Comparative efficacy and safety of acupuncture and moxibustion in the treatment of postprostatectomy urinary incontinence: A systematic review and network meta-analysis

Abstract

Background:

To compare the efficacy and safety of different acupuncture and/or moxibustion methods in the treatment of postprostatectomy urinary incontinence using frequency-based network meta-analysis (NMA) method.

Methods:

Randomized controlled trials (RCTs) on acupuncture and/or moxibustion for postprostatectomy urinary incontinence published from database inception to September 1, 2024, were identified through computerized searches of PubMed, Embase, Cochrane Library, CNKI, SinoMed, Wanfang Data, and VIP databases. Two reviewers independently screened literature, extracted data, and assessed the risk of bias of included studies. NMA was conducted using Stata 16.0 software.

Results:

A total of 17 RCTs involving 9 methods and 1286 patients were included. The NMA showed that needling-warm moxibustion was the most effective in improving the total effective rate, gentle moxibustion (GM) was the most effective in reducing International Consultation on Incontinence Questionnaire-Short Form, and standard acupuncture was the most effective in improving Incontinence Quality of Life Questionnaire. Electroacupuncture had the highest incidence of adverse events. Considering International Consultation on Incontinence Questionnaire-Short Form and Incontinence Quality of Life Questionnaire, GM may be the most ideal choice.

Conclusion:

Based on NMA, GM, needling-warm moxibustion, and standard acupuncture were found to have the highest probability of being the best therapies. Due to the limitations of this study, these results should be confirmed by detailed RCTs.

1. Introduction

Prostate cancer has the second highest incidence among all malignant tumors in men worldwide.^[1] With the advancement of modern medicine, the postoperative survival period of prostate cancer patients has gradually extended, and the quality of life of patients has attracted increasing attention. Postprostatectomy urinary incontinence refers to the uncontrolled leakage of urine from the urethra after radical prostatectomy and the use of one or more pads.^[2] Due to differences in surgical techniques, the incidence of this condition varies widely, with reported rates ranging from around 3% to 55.9%.^[3,4] The etiology of this condition includes structural or functional abnormalities of the urethral sphincter, bladder dysfunction, and urethral stricture.^[5,6] As a serious complication, postprostatectomy urinary incontinence has not been well understood or effectively addressed.

Currently, there are surgical and nonsurgical treatments for postprostatectomy urinary incontinence.^[7,8] Nonsurgical treatments mainly include physical therapy and drug therapy, but drug treatment with duloxetine may be limited in clinical practice due to indications and adverse reactions. For patients with mild to moderate symptoms, simpler options such as pelvic floor muscle training (PFMT), routine nursing care (PC), and biofeedback (BF) can be considered. For patients with severe symptoms, surgical treatments such as artificial urethral sphincter implantation and sling procedures are the last resort. However, these methods also carry risks of perioperative and long-term complications, and the availability of these surgeries may be limited. Therefore, alternative therapies are crucial.

Acupuncture and/or moxibustion, as traditional Chinese medicine therapies, are believed to effectively stimulate the nerve innervating the bladder, promoting the recovery of bladder control function and playing an important role in postoperative rehabilitation.^[9,10] However, there are currently various acupuncture techniques, each with different levels of complexity, safety, and efficacy. The selection of appropriate acupuncture methods to improve treatment outcomes is an urgent issue in clinical practice. In comparison with traditional meta-analyses, network meta-analysis (NMA) allows for the comparison of different treatment methods, quantitative analysis, and assessment and ranking of their therapeutic effects. Therefore, based on the frequency-based framework, this study conducted a NMA to compare the efficacy and safety of different acupuncture methods, and rank their advantages and disadvantages, providing a reference for clinical selection of acupuncture methods.

2. Materials and methods

This study has been registered with the PROSPERO platform, the registration number is CRD42024597744.

This study involved only meta-analysis. It was confirmed by the Institutional Review Board of Xiyuan Hospital, China Academy of Traditional Chinese Medicine, and no ethical approval was required.

2.1. Inclusion criteria

2.1.1. Types of studies

Randomized controlled trial (RCT) research, limited to studies in Chinese and English, with no restrictions on blinding methods and publication platforms.

2.1.2. Types of participants

Meeting the diagnostic criteria for moderate to severe stress urinary incontinence; postradical prostatectomy for prostate cancer; aged 50 to 85 years; voluntarily participating in this trial and signing an informed consent form.

2.1.3. Types of interventions

The experimental group used acupuncture or moxibustion, without restrictions on the specific techniques used. In the moxibustion intervention, only moxibustion was used without adding Chinese herbal medicine. The control group could receive acupuncture and/or moxibustion (different from the intervention group), or other conventional treatments such as pelvic floor muscle function training, standard care, erbium laser (EL), or biofeedback.

