Barbed versus traditional sutures for wound closure in knee arthroplasty: a systematic review and meta-analysis

Sutures are an increasing focus of research in knee arthroplasty (KA). Whether knotless barbed sutures (KBS) are safe and efficient in KA remains controversial. The objective of our study is to compare the clinical outcomes of KA according to wound closure method: KBS versus knotted traditional sutures (KTS). To clarify this, we conducted a systematic review and meta-analysis. Nine articles involving 10 studies were included in this study. The dataset consisted of 1729 patients with 1754 KA. Among these, 814 patients’ wounds were closed with KBS and 915 with KTS. Our analysis indicates that KBS is preferable for KA wound closure given its shorter wound closure time and lower total cost; postoperative Knee Society scores and complication rates were similar to those of surgeries using KTS. The subgroup analysis revealed that closure of arthrotomy with KBS appears to be associated with a lower risk of complications. This meta-analysis indicates that use of KBS in KA reduces operative time and cost. KBS is the preferred option for wound closures, including arthrotomy and reattachment of subcutaneous and subcuticular tissues. Given the possible biases, adequately powered and better-designed studies with longer follow-up are required to reach a firmer conclusion.

Inclusion criteria. Studies were identified according to the following inclusion criteria: 1) participants: human with relevant diseases requiring surgical intervention, 2) intervention: primary KA, 3) comparison: wound closure with the use of KBS (closing at least subcutaneous and subcuticular tissue) versus KTS (closing subcutaneous and subcuticular tissue as well as the arthrotomy), 4) Outcomes: at least one of the following: wound closure time, complications, cost, postoperative function, 5) Methodological criterion: a prospective study, a case-control study or a cohort study.
The following exclusion criteria were used: 1) insufficient data to estimate an odds ratio (OR) or weighted mean difference (WMD), 2) non-human or cadaver subjects, 3) redundant publications, 4) non-primary research (editorials, commentaries, etc.). Data extraction. Two authors extracted relevant data independently, including the first author's name, study design, publication year, number of patients in each group, average patient age, gender ratio, details on the suture method and technique, total complications and major complications, total cost of wound closure, and postoperative Knee Society scores (KSS). Major complications were defined as those requiring further surgical interventions. Total cost of wound closure included both suture material and operating room time. The cost of operating room time estimates are based upon the average cost for professional staff and resources required for these cases. Intention-to-treat (ITT) data were used when available. Data in other forms, such as medians, interquartile ranges, and means ± 95% confidence interval (CI), were converted to means ± SD following the guidelines in the Cochrane Handbook for Systematic Reviews of Interventions 5.0.2.
Quality assessment. Using a 12-item scale 18 , the methodological quality of each included study was assessed by two independent researchers. The 12-item scale consisted of the following: adequate randomization, concealment of allocation, patient blinding, care provider blinding, outcome assessor blinding, dropout rate, ITT analysis, avoidance of selective reporting, similarity of baseline characteristics, similarity or absence of cofactors, patient compliance, and similarity of timing. Disagreements were evaluated by kappa test and consensus was achieved by discussion with the corresponding author.
Statistical analysis. Statistical analyses were performed using Stata software (ver. 12.0; StataCorp LP, College Station, TX, USA). Weighted mean differences (WMD) with 95% confidence interval (95% CI) were calculated for continuous data and odds ratios (OR) with 95% CIs were calculated for dichotomous data. Statistical heterogeneity was assessed by Q-test and I 2 . I 2 values of 25%, 50%, and 75% were considered to indicate low, moderate, and high heterogeneity, respectively 19 . If P > 0.1 and I 2 < 50%, a fixed-effects model was used; otherwise, a random-effects model was used. For substantial heterogeneity (I 2 > 50%), a sensitivity analysis (backward elimination step-wise regression analysis) was conducted by omitting one study sequentially to examine the influence of each.
The stratified subgroup analysis compared outcomes according to suture type (within the KBS group, whether the arthrotomy was closed with KBS; in the KTS group, whether subcuticular tissue was closed with a running or an interrupted suture) or study design (randomized controlled trials (RCTs) versus non-RCTs).
Publication bias was assessed by Egger's test and Begg's test. For all statistical analyses, with the exception of heterogeneity, a value of P < 0.05 was considered to indicate statistical significance, and all tests were two-sided.

