Trigger finger is a debilitating hand condition that affects approximately 2–3% of the adult population [1]. This condition, also known as stenosing flexor tenosynovitis, causes the fingers to catch, click, or lock and can significantly impair hand function [2]. On an ultrasound examination, the A1 pulley of the intruded finger is markedly thicker and stiffer than the normal finger. Diagnosis is often based on clinical findings, with symptoms ranging from painless clicking to painful triggering and sometimes even flexed or locked digits. Conventional nonsurgical treatment options, such as splinting and local steroid injections, are often used in mild or early cases; however, they have limited long-term efficacy and may lead to recurrent attacks. Surgical intervention, which involves releasing the A1 pulley, is reserved for more severe cases; however, it has risks and complications, such as persistent triggering or unsettled flexion contracture of the affected finger, which affects 7–43% of cases.

Korean medicine therapies such as acupotomy and pharmacopuncture have shown promising results in treating trigger finger [3]. Similarly, in China, randomized controlled trials (RCTs) have demonstrated the effectiveness of acupuncture, acupotomy, and related therapies in reducing pain and dysfunction in affected fingers. However, a comprehensive review of all acupuncture and related therapies is necessary to evaluate their efficacy in treating trigger finger. The currently available systematic reviews and meta-analysis literature [4-9] regarding the effective treatment for trigger finger cover only acupotomy as an intervention, and the other Korean medicine treatments have not been discussed. In addition, these reviews treat all interventions as a single category, and different interventions have not been differentially compared. Thus, this systematic review and meta-analysis aimed to fill this gap by comprehensively evaluating various acupuncture and related therapies, differentiating between interventions, and comparing them to appropriate comparators. This review hopes to shed light on the potential of acupuncture-related therapies for improving the lives of those with trigger finger.

This systematic review and meta-analysis of the efficacy of acupuncture for treating trigger finger was conducted in accordance with the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) statement. The International Prospective Register of Systematic Reviews (PROSPERO) registry code is CRD420 23405667.

1. Types of studies

Only RCTs with available full-text journal articles were included in this review. Retrospective and quasi-randomized trial studies were excluded.

2. Type of participants

This review focused on adult individuals (aged > 18 years) who have been diagnosed with trigger finger. Studies dealing with stenosing tenosynovitis that affects other body parts such as De Quervain’s synovitis were excluded from the analysis. Other diseases that affect the fingers such as fractures, tuberculosis, tumor, metacarpophalangeal joint sprain, infection, rheumatoid arthritis, and osteoarthritis, were also excluded, in favor of the more precise diagnosis of trigger finger. Considering the wide range of terminologies used to describe this condition, this review encompasses terms such as snapping finger, tenosynovitis of the flexor tendons of the hand, stenosing flexor tenosynovitis, and flexor tendon tenosynovitis. Furthermore, this review is inclusive of various search engine languages such as Chinese, Japanese, and Korean (Supplementary Material 1).

3. Types of interventions

This review focused on acupuncture-related therapies for the treatment of trigger finger. The interventions included acupuncture, electroacupuncture, acupotomy, auricular acupuncture, fire needle, and pharmacopuncture. Studies involving herbal medicine for oral and external usage were excluded. Studies including combination and add-on treatments were included in this analysis, as long as they distinctively involved an acupuncture-related therapy. Various comparators, ranging from no intervention to block therapy, injection, and surgery, were also included. Studies that compared different types of acupuncture-related therapies and different usages of acupotomy such as blind versus ultrasound-guided techniques were not considered.

4. Outcome measures

In this study, the primary outcomes of interest were the effectiveness rate (ER) and pain intensity. The widely accepted scales, namely, visual analog scale (VAS) and Numerical Rating Scale (NRS) were used to measure pain intensity. The ER is based on the “Diagnosis and Curative Effect Criteria of Traditional Chinese Medicine Syndrome” [10,11], which is commonly used in Chinese research to standardize evaluations. The criteria consist of a three- and four-level index. In the 3-level standard, the results are classified as markedly effective, effective, or invalid. “Markedly effective” refers to a condition in which the pain and tenderness in the affected area have disappeared completely, and the patient does not experience symptoms such as snapping or inability to flex and can stretch the finger normally. “Effective” refers to a slight improvement in the pain and tenderness of the affected area, but it is not entirely healed—the finger no longer experiences catching; however, pain or snapping persists when moved. “Invalid” is defined as the lack of effective improvement in symptoms or abnormalities [10]. In the four-level standard, results are classified as cured, markedly effective, effective, or invalid. “Cured” refers to a state where clinical symptoms have disappeared, and activity has returned to normal, with complete absence of a snapping sound or locking. “Markedly effective” refers to the condition in which the patient experiences significant improvement in clinical symptoms and can perform activities normally without a snapping sound or locking. “Effective” refers to relief from clinical symptoms, but slight limitations in mobility are still present, however with no snapping sound or locking symptoms. “Invalid” refers to the lack of improvement in clinical symptoms [11]. The total ER was calculated by adding up cases that were classified as markedly effective and effective as per the three-level grade, as well as cured, markedly effective, and effective according to the four- level grade, as specified in the study’s description. The secondary outcome measures included Quinnell grade (QG) and recurrence rate (RR). Although no grading system has been universally accepted as the gold standard, the QG system is widely utilized to assess the severity of trigger finger [12]. According to QG, grade 0 indicated the presence of mild crepitus in a non-triggering finger; grade I, there was no triggering with uneven movement; grade II, the triggering was actively correctable; grade III, it is usually correctable with the help of the other hand; and grade IV, this pertains to the locked digit stage [1]. In some studies, grades were assigned on a scale from grade I to grade V, which corresponds to one level lower than the grading system described above. The latter method was used in this study, as it had already been employed in one study [13] to calculate the average.