2.1.4. Types of outcomes

The observation indicators include the overall clinical effective rate and incontinence questionnaireshort form (ICI-Q-SF), incontinence quality of life questionnaire (I-QOL) scores. The safety indicator is the occurrence of adverse reactions during the study observation process. The overall effective rate is calculated as (significant improvement + improvement)/ total number of cases × 100%.

2.2. Excluded criteria

Those with mental illness; those with severe heart, liver, or kidney dysfunction; those with communication barriers who cannot cooperate with the study; those with neurological, urological, or other diseases that may affect bladder function, including urinary tract infections, stroke, and spinal cord injury that may cause neurogenic urinary incontinence; those with organic tumor diseases who have not undergone surgical treatment; studies with missing data, inability to access the full text, duplicate publications, or data errors; experimental group sample size <20 cases.

2.3. Search strategy

A comprehensive search was conducted across multiple databases, including Web of Science, PubMed, Cochrane Library, CNKI, Wanfang Data, and VIP Database. The search period extended from the establishment of each database up to September 1, 2024, encompassing both Chinese and English languages. A combination of MeSH terms and free-text terms was employed to tailor search strategies specific to each database. The English search terms included “prostate cancer,” “urinary incontinence,” “acupuncture,” “moxibustion,” and “RCT.” Corresponding Chinese terms included “ prostate cancer “, “ postoperative “, “ urinary incontinence “, “ acupuncture “, “ moxibustion “, “ randomized controlled “, “ trial “, “clinica,” and “ efficacy “. Additionally, the references of included studies were manually checked to identify further relevant research. For instance, the specific search strategy for PubMed was as follows:

#1 (prostate cancer [MeSHT Terms]) OR (prostate cancer [Title/Abstract])

#2 (urinary incontinence [MeSHT Terms]) OR (urinary incontinence [Title/Abstract])

#3 (acupuncture [MeSHT Terms]) OR (moxibustion [MeSHT Terms]) OR (acupuncture [Title/Abstract]) OR (moxibustion [Title/Abstract])

#4 (randomized controlled trial [Publication Type]) OR (controlled clinical trial [Publication Type])

OR (randomization [Title/Abstract]) OR (randomized [Title/Abstract]) OR (RCT [Title/Abstract]) OR (trial [Title/Abstract])

#5 (“1000/1/1”[Date-Publication]”:2024/9/1”[Date-Publication])

#6 #1 AND #2 AND #3 AND #4 AND #5

2.4. Study selection and data extraction

The search results were uploaded to NoteExpress, where duplicates were removed using the software’s deduplication feature. Two researchers independently screened the literature and extracted data, performing cross-verification to ensure accuracy. In cases of disagreement, a third researcher was consulted to reach a consensus. Data extraction from the screened literature was conducted using Excel 2021. The extracted data included publication date, study title, first author, sample size, mean age, interventions, and outcome measures. If data were missing or unclear, the authors were contacted to obtain the necessary information.

2.5. Quality evaluation

The risk of bias for the included studies was assessed using the Cochrane Risk of Bias Tool provided by RevMan 5.4 software. This assessment covered 7 domains: random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, selective reporting, and other sources of bias. Each domain was evaluated and categorized as “low risk,” “unclear risk,” or “high risk.” In cases of disagreement between the 2 primary assessors, discussions were held, and if necessary, a third researcher was consulted to resolve the discrepancies. This collaborative approach ensured a rigorous and unbiased assessment of the included studies.^[11]

2.6. Statistical analysis

For categorical variables, odds ratios (OR) were used, and for continuous variables, the effect size was expressed as the SMD along with its 95% confidence interval (CI). A traditional meta-analysis was conducted using RevMan 5.4 software. Heterogeneity among studies was assessed using the chi-square test and the I² statistic. If there was no significant heterogeneity between studies (P < .05, I² < 50%), a fixed-effect model was employed; otherwise, a random-effects model was used. If heterogeneity was too substantial, a descriptive analysis was performed.

For the NMA, conducted within a frequentist framework, Stata 16.0 software with the network and mvmeta packages was utilized. This approach involved creating a visual evidence network map for each outcome and intervention, calculating the surface under the cumulative ranking curve (SUCRA) to rank the efficacy of interventions across outcomes. A “comparison-adjusted” funnel plot was used to detect small-study effects or publication bias. In cases where there were closed loops in the network, inconsistency tests were performed to evaluate the agreement between direct and indirect comparisons.^[12]

3. Results

3.1. Selection and identification of studies

A total of 557 citations were identified, including 125 duplicate studies. After excluding 112 studies that did not meet the predefined inclusion criteria based on title and abstract review, 320 full-text articles were retrieved. We screened these 320 full-text articles, and 303 were excluded due to inappropriate study design, intervention, population, or outcomes. Ultimately, 17 two-arm RCTs^[13–29] were included in the analysis. All included studies were conducted in China and were published between 2010 and 2024. The detailed screening process is presented in Figure 1.

