Efficacy and safety of barbed suture in minimally invasive radical prostatectomy: A systematic review and meta‐analysis

Abstract

As one of the earliest surgeries applying knotless barbed suture, the minimally invasive radical prostatectomy (MIRP) was reported to have various effects on the patients and the surgeons. This study reviewed the available evidence about the efficacy and safety of barbed sutures in MIRP. We searched ClinicalTrials.gov, Cochrane Register of Clinical Studies, PubMed, and Embase to identify randomized controlled trials (RCTs) and cohort studies addressing the application of barbed sutures and conventional sutures in MIRP (until August 2016). Quality assessment was performed according to Cochrane recommendations. The data were analyzed using Review Manager (Version 5.3), and sensitivity analysis was performed by sequentially omitting each study. A total of 12 studies, including three RCTs (low to moderate risk of bias, 211 patients) and nine cohort studies (low to moderate risk of bias, 698 patients), fulfilled the study criteria. The pooling of trials did not show statistical difference. Pooling data of cohort studies showed that suture time [mean difference (MD) = −8.52, 95% confidence interval (CI) = −12.60 to −4.43, p < 0.0001] and length of hospital stay (MD = −0.96, 95% CI = −1.80 to −0.11, p = 0.03) were significantly shorter in the barbed group. Results of continence rate varied according to different studies. Subgroup analysis by type of MIRP suggested that patients who underwent barbed suture during robot‐assisted surgeries had a shorter hospital stay (MD = −1.13, 95% CI = −1.82 to −0.45, p = 0.001). During the laparoscopic surgery, patients in the barbed suture group had fewer postoperative complications [odds ratio = 0.29, 95% CI = 0.08–0.98, p = 0.05). However, more evidence is needed to validate this state‐of‐the‐art technology.

Introduction

Knotless barbed suture is a particularly designed monofilament suture with barbs orientated in an opposite direction to the needle, which has been widely used in both skin and deeper structures. In general, conventional knot tying sutures required time and training, which may easily extrude with high infection rate related to knots [1]. However, the novel barbs on the ligatures make the suture grab the tissue, without allowing the suture to slide back.

Since invented in 1964 [2], this technique has been developed into three types [3], which are commercially available, including Quill SRS (Quill Self‐Retaining System; Angiotech Pharmaceuticals, Vancouver, British Columbia, Canada), the bidirectional barbed suture used in gynecologic laparoscopy; V‐Loc Absorbable Wound Closure device (Covidien, Mansfield, MA, USA), the unidirectional barbed suture that has only one needle and a loop at the end; and the STRATAFIX (STRATAFIX Knotless Tissue Control Devices, Ethicon Inc, Somerville, NJ, USA), which presents a spiral distribution of the barbs anchors.

The use of barbed sutures has been first described in plastic and gynecological surgeries previously [4]. Then various animal and cadaveric experiments confirmed the equivalence of biocompatibility and tensile strength of knotless barbed suture to conventional sutures in urological field [[5], [6]]. As one of the earliest urological surgeries that adopted this advanced technique, the minimally invasive radical prostatectomy [MIRP, including robot‐assisted radical prostatectomy (RARP) and laparoscopic radical prostatectomy (LRP)] was reported to have various effects on both patients and surgeons. Thus, a meta‐analysis and systematic review were carried out to obtain more validated results on the application of knotless barbed sutures in MIRP in comparison with the conventional sutures.

Methods

Study identification and selection

The Medline, Embase, and the Cochrane Library databases were searched using the following terms: “barbed” OR “knotless” AND “suturing” OR “suture” (last updated in August 2016). To modify the results and avoid the publication bias, we also searched clinical trials registered in ClinicalTrials.gov (last updated in August 2016).

All studies had to meet the following inclusion criteria: (1) study design had to be a randomized controlled trial (RCT) or observational studies based on human participants; (2) patients underwent RARP and LRP; (3) interventions had to be conventional suture versus barbed suture; (4) the studies provide short‐ or long‐term outcomes. The following exclusion criteria were also applied: (1) no control; (2) conventional sutures were made with other materials such as mesh or staple rather than smooth sutures; (3) abstracts, reviews, and overlapped studies; (4) studies published in languages other than English. The computer search was supplemented with manual searches for references of included studies.

