Jahitan subkutikuler untuk menutup kulit bagi pembedahan bukan obstetrik
Abstract
Background
Following surgery, surgical wounds can be closed using a variety of devices including sutures (subcuticular or transdermal), staples and tissue adhesives. Subcuticular sutures are intradermal stitches (placed immediately below the epidermal layer). The increased availability of synthetic absorbable filaments (stitches which are absorbed by the body and do not have to be removed) has led to an increased use of subcuticular sutures. However, in non‐obstetric surgery, there is still controversy about whether subcuticular sutures increase the incidence of wound complications.
Objectives
To examine the efficacy and acceptability of subcuticular sutures for skin closure in non‐obstetric surgery.
Search methods
In March 2019, we searched the Cochrane Wounds Specialised Register; the Cochrane Central Register of Controlled Trials (CENTRAL); Ovid MEDLINE (including In‐Process & Other Non‐Indexed Citations); Ovid Embase and EBSCO CINAHL Plus. We also searched clinical trials registries for ongoing and unpublished studies, and scanned reference lists of relevant included studies as well as reviews, meta‐analyses and health technology reports to identify additional studies. There were no restrictions with respect to language, date of publication or study setting.
Selection criteria
All randomised controlled trials which compared subcuticular sutures with any other methods for skin closure in non‐obstetric surgery were included in the review.
Data collection and analysis
Two review authors independently identified the trials, extracted data and carried out risk of bias and GRADE assessment of the certainty of the evidence.
Main results
We included 66 studies (7487 participants); 11 included trials had more than two arms. Most trials had poorly‐reported methodology, meaning that it is unclear whether they were at high risk of bias. Most trials compared subcuticular sutures with transdermal sutures, skin staples or tissue adhesives. Most outcomes prespecified in the review protocol were reported. The certainty of evidence varied from high to very low in the comparisons of subcuticular sutures with transdermal sutures or staples and tissue adhesives; the certainty of the evidence for the comparison with surgical tapes and zippers was low to very low. Most evidence was downgraded for imprecision or risk of bias.
Although the majority of studies enrolled people who underwent CDC class 1 (clean) surgeries, two‐thirds of participants were enrolled in studies which included CDC class 2 to 4 surgeries, such as appendectomies and gastrointestinal surgeries. Most participants were adults in a hospital setting.
Subcuticular sutures versus transdermal sutures
There may be little difference in the incidence of SSI (risk ratio (RR) 1.10; 95% confidence interval (CI) 0.80 to 1.52; 3107 participants; low‐certainty evidence).
It is uncertain whether subcuticular sutures reduce wound complications (RR 0.83; 95% CI 0.40 to 1.71; 1489 participants; very low‐certainty evidence). Subcuticular sutures probably improve patient satisfaction (score from 1 to 10) (at 30 days; MD 1.60, 95% CI 1.32 to 1.88; 290 participants; moderate‐certainty evidence). Wound closure time is probably longer when subcuticular sutures are used (MD 5.81 minutes; 95% CI 5.13 to 6.49 minutes; 585 participants; moderate‐certainty evidence).
Subcuticular sutures versus skin staples
There is moderate‐certainty evidence that, when compared with skin staples, subcuticular sutures probably have little effect on SSI (RR 0.81, 95% CI 0.64 to 1.01; 4163 participants); but probably decrease the incidence of wound complications (RR 0.79, 95% CI 0.64 to 0.98; 2973 participants). Subcuticular sutures are associated with slightly higher patient satisfaction (score from 1 to 5) (MD 0.20, 95% CI 0.10 to 0.30; 1232 participants; high‐certainty evidence). Wound closure time may also be longer compared with staples (MD 0.30 to 5.50 minutes; 1384 participants; low‐certainty evidence).
Subcuticular sutures versus tissue adhesives, surgical tapes and zippers
There is moderate‐certainty evidence showing no clear difference in the incidence of SSI between participants treated with subcuticular sutures and those treated with tissue adhesives (RR 0.77, 95% CI 0.41 to 1.45; 869 participants). There is also no clear difference in the incidence of wound complications (RR 0.62, 95% CI 0.35 to 1.11; 1058 participants; low‐certainty evidence). Subcuticular sutures may also achieve lower patient satisfaction ratings (score from 1 to 10) (MD ‐2.05, 95% CI ‐3.05 to ‐1.05; 131 participants) (low‐certainty evidence). In terms of SSI incidence, the evidence is uncertain when subcuticular sutures are compared with surgical tapes (RR 1.31, 95% CI 0.40 to 4.27; 354 participants; very low‐certainty evidence) or surgical zippers (RR 0.80, 95% CI 0.08 to 8.48; 424 participants; very low‐certainty evidence). There may be little difference in the incidence of wound complications between participants treated with subcuticular sutures and those treated with surgical tapes (RR 0.90, 95% CI 0.61 to 1.34; 492 participants; low‐certainty evidence). It is uncertain whether subcuticular sutures reduce the risk of wound complications compared with surgical zippers (RR 0.55, 95% CI 0.15 to 2.04; 424 participants; very low‐certainty evidence). It is also uncertain whether it takes longer to close a wound with subcuticular sutures compared with tissue adhesives (MD ‐0.34 to 10.39 minutes; 895 participants), surgical tapes (MD 0.74 to 6.36 minutes; 169 participants) or zippers (MD 4.38 to 8.25 minutes; 424 participants) (very low‐certainty evidence). No study reported results for patient satisfaction compared with surgical tapes or zippers.
Authors' conclusions
There is no clear difference in the incidence of SSI for subcuticular sutures in comparison with any other skin closure methods. Subcuticular sutures probably reduce wound complications compared with staples, and probably improve patient satisfaction compared with transdermal sutures or staples. However, tissue adhesives may improve patient satisfaction compared with subcuticular sutures, and transdermal sutures and skin staples may be quicker to apply than subcuticular sutures. The quality of the evidence ranged from high to very low; evidence for almost all comparisons was subject to some limitations. There seems to be no need for additional new trials to explore the comparison with staples because there are high‐quality studies with large sample sizes and some ongoing studies. However, there is a need for studies exploring the comparisons with transdermal sutures, tissue adhesives, tapes and zippers, with high‐quality studies and large sample sizes, including long‐term assessments.
PICO
Ringkasan bahasa mudah
Jahitan di bawah kulit untuk menutup luka selepas pembedahan
Apakah tujuan ulasan ini?
Ulasan ini bertujuan untuk mengetahui sama ada jahitan subkutikular (jahitan di bawah kulit) berkesan untuk menutup luka selepas pembedahan. Kami berminat dengan semua jenis pembedahan kecuali pembedahan obstetrik (pembedahan yang berkaitan dengan melahirkan anak, contohnya pembedahan caesarean). Penyelidik Cochrane telah mengumpul dan menganalisa semua kajian yang berkaitan dan mendapati 66 kajian rawak terkawal yang relevan untuk menjawab soalan ini. Kajian rawak terkawal adalah kajian perubatan di mana pesakit dipilih secara rawak untuk menerima rawatan yang berbeza. Kajian jenis ini memberikan bukti perubatan yang paling dipercayai.
Mesej utama
Dari segi jangkitan luka selepas pembedahan, tidak ada perbezaan yang jelas antara jahitan di bawah kulit dan kaedah lain untuk menutup luka pembedahan, seperti mana jahitan biasa yang melintasi kulit, pita pembedahan, dawai kokot, atau gam. Jahitan yang di bawah kulit mungkin mengurangkan komplikasi luka berbanding dengan dawai kokot dan meningkatkan kepuasan pesakit berbanding dengan jahitan yang melintasi kulit atau dawai kokot. Walau bagaimanapun, gam boleh meningkatkan kepuasan pesakit, dan jahitan yang melintasi kulit dan dawai kokot mungkin lebih cepat untuk pakar bedah.
Apakah yang dikaji dalam ulasan ini?
Pakar bedah mempunyai pelbagai pilihan untuk menutup luka pembedahan. Penutupan kulit boleh dilakukan dengan jahitan di bawah kulit, jahitan yang melintasi kulit, dawai kokot (klip), pelekat tisu (gam), pita atau kaedah yang lain. Jahitan boleh diserap (jahitan larut ke dalam tubuh sebagai sebahagian dari proses penyembuhan dan tidak perlu dikeluarkan) atau tidak boleh diserap (jahitan perlu dikeluarkan setelah luka sembuh).
Jangkitan di tapak pembedahan adalah masalah biasa selepas pembedahan dan boleh mengakibatkan pelbagai masalah bagi pesakit. Luka pembedahan juga boleh menyebabkan bekas luka yang tidak sedap dipandang jika tidak sembuh dengan betul. Kami ingin mengetahui bagaimana jahitan yang terdapat di bawah kulit berbanding dengan kaedah lain untuk menutup luka pembedahan dari segi jangkitan, parut, kepuasan pesakit, kos, kesakitan, tempoh pesakit tinggal di hospital dan kualiti hidup.
Apakah keputusan utama daripada ulasan ini?
Pada bulan Mac 2019, kami menyiasat pangkalan data perubatan dan mengenal pasti 66 kajian yang membandingkan jahitan di bawah kulit dengan kaedah penutupan kulit yang lain seperti jahitan standard, dawai kokot kulit, pelekat tisu, pita, atau zip pembedahan. Enam puluh empat kajian (yang melibatkan 7487 peserta) digunakan dalam analisa kami. Secara purata, setiap kajian melibatkan 115 orang. Kebanyakan peserta adalah orang dewasa (20 sehingga 75 tahun) yang menjalani pembedahan di hospital. Kebanyakan kajian tidak melaporkan sumber pembiayaan mereka.
Sebilangan besar kajian membandingkan jahitan yang berada di bawah kulit dengan jahitan standard, dawai kokot kulit atau pelekat tisu.
Hasil utama yang diminati adalah sama ada luka dijangkiti. Tidak ada perbezaan yang jelas antara jahitan yang berada di bawah kulit dan kaedah penutupan lain di kalangan pesakit yang mana luka mereka dijangkiti.
Berbanding dengan jahitan yang melintasi kulit, jahitan di bawah kulit mungkin meningkatkan kepuasan pesakit. Terdapat bukti bahawa jahitan di bawah kulit mungkin mencegah komplikasi luka dan meningkatkan kepuasan pesakit berbanding dengan dawai kokot kulit. Jahitan di bawah kulit mungkin boleh mencegah kerosakan luka (pemisahan kulit) dibandingkan dengan dawai kokot atau pelekat tisu, tetapi pelekat tisu dapat meningkatkan kepuasan pesakit. Walau bagaimanapun, kaedah alternatif mungkin lebih cepat digunakan oleh pakar bedah berbanding dengan jahitan di bawah kulit. Tidak ada perbezaan yang jelas antara jahitan di bawah kulit dan kaedah penutupan alternatif untuk penutupan semula, kesakitan, tempoh pesakit tinggal di hospital dan kualiti hidup.
Kajian yang kami analisa sering melibatkan sebilangan kecil peserta dan, dalam kebanyakan kes, tidak dilaporkan dengan cara yang membolehkan kami untuk memastikan bahawa kajian ini dilakukan secara berpatutan. Oleh itu, kita tidak dapat membuat pernyataan konklusif mengenai keberkesanan jahitan di bawah kulit, dan untuk semua perbandingan kecuali perbandingan dengan dawai kokot, penyelidikan berkualiti yang lebih baik diperlukan untuk membuat kesimpulan yang lebih kuat.
Sejauh manakah ulasan ini dikemaskini?
Kami mencari kajian yang telah diterbitkan sehingga Mac 2019.