Figure 1.

Flow diagram of the search for eligible studies.

3.2. Included study characteristics

A total of 17 trials^[13–29] involving 1286 patients were included in the study, all published between 2010 and 2024 (Table 1). The interventions included 5 types of acupuncture and/or moxibustion therapies, namely electroacupuncture (EA),^[13–18] standard acupuncture (SA),^[19–22] auricular acupuncture (AA),^[23] needling-warm moxibustion (NWM),^[24,25] and gentle moxibustion (GM).^[26–29] Four types of comparators were used: PFMT, usual care (UC), EL, and biofeedback therapy (BF). Further details are provided in Table 1.

Table 1.

Basic characteristics of the included studies.

Study ID	N	Age	Treatment	Primary acupoints	Retention (min)	Course	Outcomes
Chu 2023[13]	34		EA	BL35, bilateral sacrococcygeal joints	60	20 wk, TIW	*,†,‡
	31		EL			20 wk, 4w
Huang 2023[14]	51	65.53 ± 4.89	EA	SP6, ST36, RN4, KI1, BL23, and S2–S4	30	52 wk, BID	*,§
	53	64.81 ± 4.90	PFMT			52 wk, TID
Feng 2021[15]	60	68.5 ± 6.5	EA	Bilateral sacrococcygeal joints	60	8 wk, TIW	*,†,§
	29	67.0 ± 6.7	PFMT		40	8 wk, TIW
Yang 2020[16]	41	68.76 ± 10.13	EA	Bilateral sacrococcygeal joints	50	12 wk, TIW	*,†,‡
	40	69.05 ± 11.26	BF		20	12 wk, TIW
Hu 2020[17]	64	69.09 ± 6.17	EA	BL35, bilateral sacrococcygeal joints	60	8 wk, TIW	*,†
	32	67.03 ± 6.67	PFMT			8 wk, TIW
Yang 2016[18]	70	62.7 ± 12.3	EA	RN3, RN4, DU2, ST36, KI13, SP6	30	12 wk, TID	*
	70	63.1 ± 11.7	PFMT			12 wk, TID
Song 2020[19]	28	71.75 ± 7.93	SA	RN2, RN4, RN6, SP9, ST6, SP6, LR3	30	12 wk, TIW	*,†,‡
	27	72.19 ± 5.81	PFMT			12 wk, TID
Li 2020[20]	20	74.65 ± 5.35	SA	RN2, RN4, RN6, SP9, ST6, SP6, LR3	30	12 wk, TIW	*,†,‡
	20	73.50 ± 5.05	PFMT			12 wk, TID
Xia 2019[21]	20	69.60 ± 8.51	SA	RN2, RN4, RN6, SP9, ST6, SP6, LR3	30	12 wk, TIW	*,†
	20	71.65 ± 5.60	PFMT			12 wk, TID
Huang 2019[22]	20	69.60 ± 8.51	SA	RN2, RN4, RN6, SP9, ST6, SP6, LR3	30	12 wk, TIW	‡
	20	71.65 ± 5.60	PFMT			12 wk, TID
Azevedo 2023[23]	30	64.2 ± 5.8	AA	Ear pinna	15	8 wk, QIW	†
	30	64 ± 7.35	PFMT			8 wk, TID
Li 2017[24]	39	51.8 ± 7.8	NWM	RN4, RN3, RN6, ST25, ST36, SP9, SP6, LR3	10	26 wk, QD	*
	39	50.8 ± 6.9	UC		2	26 wk, TID
Yang 2012[25]	56	67.5 ± 11.2	NWM	RN4, RN3, LI11, LI4, PC6, GB34, ST36, SP6, DU20	30	3 wk, TIW	*
	42	67.5 ± 11.2	SA		30	3 wk, TIW
Lu 2021[26]	40	65 ± 9	GM	RN8, RN4, RN6, BL23	40	8 wk, TIW	*,†,‡
	40	65 ± 8	PFMT		20	8 wk, TID
Zhou 2020[27]	30	73.40 ± 9.23	GM	RN8, RN4, RN6, RN3, RN12, RN10	120	12 wk, TIW	*,†,‡,§
	30	73.20 ± 9.03	PFMT			12 wk, TID
Cheng 2019[28]	31	67.94 ± 6.28	GM	RN8, RN4, RN6, RN3	30	12 wk, QD	†
	31	68.58 ± 5.99	PFMT			12 wk, TID
Cheng 2010[29]	56		GM	RN3, RN4, LI11, LI4, PC6, ST36, GB34, SP6, DU20	30	3 wk, QD	*
	42		SA		15	3 wk, QD