5. Data sources and search methods

In the pursuit of comprehensive research, a thorough literature search was conducted using electronic databases including MEDLINE via PubMed, Cochrane, Embase, China National Knowledge Infrastructure (CNKI), Wangfang, J-STAGE, Citation Information by NII (CiNii), KoreaMed, Korean Studies Information Service System (KISS), ScienceOn, and Oriental Medicine Advanced Searching Integrated System (OASIS). Studies conducted until January 2023 were included in the review, and only RCTs were included. No language restriction was made to the search, and the studies were manually selected after sifting through the retrieved records. The search strategy for each database is available in Supplementary Material 1.

6. Data extraction and quality assessment

Two members of our team conducted a thorough search for relevant studies. They worked independently to collect and extract articles using both electronic databases and manual searches. After eliminating any duplicated records using EndNote, the team screened titles and abstracts and performed a detailed analysis of eligible full-text articles. In this stage, several factors including diagnostic criteria, study design, and outcome measures were considered. In cases of disagreement, two researchers first discussed the issue and attempted to solve it through mutual consensus. If a consensus could not be reached, a third researcher was consulted to ensure that a final decision was reached collaboratively. The data extraction process began after the studies were selected and finalized. This process involved the collection of essential information such as sample size, sex, age, disease duration, intervention type, number of treatment sessions, inserted point, details of intervention, primary and secondary outcome measurements, and adverse events (AEs). This process was accomplished by two researchers, and in the case of a difference in opinion, a third researcher also participated in the discussion.

7. Quality of evidence

Two independent researchers evaluated the risk-of-bias (ROB) in accordance with Cochrane’s ROB tool for RCTs. All decisions were taken by the two researchers through mutual consensus, and any difference of opinion was resolved with the assistance of a third researcher. The ROB summary and graph were visualized by Review Manager (RevMan) software for Windows (version 5.4, The Cochrane Collaboration, 2020). Finally, the potential for publication bias was examined using a funnel plot, with all analyses conducted using RevMan.

8. Statistical analysis

RevMan was used to synthesize the data and conduct a meta-analysis. Both dichotomous and continuous variables were analyzed based on 95% confidence intervals (CIs). The chi-square test and I-squared statistics were used for the assessment of statistical heterogeneity. Furthermore, subgroup analyses were conducted based on the intervention types. Since > 10 studies were included, the risk of publication bias was assessed using both the funnel plot and Egger’s regression test, and further correction was obtained using the Comprehensive Meta-analysis for Windows (version 4.0.000, Borenstein, M. and Rothstein, H., Biostat, Englewood Cliffs, 2022).

Before presenting the results of this review, certain important terminologies used in this review, such as trigger finger, acupotomy, acupuncture, dry needling, and block therapy, were standardized.

1. Study selection

A comprehensive search of articles yielded 1,856 articles, of which 686 duplicates were identified and were eliminated using both EndNote and manual screening methods. Subsequently, studies that were not related to acupuncture or trigger finger or were not RCTs were removed, so were cadaver studies. Studies with unavailable full-text articles were also discarded, leaving us with 156 articles for initial screening. After evaluating these articles on the basis of the eligibility criteria concerning the participants, interventions, and outcomes, several studies were further excluded, and finally, 19 studies that met the quality criteria for both qualitative and quantitative synthesis were included. A detailed PRISMA flowchart outlining the study selection process is given below in Fig. 1.

Figure 1. Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) flow diagram of the study selection process. CNKI, China National Knowledge Infrastructure; CiNii, Citation Information by NII; KISS, Korean Studies Information Service System; OASIS, Oriental Medicine Advanced Searching Integrated System; RCT, randomized controlled trial.

2. Study characteristics

Table 1 [13-31] presents the characteristics of the selected studies, and Supplementary Material 2 provides additional information about the medical agents used as local anesthesia and block therapy, disease duration, and treatment details employed in the trials. All 19 studies included in this review were conducted in China, except for one, which was from Iran [14]. These studies were published between 2016 and 2022 and involved a total of 1381 patients (507 men and 874 women), with a mean age of 47.33 years. All studies reported the heterogeneity of the general information between the experimental and control groups (p > 0.05). Two studies were three-arm trials, one study had two experimental groups [25], one study had two control groups, and the rest were two-arm parallel trials.