Authors' conclusions
Summary of findings
| Subcuticular sutures compared with transdermal sutures for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with transdermal sutures | Risk with subcuticular sutures | |||||
| Surgical site infection (SSI) Incidence of wound infection follow‐up: 7 to 42 days | 71 per 1,000 | 78 per 1,000 | RR 1.10 | 3107 | ⊕⊕⊝⊝ | There may be little difference between subcuticular and transdermal sutures groups in the incidence of SSI. |
| Wound complications Incidence of wound complications follow‐up: 5 to 42 days | 102 per 1,000 | 85 per 1,000 | RR 0.83 | 1489 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures have an effect on wound complications compared with transdermal sutures. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 7 to 42 days | 61 per 1,000 | 21 per 1,000 | RR 0.35 | 866 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with transdermal sutures (as the certainty of the evidence has been assessed as very low) . |
| Cosmesis of scar (cosmesis) follow‐up: 6 months to 12 months | Insufficient data reported. We were unable to carry out further analyses. | ‐ | 950 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures improve the cosmesis of scar compared with transdermal sutures. | |
| Patient satisfaction (at 30 days) | The mean patient satisfaction score (at 30 days) was 7.4 | The mean patient satisfaction score with subcuticular sutures was 1.6 higher (1.32 to 1.88 higher). | MD 1.60 (1.32 to 1.88) | 290 | ⊕⊕⊕⊝ | Patient satisfaction at 30 days is probably higher in subcuticular sutures group compared with transdermal sutures group. |
| Wound closure time (minutes) | The mean wound closure time was 5.40 minutes | The mean wound closure time with subcuticular sutures was 5.81 minutes longer (5.13 to 6.49 minutes longer) | MD 5.81 (5.13 to 6.49) | 585 | ⊕⊕⊕⊝ | Wound closure time is probably longer in subcuticular sutures group compared with transdermal sutures group. |
| Cost | The mean cost was 16 Naira | The mean cost with subcuticular sutures was 8 Naira lower (13.05 lower to 2.95 lower). | MD ‐8.00 (‐13.05 to ‐2.95) | 100 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the cost compared with transdermal sutures. In the study, participants used non‐absorbable (Nylon) subcuticular sutures. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded two levels: one level due to several trials at high risk of bias in at least one domain (attrition, selection, reporting and other bias); one level for imprecision as the confidence intervals overlapped 1 and 1.25. 2 Downgraded three levels: one level due to high risks of bias across varying domains (attrition, selection, reporting and other bias); one level for imprecision as the confidence intervals overlapped 1 and both 0.75 and 1.25); one level for inconsistency. 3 Downgraded three levels: one level due to risk of bias (attrition and selection bias) and two levels due to imprecision (study 95% CIs are wide). 4 Downgraded three levels: one level for high risk of attrition bias; one level for imprecision (narrative synthesis); one level for inconsistency (two reaching significance and two not). 5 Downgraded one level: one level for imprecision (low numbers of participants). 6 Downgraded one level for inconsistency. 7 Downgraded two levels: one level for risk of bias (the risk of bias in the included single study was unclear in almost every domain); one level for imprecision (low numbers of participants). | ||||||
| Subcuticular sutures compared with skin staples for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with skin staples | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 10 to 42 days | 90 per 1,000 | 73 per 1,000 | RR 0.81 (0.64 to 1.01) | 4163 | ⊕⊕⊕⊝ | There is probably little or no difference between subcuticular sutures and skin staples groups in the incidence of SSI. |
| Wound complications Incidence of wound complications follow‐up: 10 to 42 days | 110 per 1,000 | 87 per 1,000 | RR 0.79 | 2973 | ⊕⊕⊕⊝ | Subcuticular sutures probably on average decrease wound complications compared with skin staples. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 10 to 42 days | 59 per 1,000 | 37 per 1,000 | RR 0.63 | 1984 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the risk of wound dehiscence compared with skin staples. |
| Cosmesis of scar follow‐up: 6 months to 1 year | The cosmetic score in the subcuticular sutures group was on average 0.12 SDs (95% CI: 0.11 lower to 0.36 higher) higher in the patients treated with subcuticular sutures than in the patients treated with skin staples. | SMD 0.12 (‐0.11 to 0.35) | 291 | ⊕⊕⊝⊝ | As a rule of thumb, 0.2 SD represents a small effect, 0.5 a moderate effect, and 0.8 a large effect. There may be little or no difference between subcuticular sutures and skin staples groups in cosmesis of scar. | |
| Patient satisfaction (at 30 days) | The mean patient satisfaction score (at 30 days) was 4.2 | The mean patient satisfaction score with subcuticular sutures was 0.20 higher (0.10 to 0.30 higher). | MD 0.20 (0.10 to 0.30) | 1232 | ⊕⊕⊕⊕ | Patient satisfaction at 30 days after surgery is slightly higher in subcuticular sutures group compared with skin staples group. |
| Wound closure time (minutes) | The mean wound closure time ranged from 0.9 to 4.5 minutes | Mean differences ranged between 0.30 and 5.50 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 1384 | ⊕⊕⊝⊝ | Wound closure time may be a few minutes longer in subcuticular sutures group compared with skin staples group. |
| Cost | Three trials favoured subcuticular sutures. It cost almost 5 to 15 USD lower per participant than staples. Another one favoured staples because most of the cost differential was attributed to procedure times. We were unable to carry out further analyses because of insufficient data. | ‐ | 342 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the cost compared with skin staples. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level: one level for imprecision (the confidence intervals overlapped 1 and 0.75). 2 Downgraded one level: one level for imprecision (low numbers of events). 3 Downgraded two levels: one level due to high risks of bias (detection and other bias) in one trial accounting for 24% of the analysis weight; one level for imprecision (low numbers of events). 4 Downgraded two levels: one level due to high risk of bias (attrition and reporting bias); one level for imprecision (the confidence intervals overlapped 0 and minimal clinically important difference). 5 Downgraded two levels: two levels for inconsistency (I2 = 99%). 6 Downgraded three levels: one level for high risk of bias (detection and other bias); one level for imprecision (narrative synthesis); one level for inconsistency. | ||||||
| Subcuticular sutures compared with tissue adhesives for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with tissue adhesives | Risk with Subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 7 to 42 days | 50 per 1,000 | 39 per 1,000 | RR 0.77 | 869 | ⊕⊕⊕⊝ | There is no clear difference in the incidence of SSI between participants treated with subcuticular sutures and those treated with tissue adhesives. Confidence intervals are wide, spanning both appreciable benefits and harms so clear differences between treatments are not apparent. |
| Wound complications Incidence of wound complications follow‐up: 10 to 42 days | 170 per 1,000 | 106 per 1,000 | RR 0.62 | 1058 | ⊕⊕⊝⊝ | There is no clear difference in the incidence of wound complications between participants treated with subcuticular sutures and those treated with tissue adhesives. Although the point estimate on the side of a possible benefit, the 95% confidence intervals includes the possibility of both benefit and harm so clear differences between treatments are not apparent. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 10 to 42 days | 43 per 1,000 | 10 per 1,000 | RR 0.23 | 1155 | ⊕⊕⊝⊝ | Subcuticular sutures may decrease wound dehiscence in comparison with tissue adhesives. |
| Cosmesis of scar assessed with: score system scale from: 1 to 10 (best score) | The study reported there were similar outcomes between the two groups (mean score: subcuticular sutures 8.8 vs tissue adhesives 8.8). We were unable to carry out further analyses because of insufficient data. | ‐ | 99 | ⊕⊕⊝⊝ | There may be little or no difference in cosmesis of scar between subcuticular and tissue adhesives groups. | |
| Patient satisfaction (within 30 days) scale from: 1 to 10 follow up: 14 to 21 days | The mean patient satisfaction score (within 30days) was 9.5 | The mean patient satisfaction score with subcuticular sutures was 2.05 lower (3.05 to 1.05 lower). | MD ‐2.05 (‐3.05 to ‐1.05) | 131 | ⊕⊕⊝⊝ | Patient satisfaction within 30 days after surgery may be lower in subcuticular sutures group compared with tissue adhesives. |
| Wound closure time (minutes) | The mean wound closure time ranged from 0.3 to 3.7 minutes | Mean differences ranged between ‐0.34 and 10.39 across 11 studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 895 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with tissue adhesives (as the certainty of the evidence has been assessed as very low). |
| Cost | The mean cost ranged from 31.96 to 65.1 USD and from 20.3 to 34.01 EUR | Mean differences ranged between ‐57.36 and ‐4.26 USD (‐16.19 and ‐10.30 EUR). Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 422 | ⊕⊝⊝⊝ | Two studies reported the cost by using USD, the others reported the cost by using EUR. It is uncertain whether subcuticular sutures reduce the cost compared with tissue adhesives (as the certainty of the evidence has been assessed as very low). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level: one level for imprecision (the confidence intervals overlapped 1 and both 0.75 and 1.25). 2 Downgraded two levels: one level due to high risks of bias across varying domains (detection, attrition and other bias) accounting for 54% of the analysis weight; one level for imprecision (the confidence intervals overlapped 1 and 0.75). 3 Downgraded two levels: downgraded one level due to high risks of bias across varying domains (detection, attrition and other bias) accounting for 45% of the analysis weight and one level for imprecision (low numbers of events). 4 Downgraded two levels: one level due to high risks of bias (attrition and other bias); one level for imprecision (narrative synthesis). 5 Downgraded two levels: one level due to high risks of bias (attrition and other bias); one level for imprecision (low numbers of participants). 6 Downgraded three levels: one level due to high risks of bias across varying domains (detection, attrition and other bias); two levels for inconsistency (I2= 97%). 7 Downgraded three levels: one level due to high risks of bias across varying domains (attrition and other bias); two levels for inconsistency (I2= 96%). | ||||||
| Subcuticular sutures compared with surgical tapes for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with surgical tapes | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 7 to 30 days | 25 per 1,000 | 33 per 1,000 | RR 1.31 | 354 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of SSI compared with surgical tapes. |
| Wound complications Incidence of wound complications follow‐up: 5 to 42 days | 293 per 1,000 | 263 per 1,000 | RR 0.90 | 492 | ⊕⊕⊝⊝ | There may be little or no difference between subcuticular sutures and surgical tape groups in the incidence of wound complications. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 7 to 42 days | 23 per 1,000 | 2 per 1,000 | RR 0.07 | 264 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical tapes (as the certainty of the evidence has been assessed as very low). |
| Cosmesis of scar | One study reported this outcome, but the data of this trial could not be included as it was insufficient. | |||||
| Patient satisfaction | Not reported in any of the studies. | |||||
| Wound closure time (minutes) | The mean wound closure time ranged from 1.33 to 5.33 minutes | Mean differences ranged between 0.74 and 6.36 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 169 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with surgical tapes (as the certainty of the evidence has been assessed as very low). |
| Cost | Two studies reported the cost per participant was 10‐15 USD higher in subcuticular sutures. The other reported the cost was about 30 USD higher in surgical tapes. We were unable to carry out further analyses because of insufficient data. | ‐ | 315 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures increase the cost with surgical tapes. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded two levels: one level for high risk of bias (attrition bias) and two levels for imprecision (very few events and wide 95% confidence intervals). 2 Downgraded two levels: one level due to high risks of bias across varying domains (attrition, reporting and other bias) accounting for 65% of the analysis weight; one level for imprecision (the confidence intervals overlapped 1 and both 0.75 and 1.25). 3 Downgraded three levels: one level due to high risks of bias across varying domains (attrition and other bias); two levels for imprecision (very few events and wide 95% confidence intervals). 4 Downgraded three levels: one level for imprecision (low numbers of participants); two levels for inconsistency (I2= 90%). 5 Downgraded three levels: one level for high risk of bias (attrition and other bias); one level for imprecision (narrative synthesis); one level for inconsistency (the included studies reported the opposite results, leading to qualitative heterogeneity). | ||||||
| Subcuticular sutures compared with surgical zippers for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with surgical zippers | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 14 to 42 days | 14 per 1,000 | 11 per 1,000 | RR 0.80 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of SSI compared with surgical zippers. |
| Wound complications Incidence of wound complications follow‐up: 14 to 42 days | 83 per 1,000 | 45 per 1,000 | RR 0.55 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound complications compared with surgical zippers. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 14 to 42 days | 32 per 1,000 | 25 per 1,000 | RR 0.78 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical zippers. |
| Cosmesis of scar | The mean cosmesis of scar (VAS) score was 7.7 | The mean cosmetic VAS score with subcuticular sutures was 0.3 lower (0.72 lower to 0.12 higher). | MD ‐0.3 | 90 | ⊕⊕⊝⊝ | There may be little or no difference between subcuticular sutures and surgical zippers groups in the cosmesis of scar. |
| Patient satisfaction | Not reported in any of the studies. | |||||
| Wound closure time (minutes) | The mean wound closure time was 0.76 to 2.1 minutes | Mean differences ranged between 4.38 and 8.25 minutes across three studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 424 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with surgical zippers. |
| Cost | The mean cost was 13 USD | The mean cost with subcuticular sutures was 5 USD lower (8.76 lower to 1.26 lower). | MD ‐5.00 (‐8.76 to ‐1.26) | 120 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the cost compared with surgical zippers. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded three levels: one level due to high risks of bias (attrition and reporting bias) accounting for 58% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 2 Downgraded three levels: one level due to high risks of bias (detection, attrition and reporting bias) accounting for 62% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 3 Downgraded three levels: one level due to high risks of bias (attrition and reporting bias) accounting for 46% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 4 Downgraded two levels: one level due to high risk of bias in blinding of outcome assessment in the single trial; one level for imprecision (low numbers of participants). 5 Downgraded three levels: one level for high risks of bias (detection, attrition and reporting bias); two levels for inconsistency (I2= 100%). 6 Downgraded two levels: one level for risk of bias (unclear in almost every domain in included single study); one level for imprecision (low numbers of participants). | ||||||
Background
Description of the condition
Many people undergo surgical procedures in their lifetime. It is estimated that 312.9 million operations are undertaken every year worldwide (95% confidence interval (CI) 266.2 to 359.5; Weiser 2015). Since Weiser 2015 reported that 18.7 million of these are caesarean deliveries, we can estimate that approximately 250 to 300 million of them are non‐obstetric surgeries. In most operations, surgeons make an incision to gain access to the tissue or organs in which the surgery is performed. After the surgical procedure is complete, they close the incision with various wound closure materials (e.g. sutures, tissue adhesives, surgical tapes, staples) and suturing techniques (Regula 2015; Tajirian 2010).