AA = auricular acupuncture, BF = biofeedback, BID = bis in die, EA = electroacupuncture, EL = Erbium laser, GM = gentle moxibustion, NWM = needle-warming moxibustion, PFMT = pelvic floor muscle training, QD = quaque die, QIW = quarter in week, SA = simple acupuncture, TID = ter in die, TIW = three times a week, UC = usual care (comprehensive treatments including life care, health education, skin care, and so on).

^*Total efficiency ratio.

^†International consultation on incontinence questionnaireshort form.

^‡Incontinence quality of life questionnaire.

^§Adverse events.

The treatment duration ranged from 3 to 52 weeks. Among these RCTs, the most frequently used interventions were EA (6 trials, 575 patients), SA (4 trials, 175 patients), AA (1 trial, 387 patients), NWM (2 trials, 176 patients), and GM (4 trials, 300 patients). The characteristics of the 17 eligible RCTs are summarized in Table 1.

3.3. Methodological quality assessment

Two researchers separately evaluated the risk of bias of the included studies using the Cochrane Risk of Bias tool presented in the Cochrane Handbook 5.1. The results of the assessment items were as follows:Low-risk items: 3 studies in selective bias (interpretation using a random number table) and 1 study on reporting bias (description using the shakedown method); all studies in attrition bias (complete reporting of outcome data); all studies in other bias (baseline of RCTs described). High-risk events: all studies in selection bias (allocation concealment not used); all studies in performance bias (blinding of participants and personnel not used); all studies in detection bias (blinding of outcome assessments not reported). Risk items were unclear: 14 studies in selection bias (specific randomization method not stated); 16 studies in reporting bias (unclear selective reporting). Figures 2 and 3 display the risk of bias for each trial. The key limitation may be the implementation bias and measurement bias of the study, as none of the 17 included studies explicitly blinded the trial staff and outcome assessors. Among the 17 trials, 15 studies (88%) had a low risk of random sequence generation bias, 17 studies (100%) had a low risk of deviating from expected intervention measures, 17 studies (100%) had a low risk of missing outcome data,17 studies (100%) did not mention blinding the trial staff, and the risk of bias in these studies is unclear, and 17 studies (100%) had unclear risk of outcome measurement. Overall, 2 studies (13%) had 2 unclear risks of bias. However, there is no strong statistical evidence reporting global inconsistencies in most outcomes. Furthermore, we did not find any evidence of asymmetry in the funnel plot.

Figure 2.

Risk of bias graph.

Figure 3.

Risk of bias summary.

3.4. Outcomes

3.4.1. Total efficiency ratio

All 14 pieces^{[13–21,24–27,29]} of research including 4 interventions and 5 comparators recorded the total efficiency ratio of urinary incontinence rehabilitation of after radical prostatectomy (Figs. 4A and 5A).

Figure 4.

Network graph of outcomes. The blue node sizes represent the total sample sizes for treatments. Line thicknesses are associated with the number of trial comparisons. (A) Total efficiency ratio; (B) international consultation on incontinence questionnaireshort form; (C) incontinence quality of life questionnaire.

Figure 5.

Forest graphs of meta-analysis. (A) Total efficiency ratio; (B) international consultation on incontinence questionnaireshort form; (C) incontinence quality of life questionnaire.

Comparied with PFMT alone, EA (MD: 3.39, 95% CI: 1.89–6.08, low certainty), SA (MD: 3.79, 95% CI: 1.68–8.58, very low certainty), NWM (MD: 24.11, 95% CI: 3.62–160.62, low certainty), GM (MD: 5.21, 95% CI: 1.98–13.73, very low certainty) were benifical to rise total efficiency ratio (Table 2). Comparied with PFMT alone, there was no significant difference in UC (MD: 6.70, 95% CI: 0.59–75.74, low certainty), EL (MD: 0.50, 95% CI: 0.12–2.04, very low certainty), BF (MD: 0.75, 95% CI: 0.21–2.62, low certainty) in the treatment of urinary incontinence rehabilitation of after radical prostatectomy (Table 2).

Table 2.

Results of the network meta-analysis of the total efficiency ratio.