Data extraction and outcome measures

We imported the search results into a bibliographic citation management software (EndNote X7). Two reviewers independently collected the data and reached a consensus on all items. The following items were extracted from each study, if available: first author's surname, publication year, original country, study design, sample size, and postoperative complications.

The main outcome measures chosen for this meta‐analysis were operative time, suturing time, estimated blood loss or change in hemoglobin level, length of catheterization, hospital stay, postoperative complications, and continence rate. Heterogeneity of outcomes was assessed to confirm the appropriateness of combining individual studies.

Definition

Operative time was defined as the total time of surgery. Suture time was defined as the time needed for the completion of the surgical site incision, anastomosis time, and closure time. Estimated blood loss or change in hemoglobin level was defined as the blood loss during the operation and it was usually obtained from anesthesia records and/or the surgeons' operative reports. After surgeries, data on postoperative complications of the suture, length of catheterization, and hospital stay were also recorded. After discharging from the hospital, patients were asked to report their pad usage for continence rates at different follow‐up time points.

Methodological quality

The risk of bias of included RCTs was assessed following Cochrane recommendations, considering random sequence generation, allocation concealment, blinding of participants and personnel, blinding of outcome assessment, incomplete outcome data, and selective reporting [7]. Publication bias was evaluated by funnel plot.

We used the Newcastle–Ottawa Scale, a widely used and accepted instrument, for assessing the risk of bias of cohort studies [8]. We included the following items for cohort study: ascertainment of MIRP, representativeness of the barbed cohort, ascertainment of exposure to barbed suture, selection of the non‐exposed cohort, demonstration that outcome of interest (i.e., suture time) was not present at the start of study, comparability of study controls for important factors (e.g., adequate adjustment for confounders or matching for important confounding factors), assessment of outcome (e.g., blinding assessment and adjudication), and completeness of the follow‐up.

Data synthesis and analysis

The studies were divided into two subgroups, according to RARP and LRP; meanwhile, a separate meta‐analysis was conducted within different subgroups. In all analyses, we estimated the pooled mean difference (MD) to assess continuous data, whereas pooled odds ratios (ORs) were calculated for the assessment of dichotomous data (postoperative complications). We used the method to pool randomized trials given the very low event rate, and reported the pooled OR and corresponding 95% confidence interval (CI). The pooled estimations regarding outcomes expressed as either dichotomous or continuous variables were calculated with the use of a random effect model. The existence of statistical heterogeneity between the included studies was assessed with the χ² test and I ² test. In addition, we also conducted sensitivity analyses to examine the robustness of the estimates and assessed the risk of publication bias by the Begg's funnel plots. For all analyses, p < 0.05 was considered statistically significant. Statistical analyses were performed with the software program Review Manager (version 5.3; The Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen, Denmark).

Results

Study selection process and characteristics

A total of 12 studies, including three RCTs [[9], [10], [11]] (211 patients), nine cohort studies [[12], [13], [14], [15], [16], [17], [18], [19], [20]] (698 patients), and no case–control studies were found to fulfill the eligibility criteria. Table 1 presents the baseline characteristics of all studies (Figure 1).

Table 1.

Basic characteristics of all pooled studies in the meta‐analysis.