Table 1 . Characteristics of the included studies.

Study	Experimental group (M/F)	Experimental group age (y)	Control group (M/F)	Control group age (y)	Add-on treatment	Experimental group treatment	Control group treatment	Number of treatments	Outcome measures	AEs (E/C)
Azizian et al. [14] (2019)	29 (13/16)	60.24 ± 8.54	29 (11/18)	62.00 ± 7.73		Dry needling (with fast-in and fast-out technique)	None	1 session	VAS, SS, TT, and pinch grip (kg)	Not reported
Bo and Zhang [15] (2020)	15 (7/8)	67.2 ± 1.4	15 (8/7)	67.7 ± 1.3		Ultrasound-guided acupotomy with local anesthesia	Block therapy	E: 1, C: Min 1, max 3	ER, VAS, RR, and SS	Not reported
Chen et al. [16] (2022)	30 (8/22)	48.47 ± 7.52	30 (10/20)	48.40 ± 6.65		Dry needling (with TrPs stimulation)	Block therapy	E: 1, C: 2	VAS and ROM	Not reported
Chen [17] (2022)	30 (13/17)	49.08 ± 4.65	30 (11/19)	48.86 ± 4.72		Acupotomy with local anesthesia	Surgery with local anesthesia	1 session	ER (4), VAS, QG, RR, and SS	2/9
Huang et al. [18] (2021)	30 (9/21)	41.45 ± 4.26	30 (10/20)	41.42 ± 4.21		Acupotomy with local anesthesia	Surgery with local anesthesia	1 session	ER (4), VAS, QG, RR, and SS	1/10
Hu et al. [19] (2022)	29 (9/20)	52.32 ± 8.30	29 (8/21)	54.14 ± 9.61	Yes, block therapy	Ultrasound-guided acupotomy + ultrasound-guided block therapy	Ultrasound-guided block therapy	1 session	ER, VAS, QG, and TT	Not reported
Jia et al. [20] (2018)	60 (10/50)	53.28 ± 12.33	60 (13/47)	51.18 ± 10.83	Yes, block therapy	Acupotomy + block therapy	Block therapy	1 session	ER, VAS, RR, and SS	Not reported
Jiang and Zhou [21] (2022)	23 (11/12)	66.48 ± 2.41	23 (12/11)	66.71 ± 2.68		Ultrasound-guided acupotomy with local anesthesia	Block therapy	1 session, max 2	ER, VAS, RR, and SS	Not reported
Liu and Zhang [22] (2018)	52 (19/33)	49.5 ± 5.9	52 (21/31)	52.32 ± 3.5		Acupotomy with local anesthesia	Block therapy	1 session	ER and VAS	Not reported
Meng et al. [23] (2022)	36 (12/24)	41.8 ± 5.7	36 (11/25)	46.5 ± 5.3		Arc-edge acupuncture without anesthesia	Block therapy	1 session	ER (4), NRS, QG, and ROM	Not reported
Tang [24] (2016)	30 (12/18)	31 ± 10.6	30 (11/19)	32 ± 9.8		Acupotomy with local anesthesia	Block therapy	1 session	ER	Not reported
Tao et al. [13] (2021)	61 (28/33)	47.5 ± 3.1	61 (28/33)	47.5 ± 3.1		Acupotomy with local anesthesia	Block therapy	1 session	ER (4) and QG	Not reported
Wang et al. [25] (2020)	A: 40 (18/22) B: 40 (19/21)	A: 47.23 ± 8.48 B: 46.28 ± 7.86	C: 40 (16/24)	C: 46.51 ± 8.02	Yes, block therapy	A: Acupotomy with local anesthesia + block therapy B: Acupotomy with local anesthesia	C: Block therapy	A, B: Min 2, max 3, C: Min 1, max 3	ER, VAS, and blood test	Not reported
Xie et al. [26] (2017)	25 (7/18)	46.2 ± 6.9	25 (6/19)	47.5 ± 7.2		Acupotomy with local anesthesia	Block therapy	1 session	ER	Not reported
Xie [27] (2019)	30 (10/20)	36.80 ± 8.99	30 (13/17)	35.80 ± 8.98		Acupotomy + block therapy	Block therapy	1 session	ER and VAS	Not reported
Yang et al. [28] (2019)	30 (11/19)	48.21 ± 5.31	30 (10/20)	46.86 ± 5.56	Yes,antibiotic	Acupotomy with local anesthesia + roxithromycin PO	Block therapy	1 session	ER (4), VAS, and RR	Not reported
Zhang et al. [29] (2018)	39 (12/27)	38.3 ± 4.5	38 (11/27)	37.9 ± 5.1	Yes, NSAIDs	Acupotomy with local anesthesia + celecoxib PO + ice pack	Block therapy	1 session	ER, VAS, and RR	Not reported
Zhang et al. [30] (2019)	37 (16/21)	48 ± 7	37 (14/23)	43 ± 5	Yes, block therapy	Acupotomy with local anesthesia + block therapy	Block therapy with local anesthesia	1 session	ER (4) and SS	Not reported
Zhu et al. [31] (2018)	30 (14/16)	38.92 ± 19.82	Group 1: 30 (13/17) Group 2: 30 (12/18)	41.35 ± 16.35 40.42 ± 18.83		Acupotomy with local anesthesia	Group 1: Block therapy Group 2: Surgery with local anesthesia	1 session	ER, VAS, and ROM	Not reported

Values are presented as number only or mean ± standard deviation..