PFMT
3.39 (1.89,6.08)	EA
3.79 (1.68,8.58)	1.12 (0.41, 3.07)	SA
24.11 (3.62,160.62)	7.11 (0.97, 51.88)	6.35 (1.15, 35.19)	NWM
5.21 (1.98,13.73)	1.54 (0.48, 4.87)	1.37 (0.46, 4.10)	0.22 (0.03, 1.65)	GM
6.70 (0.59,75.74)	1.97 (0.16, 24.00)	1.76 (0.18, 17.33)	0.28 (0.06, 1.26)	1.29 (0.10, 16.18)	UC
0.50 (0.12,2.04)	0.15 (0.04, 0.53)	0.13 (0.03, 0.67)	0.02 (0.00, 0.22)	0.10 (0.02, 0.54)	0.08 (0.00, 1.24)	EL
0.75 (0.21,2.62)	0.22 (0.07, 0.67)	0.20 (0.04, 0.88)	0.03 (0.00, 0.30)	0.14 (0.03, 0.71)	0.11 (0.01, 1.71)	1.48 (0.27, 8.01)	BF

AA = auricular acupuncture, BF = biofeedback, EA = electroacupuncture, EL = Erbium laser, GM = gentle moxibustion, NWM = needle-warming moxibustion, PFMT = pelvic floor muscle training, SA = simple acupuncture, UC = usual care (comprehensive treatments including life care, health education, skin care, and so on).

Based on the SUCRA values, NWM had the highest probability for rise total efficiency ratio (SUCRA: 97.6%), followed by GM (SUCRA: 71.0%), and UC (SUCRA: 68.8%) (Fig. 6A and Table 5).

Figure 6.

Plots of the surface under the cumulative ranking curves for all treatments (A) Total efficiency ratio; (B) international consultation on incontinence questionnaireshort form; (C) incontinence quality of life questionnaire.

Table 5.

Surface under the cumulative ranking curve and ranking probability of different Acupuncture and/or moxibustion therapies on each outcome.

Treatment	Efficacy		ICI-Q-SF		I-QOL
	SUCRA (%)	Rank	SUCRA (%)	Rank	SUCRA (%)	Rank
PFMT	22.8	6	27.1	5	30.5	6
EA	56.9	5	86.0	2	69.6	1
SA	59.7	4	63.7	3	61.9	2
AA	–	–	50.6	4	–	–
NWM	97.6	1	–	–	–	–
GM	71.0	2	93.4	1	59.2	3
UC	68.8	3	–	–	–	–
EL	8.1	8	19.0	6	34.8	5
BF	15.3	7	10.1	7	44.0	4

The top three items are bolded. The top three items are italicized.

AA = auricular acupuncture, BF = biofeedback, EA = electroacupuncture, EL = Erbium laser, GM = gentle moxibustion, I-QOL = incontinence quality of life questionnaire, ICI-Q-SF = international consultation on incontinence questionnaire short form, NWM = needle-warming moxibustion, PFMT = pelvic floor muscle training, SA = simple acupuncture, SUCRA = surface under the cumulative ranking curve, UC = usual care (comprehensive treatments including life care, health education, skin care, and so on).

3.4.2. International consultation on incontinence questionnaireshort form

All 11 pieces^{[13,15–17,19–21,23,26–28]} of research including 4 interventions and 3 comparators recorded the international consultation on incontinence questionnaireshort form of urinary incontinence rehabilitation of after radical prostatectomy (Figs. 4B and 5B).

Comparied with PFMT alone, EA (MD: −3.25, 95% CI: −4.55 to −1.94, low certainty), SA (MD: −2.02, 95% CI: −3.55 to −0.49, very low certainty), GM (MD: −3.69, 95% CI: −4.99 to −2.39, very low certainty) were benifical to rise total efficiency ratio (Table 2). Comparied with PFMT alone, there was no significant difference in AA (MD: −1.23, 95% CI: −3.11 to 0.65, low certainty), EL (MD: 0.67, 95% CI: −2.25 to 3.60, very low certainty), BF (MD: 1.26, 95% CI: −1.28 to 3.81, low certainty) in the treatment of urinary incontinence rehabilitation of after radical prostatectomy (Table 3).

Table 3.

Results of the network meta-analysis of the ICI-Q-SF.