Author/y	Type of surgery	Country	Study design	Sample size	Age, y, mean ± SD	BMI, mean ± SD	PSA, ng/mL, mean ± SD	Suture time, min, mean ± SD	Operative time, min, mean ± SD	Estimate the intraoperative blood loss, mL, mean ± SD	Length of catheterization, d, mean ± SD	Hospital stay, days, mean ± SD	Time of follow‐up	Complications
				B/C	B/C	B/C	B/C	B/C	B/C	B/C	B/C	B/C
Williams et al. 2010 [9]	RARP	USA	RCT	45/36	59.0 ± 7.0/60.3 ± 5.1	28.4 ± 4.1/28.3 ± 5.3	6.7 (3.0)/6.1 (4.5)	9.7 ± 0.2/9.8 ± 0.2	103.8 ± 21.2/110.4 ± 19.4	181.5 ± 78.1/173.3 ± 49.7	11.1 ± 8.3/8.3 ± 3.8	Not stated	Not stated	Cystogram leak
Sammon et al. 2011 [10]	RARP	USA	RCT	31/33	60 (52.5–66)/63 (53–68)	28 (25–31)/28 (24.4–31)	4.6 (4–7)/5 (4.2–7.3)	13.4/18.1		150/100		Not stated	Median 9.1 mo	Leaked urine, urinated blood, had pain, or burning with urination
Zorn et al. 2012 [11]	RARP	Canada	RCT	33/33	59.2 ± 5.8/59.8 ± 5.6	27.7 ± 5.1/29.1 ± 4.2	5.9 ± 4.6/5.4 ± 3.7	13.1/20.8	166 ± 41/174 ± 37	167 ± 95/175 ± 110	5/5	Not stated	Mean 6.2 mo	Urinary retention, clinical urinary VUA leakage, anastomotic stricture, prolonged hematuria (>2 d)
Tewari et al. 2010 [12]	RARP	USA	Prospective cohort study	50/50	59.47 ± 7.93/59.20 ± 8.41	26.63 ± 3.13/27.14 ± 2.81	5.2 (3.87,5.85)/4.57 (3.5,6.3)	7 (6, 8.25)/12 (10,14)	106 (79,120)/114.5 (89,150.25)	150 (150,150)/150(150,162.50)	5–7/5–7	Not stated	Not stated	Postoperative clinically significant urinary leaks: 1; postoperative urinary retention
Polland et al. 2011 [13]	RARP	USA	Prospective cohort study	42/42	Not stated	Not stated	Not stated	12 (8–18)/18 (11–28)	Not stated	Not stated	7 (6–12)/7 (6–12)	Not stated	Mean 9.4 mo	Urinary leak did not occur; the use of barbed suture does not appear to cause an increased rate of bladder neck contractures or foreign body reactions
Hemal et al. 2012 [12]	RARP	USA	Cohort study	25/25	60.3 ± 7.0/62.8 ± 7.5	Not stated	7.9 ± 5.6/7.9 ± 6.8	8.4 ± 1.7/14.3 ± 4.8	73.0 ± 8.9/85.7 ± 33.2	108.0 ± 61.2/130.2 ± 83.8	Not stated	1.9 ± 0.8/2.7 ± 1.1	Minimum 6 mo	None of the patients had presented with symptomatic urinary leaks, urinary retention, or anastomotic strictures
Manganiello et al. 2012 [15]	RARP	USA	Prospective cohort study	35/35	58.7 (46.6–73.4)/61.2(48.2–71.5)	29.0 (19.7–40.5)/28.7(19.3–40.5)	6.4 (0.8–13.3)/9.3 (0.7–69)	24.8 (11–55)/26.6 (11–51)	Not stated	159 (50–300)/125 (25–450)	12.2/12.5	Not stated	Minimum 5 mo	Cystogram leak
Escudero et al. 2012 [16]	LRP	Spain	Prospective cohort study	50/50	63.44 ± 5.8/66.16 ± 5.2	Not stated	8.16 ± 3.05/9.11 ± 4.3	23.6 ± 10.1/36.1 ± 6.7	179.8 ± 40.8/166.4 ± 39.5	Not stated	12.2 ± 3/13.7 ± 5.2	3.1 ± 2.4/3.5 ± 2.1	Not stated	No urinary leakage was observed; enlarged catheter stay (>3 d) in seven patients due to high debit because of bleeding and of a rectal injury
Massoud et al. 2013 [17]	RARP	France	Prospective cohort study	40/40	60.7 ± 6.25/64.5 ± 5.31	Not stated	Not stated	8.5 ± 3.2/11.5 ± 4	Not stated	Not stated	Not stated	4.1 ± 1.8/5.6 ± 0.9	12 mo	Urethral stenosis
Arslan et al. 2014 [18]	LRP	Turkey	Cohort study	45/47	65.2 ± 5.7/63.1 ± 6.3	26.9 ± 4.5/27.0 ± 5	11.0 ± 11.5/9.0 ± 5.9	24 ± 5.5/40 ± 7.1	179.7 ± 20.5/196.1 ± 24.2	415.1 ± 223.1/424.6 ± 172.7	9.0 ± 1.9/14.0 ± 1.7	2.1 ± 0.4/4.2 ± 1.5	Mean 4.1 mo	Cystogram leak
Gozen et al. 2015 [19]	LRP	Germany	Prospective cohort study	31/31	65.7 ± 7.5/65.1 ± 5.1	28.5 ± 4.3/29 ± 4.5	10.6 ± 4.8/8.5 ± 4.2	Not stated	205 ± 31.1/219.4 ± 48.8	654.8 ± 457.5/670.9 ± 233.2	Not stated	Not stated	Mean 20.2 ± 3.3 mo	Not stated
Takeda et al. 2014 [20]	LRP	Japan	Cohort study	30/30	66.3 ± 5.4/64.7 ± 4.9	23.5 ± 2.7/24.4 ± 2.3	9.0 ± 10.1/8.4 ± 5.4	13.2 ± 2.3/19.1 ± 3.3	177.1 ± 29.4/184.2 ± 40.3	182.7 ± 183.9/182.2 ± 209.5	3.0 ± 1.0/3.0 ± 0.4	5.5 ± 1.1/5.4 ± 1.1	3–9.5 mo	Cystogram leak, urinary retention