M, male; F, female; C, control group; E, experimental group; ER, effectiveness rate; QG, Quinnell grade; ROM, range of motion; RR, recurrence rate; SS, symptom and sign; TT, tendon thickness; VAS, visual analog scale; TrP, trigger point; PO, per oral; NRS, Numerical Rating Scale; NSAID, nonsteroidal anti-inflammatory drug..

In this review, the 19 studies used the following interventions: two studies used acupuncture [14,16]; one, arc-edge acupuncture [23]; two, ultrasound-guided acupotomy [15,21]; eight, acupotomy [13,17,18,22,24-26,31]; one, ultrasound-guided acupotomy with add-on treatment [19]; and six, acupotomy with add-on treatment [20,25,27-30]. In the studies that used acupuncture, the criteria for comparison were “no intervention” [14] and block therapy [16]. In studies using arc-edge acupuncture, the control group used block therapy [23]. In studies that used ultrasound-guided acupotomy, the comparison was against block therapy [15,21]. In eight studies with acupotomy, the main comparison was with block therapy in six studies [13,22,24-26,31], and the secondary comparison was with conventional surgery [17,18,31]. Regarding ultrasound-guided acupotomy add-on, only one study [19] combined ultrasound-guided block therapy and compared it with ultrasound-guided block therapy. For the acupotomy add-on treatment, the add-on treatment included blind block therapy [20,25,27,30], and in one study, only oral medication [28] along with an ice pack [29] was used as add-on treatments.

The number of treatments varied from one to a maximum of three sessions; however, most of the studies reported only one treatment session. The primary outcome measures were the ER and pain levels, and the ER was the most dominant outcome in all studies, except two [14,16]. If the ER was assessed using the four-level standard, it is marked 4 in parentheses. Pain intensity was the secondary dominant outcome, in nearly all, except four studies [13,24,26,30]. All studies used the VAS scale to assess pain level, except one study which used the NRS scale [23]. The ER was expressed as odds ratio (OR), and the VAS and NRS were expressed as mean difference (MD). The secondary outcomes included QG and RR. QG was utilized in five studies [13,17-19,23], and RR was reported in seven studies [15,17,18,20,21,28,29]. QG was expressed as the MD, and RR was expressed as an OR. Unfortunately, owing to the varying standards of assessment across studies and the small number of studies available, we could not analyze other outcome measures, such as pinch grip, tendon thickness, blood tests, and range of motion.

The majority of the included studies utilized the Ashi point or the anatomic A1 pulley position where hypertrophic nodules or indurations were detected. One study [14] specified a retention time of 1 minute, whereas the rest of the studies did not provide a precise treatment time. Instead, the operation was repeated until the patient’s finger was free from dorsal movement restriction or resistance (Supplementary Material 2).

3. Risk-of-bias assessment

In the majority of the studies included in the analysis, the randomization principle was implemented using a “random number table method” or “group names picked from a bowl,” resulting in a low ROB in terms of random sequence generation criteria. However, in seven studies [13,19,20,22,26,28,30], the risk was unclear, as they only mentioned “random allocation” without describing the exact method used to achieve the randomization. Regarding allocation concealment, most studies did not report this information, leading to an unclear ROB. In three studies [13,19,30], random allocation was conducted based on the order of hospital visits, which could be considered to have a high ROB (Fig. 2).

Figure 2. Risk-of-bias summary.

As none of the included studies used sham acupuncture or a sham acupotomy for comparison, no studies could be considered complete blind studies, resulting in a high ROB. A study [18] compared acupotomy and conventional surgery, making it difficult to compare against sham acupotomy because of a lack of information recording unclear risk. All the selected studies recorded an unclear risk of detection bias because none reported details of blinding outcome assessment.

Regarding incomplete outcome data, only one study [29] reported missing data, with 37 participants missing RR data (11 in the experimental group and 26 in the control group), accounting for 48.05% of the total participants. This is a high percentage of missing data, as anything > 20% presents a high ROB [32]. However, none of the other studies have reported any missing data, resulting in a low ROB. Most of the included studies were associated with a low ROB concerning selective report bias because they all reported estimated results. However, one study [14] mentioned a pilot study that could not be assessed, resulting in an unclear ROB. All studies had a low ROB for other risks because other biases could be possible; however, no information was available for assessment. The ROB is visualized and summarized in Figs. 2 and 3.

Figure 3. Risk-of-bias graph.