PFMT
−3.25 (−4.55, −1.94)	EA
−2.02 (−3.55, −0.49)	1.23 (−0.78, 3.24)	SA
−1.23 (−3.11, 0.65)	2.02 (−0.28, 4.31)	0.79 (−1.64, 3.21)	AA
−3.69 (−4.99, −2.39)	−0.45 (−2.29, 1.39)	−1.67 (−3.68, 0.33)	−2.46 (−4.75, −0.17)	GM
0.67 (−2.25, 3.60)	3.92 (1.30, 6.54)	2.69 (−0.61, 5.99)	1.90 (−1.58, 5.38)	4.37 (1.16, 7.57)	EL
1.26 (−1.28, 3.81)	4.51 (2.33, 6.69)	3.28 (0.31, 6.25)	2.49 (−0.67, 5.66)	4.96 (2.10, 7.81)	0.59 (−2.82, 4.00)	BF

AA = auricular acupuncture, BF = biofeedback, EA = electroacupuncture, EL = Erbium laser, GM = gentle moxibustion, ICI-Q-SF = international consultation on incontinence questionnaire short form, NWM = needle-warming moxibustion, PFMT = pelvic floor muscle training, SA = simple acupuncture, UC = usual care (comprehensive treatments including life care, health education, skin care, and so on).

According to the SUCRA value, GM has the highest probability of reducing ICI-Q-SF (SUCRA: 93.4%), followed by EA (SUCRA: 86.0%), and SA (SUCRA: 63.4%) (Fig. 6B and Table 5).

3.4.3. Incontinence quality of life questionnaire

All 7 pieces^{[13,16,19,20,22,26,27]} of research including 3 interventions and 3 comparators recorded the incontinence quality of life questionnaire of urinary incontinence rehabilitation of after radical prostatectomy (Figs. 4C and 5C).

Comparied with PFMT alone, SA (MD: 8.66, 95% CI: 5.21–12.12, very low certainty), GM (MD: 8.29, 95% CI: 4.91–11.67, very low certainty) were benifical to rise total efficiency ratio (Table 2). Comparied with PFMT alone, there was no significant difference in EA (MD: 9.67, 95% CI: −3640.37 to 3659.71, low certainty), EL (MD: −11.75, 95% CI: −3661.80 to 3638.30, very low certainty), BF (MD: −9.57, 95% CI: −3659.61 to 3640.47, low certainty) in the treatment of urinary incontinence rehabilitation of after radical prostatectomy (Table 4).

Table 4.

Results of the network meta-analysis of the I-QOL.

PFMT
9.67 (−3640.37, 3659.71)	EA
8.66 (5.21, 12.12)	−1.00 (−3651.05, 3649.04)	SA
8.29 (4.91, 11.67)	−1.38 (−3651.42, 3648.67)	−0.37 (−5.30, 4.56)	GM
−11.75 (−3661.80, 3638.30)	−21.42 (−27.77, −15.07)	−20.42 (−3670.47, 3629.63)	−20.04 (−3670.09, 3630.00)	EL
−9.57 (−3659.61, 3640.47)	−19.24 (−22.80, −15.68)	−18.24 (−3668.28, 3631.81)	−17.86 (−3667.91, 3632.18)	2.18 (−5.09, 9.45)	BF

AA = auricular acupuncture, BF = biofeedback, EA = electroacupuncture, EL = Erbium laser, GM = gentle moxibustion, I-QOL = incontinence quality of life questionnaire, NWM = needle-warming moxibustion, PFMT = pelvic floor muscle training, SA = simple acupuncture, UC = usual care (comprehensive treatments including life care, health education, skin care, and so on).

According to the SUCRA values, SA had the highest probability for rise I-QOL (SUCRA: 61.9%), followed by GM (SUCRA: 59.2%) (Fig. 6C and Table 5).

3.4.4. Adverse reactions

Three studies^[14,15,27] reported adverse events in 9 patients, of which 6 were related to acupuncture treatment. Specific adverse events included local skin redness (2 cases) and premature treatment cessation (4 cases). Four patients withdrew from the study due to adverse reactions, and the score from their last treatment session was used as the posttreatment score for statistical analysis. We found that the electroacupuncture (EA) method had the highest probability of adverse events occurring.

3.5. Cluster analysis

The influences of interventions in 2 diverse results were synthetically contrasted by cluster analysis. Two sets of cluster analyses were conducted in our study, containing the ICI-Q-SF and I-QOL of urinary incontinence rehabilitation of after radical prostatectomy, respectively. The results are shown in Figure 7. Through synthetical analysis by means of cluster analysis, EA has more advantages in improving ICI-Q-SF, GM has more advantages in improving I-QOL. Consider the combination of effectiveness and safety, GM might possess the good therapeutic results and high safety.

Figure 7.

Cluster analysis plots. Interventions located in the upper right corner indicate optimal combination therapy for 2 different outcomes.