B = barbed group; BMI = body mass index; C = control group; LRP = laparoscopic radical prostatectomy; RARP = robot‐assisted radical prostatectomy; RCT = randomized controlled trial; PSA = prostate‐specific antigen; SD = standard deviation; VUA = vesicourethral anastomosis.

Figure 1.

Flow diagram of the detailed selection process.

Of the three trials, all conducted computer‐generated randomization; however, there was no statement of allocation concealment, blinding, and selective reporting. The loss to follow‐up occurred in 0–2.2% of patients. In general, risk of bias was low to moderate in RCTs (see Table S1 in the supplementary material online).

Of the nine cohort studies, the patient population within a study, regardless of exposed patients (i.e., those using barbed suture) or unexposed patients (i.e., those not using barbed suture), was identified from a same clinical setting. Ascertainment of patients with prostate cancer in all studies was based on surgical records. Only one study adjusted important confounding factors by matching pairs (body mass index and preoperative prostate volume) and one study performed multivariate analysis of the outcomes to adjust the results. All of studies reported information regarding loss to follow‐up (0–4.29%). In general, risk of bias in cohort studies was low to moderate (see Table S2 in the supplementary material online).

Evidence from RCTs

The pooling of trials [[9], [10], [11]] (all focused on the RARP) did not show statistically significant difference in operative time (MD = −6.85, 95% CI = −14.87 to 1.17, p = 0.09, I ² = 0%), estimated blood loss (MD = 4.28, 95% CI = −20.10 to 28.67, p = 0.73, I ² = 0%), and postoperative complications (MD = 2.79, 95% CI = 0.89–8.79, p = 0.08, I ² = 62%). In addition, suture time, although not readily available for pooling, was demonstrated to be significantly lesser in the barbed suture group.

Evidence from cohort studies

Suturing time of urethrovesical anastomosis

Of the nine cohort studies, five reported the suturing time of urethrovesical anastomosis [[14], [16], [17], [18], [20]] (Figure 2A). Although the heterogeneity was high (p < 0.00001, I ² = 95%), the pooled and individual outcomes all showed significantly shorter time in the barbed group. (MD = −8.52, 95% CI = −12.60 to −4.43, p < 0.0001). In the subgroup analysis, pooled outcomes were similar based on RARP or LRP (LRP; Table 2 and Figure S1 in the supplementary material online).