4. Quantitative synthesis

All outcomes were classified into subgroups based on the type of intervention used. The primary outcomes were the ER and pain intensity. Arc-edge acupuncture, acupotomy, and acupotomy add-on groups were listed in the ER subgroup. Acupuncture, acupotomy, and acupotomy add-on groups belonged to the VAS subgroup. Only one arc-edge acupuncture study reported NRS, and it was analyzed separately. The secondary outcome was divided into the QG and RR. The QG was divided into arc-edge acupuncture, acupotomy, and acupotomy add-on groups, whereas the RR was divided into acupotomy and acupotomy add-on groups (Figs. 4–8).

Figure 4. Effectiveness rate of (A) arc-edge acupuncture, (B) acupotomy add-on, and (C) acupotomy add-on. CI, confidence interval; NSAID, nonsteroidal anti-inflammatory drug.

5. Effectiveness rate

The ER analysis subgroup included 17 studies (Fig. 4). The arc-edge acupuncture group utilized a fixed-effect model, and both the acupotomy and acupotomy add-on groups utilized a random-effects model because of inconsistencies in the study design and intervention. Only one study utilized arc-edge acupuncture, necessitating the use of a fixed-effects model, as the variation between studies could not be estimated. Within this group, one study [23] yielded significant results (OR, 22.27; 95% CI, 2.73–181.46; p = 0.004; I² = not applicable) when compared with block therapy (Fig. 4A).

In the acupotomy group, two studies [15,21] showed no significant difference between ultrasonography-guided acupotomy and block therapy (OR, 4.85; 95% CI, 0.95–24.70; p = 0.06; I² = 0%), whereas six studies [13,22,24-26,31] demonstrated a significant difference with substantial heterogeneity between acupotomy and block therapy (OR, 19.80; 95% CI, 3.36–116.60; p = 0.0010; I² = 85%). Three studies [17,18,31] demonstrated a significant difference between acupotomy and surgery (OR, 4.22; 95% CI, 1.67–10.70; p = 0.002; I² = 0%). The overall pooled result demonstrated a statistically significant difference with substantial heterogeneity (OR, 9.59; 95% CI, 3.52–26.11; p < 0.00001; I² = 72%). However, in the subgroup difference test, heterogeneity was comparatively low (I² = 14.2%) (Fig. 4B).

In the acupotomy add-on group, one study [19] demonstrated no significant difference between ultrasound- guided acupotomy plus ultrasound-guided block therapy and ultrasound-guided block therapy (OR, 16.32; 95% CI, 0.87–304.72; p = 0.06; I² = not applicable). Four studies [20,25,27,30] showed significant difference between acupotomy plus block therapy and block therapy (OR, 23.09; 95% CI, 5.61–95.01; p < 0.00001; I² = 24%). One study [28] also showed significant difference between acupotomy plus antibiotics and block therapy (OR, 10.55; 95% CI, 1.23–90.66; p = 0.03; I² = not applicable), while one study [29] showed no significant difference between acupotomy plus nonsteroidal anti-inflammatory drugs (NSAIDs) and ice pack and block therapy (OR, 5.41; 95% CI, 0.25–116.51; p = 0.28; I² = not applicable). The total pooled result showed a statistically significant difference (OR, 16.12; 95% CI, 6.26–41.53; p < 0.00001; I² = 0%) (Fig. 4C).

6. Effects on pain intensity in trigger finger

The analysis of VAS included a total of 14 studies, using both fixed-effects and random-effects models for the reasons mentioned previously (Fig. 5). One study [14] showed a significant difference between acupuncture and no intervention (MD, −1.80; 95% CI, −2.18 to −1.42; p < 0.00001; I² = not applicable). Another study [16] compared acupuncture with block therapy and demonstrated a significant difference (MD, −0.70; 95% CI, −1.22 to −0.18; p = 0.008; I² = not applicable). The overall pooled result indicated a statistically significant difference and was considerably heterogeneous (MD, −1.41; 95% CI, −1.72 to −1.11; p < 0.00001; I² = 91%). This high heterogeneity is most likely due to the mixture of interventions (Fig. 5A).

Figure 5. Visual analog scale of (A) acupuncture, (B) acupotomy, and (C) acupotomy add-on. SD, standard deviation; CI, confidence interval; NSAID, nonsteroidal anti-inflammatory drug.

When analyzing the effects of acupotomy as the only intervention, ultrasound-guided acupotomy compared with block therapy in two studies [15,21] demonstrated a significant difference with considerable heterogeneity (MD, −1.40; 95% CI, −2.38 to −0.43; p = 0.005; I² = 99%). However, three studies [22,25,31] showed no significant difference between acupotomy and block therapy (MD, −0.67; 95% CI, −2.05 to 0.70; p = 0.34; I² = 97%). Three different studies [17,18,31] that compared acupotomy with surgery showed a significant difference with considerable heterogeneity (MD, −0.97; 95% CI, −1.55 to −0.39; p = 0.0010; I² = 78%). The overall pooled results indicated a significant difference (MD, −0.96; 95% CI, −1.54 to −0.38; p = 0.001; I² = 98%). Although the heterogeneity of each subgroup and the overall pooled results were high, the subgroup difference test demonstrated a low heterogeneity (I² = 0%) (Fig. 5B).