Wound complications such as surgical site infections (SSI) are among the most common issues reported after surgery, and are often very problematic for patients in terms of cosmetic appearance, decreased quality of life, prolonged hospital stays, and increased healthcare costs (De Lissovoy 2009; Perencevich 2003; Zimlichman 2013).
Incidence of wound complications depends on various risk factors including those related to patients (e.g. comorbidities, medications), those related to operations (e.g. the type of surgery, duration of operation and method of wound closure), and preventive measures (Cardo 2004; Gaynes 2001; Kwon 2013; Mangram 1999; Pull ter Gunne 2012; Talbot 2005; Zhang 2014; Zhang 2015).
In the USA, the Centers for Disease Control and Prevention (CDC) provide guidelines and tools for the healthcare community to help prevent SSI, together with resources to help the public understand these infections and take measures to safeguard their own health when possible. Many preventive measures against SSI are recommended and have spread globally. The incidence of SSI varies and depends on the classification of surgical wounds (Garner 1986). The Garner 1986 guideline categorises operative wound sites into four classes (classes 1 to 4) according to the degree of contamination, that is: clean (class 1), clean‐contaminated (class 2), contaminated (class 3), and dirty or infected (class 4) (Garner 1986). This classification is shown in more detail in Appendix 1. CDC recommend taking different preventive approaches according to each class (Mangram 1999).
Description of the intervention
There are many ways to close surgical incisions, for example, using sutures, staples, and other devices (e.g. tissue adhesives, tapes) (Dumville 2014; Regula 2015; Tajirian 2010). Conventional sutures are usually non‐absorbable interrupted sutures (individual stitches, typically placed transdermally) (Pauniaho 2010). Staples are usually non‐absorbable skin closure clips placed transdermally. Other devices for wound closure include tissue adhesives or tapes, but their use is less widespread due to problems with wound dehiscence (breakdown) (Dumville 2014). In addition, costs are increased because of the high price of adhesive compared with that for subcuticular and other sutures. Brown 2009 reported that, for closure of paediatric hernia incisions, material costs related to skin closure were higher for skin adhesive than for suturing (suture materials USD 11.70 versus skin adhesive USD 22.63; P value < 0.001).
Subcuticular suturing was introduced by Carl Thiersch in 1874. The development of the subcuticular suture sprang from concepts for improving wound healing and avoiding infection (Fisher 1980). Subcuticular suturing became known in the field of plastic surgery in the early 1900s through the efforts of Dr Halsted and Dr Davis (Fisher 1980). 'Subcuticular' means intradermal; i.e. within the layer of the skin (immediately below the epidermal layer). Subcuticular sutures can be either absorbable or non‐absorbable. When non‐absorbable filaments are used, the suture ends are not buried in the skin but exposed outside, which can increase the risk of contamination (Stanec 1997). On the other hand, when absorbable sutures are used, they can be completely buried and retained at or near wound ends (La Paudula 1995; Ranaboldo 1992; Singh‐Ranger 2003; Smoot 1998). Synthetic absorbable filaments (e.g. polyglecaprone, polydioxanone, polyglactin) have only recently become available and are now used widely. Prior to this only natural absorbable filaments (e.g. catgut) were available, but they were rarely used for skin closure due to the risk of infection. With subcuticular sutures, no foreign material reaches beyond the epidermis except for the suture ends. This does not leave any mark points (Kobayashi 2015).
Subcuticular sutures were not previously the preferred method of skin closure except in clean surgery, because of the risk of infection. Since the arrival of synthetic absorbable sutures, their use has been spreading rapidly, not only for CDC class 1 (clean) surgery, but also for class 2 and 3 procedures, partly because wound cosmesis (cosmetic appearance) is currently considered more important than it was previously (Tanaka 2014; Taube 1983). The recent development of suture filaments and surgical devices, and the fact that endoscopic surgery is now more widely performed also lie behind the trend.
How the intervention might work
The use of subcuticular sutures for skin closure is an attractive alternative closure method because of the low incidence of wound complications and good cosmetic appearance it produces (Fisher 1980). With subcuticular sutures, no foreign material reaches beyond the epidermis except for the suture ends. This can obviate the need for postoperative suture removal except for the suture ends and does not leave any mark points (Kobayashi 2015).
Common alternatives to subcuticular sutures are conventional transdermal sutures and staples, both of which have to be removed. Staples are attractive because of speed of application (Gatt 1985; Tajirian 2010), however their cost is higher than that of suture filaments in general.
Compared with staples or conventional transdermal sutures, some clinical trials have shown that subcuticular sutures are associated with a lower incidence of wound complications and better cosmetic results after CDC class 1 (clean) surgery such as: orthopaedic procedures (Shetty 2004), cardiovascular surgery (Angelini 1984; Johnson 1997), and obstetric surgery (Ibrahim 2014; Mackeen 2012; Mackeen 2015). For closure of hip wounds, a cost‐effectiveness study showed that subcuticular sutures were significantly better than clips in terms of wound healing and also in terms of cost (Singh 2006). It has also been reported that the cost incurred for closure of sternal (chest bone) and leg incisions in coronary arterial bypass grafting (CABG) patients was significantly greater when skin clips were used for closure than when sutures were used (Angelini 1984; Chughtai 2000; Johnson 1997). Chughtai 2000 reported a cost of USD 4.5 for each wound closed with sutures and USD 15 for each wound closed with staples. In CDC class 2 (clean‐contaminated) surgery such as gastrointestinal procedures, several randomised controlled trials have shown that subcuticular sutures do not increase the incidence of wound complications (Tsujinaka 2013), and that patients prefer this closure technique because it produces better cosmetic results and less pain (Tanaka 2014).
The advantage of subcuticular sutures may be partly attributable to the use of absorbable sutures (Gurusamy 2014); the advantage of absorbable suture materials is that they do not have to be removed later, which saves surgeons time and decreases the anxiety and discomfort of patients (Parell 2003).
Absorbable sutures may, however, lead to an increased inflammatory response (Parell 2003), and it should be noted that the cost of absorbable suture filaments is higher than that for non‐absorbable filaments.
Why it is important to do this review
Two systematic reviews and two meta‐analyses that evaluated subcuticular sutures in cesarean deliveries have been published. One systematic review did not find conclusive evidence about how the skin should be closed (Mackeen 2012), but the others concluded that there was a possible benefit with subcuticular sutures compared with skin staples, because of a lower incidence of wound complications (Clay 2011; Mackeen 2015; Tuuli 2011).
In the field of non‐obstetric surgery however, there is still controversy about whether subcuticular sutures increase the incidence of wound complications, and, to date, no systematic review has been conducted on this important topic.
One related systematic review entitled 'Continuous versus interrupted skin sutures for non‐obstetric surgery' showed that superficial wound dehiscence (wound separation) may be reduced by using continuous subcuticular sutures (Gurusamy 2014). The authors suggested that this difference might depend on whether sutures were absorbable or not, because most of these wound dehiscences were reported in two recent trials in which the continuous skin suture groups received absorbable subcuticular sutures, while the interrupted skin suture groups received non‐absorbable transcutaneous sutures. In this review, we have focussed on investigating the advantages of subcuticular sutures regardless of whether they are continuous or interrupted.
Objectives
To examine the efficacy and acceptability of subcuticular sutures for skin closure in non‐obstetric surgery.
Methods
Criteria for considering studies for this review
Types of studies
We included all relevant published and unpublished RCTs that compared subcuticular sutures with any other sutures or devices for skin closure in non‐obstetric surgery, irrespective of their sample sizes and language of report.
We had planned to include cluster‐randomised trials when effects of clustering were taken into account (however, we found no such cases). We excluded quasi‐randomised controlled trials (in which treatment assignment is decided through methods such as alternate days of the week). No language or publication status restrictions were imposed.
Types of participants
We included patients of any age and sex undergoing non‐obstetric surgery. We included both outpatients and inpatients with any type of disease and with any comorbidities.
We excluded obstetric operations because there is already a Cochrane Review that addresses methods of skin closure after caesarean sections (Mackeen 2012).
Types of interventions
Subcuticular sutures versus any other sutures or devices for skin closure in non‐obstetric surgery. We made a post hoc decision to exclude studies in which tissue adhesives were used in addition to subcuticular sutures as this represented an additional difference between the groups (see Differences between protocol and review).
Experimental interventions
We included studies that used absorbable and non‐absorbable subcuticular sutures for skin closure, irrespective of whether the sutures were continuous or interrupted.
Comparator interventions
We included studies in which a conventional suture (e.g. transdermal interrupted suture) or a device for skin closure (e.g. staples and other skin closure devices) was used as a control intervention.
Types of outcome measures
Primary outcomes
-
Incidence of surgical site infection (SSI) within 30 days of the operation.
Secondary outcomes
-
Incidence of wound complications (e.g. haematoma, seroma, skin separation) within 30 days of the operation.
When the data allowed, we also presented the results for specific outcome subcategories, such as complications of higher severity or specific type.
-
Incidence of wound dehiscence (skin separation). We added this outcome as a post hoc decision. See Differences between protocol and review.
-
Proportion of re‐closure of the skin incision required within 60 days of the operation.
-
Incidence of hypertrophic scar at maximal follow‐up.
-
Incidence of keloid scar at maximal follow‐up.
-
Wound pain intensity within seven days, and at or after 30 days of the operation (as measured on visual analogue scale, numerical rating scale or other valid instruments).
-
Length of hospital stay (for inpatient surgery, this included any readmissions for wound‐related complications as defined by the authors for a period of one year).
-
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months).
If both self and observer‐rated assessments were available, we gave preference to the latter.
-
Patient satisfaction as defined by the authors within 30 days, and at or after 60 days of the operation.
-
Quality of Life (QoL; short‐term and long‐term as defined by the authors).
-
Wound closure time in the operation (minutes).
-
Cost at maximal follow‐up (as reported by authors).
If both total cost (including time cost) and material cost per patient were available, we gave preference to the latter.
Search methods for identification of studies
Electronic searches
We searched the following electronic databases to identify reports of relevant clinical trials:
-
the Cochrane Wounds Specialised Register (searched 26 March 2019);
-
the Cochrane Central Register of Controlled Trials (CENTRAL; 2019, Issue 2) in the Cochrane Library (searched 26 March 2019);
-
Ovid MEDLINE including In‐Process & Other Non‐Indexed Citations (1946 to 26 March 2019);
-
Ovid Embase (1974 to 26 March 2019);
-
EBSCO CINAHL Plus (Cumulative Index to Nursing and Allied Health Literature; 1937 to 26 March 2019).
The search strategies for the Cochrane Wounds Specialised Register, CENTRAL, Ovid MEDLINE, Ovid Embase and EBSCO CINAHL Plus can be found in Appendix 2. We combined the Ovid MEDLINE search with the Cochrane Highly Sensitive Search Strategy for identifying randomised trials in MEDLINE: sensitivity‐ and precision‐maximising version (2008 revision) (Lefebvre 2011). We combined the Embase search with the Ovid Embase filter developed by the UK Cochrane Centre (Lefebvre 2011). We combined the CINAHL Plus searches with the trial filters developed by the Scottish Intercollegiate Guidelines Network (SIGN 2018). There were no restrictions with respect to language, date of publication or study setting.