Figure 2.

Forest plots of all the results comparing barbed suture with conventional suture: (A) a forest plot of suturing time with or without barbed suture; (B) a forest plot of operative time with or without barbed suture; (C) a forest plot of estimated blood loss with or without barbed suture; (D) a forest plot of length of catheterization with or without barbed suture; (E) a forest plot of hospital stay with or without barbed suture; (F) a forest plot of postoperative complications with or without barbed suture. CI = confidence interval; SD = standard deviation.

Table 2.

Pooled outcomes of all the subgroups.

Outcomes	No. of studies	No. of cases: barbed/control	MD/OR	95% CI	Heterogeneity	p for effect size
Urethrovesical anastomosis time
RARP	2	65/65	−4.38	[−7.22, −1.55]	p = 0.03; I 2 = 80%	Z = 3.03 (p = 0.002)
LRP	3	125/127	−11.39	[−18.26, −4.52]	p < 0.00001; I 2 = 96%	Z = 3.25 (p = 0.001)
Operative time
RARP	1	25/25	−12.7	[−26.17, 0.77]	N/A	Z = 1.85 (p = 0.06)
LRP	4	156/158	−6.4	[−20.67, 7.86]	p = 0.01; I 2 = 72%	Z = 0.88 (p = 0.38)
Estimated blood loss
RARP	1	25/25	−22.20	[−62.72, 18.32]	N/A	Z = 1.07 (p = 0.28)
LRP	3	106/108	−6.55	[−67.14, 54.03]	p = 0.98; I 2 = 0%	Z = 0.21 (p = 0.83)
Hospital stay
RARP	2	65/65	−1.13	[−1.82, −0.45]	p = 0.09; I 2 = 64%	Z = 3.24 (p = 0.001)
LRP	3	125/127	−0.82	[−2.36, 0.72]	p < 0.00001; I 2 = 95%	Z = 1.04 (p = 0.30)
Postoperative complications a
RARP	5	192/192	1.61	[0.41, 6.35]	p = 0.88; I 2 = 0%	Z = 0.68 (p = 0.50)
LRP	3	125/127	0.29	[0.08, 0.98]	p = 0.06; I 2 = 71%	Z = 1.99 (p = 0.05)

CI = confidence interval; LRP = laparoscopic radical prostatectomy; MD = mean difference; N/A = not applicable; OR = odds ratio; RARP = robot‐assisted radical prostatectomy.

^aOdds ratio.

Operative time

Five studies [[14], [16], [18], [19], [20]] reported data on operative time (Figure 2B). The pooling of raw data did not present a statistically significant association between the suture types and operative time (MD = −7.93, 95% CI = −18.68 to 2.81, p = 0.15), whereas the studies were relatively statistically heterogeneous (p = 0.03, I ² = 63%). In addition, the outcomes of subgroup analysis were similar based on RARP or LRP (Table 2 and Figure S1 in the supplementary material online). Because Escudero's study [16] accounts for the learning curve of LRP and it also compared a novel surgeon with an expert surgeon, we excluded this study from sensitivity analysis. In addition, the pooled results showed significant reduction of operative time in the barbed suture groups (MD = −14.04, 95% CI = −20.63 to −7.44, p < 0.0001, I ² = 0%; see Figure S3 in the supplementary material online).

Estimated blood loss

Four studies [[14], [18], [19], [20]] reported outcomes of estimated blood loss (Figure 2C). The pooling outcomes also did not present significant benefits of barbed suture over conventional suture (MD = −17.36, 95% CI = −51.05 to −16.32, p = 0.31, I ² = 0%). In addition, the outcomes of subgroup analysis were similar based on RARP or LRP (Table 2 and Figure S4 in the supplementary material online).

