AIM: To report a systematic review of published randomized controlled trials (RCTs) investigating the role of absorbable suture (AS) against non-AS (NAS) used for the closure of surgical incisions.

METHODS: RCTs investigating the use of AS vs NAS for the closure of surgical incisions were statistically analysed based upon the principles of meta-analysis and the summated outcomes were represented as OR.

RESULTS: The systematic search of medical literature yielded 10 RCTs on 1354 patients. Prevalence of wound infection (OR = 0.97; 95%CI: 0.56, 1.69; Z = 0.11; P = 0.92) and operative morbidity (P = 0.45) was comparable in both groups. Nonetheless, the use of AS lead to lower risk of wound break-down (OR = 0.12; 95%CI: 0.04, 0.39; Z = 3.52; P < 0.0004).

CONCLUSION: This meta-analysis of 10 RCTs demonstrates that the use of AS is similar to NAS for skin closure for surgical site infection and other operative morbidities. AS do not increase the risk of skin wound dehiscence, rather lead to a reduced risk of wound dehiscence compared to NAS.

A number of studies have been reported in search of improving the skin closure related outcome measures following various surgical procedures, and due to this fact the skin closure techniques are evolving vastly and immensely, predominantly over the last few decades. Innumerable skin closure methods reported in medical literature include continuous stitch closure, interrupted stitch closure, full thickness closure, sub-cuticular closure, primary closure, secondary closure, vacuum assisted closure, glue assisted closure, skin clips or staples closure, simple suture vs mattress sutures, steri-strips closure, absorbable or non-absorbable suture (NAS) closure and other innovative methods[1-13]. These manifold practices of skin approximation after surgical procedures can jointly be classified into two groups. Group I includes the use of NAS for skin closure requiring additional clinical care due to the need of removal of stitches or metallic staples. Group II includes the use of absorbable stitches (AS) or glue which does not require additional clinical care like the group I. The proponents of the use of NAS for skin closure claim that an increased tensile strength of NAS keep wound margins adequately coapted resulting in optimal wound and skin healing[14-16]. The supporters of AS advocate similar effectiveness in wound healing without the requirement of additional clinical care in addition to the benefits of an improved cosmetic outcome and the reduced risk of surgical site infection[17-21]. Due to significant differences in the opinion, the general consensus about the use of either absorbable AS or NAS is still lacking.

The aim of this study is to report a systematic review of published randomized controlled trials (RCTs) on the use of AS against NAS for skin closure.

The PRISMA diagram flow chart explaining the trial selection approach, filtration of trials and eventual study inclusion for quantitative and qualitative analysis is shown in Figure 1. Ten RCTs[31-41] on 1354 patients were found suitable for inclusion and for final analysis. Six hundred and sixty-three were investigated in the AS arm and 691 in NAS arm of the included RCTs. Table 1 depicts the characteristics of the included RCTs. Table 2 is showing the various procedures, type of sutures, type of stiches used in included RCTs.

Open in New Tab   Full Size Figure   Download Figure

Figure 1 PRISMA flow chart showing trial selection methodology. RCT: Randomized controlled trial.

Methodological quality of included studies

Inadequate randomization approach, improper concealment in the process of allocation, absence of power calculations, lack of utilization of single or double blinding and lastly lack of reporting of IIT were major factors responsible for scoring the majority of included RTCs of poor quality (Table 3). GRADE[31] quality of evidence is shown in Figure 2.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Strength and summary of the evidence analysed on GradePro^®.

Ref.	Randomisation technique	Power calculations	Blinding	Intention-to-treat analysis	Concealment
Dørflinger et al[32]	Consecutive patients	No	Yes	No	Inadequate
Foster et al[33]	Consecutive patients	No	No	No	Inadequate
Glough et al[34]	Consecutive patients	No	No	No	Inadequate
Harimoto et al[35]	Sealed envelop	Yes	No	No	Adequate
Kotaluoto et al[36]	Consecutive patients	Yes	No	Yes	Inadequate
Lundblad et al[37]	Consecutive patients	No	No	No	Inadequate
Pauniaho et al[38]	Consecutive patients	Yes	No	No	Inadequate
Ralphs et al[39]	Consecutive patients	No	No	No	Inadequate
Szabó et al[40]	No	No	No	No	Inadequate
Tan et al[41]	Consecutive patients	Yes	No	Yes	Inadequate

Surgical site infection

There was no heterogeneity [Tau² = 0.23, χ² = 12.12, γ = 8, (P = 0.15); I² = 34%] among RCTs that contributed to the combined calculation of this variable. In the random effects model (OR = 0.97; 95%CI: 0.56, 1.69; Z = 0.11; P = 0.92; Figure 3), the risk of surgical site infection was statistically similar in both groups. Although the AS lead to lower incidence of wound infection but it failed to reach at statistical significance.

