B) Comparison of add-on acupuncture treatment (acupuncture + Western medicine) and Western medicine treatment. M–H, Mantel–Haenszel; CI, confidence interval.

In add-on acupuncture treatment, four results with acupuncture showed a significant difference (OR 0.42; 95% CI 0.25–0.69; p = 0.0007; I² = 35%) (Fig. 9B).

13. Adverse events (safety)

Eight studies reported information on safety [16,17,22-24,30,32,35].

In treatment with acupuncture alone, three results with acupuncture showed a significant difference (OR 0.15; 95% CI 0.05–0.52; p = 0.003; I² = 0%). EA treatment showed a significant difference in the pooled result (OR 0.11; 95% CI 0.02–0.61; p = 0.01; I² = 0%). The overall pooled result showed a statistically significant difference (OR 0.14; 95% CI 0.05–0.37; p < 0.0001; I² = 0%) (Fig. 10).

Figure 10. Fig. 10 . Incidence of adverse events. M–H, Mantel–Haenszel; CI, confidence interval.

In add-on acupuncture treatment, only one study (Zhu 2019 [32]) with acupuncture showed no adverse events.

14. Heterogeneity of results

When results showed significant heterogeneity, we performed a sensitivity analysis based on the interventions and number of treatments. In the VAS for acupuncture alone, a too-large initial value (Li et al. 2009 [28]) or variation outside the average (Zhang et al. 2009 [21], Liu et al. 2013 [19]) affected the heterogeneity (Table 2).

Table 2 . Sensitivity analysis of visual analog scale for pain in treatment with acupuncture alone.

Study	SMD	95% CI	I2 (%)	p-value
Li 2012 [30]	0.78	0.51–1.05	61	<0.0001
Su 2017 [16]	0.78	0.52–1.05	61	<0.0001
Su 2017 [16]	0.76	0.50–1.03	62	<0.0001
Zhang et al. 2009 [21]	0.67	0.47–0.87	32	<0.0001
Zhou et al. 2020 [31]	0.76	0.50–1.03	62	<0.0001
Bian 2010 [24]	0.79	0.52–1.06	60	<0.0001
Li et al. 2009 [28]	0.66	0.45–0.87	36	<0.0001
Liu et al. 2013 [19]	0.81	0.54–1.07	54	<0.0001
Song 2009 [23]	0.78	0.51–1.05	62	<0.0001
Sun et al. 2015 [17]	0.76	0.49–1.03	62	<0.0001
Sun et al. 2015 [17]	0.76	0.50–1.03	62	<0.0001

SMD, standardized mean difference; CI, confidence interval..

15. Publication bias

The funnel plot for the total effective rate of acupuncture alone in the treatment of HZ showed asymmetry, and the Egger regression p value was 0.0149, which suggest the presence of publication bias in the literature included in this meta-analysis. After applying the trim-and-fill method, the results were corrected from OR 3.54, 95% CI 2.32 to 5.40, p < 0.00001, to OR 2.74, 95% CI 1.82 to 4.14, p < 0.0001, which showed a decrease in effectiveness.

The funnel plot for the total VAS with acupuncture alone in the treatment of HZ showed no obvious asymmetry, and the Egger regression p-value was 0.1826, suggesting a lack of any obvious publication bias in the literature included in this meta-analysis. The results for other outcomes also showed no obvious publication bias.

The results of the systematic review are reliable.

Virus dormant in the human ganglion reemerges when immunity is weakened due to age, stress, and use of pharmaceuticals. This causes nerve damage and evokes severe pain together with skin lesions consisting of rash, blisters, and crustation. Western medicine treats this condition with antiviral and analgesic medications. Antiviral medication relieves pain and shortens the acute phase by depressing virus replication. The number of patients experiencing PHN even after treatment and the side effects of medications suggest a need for alternative or accompaniment treatments. This study attempted to reveal the efficacy and safety of acupuncture in the acute phase of HZ.

Our meta-analysis on the effective rate, VAS, time to pain reduction, time to skin symptoms reduction, and incidence of PHN from 22 studies showed that acupuncture has advantages in all categories. Additionally, the risk of adverse effects is lower with acupuncture alone than with either Western medicine alone or Western medicine combined with acupuncture. Regarding heterogeneity, we performed a sensitivity analysis for the primary outcome (VAS) and found that cases with a high initial value or a value difference outside the average created heterogeneity. In secondary outcomes, considering that the results showing high heterogeneity were time-related items, the heterogeneity was thought to be influenced by the patients’ stage at the time of study enrollment. Predicting the exact stage of a patient is difficult, and studies stated only that the patient’s condition with the virus at the time of enrollment was 0–7. Furthermore, considering that HZ is a disease with changing symptoms, it might be a cause of the heterogeneity related to time categories.

Taken together, these results suggest that acupuncture could be a reasonable treatment for patients with HZ. Acupuncture is known to be safe because it rarely has adverse effects. Both acupuncture and EA can be used for treatment. Either technique can be used mainly by stimulating the area around the affected area or by selecting a meridian such as Governor Vessel, Bladder, or Extra point B2, which is the spinal segmental approach. Acupuncture can be applied for people who cannot take medicine for some reason or who are already using a medication. Considering that people develop HZ due to weakened immunity, old age, and complications of existing diseases, this external treatment can be considered effective. It not only reduces the duration of illness, but also the occurrence of PHN.

However, this study has some limitations. First, no uniformity in the protocols of the included studies was evident. The method of acupuncture treatment, number of treatments, sex ratio, and age criteria will all affect the results depending on the study. Second, the included studies were conducted mostly in China, which results in a lack of regional diversity. Third, although causes for the heterogeneity in the included studies have been hypothesized, the risk of heterogeneity remains. Fourth, no study satisfied the criteria of blinding participants and personnel, which affects the results. Lastly, we analyzed acupuncture and EA at the same time, which could have resulted in methodologic error and increased heterogeneity. However, the included studies were insufficient to be compared separately. So, related studies must be carried out in future, and a more detailed analysis is needed.

Several systematic reviews that include various acupuncture treatment options or herbal treatments have been reported. Many included patients with PHN. However, we focused on acupuncture and EA, which doctors of Korean medicine use most for treatment as an intervention. We also limited our analysis to patients with HZ, excluding those with PHN. And we included studies from countries other than China. Further research is needed to address the limitations observed in our analysis. Although this study examining the effectiveness and safety of acupuncture in acute HZ has some limitations, it could be used as basic data for future studies of acupuncture in HZ.

The results of our meta-analysis suggest that acupuncture could be a reasonable treatment for patients with HZ who suffer from pain and have accompanying symptoms. Acupuncture treatment might also lower the possibility of PHN. However, there are some limitations related to the included studies, and further studies are needed to confirm our results. Nevertheless, this systematic review suggests that clinical evidence supports acupuncture as a treatment for HZ.

Conceptualization: EJK. Data curation: JHM, SKC. Formal analysis: WSS. Funding acquisition: EJK. Investigation: SDL, KHK. Methodology: WSS. Project administration: JHM, SKC. Supervision: EJK. Visualization: SKC. Writing – original draft: SKC. Writing – review & editing: SKC, JHM, WSJ, JEJ, SHP, CYJ, BKS, SDL, KHK, EJK.

The authors have no conflicts of interest to declare.

This study was supported by the Traditional Korean Medicine R&D program funded by the Ministry of Health and Welfare through the Korean Health Industry Development Institute (grant No. HF22C0071).

This research did not involve any human or animal experiment.