Length of catheterization

Three studies, all related to LRP, reported data on catheterization duration [[16], [18], [20]] (Figure 2D). Pooling of these cohort studies did not suggest a statistically significant association between the suture types and catheterization duration (MD = −2.17, 95% CI = −5.86 to 1.52). This finding, however, was statistically heterogeneous (I ² = 99%).

Hospital stay

Five studies [[14], [16], [17], [18], [20]] reported about hospital stay (Figure 2E). Pooling of those five cohort studies indicated that patients using barbed suture stayed significantly longer in hospital (MD = −0.96, 95% CI = −1.80 to −0.11, p = 0.03). This finding was, however, highly limited due to a very high level of heterogeneity (I ² = 98%). Then subgroup analysis (Table 2 and Figure S5 in the supplementary material online) by surgery types gave the following results: compared with RARP, barbed suture significantly decreased the hospital stay (MD = −1.13, 95% CI = −1.82 to −0.45, p = 0.001, I ² = 64%); compared with LRP, the use of barbed suture did not appear to significantly decrease the hospital length (MD = −0.82, 95% CI –2.36 to 0.72, p = 0.30, I ² = 95%).

Postoperative complications

Eight cohort studies [[12], [13], [14], [15], [16], [17], [18], [20]] reported raw event data on postoperative complications (Figure 2F). Pooling data did not suggest statistically significant differences between both groups (OR = 0.59, 95% CI = 0.25–1.38, p = 0.23, I ² = 25%). Then, subgroup analysis (Table 2 and Figure S6 in the supplementary material online) by surgery types gave the following results: compared with RARP, pooled data did not show any significant difference (OR = 1.61; 95% CI = 0.41–6.35, p = 0.50, I ² = 0%); compared with LRP, the use of barbed suture statistically decreased the postoperative complications (OR = 0.29, 95% CI = 0.08–0.98, p = 0.05, I ² = 71%).

Continence rate

As a reflection of continence rate, pad usage per day was applied for measurement. Five additional cohort studies reported continence rate after the operations; the data were, however, not ready for pooling. Four of them suggested no difference of overall pad usage between two groups at 4 weeks [20], 6 weeks [13], 5 months [15], 6 months [13], and 12 months [17].

Further, two studies demonstrated that patients using conventional suture had higher continence rate in 2 months [15] (p = 0.035) and 6 months [18] (p = 0.002), respectively. However, Takeda et al. [20] found that the percentage of patients who required no more than one pad per day was significantly greater in the barbed group at 3 months (p = 0.020).

Publication bias

Publication bias was assessed by the Begg's funnel plots. The shape of the funnel plots seemed symmetric in the barbed versus conventional suture, suggesting no evidence of publication bias (see Figures S7–S12 in the supplementary material online).

Discussion

In this systematic review, we have included randomized and nonrandomized studies to test the effect of barbed suture on MIRP. Because most of the randomized trials were limited with relatively short exposure to barbed suture (i.e., short follow‐up) and only focused on RARP with small number of patients, the findings provide little insights regarding the impact of barbed suture on status of patients during MIRP. The pooling of those trials showed no significant difference in outcomes between the use of barbed suture and control; however, the wide CIs suggested substantial uncertainty of the effects.

By contrast, evidence from observational studies—all were cohort studies—provided more information to assess the effects of barbed suture. In general, barbed suture reduced suture time, operative time, and hospital stay compared with conventional suture. There was no significant change concerning estimated blood loss, length of catheterization, and postoperative complications. Moreover, results differed in different types of MIRP.

The urethrovesical anastomosis is a technically challenging aspect of RARP and LRP [[21], [22], [23]]. To facilitate the urethrovesical anastomosis technique, various techniques using monofilament sutures have been advised while drawbacks were also apparent [[24], [25]]. The pooled outcome of all the cohort studies showed significant reduction of anastomosis time in the barbed suture groups, as well as the operative time after analysis of sensitivity. Different hospitals or surgeons from different countries may be the reason for the high heterogeneity of suture time and the learning curve of RARP or LRP from some teaching hospitals could also be an important factor [[15], [16]]. In addition, subgroup analysis suggested the same results. Further, the raw data of RCTs, though cannot be pooled, indicated similar results individually.