Results
Study selection. The article selection process is shown in Fig. 1. The search yielded 88 potentially relevant articles: 46 from PubMed, 32 from Embase, and 10 from the Cochrane library. Of these, 32 duplicates were removed using Endnote software. Upon review of titles and abstracts of the 56 remaining articles, the full text of 11 articles was retrieved. Because sufficient data were not available in two articles, they were excluded 20,21 , leaving a total of nine articles included in this study 2,4,11-16 . Study characteristics. The characteristics of the nine articles are presented in Table 1. These 9 articles, published between 2010 and 2015, actually include 10 studies; that by Smith et al. 16 included a RCT and a retrospective study. Of these, four were RCTs, one was a prospective cohort study, and five were retrospective studies. The dataset consisted of 1729 patients, including 1754 KA. Among these, 814 patients' wounds were closed with KBS and 915 with KTS. Each study included between 18 and 416 patients. The average age, gender ratio, and surgical site were also noted. In each study, the demographic characteristics of the two groups were similar.
For KA, details on the exact type of suture used and the method of placement are in Table 2. For surgeries involving both of KBS and KTS, placement and type of stitches varied. Among surgeries using the KBS method, the arthrotomy was closed with an interrupted knotted suture in two studies 15,22 , while others used a running KBS. In the KTS group, subcuticular tissue was closed with a running suture in three studies, while others used the interrupted suture technique.   Study quality. Table 3 shows the quality of the included studies. Of these, only one study was high quality; the others were of moderate quality. There was excellent inter-rater agreement between the investigators regarding eligibility (κ = 0.78).

Meta-analysis results. Wound closure time.
In KA using the KBS method, wound closure times were on average 3.56 minutes shorter than in those using KTS (n = 971, WMD = − 3.56, 95% CI = − 5.05 to − 2.08, P < 0.01, I 2 = 94%, P < 0.01). We could not eliminate heterogeneity through a sensitivity analysis, and thus a random-effects model was used.
Risk of total complications. No significant difference was detected in total complication rate between the two groups (n = 1729, OR = 0.98, 95% CI = 0.51 to 1.87, P = 0.95, I 2 = 56%, P = 0.02). Heterogeneity was moderate in the pooled result (I 2 = 56%); a sensitivity analysis was thus performed. Exclusion of data from the Campbell et al. study 12 decreased heterogeneity significantly from 56% to 6%. However, there was still no difference in total risk of complications between the two groups (n = 1313, OR = 0.73, 95% CI = 0.46 to 1.15, P = 0.17, I 2 = 6%, P = 0.38).
Major complications and other complications. Patients in both groups experienced similar rates of major complications (n = 1634, OR = 1.17, 95% CI = 0.67 to 2.18, P = 0.62, I 2 = 17%, P = 0.30). Full details of complications are summarized in Table 4. There were no differences between the two groups in rates of superficial infection, deep infection, wound dehiscence, arthrofibrosis, hematoma, or suture abscess.   Cost. The cost differences in terms of material between the two groups were summarized in Table 5. Upon analysis of the pooled cost data, KBS was associated with 290.72 USD lower costs than KTS (n = 871, WMD = − 290.72, 95% CI = − 474.00 to − 107.45, P = 0.002, I 2 = 99.1%, P < 0.01, Fig. 2).  Table 3. Study quality. 1 Only if the method of sequence made was explicitly introduced could get a 'Yes' . 2 Dropout rate < 20% could get a 'Yes' , otherwise 'No' . 3 ITT = intention-to-treat, only if all randomized participants were analyzed in the group they were allocated to could receive a 'Yes' . 4 "Yes" items more than 7 means 'High'; more than 4 but no more than 7 means 'Moderate'; no more than 4 means 'Low' .   Result of the subgroup analysis. Table 6 shows the results of the subgroup analysis. No significant difference was observed in risk of complications according to different types of study design (Supplement materials 1-3). Using KBS to close arthrotomies appears to lead to a lower total risk of complications relative to KTS. However, among surgeries employing KBS, rates of all complications, major complications (deep infection and others complications which required further surgical interventions, including polyethylene exchange, irrigation and debridement), and superficial infection were greater if KBS was not used for arthrotomy closure. Other factors did not differ significantly between the two groups.
Publication bias. Begg's test (P = 0.89, continuity corrected) and Egger's test (P = 0.108) indicated that publication bias did not affect our results.