We also searched the following clinical trials registries:
-
ClinicalTrials.gov (www.clinicaltrials.gov) (searched 22 March 2019);
-
World Health Organization (WHO) International Clinical Trials Registry Platform (http://apps.who.int/trialsearch/Default.aspx) (searched 22 March 2019);
-
EU Clinical Trials Register (https://www.clinicaltrialsregister.eu/ctr‐search/search) (searched 22 March 2019);
-
University Hospital Medical Information Network Clinical Trials Registry (UMIN‐CTR) (www.umin.ac.jp/ctr/index‐j.htm) (searched 22 March 2019).
Search strategies for clinical trial registries can be found in Appendix 2.
Searching other resources
We searched the bibliographies of all retrieved and relevant publications identified by these strategies for further studies. We checked the reference lists of all included studies and relevant systematic reviews to identify additional studies missed from the original electronic searches.
A citation search was also conducted on the Web of Science to identify articles that cited any of the included studies.
We contacted experts and industry representatives to enquire about unpublished or ongoing studies. We contacted suture manufactures, such as Ethicon and Covidien, but received no response.
Data collection and analysis
Data collection and analysis was carried out according to the methods stated in the published protocol (Goto 2016) which were based on the Cochrane Handbook for Systematic reviews of Interventions (Higgins 2011).
Selection of studies
Two review authors (SG and KH) examined the titles and abstracts of references identified by the electronic search strategies described above to determine which were likely to be relevant. We obtained the full text for each potentially relevant study. These two authors assessed each article independently, and decided whether to include the study in the meta‐analysis. Disagreement between authors was resolved by discussion. Arbitration was provided by a third author (TAF). Agreement between review authors in the study selection was reported. The disagreement in the selection of studies was evaluated by quantifying both the percentage of agreement and Cohen's kappa (k) (Cohen 1960). These are the methods to measure interrater reliability (McHugh 2012). Cohen's kappa gives a score of how much homogeneity or consensus there is in the ratings given by judges. Cohen suggested the Kappa result be interpreted as follows: values ≤ 0 as indicating no agreement and 0.01 to 0.20 as none to slight, 0.21 to 0.40 as fair, 0.41 to 0.60 as moderate, 0.61 to 0.80 as substantial, and 0.81 to 1.00 as almost perfect agreement.
When missing information inhibited the evaluation of a study, we classified the study as a 'study awaiting assessment' and sought further information from the original authors or other possible sources. We described the reasons for exclusion of studies for which we obtained full copies of the text in the 'Characteristics of excluded studies' table. The study selection process is reported in a PRISMA flow diagram to summarise this process (Liberati 2009).
When studies were reported in multiple publications/reports, we obtained all publications. Whilst the study was included only once in the review, we extracted data from all reports to ensure all available relevant data were obtained.
Data extraction and management
Independently, at least two of three review authors (SG and TS or RG) extracted information from the included trials using a structured, pilot‐tested, Excel data extraction form. Any disagreement was resolved either by discussion or by consultation with a fourth author (KH). If necessary, authors of studies were contacted to obtain further clarification. Agreement between the data extractors with regard to the primary outcome was reported. This is one of the methods to measure interrater reliability (McHugh 2012).
This data extraction form included the following items:
-
general information: title, authors, and year of publication of the first report;
-
study characteristics including design, setting, country, and duration of the study;
-
participants: total number; number of each age, sex, and comorbidity; type of surgery; and wound class;
-
interventions and comparisons: total number of intervention groups, type of interventions, and type of suture materials and suturing method in each arm;
-
outcomes: definition of outcomes, number of participants allocated to each intervention group, sample size, number of missing participants, number of events (dichotomous outcomes), standard deviation (SD) and mean (continuous outcomes), timing of assessment, and duration of follow‐up;
-
risk of bias and publication status.
Assessment of risk of bias in included studies
Independently, at least two out of three review authors (SG and TS or RG) assessed the risk of bias of the included studies using the criteria described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). If assessors disagreed, the final rating was made by discussion or with the involvement of an additional assessor (TAF), if necessary. Agreement between the two independent raters in the 'Risk of bias' assessment was reported as percentage agreement and weighted kappa. The following domains were assessed (see Appendix 3):
-
random sequence generation;
-
allocation concealment;
-
blinding of participants and personnel;
-
blinding of outcome assessment;
-
incomplete outcome data;
-
selective outcome reporting;
-
other bias (distribution of baseline characteristics, industry funding etc.).
We assessed blinding and incomplete outcome data for each of the review outcomes separately. In the 'Risk of bias' table, we have presented the risk of blinding and incomplete outcome data mainly focussing on the short‐term postoperative outcomes including SSI, wound complications and wound dehiscence. For the GRADE assessment for the long‐term outcomes such as cosmesis of scar, patient satisfaction and QOL, we evaluated relevant 'Risk of bias' domains. In this review, we anticipated that blinding of participants and personnel may not be possible. For this reason, the assessment of the risk of blinding focused on whether blinded outcome assessment was reported: blinding of assessment is especially important because assessment of wound outcomes, such as SSI, wound complications and dehiscence, can be subjective. We used blinding of outcome assessment to determine risk of bias from blinding in these instances. Although we recorded risk of bias for blinding of personnel and participants, we did not downgrade the certainty of the evidence for this alone, where the nature of the comparison made it highly likely.
The risk of bias in each domain was assessed and categorised into:
-
low risk of bias, i.e. plausible bias that is unlikely to alter the results seriously;
-
high risk of bias, i.e. plausible bias that seriously weakens confidence in the results;
-
unclear risk of bias, i.e. plausible bias that raises some doubt about the results.
Where inadequate details of randomisation and other characteristics of trials were provided, the risk of bias was classified as unclear, unless further information could be obtained by contacting the authors. We provided a quote from the study report together with a justification for our judgement in the 'Risk of bias' table. We summarised the 'Risk of bias' judgements across different studies for each of the domains listed. Where information on risk of bias related to unpublished data or correspondence with a trialist, we have noted this in the 'Risk of bias' table.
Measures of treatment effect
We used Review Manager 5 (Review Manager 2014) to analyse the data. We identified both dichotomous data and continuous data.
Dichotomous data
For binary outcomes, we presented results as the risk ratio (RR) with 95% CI, because risk is a concept that is more familiar and simpler to understand for clinicians than odds.
Continuous data
Wherever possible, we expressed continuous data as mean difference (MD) with 95% CI. In cases where different scales were used to measure the same or similar construct, we used the standardised mean difference (SMD) with 95% CI for continuous outcomes.
Endpoint versus change data
We used endpoint data, which typically cannot have negative values and are easier to interpret from the clinical point of view. If endpoint data were not available, we had planned to use the change data, however, we found no such cases. We considered this strategy to be less prone to selective reporting.
Time‐to‐event data
For time‐to‐event data, we planned that our primary effect measure would be the hazard ratio (HR) with 95% CI. However, we found no studies which reported this type of data.
Skewed data
To avoid analysing skewed data as normally distributed data, we applied the following standards to all data before inclusion.
-
We entered data from studies of at least 100 participants into the analysis irrespective of the following rules, because skewed data pose less of a problem in large studies.
-
Endpoint data: when a scale started from the finite number zero, we subtracted the lowest possible value from the mean and divided this by the standard deviation if the data were reported.
-
If this value was lower than 1.0, it strongly suggested a skew and we excluded the study from meta‐analytic pooling and presented it narratively.
-
If this ratio was higher than 1.0 but below 2.0, there was suggestion of a skew. We entered the study in the analysis and tested whether its inclusion or exclusion changed the results substantially.
-
If the ratio was larger than 2.0, the study was included in the analysis because skew was less likely (Altman 1996; Higgins 2011).
-
-
When continuous data are presented on a scale that includes the possibility of negative values (such as change data), it is difficult to tell whether data are skewed or not. We had planned to enter such studies into the analysis because change data tend to be less skewed than other data and because excluding studies also leads to bias, as not all the available information is used. However, we found no such cases.
-
A common way that trialists indicate that they have skewed data is by reporting medians and interquartile ranges. When we encountered this, we noted that the data were skewed and the study was excluded from meta‐analytic pooling and was summarised narratively.
Unit of analysis issues
Cluster‐randomised trials
In cluster‐randomised trials, groups of individuals rather than individuals are randomised to different interventions (Higgins 2011). In this review, no cluster‐randomised trials were identified.
In future versions of this review, when cluster‐randomised trials are analysed as if the randomisation was performed on the individuals rather than the clusters, we will perform approximately correct analyses (Higgins 2011). The idea is to reduce the size of each trial to its 'effective sample size' (Rao 1992). The effective sample size of a single intervention group in a cluster‐randomised trial is its original sample size divided by a quantity called the 'design effect'. The design effect calculated by the equation: 1 + (M – 1) ICC, where M is the average cluster size and ICC is the intra‐cluster correlation coefficient. A common design effect is usually assumed across intervention groups. For dichotomous data, both the number of participants and the number experiencing the event will be divided by the same design effect. For continuous data, only the sample size will be reduced; means and standard deviations should remain unchanged.
Multiple body parts Ⅰ: body parts receive the same intervention
Where studies were randomised at the participant level and outcomes measured at the wound level, we treated the participant as the unit of analysis when the number of incisions (wounds) assessed appeared to be equal to the number of participants (e.g. one wound per person).
In some studies where people were randomised and multiple wounds of the body received the same intervention (e.g. multiple wounds per participant or perhaps only on some participants), a separate outcome judgement was made for each wound, and the number of wounds was used as the denominator in the analysis. Since not all participants had multiple wounds, this was not a cluster trial per se but rather a trial that incorrectly included a mixture of individual and clustered data.
In cases where included studies contained some or all clustered data, we reported this, noting whether data had been (incorrectly) treated as independent. We noted this situation in the other risk of bias of Characteristics of included studies and performed a post hoc sensitivity analysis excluding these studies.
Multiple body parts Ⅱ: body parts receive different interventions
If multiple wounds were randomised to different groups, we had planned to include the trial only if appropriate analysis was undertaken to take within‐subject correlation into account (paired data), or if it was possible to perform such an analysis using the available data. However, it was often not clear whether such analysis had been undertaken. We noted this situation in the 'other risk of bias' of Characteristics of included studies and performed a post hoc sensitivity analysis excluding these studies.
While we accepted the results from trials in which multiple wounds were randomised to different intervention groups (split‐body design), we excluded trials in which a part of the wound was randomised to one intervention and the rest of the wound to another intervention (split‐wound design).
These trials have similarities to cross‐over trials: in cross‐over trials, individuals receive multiple treatments at different times, while in these trials they receive multiple treatments at different sites.
Multiple intervention groups
The studies that compared more than two intervention groups were included in meta‐analysis by making multiple pairwise comparisons between all possible pairs of intervention groups. If two or more interventions were compared with control and were eligible for the same meta‐ analysis, we pooled the intervention arms and compared these with control. We used combined group data, as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011).
Dealing with missing data
We tried to contact authors of trials to obtain any missing data.
Missing participants
Dichotomous data
All dichotomous data were analysed on the basis of the intention‐to‐treat (ITT) principle. When participants had been withdrawn from a trial and the original authors had not imputed the data appropriately, we assumed that the condition of these participants would have remained unchanged if they had stayed in the trial, or we treated them as treatment failures (a 'worst‐case' scenario). We performed sensitivity analyses to assess how sensitive results were to reasonable changes in the assumptions. We addressed the potential impact of missing data on the findings of the review in the Discussion.
Continuous data
We used continuous data as they were presented by the original authors, without any imputations according to the recommendation in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Whenever ITT data were presented by the authors, we preferred these to 'per protocol or completer' data sets.
Missing data
We contacted investigators or study sponsors in order to obtain numerical outcome data where possible (e.g. when a study was identified as abstract only).
Missing statistics
When only P values or standard error (SE) values were reported, we calculated standard deviations (SDs) (Altman 1996). In the absence of supplemental data after requests to the original authors, we calculated SDs from CIs, T values, or P values as described in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). Otherwise, we had planned to impute them from other studies in the meta‐analysis according to a validated method developed by Furukawa (Furukawa 2000). However, we could not use this method because SDs of wound closure time and length of hospital stays varied greatly by study.