With regard to estimated blood loss, length of catheterization, and postoperative complications, the pooled outcomes were all comparable, respectively. However, the subgroup analysis suggested possible differential effect by type of MIRP. Specifically, postoperative complications occurred less often in the barbed suture groups in LRPs, whereas in RARP the outcomes were equivalent in both groups.

It seems that barbed suture has a potentiality to decrease postoperative complications (i.e., urine leak). As a result, the length of catheterization was supposed to be longer, which is consistent with Arslan and co‐workers' study [18], even though the pooled outcomes show no differences in catheterization duration. Nevertheless, other researchers [15] proposed that tear through urethral mucosa, anastomotic stricture, and secondary ischemia from high pressure of the suture could be found at long follow‐up even though none of the recruited studies encountered these events in our case. Moreover, it is possible that there were asymptomatic complications that went undetected.

Meanwhile, shortened hospital stay was also reported in the barbed suture group, especially in the RARP subgroup, which indicated that such patients had a better prognosis. However, Hemal et al. [14] reported that reduction of hospital stay duration may result from other personal or social factors, for example, patient coming from far off places.

After discharge, continence rates play an essential role in predicting patients' urine loss ratio. Although continence rates in both groups continue to improve at longer follow‐up, the results varied a lot at different time points. On the one hand, patient reports of daily pad usage are subjective and dependent on individual preference as to when a pad should be changed [15]. On the other hand, other factors such as patient's baseline function, the degree of bladder neck preservation, and nerve sparing other than the technical nuances of the anastomosis itself were determined [26].

Our pooled outcome provides a convincing evidence for the relationship between the barbed suture and some important surgical indicators for MIRP. However, there are limitations of this study: (1) only few RCTs were recruited and the pooled data were not persuasive; (2) relatively high heterogeneity among studies was estimated for surgery‐related outcomes; (3) although our literature search was extensive, it did not cover conference publications, letters to the editor, and animal studies.

Nevertheless, our results renew a latest meta‐analysis on barbed suture, which was limited by inadequately pooled data. To the best of our knowledge, this is the most comprehensive meta‐analysis to date investigating the association between barbed and traditional sutures.

Conclusions

Significant decline of suture time, operative stay, and hospital stay were found using barbed suture during MIRP. LRP seemed to be safer in the application of the barbed suture. Furthermore, more evidence is need to validate this state‐of‐the‐art technology and future studies should be conducted to analyze its impact on cost‐effectiveness.

Acknowledgments

Supported by Grant No. 81403276 from the National Natural Science Foundation of China and Grant No. JH20140066 from the Technology Support Program of Science and Technology Department of Sichuan Province.

Supporting information

The following is the supplementary data related to this article:

Table 1: Quality assessment of studies in the meta‐analysis based on Cochrane recommendations.

Table 2: Quality assessment of studies in the meta‐analysis based on Newcastle‐Ottawa Scale (NOS).

Figure 1: A forest plot of suturing time with or without barbed suture.

Figure 2: A forest plot of operative time with or without barbed suture.

Figure 3: A forest plot of sensitivity analyses of operative time with or without barbed suture.

Figure 4: A forest plot of estimated blood loss with or without barbed suture.

Figure 5: A forest plot of hospital stay with or without barbed suture.

Figure 6: A forest plot of postoperative complication with or without barbed suture.

Figure 7: Funnel plot of suturing time in all included studies.

Figure 8: Funnel plot of operative time in all included studies.

Figure 9: Funnel plot of estimate blood loss in all included studies.

Figure 10: Funnel plot of length of catheterization in all included studies.

Figure 11: Funnel plot of hospital stay in all included studies.

Figure 12: Funnel plot of postoperative complications in all included studies.

Supplementary data

Supplementary data related to this article can be found at https://doi.org/10.1016/j.kjms.2016.12.005.