In the acupotomy add-on group, one study [19] showed a significant difference between ultrasound-guided acupotomy plus ultrasound-guided block therapy and ultrasound-guided block therapy (MD, −0.83; 95% CI, −1.29 to −0.37; p = 0.0004; I² = not applicable). Three studies [20,25,27] showed lack of significant difference but considerable heterogeneity between acupotomy plus block therapy and block therapy (MD, −1.32; 95% CI, −2.70 to 0.05; p = 0.06; I² = 99%). In one study, a significant difference was observed (MD, −2.80; 95% CI, −3.53 to −2.07; p < 0.00001; I² = not applicable) between acupotomy plus antibiotics and block therapy [28]. The study that compared acupotomy plus NSAIDs and ice pack with block therapy [29] revealed a significant difference (MD, −0.82; 95% CI, −1.32 to −0.32; p = 0.001; I² = not applicable). The overall pooled results demonstrated a statistically significant difference with considerable heterogeneity (MD, −1.39; 95% CI, −2.30 to −0.48; p = 0.003; I² = 98%) (Fig. 5C). Given the significant heterogeneity, sensitivity analysis was conducted (refer to heterogeneity of results).

Only one study [23] was considered for the NRS category, where arc-edge acupuncture was compared with block therapy. The results indicated a significant difference (MD, −0.60; 95% CI, −0.92 to −0.28; p = 0.0003; I² = not applicable), as shown in Fig. 6.

Figure 6. Quinnell grade of (A) arc-edge acupuncture, (B) acupotomy, and (C) acupotomy add-on. SD, standard deviation; CI, confidence interval.

7. Quinnell grade

The QG subgroup analysis included a total of five studies that used the QG as the secondary outcome (Fig. 7). The only study [25] that compared arc-edge acupuncture with block therapy revealed a significant difference (MD, −1.39; 95% CI, −1.71 to −1.07; p < 0.00001; I² = not applicable) (Fig. 7A). The study [13] that compared acupotomy with block therapy demonstrated a significant difference (MD, −1.41; 95% CI, −1.74 to −1.08; p < 0.00001; I² = not applicable). Acupotomy was compared with surgery in two studies [17,18], where a significant difference was observed (MD, −0.82; 95% CI, −1.14 to −0.49; p < 0.0001; I² = 0%). The overall pooled results demonstrated a statistically significant difference with moderate heterogeneity (MD, −1.04; 95% CI, −1.47 to −0.61; p < 0.00001; I² = 69%) (Fig. 7B). While heterogeneity was 0% when compared with surgery, the overall heterogeneity increased to 69%. This increase in heterogeneity was thought to be due to the inclusion of a block therapy subgroup; therefore, it appears to be caused by the heterogeneity of the different control group interventions. A study [19] demonstrated a significant difference between ultrasound-guided acupotomy plus ultrasound-guided block therapy and ultrasound-guided block therapy (MD, −0.66; 95% CI, −1.15 to −0.16; p = 0.010; I² = not applicable) (Fig. 7C).

Figure 7. Numerical Rating Scale of arc-edge acupuncture. SD, standard deviation; CI, confidence interval.

8. Recurrence rate

Five studies evaluated the RR (Fig. 8). Two studies [15, 21] that used acupotomy alone and compared ultrasound-guided acupotomy with block therapy showed a significant difference (OR, 0.18; 95% CI, 0.04–0.90; p = 0.04; I² = 0%). One study [20] that compared acupotomy with block therapy demonstrated a significant difference (OR, 0.10; 95% CI, 0.02–0.48; p = 0.004; I² = not applicable). Overall, the pooled results showed a statistically significant difference (OR, 0.13; 95% CI, 0.04–0.41; p = 0.0004; I² = 0%) (Fig. 8A).

Figure 8. Recurrence rate of (A) acupotomy and (B) acupotomy add-on. CI, confidence interval.

In the acupotomy add-on group, one study [28] did not find a significant difference when acupotomy plus antibiotics was compared with block therapy (OR, 0.11; 95% CI, 0.01–0.99; p = 0.05; I² = not applicable). A study [29] that was included for comparing acupotomy plus NSAIDs and icepack with block therapy showed a significant difference (OR, 0.09; 95% CI, 0.02–0.45; p = 0.004; I² = not applicable). The total pooled results showed a statistically significant difference (OR, 0.10; 95% CI, 0.03–0.35; p = 0.0005; I² = 0%) (Fig. 8B).