3.6. Inconsistency test

Inconsistency testing is primarily used to assess the degree of consistency between direct and indirect comparisons of results, particularly when there are closed loops present. Various interventions showed direct comparisons in terms of overall efficacy, and upon identifying a closed loop in the NMA, we conducted an inconsistency test and found no evidence of inconsistency. Consequently, we utilized a consistency model for further analysis.

3.7. Publication bias

Figure 8 presents funnel plots for the 3 primary outcomes to assess publication bias. Visually, none of the funnel plots are perfectly symmetrical, and each adjusted reference line is not perpendicular to the central line. Therefore, there may be significant publication bias present.

Figure 8.

Funnel plots. (A) Total efficiency ratio; (B) international consultation on incontinence questionnaireshort form; (C) incontinence quality of life questionnaire.

4. Discussion

4.1. Summary of main findings

This study systematically evaluated the efficacy of 9 common methods (EA, SA, AA, NWM, GM, PFMT, UC, EL, BF) for treating postprostatectomy urinary incontinence using data from 17 relevant studies through NMA. The results indicated that most acupuncture therapies outperformed control treatments across all outcomes, with statistically significant differences observed.

Considering statistical variability and SUCRA rankings, NWM was most likely to be the optimal treatment regarding overall efficacy; GM showed the highest potential for reducing ICI-Q-SF scores; SA was most likely to enhance I-QOL scores; however, EA was associated with the highest incidence of adverse events.

Overall, integrating ICI-Q-SF and I-QOL scores suggests that GM is the best treatment option for postprostatectomy urinary incontinence. Therefore, while GM’s efficacy warrants attention, clinicians should select appropriate methods based on individual patient circumstances.

4.2. Research significance and importance

During radical prostatectomy for prostate cancer, the proximal urethral sphincter may be excised, leading to local scar adhesions, ischemia, and potential damage to the pudendal nerve and bladder sphincter, resulting in bladder instability and reduced compliance, which adversely affects urination. Patients in this cohort are often older with multiple comorbidities, and urinary incontinence can persist even 1 year postsurgery, impacting their quality of life and causing a loss of self-esteem due to the prolonged recovery of neurological and urinary functions.^[30,31]

Studies indicate that low-frequency continuous wave stimulation can effectively stimulate the pelvic floor muscles, restoring muscle function and alleviating incontinence symptoms.^[32] Additionally, 2.0 Hz low-frequency pulse electrical stimulation can excite the pudendal nerve, regulating the function of its target organ systems and improving coordination of the urethral sphincter.^[33] Regular, continuous electrical acupuncture stimulation of the pudendal nerve may also strengthen the pelvic floor muscles through rhythmic contractions, enhancing muscle strength and improving leakage symptoms.^[34] In this study, EA demonstrated notable clinical effectiveness, with ICI-Q-SF outcomes ranking just below GM; however, although EA showed superior performance in I-QOL scores, the presence of a lower bound of the CI suggests it should not be prioritized.

SA is a classic acupuncture method known for its significant efficacy and ease of application, characterized by its green and safe nature. In treating urinary incontinence, it primarily targets acupuncture points along the Ren, Du, bladder, kidney, spleen, lung, heart, and liver meridians.^[35] In this study, SA demonstrated effectiveness on the ICI-Q-SF scale, ranking just below GM and electroacupuncture (EA). Additionally, SA outperformed other treatment methods on the I-QOL scale, making it the most effective intervention.

AA is closely related to the visceral meridians, with corresponding reflex zones for various organ systems located on the auricle. When visceral tissues are damaged or when a specific organ undergoes pathological changes, corresponding local responses may occur on the auricle. Based on traditional Chinese medicine principles, points such as those related to the bladder and small intestine can be stimulated to facilitate the drainage of fluids and regulate water metabolism. Due to its minimally invasive nature and low side effects, AA is generally simpler to perform than SA, leading to its widespread clinical application.^[36] In this study, AA was only compared on the ICI-Q-SF scale, ranking fourth in clinical effectiveness. However, due to the limited number of AA studies included – only one – the specific efficacy of AA may be subject to bias.

Acupuncture can promote yang, invigorate qi, and facilitate the flow of fluids, while mugwort, known for its pure yang properties, traverses the 12 meridians to regulate qi and blood, dispelling cold and dampness. Warm needling moxibustion complements acupuncture, contributing to smoother urination.^[25,37] In this study, AA showed the best overall efficacy; however, due to a lack of comparative data on the ICI-Q-SF and I-QOL scales, further in-depth analysis was limited.