Assessment of heterogeneity
Initially, we investigated heterogeneity by visual inspection of the forest plots. We performed the Chi2 test in order to detect the presence of heterogeneity. We regarded heterogeneity as present if there was a low P value (less than 0.10) in the Chi2 test for heterogeneity. Since the Chi2 test has low power to assess heterogeneity when a small number of participants or studies are included, we set the probability at the 10% level of significance. We also calculated the I2 statistic in order to assess the degree of heterogeneity (Higgins 2002). The I2 statistic is defined as the proportion of total heterogeneity that exceeds what would be expected due to chance (Higgins 2003). It is interpreted as recommended in the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011), as follows: 0% to 40% may not be important; 30% to 60% may represent moderate heterogeneity; 50% to 90% may represent substantial heterogeneity; 75% to 100% represents considerable heterogeneity.
We also reported Tau2, that is, the between study variance in random‐effects model meta‐analyses.
If apparent upon visual inspection of the forest plots or if there was statistically substantive heterogeneity (I2≥50%), we investigated its potential sources through subgroup and sensitivity analyses.
Assessment of reporting biases
We assessed publication bias by a funnel plot if the number of studies included was 10 or more. We investigated the presence of small study effects for the primary outcome only; along with visual inspection of the plots, we used Egger's test to examine whether the association between estimated intervention effects and the study size was greater than might be expected to occur by chance (Egger 1997).
Data synthesis
We performed meta‐analyses according to the recommendations of the Cochrane Handbook for Systematic Reviews of Interventions (Higgins 2011). We used a random‐effects model in the analysis because we considered that the different studies would estimate different, yet related, intervention effects (DerSimonian 1986). We used a fixed‐effect model for the sensitivity analysis.
With regard to dichotomous outcomes, risk ratio calculations did not include trials in which no events occurred in either group in the meta‐analysis, whereas risk difference calculations did. We reported the risk difference (RD) if the results using this association measure were different from the risk ratio in terms of statistical significance. However, the risk ratio is the measure that we planned to use to arrive at conclusions, since risk ratios perform better when there are differences in the control event rate (proportion of participants who develop the event in the control groups) (Furukawa 2002).
When there was extreme heterogeneity (I2≥90%), we did not pool the data and presented the results narratively.
Subgroup analysis and investigation of heterogeneity
Subgroup analyses are often exploratory in nature and should be interpreted cautiously. Firstly, because these analyses often involve multiple analyses, they may yield false positive results; secondly, these analyses lack power and are more likely to result in false negative results. With these reservations in mind, we performed the following subgroup analysis for the primary outcome only and only where there were sufficient studies:
-
absorbable versus non‐absorbable subcuticular sutures;
-
location of surgery on the body (trunk, extremities, and face) as wound healing rates may be different;
-
CDC class 1 (clean) versus class 2 (clean‐contaminated) versus class 3 (contaminated) surgery;
-
continuous versus interrupted skin sutures;
-
endoscopic (e.g. laparoscopic, thoracoscopic and arthroscopic) versus open surgery.
Sensitivity analysis
The process of undertaking a systematic review and meta‐analyses involves a sequence of decisions that may be somewhat arbitrary or unclear (Higgins 2011). A sensitivity analysis is a repeat of the primary analysis in which alternative decisions or ranges of values are substituted for decisions that were arbitrary or unclear. We performed the following sensitivity analyses for the primary outcome only:
-
restricted inclusion in the analysis to only those studies that are considered to be at a low risk of selection bias (i.e. adequate allocation sequence generation and adequate allocation concealment). Since it was impossible for both the operators of the procedure and assessors of postoperative short‐term outcomes to be blinded to the intervention, we did not use blinding of personnel and outcome assessment as a marker of trial quality.
-
examined handling of missing participants firstly by ITT analysis based on the worst‐worst scenario assuming that the dropouts in both the intervention and the control groups had the event of interest, and secondly by the worst‐best scenario, assuming that the dropouts in the intervention had the event of interest while those in the control did not.
-
excluded studies sponsored by companies that produce suture devices, as they have an inevitable conflict of interest.
-
for meta‐analyses, use a fixed‐effect model instead of a random‐effects model.
-
excluded studies that had unit of analysis issues. See 'Differences between protocol and review'.
'Summary of findings' tables and assessment of the quality of the evidence using the GRADE approach
We have presented the main results of the review in 'Summary of findings' tables, which provide key information concerning the quality of evidence, the magnitude of the effect of the interventions examined, and the sum of the available data on the main outcomes, as recommended by Cochrane (Schünemann 2011a). The 'Summary of findings' tables also included an overall grading of the body of evidence related to each of the main outcomes using the GRADE (Grading of Recommendations Assessment, Development and Evaluation) approach (Schünemann 2011b).
The GRADE approach defines the quality of a body of evidence as the extent to which one can be confident that an estimate of effect or association is close to the true quantity of specific interest. The quality of a body of evidence involves consideration of within‐trial risk of bias (methodological quality), directness of evidence, heterogeneity, precision of effect estimates, and risk of publication bias (Schünemann 2011b). We presented the following outcomes in the 'Summary of findings' tables:
-
incidence of surgical site infection (SSI) within 30 days of the operation;
-
incidence of wound complications (e.g. haematoma, seroma, skin separation) within 30 days of the operation;
-
incidence of wound dehiscence (skin separation) (we added this outcome as a post hoc decision);
-
cosmesis of scar (as defined by the authors for a minimum follow‐up of six months);
-
patient satisfaction (as defined by the authors, within 30 days, and at, or after, 60 days of the operation);
-
wound closure time in the operation (minutes);
-
cost at maximal follow‐up (as reported by authors).
Please see Differences between protocol and review for changes to this section.
For relevant outcomes reported for comparisons not listed above, we presented GRADE assessments narratively within the Results section without inclusion in a 'Summary of findings' table.
In terms of the GRADE assessment, when making decisions for the risk of bias domain, we downgraded only when studies were classed at high risk of bias for one or more domains. We did not downgrade for unclear risk of bias assessments. In assessing the precision of effect estimates, we also followed GRADE guidance (GRADE 2013); we assessed the size of confidence intervals, downgrading twice for imprecision when there were very few events and CIs around effects included both appreciable benefit and appreciable harm.
Results
Description of studies
See Characteristics of included studies; Characteristics of excluded studies; Characteristics of studies awaiting classification and Characteristics of ongoing studies.
Results of the search
The number of references identified by the searches was 1606. Of these, 1480 remained after de‐duplication. We excluded 1321 references after assessment of the titles and abstracts. We retrieved a total of 159 full‐text papers for full inspection. Of these studies, we excluded 64 (65 full‐text articles) with reasons, sixteen were ongoing trials and five presented too little information to be classified. We included the remaining 66 studies (73 full‐text) in the final qualitative analyses; among these, we included 64 in the final quantitative analyses (50 for the primary outcome analyses). See Figure 1 for a PRISMA flow diagram depicting the study selection process. Cohen's weighted kappa among assessors for the selection was 0.963 (percentage of agreement = 96.2%). This is the method to measure interrater reliability (McHugh 2012). It was interpreted as almost perfect agreement.
Study flow diagram.
Included studies
We included 66 studies in this review, of which we included 64 in quantitative analyses (50 for the primary outcome analyses). We could not include two studies (Fiennes 1985; Sakka 1995) in quantitative analyses because no data were presented in one study (Fiennes 1985) and the other did not report the number randomised in each group (Sakka 1995); we attempted to contact the authors, but received no reply. The characteristics of the included studies can be summarised as follows (see Characteristics of included studies).
Design
All included studies were randomised controlled trials. Only seven studies had a multicentre design (Imamura 2016; Khan 2006; Kobayashi 2015; Liu 2017; Maartense 2002; Obermair 2007; Tsujinaka 2013). All trials were of parallel group design except for five that had a split‐body design (different wounds on same participant randomised (Anatol 1997; Buchweitz 2005; Rebello 2009; Rosen 1997; Subramanian 2005).
Eleven included trials had more than two arms: eight studies had three arms (Anatol 1997; Barker 1984; Khan 2006; Maartense 2002; Obermair 2007; Rosen 1997; Simpson 1979; Zwart 1989) and three studies had four arms (Eggers 2011; Mullen 1999; Murphy 1995).
Country
Twenty‐four studies were conducted in the United Kingdom. Eight studies were conducted in the United States, five studies were conducted in Japan and the Netherlands, three studies were conducted in Australia, China, Iran and two studies were conducted in Canada, Finland and Nigeria. Other studies were conducted in Denmark, Germany, India, Italy, Malaysia, Mexico, Singapore, Turkey and Trinidad.
Sample sizes
The total sample size varied between eight (50 incisions) (Rebello 2009) and 1264 participants (Kobayashi 2015), with a mean sample size of 115 participants per study. The total number of participants included in the analyses is 7487.
Participants
The majority of studies included only adults, except for nine studies that included adults and children (Anatol 1997; Brown 2009; Hopkinson 1982; Keng 1989; Khajouei 2007; McGreal 2002; Onwuanyi 1990; Tanaka 2016; Taube 1983;) and seven studies that included only children (Ademuyiwa 2009; Grottkau 2010; Ong 2002; Pauniaho 2010; Rebello 2009; Van den Ende 2004; Xu 2014).
Thirty‐seven studies included only CDC class 1 (clean) surgeries such as: orthopaedic procedures (Baek 2009; Eggers 2011; Grottkau 2010; Khan 2006; Rebello 2009; Roolker 2002; Sakka 1995; Shetty 2004; Xu 2014), cardiovascular surgery (Chughtai 2000; Corder 1991; Karabay 2005; Krishnamoorthy 2009; Lazar 2011; Mullen 1999; Tanaka 2016), surgeries through groin crease incision (Ademuyiwa 2009; Anatol 1997; Brown 2009; Keng 1989; Murphy 1995; Ong 2002; Subramanian 2005; Swtizer 2003; Van den Ende 2004), breast surgeries (Barker 1984; Gennari 2004; Steele 1983), port or pacemaker implantation (Chen 2013; Martin 2017; Pitcher 1983), and neck or facial surgeries (Liu 2017; O'Leary 2013; Reed 1997; Selvadurai 1997; Soni 2013; Teoh 2018).
Eleven studies included only CDC class 2 (clean‐contaminated) surgeries such as: gastrointestinal surgeries (Jallali 2004; Kobayashi 2015; Tanaka 2014; Tsujinaka 2013), hepatectomy (Chen 2018), urological surgeries (Sebesta 2004), and gynaecological surgeries (Buchweitz 2005; Jan 2013; Kuroki 2017; Obermair 2007; Rosen 1997).
Ten studies included appendectomies (Andrade 2016; Foster 1977; Ghaderi 2010; Hopkinson 1982; Javadi 2018; Khajouei 2007; Kotaluoto 2012; McGreal 2002; Onwuanyi 1990; Pauniaho 2010). Appendectomies could be classified CDC class 2 to 4. In eight other trials, the contamination level in the surgeries included was variable (Clough 1975; Fiennes 1985; Imamura 2016; Maartense 2002; Ranaboldo 1992; Simpson 1979; Taube 1983; Zwart 1989).
Although the majority of studies enrolled people who underwent CDC class 1 (clean) surgeries, two‐thirds of participants were enrolled in the studies which included CDC class 2 to 4 surgeries such as appendectomies and gastrointestinal surgeries.
Characteristics of interventions
All the included studies used absorbable subcuticular sutures except for eleven studies that used non‐absorbable subcuticular sutures (Ghaderi 2010; Hopkinson 1982; Khajouei 2007; Liu 2017; Onwuanyi 1990; Selvadurai 1997; Soni 2013; Steele 1983; Subramanian 2005; Tanaka 2016; Taube 1983). Two studies used absorbable and non‐absorbable subcuticular sutures (Barker 1984; Simpson 1979). Twenty‐three studies also used subcuticular continuous sutures (Anatol 1997; Andrade 2016; Baek 2009; Chughtai 2000; Eggers 2011; Gennari 2004; Hopkinson 1982; Khan 2006; Kotaluoto 2012; Krishnamoorthy 2009; Lazar 2011; Liu 2017; Martin 2017; McGreal 2002; Murphy 1995; O'Leary 2013; Obermair 2007; Pauniaho 2010; Pitcher 1983; Ranaboldo 1992; Selvadurai 1997; Swtizer 2003; Tanaka 2016) and eight studies used subcuticular absorbable interrupted sutures (Anatol 1997; Chen 2018; Fiennes 1985; Imamura 2016; Kobayashi 2015; Maartense 2002; Tanaka 2014; Tsujinaka 2013). The remaining studies did not report the method of continuous or interrupted sutures. Two studies did not report the nature of the suture material used (Gennari 2004; Mullen 1999).