9. Adverse events

Two studies [17,18] reported the occurrence of AEs when acupotomy was compared with conventional surgery. Chen [17] reported that 2 out of 30 participants in the experimental group and 9 out of 30 participants in the control group experienced AEs, with an OR of 0.17 (95% CI, 0.03–0.85). Similarly, in the study by Huang et al. [18], only 1 out of 30 participants in the experimental group reported AEs compared with 10 out of 30 in the control group, with an OR of 0.07 (95% CI, 0.01–0.58). The types of AEs reported varied; the acupotomy group experienced one case of wound infection and two cases of tenderness, whereas the surgery group experienced six cases of tenderness, four cases of wound infection, four cases of scar hyperplasia and adhesion, three cases of tendon adhesion, and two cases of neurovascular injury. The overall results showed a statistically significant difference in the occurrence of AEs between the two groups (OR, 0.11; 95% CI, 0.03–0.41; p = 0.0009; I² = 0%) (Fig. 9). The remaining studies did not report any AEs.

Figure 9. Adverse events of acupotomy compared with conventional surgery. CI, confidence interval.

10. Publication bias

The inclusion of 11 studies in the acupotomy ER analysis made it necessary to assess publication bias. The funnel plot of the overall ER of acupotomy (Fig. 10) exhibited significant asymmetry. The Egger regression p-value of 0.0073 indicated the presence of publication bias. The trim-and-fill method was employed to correct this asymmetry. Following the correction, the OR increased from 9.59 (95% CI, 3.52–26.11; p < 0.00001) to 15.36 (95% CI, 5.32–44.36; p < 0.00001).

Figure 10. Funnel plot of the effectiveness rate of acupotomy. SE, standard error; OR, odds ratio.

11. Heterogeneity of results

In the analysis of the VAS scores of the acupotomy add-on group, both the general subgroup and pooled results exhibited significant heterogeneity. Therefore, an additional sensitivity analysis was conducted (Table 2). The MD of exception and I² of exception values were measured by subtracting the following each study. First, we observed differences in weights, where Yang et al. [28] had the least weight, whereas Jia et al. [20] had the highest weight. However, the differences were not significant. Among the MDs of exception, Jia et al. [20] and Yang et al. [28] showed maximum difference (0.306 and 0.261, respectively) compared with other results. Regarding the I² of exception, Jia et al. [20] had significant differences from the original total. The heterogeneity of the ER in acupuncture, VAS scores, and QG in the acupotomy group has already been mentioned above.

Table 2 . Sensitivity analysis for visual analog scale on the acupotomy add-on group.

Study	MD of exception	Weight (random)	I² of exception
Hu et al. [19] (2022)	−1.502	16.6	98.538
Jia et al. [20] (2018)	−1.084	17.26	86.524
Wang et al. [25] (2020)	−1.533	17.21	97.214
Xie [27] (2019)	−1.524	16.86	98.461
Yang et al. [28] (2019)	−1.129	15.58	98.561
Zhang et al. [30] (2019)	−1.503	16.49	98.547
Original total	−1.39		98.259

MD, mean difference..

Trigger finger is a widely recognized condition that impedes the movement of the affected digit, causing considerable disturbance. This disease typically affects the dominant hand, and the fourth finger is most commonly affected. Although trigger finger has traditionally been caused by aseptic inflammation resulting from repetitive movements, a recent study suggested the lack of inflammation-related agents at the site [2]. Given its higher incidence in patients with diabetes, microvascular complications may have an important role in the pathology of trigger finger [12]. However, the exact etiology is not completely known.

Several systematic reviews and meta-analyses have focused on trigger finger [4-9]. However, these reviews have certain limitations. First, all these reviews have mainly analyzed the efficacy of acupotomy rather than other acupuncture modalities and related interventions. Thus, while acupotomy has been thoroughly investigated, the efficacy of other methods remains unknown. Second, most of them lacked subgroup analyses, whereas some had limitations including misrepresented forest plots and omitted ROB summary or graph. Reviewing acupotomy or acupotomy add-on treatment without comparator subgroup analysis can verify the overall efficacy of acupotomy compared with other conventional treatments. However, this cannot convey the superiority of acupotomy to other conventional treatments; thus, further reviews that include subgroup analyses are needed. Moreover, all the aforementioned reviews have included studies published up to 2020. Our review offers an updated overview, including five trials published in 2022.

A total of 19 studies were included in this systematic review and meta-analysis. The experimental group consisted of various types of acupuncture and related interventions. Arc-edge acupuncture is a new acupuncture technique invented by Dr. Wang [33] from China, which involves the use of a needle with a hollow body and V-shaped arc-edged structure at the distal end, combining aspects of traditional acupotomy, injection needle, and acupuncture. Results indicated a significant difference in the ER, QG, and lowering pain intensity as measured by the NRS. Since arc-edge acupuncture is a relatively new addition to the existing armamentarium of available treatment modalities, our analysis included only one study. Further studies are needed to verify the efficacy of this intervention. On the contrary, acupuncture demonstrated an overall significant efficacy in lowering pain intensity. However, caution must be exercised with this result, as only one index was included because it was the common factor that facilitated the comparison of two studies. In addition, the measured VAS was only a short-term outcome.