GM is simple to perform, safe, and effective. It is particularly beneficial for individuals with deficiency, as moxibustion enhances warmth, strengthens the kidneys, and helps reduce incontinence. Modern research indicates that moxibustion can improve blood circulation and metabolism in the pelvic floor muscles of patients with urinary incontinence, enhance tissue nutrition, increase muscle excitability, and strengthen pelvic floor muscle strength, thereby alleviating overactive bladder symptoms.^[38] In this study, GM ranked second in clinical efficacy behind NWM for overall effectiveness; it achieved the best results on the ICI-Q-SF scale and was second only to SA on the I-QOL scale.

PFMT, UC, and biofeedback (BF) are common rehabilitation treatments in modern medicine, while EL, recognized for its efficacy, is often used for stress urinary incontinence and is gradually being explored for postprostatectomy urinary incontinence.^[39,40] This study also included this method. The results indicate that, regarding overall efficacy, UC ranks just below NWM and GM; however, in other outcome measures, all comparator methods were less effective than acupuncture and/or moxibustion therapies.

4.3. Strengths and limitations

This study aims to explore the most effective and safest acupuncture therapies for treating postprostatectomy urinary incontinence, which may facilitate rational treatment decisions in the absence of a gold standard treatment and amidst significant healthcare burdens. Additionally, the acupuncture methods in this study were named according to the international standard terminology for traditional Chinese medicine released by the World Health Organization in 2022.^[41] Furthermore, to limit the impact of other interventions on the outcomes, this study excluded any combination of acupuncture and/or moxibustion with herbal treatments.

However, there are several limitations to this study. First, the number of databases searched was limited, resulting in the inclusion of only 17 trials and 9 treatment methods, with most comparisons relying on 1 to 2 studies, which may influence the results. Second, the lack of direct comparisons among different acupuncture methods may affect the validity of the findings. Third, the quality of the included studies was variable; some studies did not specify the randomization methods, and most failed to describe the implementation of allocation concealment and blinding. As a result, the effect sizes reported in these trials may have been exaggerated. All the included literature did not explicitly state the blinding of staff during implementation and measurement processes, so there may be some implementation bias and measurement bias. Fortunately, other bias risks are relatively low. Although there are few included literature, the overall quality of the literature is high and the risk of bias is generally low. Therefore, the reliability of this NMA is good. Lastly, the small number of studies included makes it challenging to conduct subgroup analyses.

4.4. Prospects

In our NMA, GM emerged as the most effective method, this also gives us quite a promising treatment in the clinic, for patients who come to us for treatment of postoperative urinary incontinence after prostate cancer, in the process of adjuvant postoperative rehabilitation, given the improvement of NMW on the 2 indexes of ICI-Q-SF and I-QOL, coupled with the absence of any significant adverse effects, we can prefer to choose NMW treatment, which becomes, due to the fact that it is a non-pharmacological treatment, to many patients more acceptable.

However, beyond the choice of method, the selection and stimulation of acupuncture points are also crucial. Furthermore, the choice of acupuncture points varies according to patient conditions, necessitating further research on specific point combinations and their mechanisms. Additionally, due to the differing efficacy evaluations in studies on acupuncture for postprostatectomy urinary incontinence, some studies used a 24-hour or 72-hour pad test as outcomes, while others employed visual analog scales for incontinence severity and pelvic floor muscle strength assessments using Glazer scoring. This variability makes it challenging to conduct a meta-analysis. It is recommended that standardized methods for measuring incontinence severity be established for future research.

Postprostatectomy urinary incontinence typically has a long chronic progression; thus, beyond short-term clinical improvements, more evidence is needed to confirm whether acupuncture and/or moxibustion therapies have long-term effects. Future researchers are encouraged to conduct high-quality, large-sample, multicenter RCTs to validate the existing evidence.

5. Conclusion

Based on the results of this study, acupuncture interventions hold significant potential for improving postprostatectomy urinary incontinence in patients. Notably, GM demonstrates substantial advantages in enhancing clinical efficacy, reducing ICI-Q-SF scores, and improving I-QOL scores. As a classic method of moxibustion, GM is simple, economical, safe, and effective. By stimulating specific acupuncture points while simultaneously applying heat to the deeper layers of the skin, GM can improve blood circulation and metabolism in the pelvic floor muscles, strengthen muscle strength, and alleviate overactive bladder symptoms, leading to favorable clinical outcomes.

Author contributions

Data curation: Yiwen Tang.

Formal analysis: Yuyang Cai.

Funding acquisition: Yiwen Tang, Yuyang Cai, Zhan Gao.

Methodology: Xiong Wang, Ruoyu Qiao, Pengyu Cheng.

Project administration: Jiasen Ding, Ruoyu Qiao, Pengyu Cheng.

Resources: Jiasen Ding.

Software: Yiwen Tang, Xiong Wang.

Visualization: Yuyang Cai, Feng Xu.