Twenty‐five of 66 included studies compared subcuticular sutures with transdermal sutures for skin closure (Andrade 2016; Baek 2009; Buchweitz 2005; Chen 2013; Clough 1975; Corder 1991; Fiennes 1985; Foster 1977; Ghaderi 2010; Hopkinson 1982; Javadi 2018; Karabay 2005; Khajouei 2007; Kotaluoto 2012;Liu 2017; McGreal 2002;Murphy 1995; Onwuanyi 1990; Pauniaho 2010; Rosen 1997; Sakka 1995; Simpson 1979; Taube 1983; Tanaka 2014; Zwart 1989). Two of these studies had a third arm using staples (Murphy 1995; Zwart 1989) and one of these studies had a third arm using surgical tapes (Rosen 1997). All studies used non‐absorbable transdermal sutures except for two studies (Buchweitz 2005; Chen 2013). All studies also used transdermal interrupted sutures except for three studies (Murphy 1995; Sakka 1995; Zwart 1989).
Eighteen studies compared subcuticular sutures with staples (Chen 2018; Chughtai 2000; Eggers 2011; Imamura 2016; Khan 2006; Kobayashi 2015; Kuroki 2017; Mullen 1999; Murphy 1995; Obermair 2007; Ranaboldo 1992; Reed 1997; Selvadurai 1997; Shetty 2004; Steele 1983; Subramanian 2005; Tsujinaka 2013; Zwart 1989) and two of these studies also had a third arm using tissue adhesives (Eggers 2011; Khan 2006).
Seventeen trials compared subcuticular sutures with tissue adhesives (Ademuyiwa 2009; Brown 2009; Eggers 2011; Gennari 2004; Jallali 2004; Jan 2013; Khan 2006; Keng 1989; Krishnamoorthy 2009; Maartense 2002; Martin 2017; Ong 2002; Sebesta 2004; Soni 2013; Swtizer 2003; Teoh 2018; Van den Ende 2004) and one of these studies also had a third arm using surgical tapes (Maartense 2002).
Nine studies compared subcuticular sutures with surgical tapes (Anatol 1997; Barker 1984; Grottkau 2010; Lazar 2011; Maartense 2002; O'Leary 2013; Pitcher 1983; Rebello 2009; Rosen 1997). Two of these studies used Steri‐strips® (Maartense 2002; O'Leary 2013; Rosen 1997), three of these studies used Steri‐Strip S® (Grottkau 2010; Lazar 2011; Rebello 2009) and two studies used Opsite®(Barker 1984; Pitcher 1983). Anatol 1997 did not report details of the tapes.
Three studies (Roolker 2002; Tanaka 2016; Xu 2014) compared subcuticular sutures with a new skin closure device, namely surgical zipper.
Funding source
The majority of included studies did not report funding or conflict of interest. Fifteen studies reported the funding of source clearly (Ademuyiwa 2009; Chen 2018; Eggers 2011; Grottkau 2010; Imamura 2016; Jan 2013; Kobayashi 2015; Krishnamoorthy 2009; Kuroki 2017; Lazar 2011; Liu 2017; Martin 2017; Tanaka 2014; Tsujinaka 2013; Xu 2014). Of these studies, four studies received corporate/industry funds (Grottkau 2010; Lazar 2011; Liu 2017; Tsujinaka 2013).
Primary outcome measures
Fifty of the included trials reported incidence of SSI. Agreement between the data extractors with regard to the primary outcome was 97%. The definition of infection varied and the time of postoperative wound examination varied between studies. Only seven trials defined infection according to the CDC criteria for SSI (Chen 2018; Imamura 2016; Kobayashi 2015; Kotaluoto 2012; Kuroki 2017; Tanaka 2014; Tsujinaka 2013), which is considered to be the gold standard definition for wound infection (Mangram 1999). Eleven trials used author‐defined clinical criteria (Andrade 2016; Buchweitz 2005; Chughtai 2000; Eggers 2011; Javadi 2018; Khan 2006; Liu 2017; Maartense 2002; Mullen 1999; Pauniaho 2010; Roolker 2002), and one trial used a self‐devised wound scale to define infection (Karabay 2005). Seven trials defined infection if pus was discharged (Clough 1975; Corder 1991; Foster 1977; Hopkinson 1982: Keng 1989; Ranaboldo 1992; Taube 1983), and three trials required positive wound swabs to define infection (Lazar 2011; Murphy 1995; Shetty 2004). The remaining trials did not record the definition of infection used.
Excluded studies
The characteristics of the excluded studies can be summarised as follows (see Characteristics of excluded studies).
We excluded a total of 65 studies (66 references). Five studies were not RCTs (Bernstein 2001; Navali 2014; Serour 1996; Singh 2006; Watson 1983). Seven studies were quasi‐randomised (Angelini 1984; Cassie 1988; Clayer 1991; Davies 1995; Elliot 1989; Matin 2003; Ralphs 1982). Three references were letters or comments on excluded trials (Cordova 2013; Ries 2016; Watts 1982).
Three studies investigated methods of fascial closure (layers of connective tissue that surround muscles and other structures) (Erel 2001; Greene 1999; Kharwadkar 2005). Five studies investigated methods of closure of other layers of the wound (Cameron 1987; Chan 2017; Leaper 1985; Liang 2015; Nair 1988). Shanahan 1990 investigated methods of dressing. Four studies were not comparisons of subcuticular sutures with any other skin closure methods (Meinke 1996; Nipshagen 2008; Pickford 1983; Plotner 2011). In 15 studies, it was not clear whether the study involved a comparison of subcuticular sutures versus any other skin closure methods (Bernard 2001; Cheng 1997; Eldrup 1981; Gatt 1985; Handschel 2006; Harvey 1986; McLean 1980; Menovsky 2004; Risnes 2001; Sadick 1994; Selo‐Ojeme 2002; Shamiyeh 2001; Sinha 2001; Szabó 2002; Van de Gevel 2010).
Two studies did not assess a relevant wound type (Milone 2014; You 2016), five studies did not assess eligible outcomes of our review, although these studies reported their own outcomes (Alicandri‐Ciufelli 2014; Consorti 2013; Lombardi 2011; Rizvi 2018; Wyles 2016), and three studies were split‐wound design (Johnson 1997; Kerrigan 2010; Richter 2012).
We excluded seven studies because both arms in the study received subcuticular sutures (Blondeel 2014; Buttaro 2015; Park 2015) or dermal sutures (Koonce 2015; Nahas 2004; Parvizi 2013; Rui 2017). We judged that these studies did not focus on subcuticular sutures for skin closure, but on superficial adjunct wound closure methods. In four studies (Glennie 2017; Lalani 2016; Lazar 2008; Mudd 2013), all participants in the intervention group had tissue adhesives (Dermabond) placed. We excluded these studies because they assessed a mixture of subcuticular sutures and tissue adhesives within the same intervention group. See also Differences between protocol and review.
Ongoing studies and studies awaiting classification
We identified 16 ongoing studies. Eight studies are exploring subcuticular suture versus skin staples (one of these, absorbable staples) (ACTRN12611000399998; CTRI/2018/08/015470; NCT02046239; NCT02936063; NCT03108742; NCT03788239; UMIN000002873; UMIN000003235), five studies are exploring subcuticular suture versus tissue adhesives (CTRI/2018/02/011698; ISRCTN80786695; ISRCTN96030942; NCT01996917; NCT02551510) and three studies are exploring subcuticular suture versus transdermal suture (Maschuw 2014; IRCT20161217031440N1; IRCT20180820040840N1). For further details, see Characteristics of ongoing studies. Five potentially eligible studies have not yet been incorporated into the review. We are awaiting full text for three of these studies (Choudry 1996; Rubio‐Perez 2014; Zhang 2011) and awaiting data for another two studies. In these two studies, we could not use the data for skin closure because they were combined with data for other methods of sutures (Lubowski 1985) and for laceration closure (Singer 2002); we attempted to contact the authors to request data by subgroup, but received no reply. Details of these studies are presented in the table of Characteristics of studies awaiting classification.
Risk of bias in included studies
None of the included studies were at low risk of bias for all domains. All the studies had an unclear or high risk of bias for two or more domains. Given the nature of the intervention, participants and caregivers may not be blinded and therefore almost all of the studies were rated at high or unclear risk of bias for this domain (performance bias). With regard to the other domains where risk of bias could be avoided, there were 32 studies with one or more domains classed at high risk of avoidable bias and we rated seven of these studies as being at high risk of avoidable bias in more than one domain. Most studies had multiple domains which were at unclear risk of bias. For only one domain (reporting bias), we considered the majority of the studies to be at low risk of bias. For details of the 'Risk of bias' judgements for each study, see Characteristics of included studies. A graphical representation of the overall risk of bias in included studies is presented in Figure 2 and Figure 3. The agreement between the two independent raters in the 'Risk of bias' assessment ranged between 62% and 96%, with weighted Kappa between 0.88 to 0.99.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Allocation
Sequence generation was performed adequately in 31 studies. Almost all these studies used random sequences generated by computerised randomisation programmes except for Anatol 1997, Chen 2013, Pauniaho 2010, Rosen 1997 and Selvadurai 1997. Anatol 1997 used throwing of a dice, Chen 2013 used envelopes, Pauniaho 2010 used coin tossing, Rosen 1997 used random number draws and Selvadurai 1997 used a table of random numbers. The remaining studies did not report enough information to enable a judgement. We considered these studies to be at unclear risk of bias (sequence generation).
The majority of studies did not report the details about allocation concealment. Some studies gave evidence that they used envelopes that were sealed and opaque, but lacked sufficient evidence that they were sequentially numbered. Many other studies provided no information on allocation of the randomisation sequence at all. We assessed only five studies as being at low risk of bias for both sequence generation and allocation concealment (Imamura 2016; Kobayashi 2015; Kuroki 2017; Tanaka 2014; Tsujinaka 2013). Agreement between the two independent raters in the risk of allocation bias assessment was 96% (weighted Kappa 0.99 for sequence generation and 0.99 for allocation concealment).
Blinding
It is difficult to blind the caregivers, who normally check the wounds every day, and participants, who have them checked every day. Almost all the trials were at high or unclear risk of bias due to lack of blinding of participants and caregivers. However, in three (Khan 2006; Subramanian 2005; Teoh 2018) of 66 trials, the researchers paid particular care to blind the caregivers and the participants by using occlusive dressings. In seven trials, outcome assessors were blinded (Anatol 1997; Barker 1984; Brown 2009; Khan 2006; Maartense 2002; Teoh 2018; Tsujinaka 2013). We considered four studies to be at high risk of detection bias (Chughtai 2000; Jan 2013; Martin 2017; Xu 2014). The remaining trials did not report enough information to enable a judgement. We considered these studies to be at unclear risk of bias. Agreement between the two independent raters in the risk of detection bias assessment was 94% (weighted Kappa 0.99).
Incomplete outcome data
We rated the risk of incomplete outcome reporting for short‐term outcomes. We rated 29 studies as being at low risk of attrition bias and 16 studies as being at high risk of attrition bias. This was often due to high proportional rates of dropout, or unclear reasons for large numbers of dropouts, or both. The remaining 21 studies were judged as being at unclear risk of bias for this domain, as reasons for dropout were not clear, or the extent to which the dropout rates affected the results was unclear. Agreement between the two independent raters in the risk of attrition bias assessment was 62% (weighted Kappa 0.88).
Selective reporting
None of the included trials had a published protocol. We rated 38 studies as being at low risk of selective outcome reporting as they fully reported all the outcomes they had planned to. We rated nine studies as being at high risk, because four studies reported the outcomes they did not pre specify in the methods section (Ranaboldo 1992; Selvadurai 1997; Tanaka 2016; Zwart 1989), four studies reported the important outcomes incompletely (Rebello 2009; Rosen 1997; Sakka 1995; Steele 1983), and one study did not report the outcome prespecified in the trial registry (Kuroki 2017). In the remaining 19 studies, the available information was not enough to make a judgement. Agreement between the two independent raters in the risk of reporting bias assessment was 68% (weighted Kappa 0.88).