The outcome of acupotomy varied between subgroups. The validity of ultrasound-guided acupotomy concerning the efficacy in reducing pain intensity and low RR was evident; however, its efficacy as regards ER compared with block therapy could not be effectively proven. However, further analysis of the increasing number of randomized trials currently being conducted in China could provide better insights into the matter. Acupotomy showed significant effects on the ER, QG, and low RR, but no significant effect on lowering pain intensity (VAS) compared with block therapy. In comparison with surgery, acupotomy showed a significant effect on the ER, VAS, and QG. Overall, the pooled result showed that acupotomy has a significant effect on the ER, VAS, QG, and RR.

The acupotomy add-on treatment also showed significant results with regard to the ER, QG, RR, and pain intensity. The combination of ultrasound-guided acupotomy and ultrasound-guided block therapy demonstrated significant efficacy in QG and lowering pain intensity and was associated with a low RR; however, the ER is not significant. The combination of acupotomy and block therapy demonstrated a significant effect on the ER, but no significant effect on pain intensity. The combination of acupotomy plus antibiotics was efficacious in the ER and pain intensity but was insignificant in RR. Acupotomy in combination with NSAIDs and ice pack displayed a significant result in lowering pain intensity and a low RR but was insignificant in the ER. However, regarding pain intensity, heterogeneity needs to be considered carefully. Sensitivity analysis revealed that the studies by Jia et al. [20] and Yang et al. [28] may be influential in contributing to the heterogeneity. Nevertheless, the analysis also indicated that general studies could be substantial contributors, which could be attributed to various subgroups with lesser investigations. Furthermore, the overall results for acupotomy demonstrated a notable significant effect of the add-on treatment compared with acupotomy alone.

Like with many diseases, the initial approach to treating trigger finger includes conservative measures before considering surgical treatment. Few conventional measures considered for treating trigger finger include immobilization using orthoses, corticosteriod injections, surgical division of the A1 pulley [34], medications (typically NSAIDs), and extracorporeal shock wave therapy. However, these treatment modalities are associated with limitations, AEs, recurrence or recalcitrance (often necessitating additional interventions), suture-related problems, and symptoms regarding steroid injections [35-37]. The minimally invasive treatment options, acupotomy and acupotomy add-on treatments, have demonstrated low RR than the control groups. In addition, AEs associated with acupotomy were reported to be much lesser than those due to surgery. Unfortunately, data on RR and AEs caused by acupotomy performed with other types of interventions are lacking, emphasizing the need for further clinical trials in this area.

Traditional Korean medicine has been used to treat trigger finger, in conjunction with acupuncture and related interventions such as electroacupunture, acupotomy, warm acupuncture, triangular needle, and other new treatment modalities being developed continuously [38]. In particular, acupotomy is an actively developing treatment option with variations being introduced. Many clinical trials are comparing the efficacies of these treatment modalities. Furthermore, the use of sonograms for minimizing AEs has also been explored. However, due to existing inconsistencies in the study design and the associated quality issues such as blinding defects, variety of treatment methods, and lack of description of patient inclusion/exclusion criteria, more studies could not be included in this review, leading to enhanced heterogeneity and inconsistent results. Despite these limitations, this review suggests that traditional Korean medicine may be effective for treating trigger finger, and the addition of conventional Western medicine may enhance its effectiveness. Therefore, based on these findings, clinicians can adjust their approach to encourage patients to receive the best treatment for pain relief and restoration of function.

This review sought to explore the effectiveness of acupuncture and related interventions for treating trigger finger, has provided a detailed description of the interventions, divided them into subgroups, and conducted heterogeneity, funnel plot, and sensitivity analyses where necessary. However, some limitations exist. First, some defects exist in the current study design, especially blinding, warranting a cautious interpretation of the results of this review. Second, the number of studies selected for certain interventions was small. Furthermore, long-term effects could not be confirmed in studies where the RR was not reported, as the follow-up period was not subdivided. In addition, most of the included studies mentioned only one treatment session, which is very different from the normal clinical practice of repeated treatments, and provides limited information on complications or effects of repeated treatment. Third, efforts were made to overcome heterogeneity through various methods; however, it could not be resolved in some cases. Finally, most studies, except for one, were conducted in China, which limits national diversity. Therefore, further high-quality RCTs that supplement research design and methodology are needed. A review that synthesizes the results for each treatment method is also necessary.

Notwithstanding some limitations, the results of this review indicate that acupuncture and related interventions can effectively reduce pain and alleviate symptoms associated with trigger finger. Furthermore, acupotomy add-on treatment might be more effective than acupotomy alone. From the perspective of minimally invasive treatments, acupuncture and related interventions hold promise as potential treatment options with low AEs and RR.

Supplementary data is available at https://doi.org/10.13045/jar.2023.00066.

Conceptualization: SHL. Data curation: HWH, KHK. Formal analysis: HWH, MIJ. Investigation: MIJ, HIJ. Methodology: JWP. Project administration: HWH. Resources: SWC. Software: JSH. Supervision: SHL. Validation: JWP. Writing – original draft: HWH. Writing – review & editing: HWH, KHK.

The authors have no conflicts of interest to declare.

None.

This research did not involve any human or animal experiment.