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Forest plot for surgical site infection following the use of absorbable suture and non-absorbable suture for skin closure. Odds ratios are shown with 95%CI. AS: Absorbable stitch; NAS: Non-absorbable stitch.

Postoperative complications

Combined analysis showed significant statistical heterogeneity [Tau² = 0.61, χ² = 20.57, γ = 9, (P = 0.01); I² = 56%] among included RCTs. Therefore, in the random effects model (OR = 0.77; 95%CI: 0.39, 1.52; Z = 0.75; P = 0.45; Figure 4), the incidence of operative morbidity was statistically comparable in both arms of included RCTs. Although the AS was associated with the reduced risk of developing postoperative complications but statistically it was not significant.

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Forest plot for postoperative complications following the use of absorbable suture and non-absorbable suture for skin closure. Odds ratios are shown with 95%CI. AS: Absorbable stitch; NAS: Non-absorbable stitch.

Risk of wound dehiscence

There was no heterogeneity [χ² = 3.64, γ = 4, (P = 0.46); I² = 0%] among included RCTs. Six trials[32,36-40] contributed to the combined calculation of this variable. Therefore, in the random effects model (OR = 0.12; 95%CI: 0.04, 0.39; Z = 3.52; P < 0.0004; Figure 5), the use of AS was associated with the reduced risk of developing wound break-down.

Open in New Tab   Full Size Figure   Download Figure

Figure 5 Forest plot for the risk of wound dehiscence following the use of absorbable suture and non-absorbable suture for skin closure. Odds ratios are shown with 95%CI. AS: Absorbable stitch; NAS: Non-absorbable stitch.

The findings of this review article demonstrate that the use of AS is similar to NAS for skin closure for surgical site infection and other operative morbidities. AS do not increase the risk of skin wound dehiscence, rather lead to a reduced risk of wound dehiscence compared to NAS.

The conclusions of this study are consistent with the previously reported several RCTs[32,34-41] and comparative studies[32,34,35,37-41]. Majority of these studies compared the usage of AS against NAS by continuous skin closure stitches. Two trials[33,36] compared the use of AS with NAS by interrupted skin closure stiches. Their outcome was also in favour of AS as for as surgical site infection and postoperative complications are concerned. The comparison between continuous stitch vs interrupted stitch closure of skin by using absorbable or non-absorbable sutures could not be performed in this review due to scarcity of trials and number of patients. Therefore, it is difficult to analyse and conclude the superiority of any technique of skin closure.

Current study has many limitations. There were substantial variances in the inclusion criteria such as RCTs on general surgical patients, plastic surgical patients and gynaecology were jointly analysed in this review which may be a potential source of bias due to diversity of patients. Further sub-classification of patients in the form of clean and contaminated wounds was not reported and therefore subgroup analysis was not possible to detect the difference in complications and wound infection following the use of AS and NAS. Varying degrees of differences also existed among included RCTs in terms of the definitions of “wound infection” and “wound dehiscence”. RCTs with fewer patients may not have been sufficient power to recognise small differences in primary and secondary outcomes. Different skin closure techniques like interrupted, subcuticular and continuous suturing were reported in included trials. In addition, different types of absorbable sutures were used in the included studies and one may consider this biased. Inadequate randomization approach, improper concealment in the process of allocation, absence of power calculations, lack of utilization of single or double blinding and lastly lack of reporting of IIT were major factors responsible for scoring the majority of included RTCs of poor quality. Variables like foreign body sensation, stitch granulomas, cosmetic score, health-related quality of life measurement and cost effectiveness should have been considered too. Due to significant clinical and methodological diversity among included studies in addition to aforementioned several limitations, a major, multicentre and high quality randomized, controlled trial is required to validate these findings before recommending the routine use of AS for skin closure.