Other potential sources of bias
We rated 13 of the included studies as being at low risk of other bias and nine studies as being at high risk. In the remaining 44 studies, the available information was not enough for us to make a judgement. The majority of these studies did not report funding or conflict of interest, thus, we could not judge whether the study had an inappropriate influence of funders. Fifteen studies were identified as having potential unit of analysis issues as it did not appear that paired data or clustered data were accounted for in the analysis (paired data: Anatol 1997; Rebello 2009; Rosen 1997; Subramanian 2005; clustered data: Ademuyiwa 2009; Anatol 1997; Baek 2009; Barker 1984; Chughtai 2000; Corder 1991; Jallali 2004; Jan 2013; Keng 1989; Murphy 1995; Swtizer 2003). We rated these studies as being at unclear or high risk of bias. Agreement between the independent raters in the risk of this bias assessment was 65% (weighted Kappa 0.91).
Effects of interventions
See: Summary of findings for the main comparison Subcuticular sutures compared with transdermal sutures for skin closure in non‐obstetric surgery; Summary of findings 2 Subcuticular sutures compared with skin staples for skin closure in non‐obstetric surgery; Summary of findings 3 Subcuticular sutures compared with tissue adhesives for skin closure in non‐obstetric surgery; Summary of findings 4 Subcuticular sutures compared with surgical tapes for skin closure in non‐obstetric surgery; Summary of findings 5 Subcuticular sutures compared with surgical zippers for skin closure in non‐obstetric surgery
See: summary of findings Table for the main comparison; summary of findings Table 2; summary of findings Table 3; summary of findings Table 4; summary of findings Table 5.
Comparison 1. Subcuticular sutures compared with transdermal sutures (23 studies, 3698 participants)
Primary outcome: SSI
Data from 20 trials compared the use of subcuticular sutures with transdermal sutures for SSI. However, as one trial had no cases of infection (Baek 2009), only data from the remaining 19 trials contributed to the meta‐analysis (Andrade 2016; Buchweitz 2005; Chen 2013; Clough 1975; Corder 1991; Foster 1977; Hopkinson 1982; Javadi 2018; Karabay 2005; Khajouei 2007; Kotaluoto 2012; Liu 2017; McGreal 2002; Murphy 1995; Onwuanyi 1990; Pauniaho 2010; Tanaka 2014; Taube 1983; Zwart 1989). Overall, 7.7% (238/3107 participants) developed an SSI. There is no clear difference in the incidence of SSI between the two groups (RR 1.10; 95% CI 0.80 to 1.52; Analysis 1.1) with no important heterogeneity (Tau2= 0.11, I2= 24%). Using risk difference, there were 0 more SSIs per 1000 with subcuticular sutures compared with transdermal sutures (10 fewer to 20 more) (RD 0.00, 95% CI ‐0.01 to 0.02; Tau2= 0, I2= 36%). There may be little difference in the incidence of SSI. No publication bias was evident in the funnel plot (Figure 4) and small size effect was not apparent. The evidence was downgraded to low certainty due to high risk of bias across varying domains (attrition, selection, reporting and other bias affecting 40% of the analysis weight across nine of 19 studies) and imprecision.
Funnel plot of comparison: 1 Subcuticular sutures compared with transdermal sutures, outcome: 1.1 Surgical site infection.
In the sensitivity analysis, there was no change in the interpretation of results by adopting the fixed‐effects model. We also conducted a sensitivity analysis in which we restricted inclusion in the analysis to only those studies that were considered to be at a low risk of selection bias (Tanaka 2014), there was no change in the interpretation of results (RR 0.93, 95% CI 0.57 to 1.50; Analysis 1.2). Three studies were considered to have potential unit of analysis issues (Baek 2009; Corder 1991; Murphy 1995). We conducted a post hoc sensitivity analysis removing these studies from the meta‐analysis. Again there was no evidence of a difference in the proportion of participants developing infection between groups (RR 1.04, 95% CI 0.75 to 1.42; Analysis 1.2). In an ITT sensitivity analysis based on the worst‐worst scenario, there was also no change in the interpretation of results (RR 1.23, 95% CI 0.97 to 1.55; Analysis 1.2). However, an ITT sensitivity analysis based on the worst‐best scenario showed that the results changed in favour of the transdermal sutures group (RR 1.59, 95% CI 1.05 to 2.42; Analysis 1.2).
Secondary outcomes
Wound complications
Data from nine trials compared the use of subcuticular sutures with transdermal sutures for wound complications and contributed to the meta‐analysis (Andrade 2016; Buchweitz 2005; Corder 1991; Ghaderi 2010; Khajouei 2007; Kotaluoto 2012; Onwuanyi 1990; Pauniaho 2010; Rosen 1997). It is uncertain whether subcuticular sutures have an effect on wound complications compared with transdermal sutures because the certainty of the evidence is very low (RR 0.83; 95% CI 0.40 to 1.71; 1489 participants; Tau2= 0.79, I2= 70%) (Analysis 1.3) (very low‐certainty evidence, downgraded once each for risk of bias (variously attrition, selection, reporting and other bias affecting 62% of the analysis weight), for imprecision and for inconsistency).
Wound dehiscence
Six trials reported wound dehiscence (Andrade 2016; Buchweitz 2005; Javadi 2018; Kotaluoto 2012; Liu 2017; Pauniaho 2010). A total of 33 participants (33/866 (3.8%)) developed wound dehiscence. Of these, 27 participants belonged to non‐absorbable interrupted transdermal sutures group. It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with transdermal sutures because the certainty of the evidence is very low (RR 0.35; 95% CI 0.08 to 1.54; Tau2= 1.59, I2= 49%: Analysis 1.4) (very low‐certainty evidence, downgraded once due to risk of bias (attrition and selection bias) and twice due to imprecision).
Proportion of re‐closure of the skin incision required
Only two trials reported the proportion of re‐closure of the skin incision (Hopkinson 1982; Karabay 2005). It is uncertain whether there is a difference in the proportion of re‐closure between the two groups because the certainty of the evidence is very low (RR 1.16; 95% CI 0.09 to 14.57; Tau2= 1.47, I2= 43%; Analysis 1.5). The evidence was downgraded to very low certainty once due to high risk of bias (attrition and other bias) and twice due to imprecision (only five events in total).
Hypertrophic scar
Only two trials reported this outcome (Onwuanyi 1990; Simpson 1979). A total of 56 participants (56/233 (24%)) developed hypertrophic scar. Of these, 54 participants developed hypertrophic scar in one old trial (Simpson 1979). It is uncertain whether subcuticular sutures increase or reduce the risk of hypertrophic scar compared with transdermal sutures because the certainty of the evidence is very low (RR 0.91; 95% CI 0.25 to 3.39; Tau2= 0.48, I2= 28%; Analysis 1.6) (very low‐certainty evidence, downgraded once due to risk of bias (attrition bias) and twice due to imprecision (wide 95% CI).
Keloid scar
Clough 1975 (143 participants) reported this outcome, but had no cases with this event. It is uncertain whether subcuticular sutures increase or reduce the risk of keloid scar compared with transdermal sutures because the certainty of the evidence is very low (downgraded once due to risk of bias (unclear in all domains in included single study), and twice due to imprecision (small sample size and no events)).
Wound pain intensity
Four trials (372 participants) reported this outcome (Baek 2009; Javadi 2018; Pauniaho 2010; Rosen 1997). However, the data of three trials could not be combined because of missing data or statistics (see also Characteristics of included studies). In addition, as the data of the other trial (Javadi 2018) was considered to be skewed, we summarised this narratively. Three trials (Baek 2009; Javadi 2018; Pauniaho 2010) reported the mean of severity of pain score using the visual analogue scale (VAS) scored between 0 for the lowest level of pain, and 10 for the most severe pain. Rosen 1997 reported the pain score using VAS scored from 1 (lowest pain) to 5 (most severe pain). All trials reported lower pain scores (range from 0.2 to 1.4) in subcuticular sutures compared with that in transdermal sutures, however, Pauniaho 2010 did not show statistical significance. Javadi 2018 reported the mean scores at postoperative day seven in the subcuticular suture group were significantly lower than the control group (mean 0.86, 95% CI 0.05 to 1.67 (median 1, IQR 0‐1) versus mean 1.40, 95% CI 0.55 to 2.25, (median 1, IQR 1‐2), P value = 0.008). It is uncertain whether subcuticular sutures reduce the pain compared with transdermal sutures because the certainty of the evidence is very low (downgraded due to risk of bias, imprecision and inconsistency).
Length of hospital stay
Four trials reported this outcome (Chen 2013; Karabay 2005; McGreal 2002; Onwuanyi 1990), but the data from three trials (Karabay 2005; McGreal 2002; Onwuanyi 1990) could not be combined in the meta‐analysis because of missing statistics (see also Characteristics of included studies) (374 participants).
All the trials reported there was no significant difference in the length of hospital stay between the two groups. There is probably little or no difference in the length of hospital stay between subcuticular and transdermal sutures groups (MD 0.40 days, 95% CI ‐0.44 to 1.24; 292 participants; Analysis 1.7) because the certainty of the evidence is moderate; downgraded due to imprecision.
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months)
Five trials (about 950 participants) reported this outcome (Clough 1975; Kotaluoto 2012; Liu 2017; Tanaka 2014; Zwart 1989). Because of the differences in the scale used, we did not conduct meta‐analytic pooling and we presented the results narratively.
Clough 1975 reported surgeon‐assessed cosmetic appearance at three to four years. The cosmetic presence of the wound was classified as good (it was neat and uniform and not more than 3 mm wide) or poor (if it was hypertrophic or greater than 3 mm wide). The proportion of cosmetically poor scars was similar between groups. However, the data could not be used as the group(s) from which 18 participants had dropped out was not clear and the number of participants was imbalanced between two groups.
Kotaluoto 2012 reported cosmesis after a median of 14 months by blinded assessment of photographs. For subjective scar assessment, the Vancouver scar scale, the patient and observer scar assessment scale (POSAS), and a visual analogue scale (VAS) were used. Objective evaluation was carried out by measuring surface area, average width, and estimated concentration change (ECC) of haemoglobin and melanin in the scar using spectrocutometry. Kotaluoto 2012 concluded both objective and subjective analyses showed better cosmetic results for subcuticular suturing. The difference between the two groups was statistically significant as regards POSAS in both patient and observer scales, VAS, surface area, average width, and estimated concentration change (ECC) of melanin. However, there were many dropouts (33%) in the analyses. Liu 2017 reported the cosmetic outcome as assessed by the overall impression on the Dutch Patient and Observer Scar Assessment Scale (POSAS) version 2.0, the POSAS, the 4‐point scale (excellent, good, fair, bad) by the patient and observer, and the measurement with a colorimeter. The cosmetic result was evaluated at three and 12 months after surgery. A observer, blinded to the suturing technique, assessed the scars in person by using the Observer Scar Assessment Scale (OSAS) and the 4‐point scale. At 12 months, no significant differences were found. Tanaka 2014 reported cosmesis using a modified scar assessment scale based on the Hollander Wound Evaluation (Hollander 1995) and POSAS at six months. They showed significantly better cosmetic results for subcuticular suturing as regards to pain, scar vascularity and width. Zwart 1989 reported the objective (assessed by the head nurse) and subjective (assessed by patient) cosmetic results at one, three and six months. The cosmetic presence of the wound was classified as excellent, good, fair or poor. At six months, no statistically significant differences were found between the groups.
Two studies showed there were significant differences in the cosmetic outcomes in favour of subcuticular sutures, and the other studies reported no significant difference. It is uncertain whether subcuticular sutures improve the cosmesis of scar compared with transdermal sutures because the certainty of evidence is very low: downgraded due to high risk of attrition bias, imprecision (narrative synthesis) and inconsistency (two reaching significance and the others not).
Several studies did not contribute to this outcome because, although they evaluated cosmetic appearance, they did so less than six months after surgery (Baek 2009; Buchweitz 2005; Karabay 2005; Taube 1983).