Discussion
Our study found that in KA, KBS was associated with shorter wound closure times and lower costs than KTS. Meanwhile, closure of arthrotomies with KBS led to similar postoperative function and lower risk of all complications within the KBS group. KBS was associated with a shorter time to wound closure in our study, confirming results of previous studies 20 23 . KBS is self-anchoring, requiring no knots, thus allowing faster closure 13 . However, we observed substantial heterogeneity in wound closure time, likely resulting from variation in KBS and KTS technique ( Table 2). Among included studies, surgical approaches employing KBS varied with regard to suture method for arthrotomy and superficial skin closure, making significant heterogeneity in wound closure time unavoidable.   We observed no difference in complication rate between surgeries employing KBS versus KTS, consistent with previous studies 20,21 . Theoretically, knots may place uneven pressure on soft tissue, resulting in ischemia, while adsorption of bulky knots may cause local tissue inflammation and scarring and serve as a potential nidus for infection. Moreover, KBS provides more uniform tissue tension to reduce local ischemia, thereby decreasing risk of wound complications 12 . However, KBS is a running suture, which can strangulate the vascular supply and inhibit soft tissue healing 16 . Additionally, Shermak et al. found that KBS increases risk of wound healing complications in the arm 25 , which they speculated results from increased surface area caused by barbs and continuous suturing, promoting spread of inflammation along the length of the closure 25 .
Moreover, we also found that closure of arthrotomy, subcutaneous, and subcuticular tissues with KBS appeared to decrease the total complication rate in KA. Conversely, arthrotomy closure by other methods in KBS group appeared to lead to a higher complication rate. A cadaver study simulating tense hemarthrosis found 74% lower leakage from a barbed suture arthrotomy closure than from KTS 9 . Arthrotomy leakage was minimal when distal arthrotomy closure was tight. Moreover, a biomechanical study found that KBS arthrotomy closure provides similar performance to interrupted KTS upon cyclical loading. Furthermore, KBS performed better than interrupted sutures when the repair was intentionally damaged 8 . Thus, using KBS to close arthrotomy is very important. Whereby achieving a more watertight wound closure may limit the risk of complications 9 .
At early stages (< 6 weeks), patients receiving KBS had 0.98 higher postoperative KSS than those receiving KTSs in our study; this difference was not clinically significant. Likewise, Sah et al. found that range of knee motion is similar for both suture methods for up to one year 2 . An adequately powered RCT with long-term follow-up is necessary to determine the effect of suture method on recovery of function.
Our analysis found that KBS to significant cost savings. Similarly, Mansour et al. found that KBS closure of spinal fusion incisions resulted in hospital charges for operation time that were 884.60 USD lower than those of surgeries using KTS 23 . The cost of operating room time estimates are based upon the average cost for professional staff and resources required for these cases. The average cost of operating room was USD 62 per minute (range, USD 22-133/minute) in 100 United States hospitals 26 . In our study, KBS were associated with a mean savings of 3.56 minutes, which is in the range of 2.08 to 5.05 minutes with use of this suture device. Of note, the material cost with KTS could be saved USD 91.93, at most, in these included studies. Though barbed closure materials are more expensive than those for KTS, shorter surgery time leads to a reduced total cost.
To our knowledge, this is the first meta-analysis comparing the risk of complications and postoperative function between KBS and KTS for KA that includes all available comparative evidence and comprehensively investigates differences in the clinical outcomes. However, it has the following limitations. Most importantly, surgeries classified as KBS and KTS included multiple methods for closing superficial skin (Table 2), making significant heterogeneity in wound closure time and total cost unavoidable. Moreover, there is a lack of uniform technique of KTS. Second, as few RCTs in this area have been performed (perhaps because KBS remains relatively new in KA), our study included several non-RCTs, which inevitably involved recall and interviewer bias, which likely weakened our analysis. Nonetheless, excluding non-RCTs would have underpowered the analysis, increase the risk of false-negative errors and influence the accuracy of our findings. Demographic characteristics were similar between the two groups in all included studies, suggesting that selection bias was limited. Third, follow-up duration was relatively short, preventing examination of long-term outcomes, especially postoperative function. Furthermore, the bias might also be produced by the variations of stitches, such as V-Loc stitch and Quill stitch in KBS.

Conclusions
Based on available evidence, use of KBS in KA leads to shorter operation times and lower total costs. Closure of arthrotomy and subcutaneous and subcuticular tissues by KBS yields similar postoperative function and lower total complication risk when compared with KTS. We thus conclude that KBS is an optimal approach for closure of arthrotomies and subcutaneous and subcuticular tissues in KA. Given the relevant possible biases in our study, adequately powered and better-designed studies with long-term follow-up are required to reach a firmer conclusion.