Patient satisfaction
Only two trials reported this outcome (Javadi 2018; Tanaka 2014), but the data from Javadi 2018 could not be combined in the meta‐analysis as they used neither scale or score. Therefore, we presented results narratively. Javadi 2018 (70 participants) reported patient satisfaction with the surgical site scar at 90 days after operation. They concluded that a significantly greater number of patients in the subcuticular suture group were satisfied with their wound healing and scar status compared with the control group (91.42% vs. 71.42%). Tanaka 2014 (290 participants) assessed patient satisfaction at seven days, 30 days, three months and six months using scores on a scale of 1 (lowest) to 10 (best possible score). We used the data at 30 days and six months. Patient satisfaction at 30 days is probably higher in subcuticular sutures group (score 9) compared with transdermal sutures group (score 7.4) with a difference in means of 1.60 (95% CI 1.32 to 1.88; Analysis 1.8). This is moderate‐certainty evidence downgraded once due to imprecision. In addition, patient satisfaction at six months is also probably higher in the subcuticular sutures group (score 9) compared with the transdermal sutures group (score 7.3) with a difference in means of 1.70 (95% CI 1.37 to 2.03; Analysis 1.9) (moderate‐certainty evidence downgraded once due to imprecision).
Quality of Life
None of the trials reported quality of life.
Wound closure time in the operation (minutes)
Only two trials reported this outcome (Chen 2013; Tanaka 2014). Wound closure time is probably longer in the subcuticular sutures group compared with the transdermal sutures group with a difference in means of 5.81 minutes (95% CI 5.13 to 6.49; Tau2= 0.15, I2= 61%; 585 participants; Analysis 1.10) (moderate‐certainty evidence downgraded once due to inconsistency).
Cost
Only two trials reported cost (Murphy 1995; Onwuanyi 1990), however the data of Murphy 1995 were excluded as they were insufficient because of missing statistics (see also Characteristics of included studies), and only data from Onwuanyi 1990 contributed to the meta‐analysis. In this study, participants used non‐absorbable (Nylon) subcuticular sutures. We calculated the SD from the reported P value. Subcuticular sutures may reduce the cost compared with transdermal sutures. The mean cost was 8 Naira in the subcuticular sutures group compared with 16 Naira in the transdermal sutures group with a difference in means of ‐8.00 Naira (95% CI ‐13.05 to ‐2.95; 100 participants; Analysis 1.11). The evidence was downgraded to low certainty due to risk of bias and imprecision.
Summary of comparison
Low‐certainty evidence suggests that there may be little difference between subcuticular sutures and transdermal sutures groups in the incidence of SSI. There is very low‐certainty evidence that it is uncertain whether subcuticular sutures have an effect on wound complications and dehiscence compared with transdermal sutures. Moderate‐certainty evidence shows that subcuticular sutures probably improve patient satisfaction compared with transdermal sutures. Additionally, subcuticular sutures may reduce the cost in materials (low‐certainty evidence), which may, however, be offset by increase in costs necessary for longer operations. Wound closure time is probably longer when subcuticular sutures are used (moderate‐certainty evidence). See summary of findings Table for the main comparison.
Comparison 2. Subcuticular sutures compared with skin staples (18 studies, 4428 participants)
None of the trials reported the proportion of re‐closure or the incidence of keloid scar.
Primary outcome: SSI
Fifteen trials compared the use of subcuticular sutures with staples for SSI. However, as one trial had no cases of infection (Selvadurai 1997), only data from the remaining 14 trials contributed to the meta‐analysis (Chen 2018; Chughtai 2000; Eggers 2011; Imamura 2016; Khan 2006; Kobayashi 2015; Kuroki 2017; Mullen 1999; Murphy 1995; Ranaboldo 1992; Shetty 2004; Subramanian 2005; Tsujinaka 2013; Zwart 1989). Overall, 8.1% (337/4163 participants) developed an SSI. There is probably little or no difference in the incidence of SSI between the two groups (RR 0.81, 95% CI 0.64 to 1.01; Analysis 2.1) with no important evidence of heterogeneity (Tau2= 0.01, I2= 5%). No publication bias was evident in the funnel plot (Figure 5) and the small study size effect was not apparent. The evidence was downgraded to moderate certainty due to imprecision. In the sensitivity analysis, there was no change in the interpretation of results by adopting the fixed‐effects model. We also conducted a sensitivity analysis in which we restricted inclusion in the analysis to only those studies that were considered to be at a low risk of selection bias (Imamura 2016; Kobayashi 2015; Kuroki 2017; Tsujinaka 2013), but there was no change in the interpretation of results (RR 0.91, 95% CI 0.72 to 1.16; Analysis 2.2). Three studies were considered to have potential unit of analysis issues (Chughtai 2000; Murphy 1995; Subramanian 2005). We conducted a post hoc sensitivity analysis removing these studies from the meta‐analysis. Again there was no evidence of a difference in the proportion of participants developing infection between the groups (RR 0.81, 95% CI 0.61 to 1.09; Analysis 2.2). ITT sensitivity analyses showed that the results did not change depending on the methods of imputation of the missing data (Analysis 2.2).
Funnel plot of comparison: 2 Subcuticular sutures compared with skin staples, outcome: 2.1 Surgical site infection.
Secondary outcomes
Wound complications
Nine trials compared the use of subcuticular sutures with staples for wound complications. However, as two trials had no cases of complications (Reed 1997; Selvadurai 1997), only data from the remaining seven trials contributed to the meta‐analysis (Khan 2006; Kobayashi 2015; Kuroki 2017; Obermair 2007; Shetty 2004; Steele 1983; Tsujinaka 2013). A total of 291 participants (291/2973 (9.8%)) developed wound complications. Subcuticular sutures probably on average decrease wound complications in comparison with staples (RR 0.79, 95% CI 0.64 to 0.98; Tau2= 0, I2= 0%; Analysis 2.3) (moderate‐certainty evidence downgraded once due to imprecision).
Wound dehiscence
Seven trials reported wound dehiscence and contributed to the meta‐analysis (Chen 2018; Chughtai 2000; Eggers 2011; Kuroki 2017; Obermair 2007; Shetty 2004; Tsujinaka 2013). Overall, 4.7% (93/1984 participants) developed wound dehiscence. Subcuticular sutures may reduce the risk of wound dehiscence compared with skin staples (RR 0.63, 95% CI 0.43 to 0.94; Analysis 2.4) with no evidence of heterogeneity (Tau2= 0, I2= 0%). The evidence was downgraded to low certainty due to high risk of bias (detection and other bias affecting 24% of the analysis weight in one (Chughtai 2000) of six studies) and imprecision.
Hypertrophic scar
Only three trials reported this outcome (Ranaboldo 1992; Selvadurai 1997; Tsujinaka 2013). However, as one trial had no cases of hypertrophic scar (Ranaboldo 1992), only data from the remaining two trials contributed to the meta‐analysis. A total of 207 participants (207/1195 (17%)) developed hypertrophic scar. Subcuticular sutures probably on average decrease hypertrophic scar in comparison with skin staples (RR 0.77, 95% CI 0.60 to 0.98; Tau2= 0, I2= 0%; Analysis 2.5) (moderate‐certainty evidence downgraded once due to imprecision).
Wound pain intensity
Six trials reported this outcome (Eggers 2011; Obermair 2007; Ranaboldo 1992; Selvadurai 1997; Subramanian 2005; Reed 1997). However, as the data of one trial were insufficiently reported because of missing data (Subramanian 2005), only data from the remaining five trials contributed to the meta‐analysis. Three studies (Obermair 2007; Ranaboldo 1992; Selvadurai 1997) assessed pain intensity using a VAS scale that ran from 0 to 100, where 100 represented maximal pain and two studies (Eggers 2011; Reed 1997) used a VAS scale that ran from 0 to 10, where 10 represented maximal pain. Eggers 2011 reported pain intensity at three and six weeks, Reed 1997 assessed pain intensity by patients about 320 days (mean) after operation, and Obermair 2007 reported pain intensity assessed by patients and surgeons at one and six weeks, and three months. We used the data of pain intensity assessed by patients at one and three months and calculated the SD from the reported 95% CI. Selvadurai 1997 reported pain scores as measured at the first three postoperative days, and we included VAS data at the second postoperative day and calculated SDs from the SEs. In one study (Ranaboldo 1992), we calculated the SD from the reported P value. It is uncertain whether there is a difference in the pain intensity between the two groups because the certainty of the evidence is very low: pain intensity within seven days (scale from 0 to 100, 100 represented the maximum pain) (MD ‐1.86, 95%CI ‐10.37 to 6.65, Tau2= 42.14, I2= 76%; Analysis 2.6) (very low‐certainty evidence, downgraded due to high risk of bias, imprecision (low numbers of participants (218) and wide 95% CI) and inconsistency), pain intensity after 30 days (SMD 0.18, 95%CI ‐0.30 to 0.66, Tau2= 0.11, I2= 61%; Analysis 2.7) (very low‐certainty evidence: downgraded due to high risk of bias (attrition bias affecting 36% of the analysis weight), imprecision (low numbers of participants (196)) and inconsistency).
Length of hospital stay
Six trials reported this outcome (Chen 2018; Eggers 2011; Khan 2006; Kobayashi 2015; Reed 1997; Tsujinaka 2013). However, as the data of one trial were considered to be skewed (Khan 2006), only data from the remaining five trials contributed to the meta‐analysis. Khan 2006 reported there was no significant difference in this outcome. In the meta‐analysis, it is uncertain whether subcuticular sutures shorten the length of hospital stay compared with skin staples (MD ‐0.58 days, 95% CI ‐1.57 to 0.42, Tau2= 0.78, I2= 77%; 2794 participants; Analysis 2.8) because the certainty of the evidence is very low: downgraded due to high risk of bias (attrition bias affecting 27% of the analysis weight), imprecision and inconsistency.
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months)
Only three trials reported this outcome. Reed 1997 assessed cosmesis by patients using a VAS scale that ran from 0 to 10, where 10 represented the best score at about 320 days (mean) after operation. Selvadurai 1997 assessed cosmesis by independent observer using a VAS scale that ran from 0 to 100, where 100 represented maximal score at six months and SD was calculated from SE. Zwart 1989 reported the objective (assessed by the head nurse) and subjective (assessed by patient) cosmetic results at one, three and six months. The cosmetic presence of the wound was classified as excellent, good, fair or poor. We used the objective results and calculated the mean and SD using a scoring scale that ran from 1 (poor) to 4 (excellent). There may be little or no difference in the cosmesis between the two groups (SMD 0.12, 95% CI ‐0.11 to 0.35; Analysis 2.9; 291 participants) with no evidence of heterogeneity (Tau2= 0, I2= 0%). The evidence was downgraded to low certainty due to high risk of bias (attrition and reporting bias) and imprecision.
Several studies did not contribute to this outcome because although they evaluated cosmetic appearance they did so less than six months after surgery (Chughtai 2000; Eggers 2011; Khan 2006; Kobayashi 2015; Kuroki 2017; Obermair 2007; Ranaboldo 1992; Steele 1983).
Patient satisfaction
Only three trials reported this outcome (Khan 2006; Kobayashi 2015; Kuroki 2017). However, as the data of two trials were considered to be skewed (Khan 2006; Kuroki 2017), only Kobayashi 2015 contributed to the meta‐analysis. Khan 2006 (127 participants) assessed with a VAS between 0 and 100, where 100 represented maximal satisfaction at eight to 12 weeks after operation and reported there was no significant difference in the patient satisfaction. Kuroki 2017 (163 participants) assessed patient satisfaction with appearance of scar, location of scar and discomfort at scar using scores (represented by per cent, but the details were not available). There were no significant differences in median satisfaction scores of the scar appearance (suture 77% compared with staples 68%, P = 0.11) nor in the satisfaction with the discomfort at the incision site (staples 71% compared with suture 77%, P = 0.20).
Kobayashi 2015 reported patient satisfaction using a scale that ran from 1 to 5, where 5 represented the best score at 30 days after operation. Patient satisfaction at 30 days was slightly higher in the subcuticular sutures group (score 4.4) compared with the skin staples group (score 4.2) with a difference in means of 0.20 (95% CI 0.10 to 0.30; Analysis 2.10; 1232 participants). This is high‐certainty evidence.
Quality of Life
Eggers 2011 reported general health as judged by responses to the SF‐12 v2 (QualityMetric Inc., Lincoln, RI) survey of the physical composite score (PCS) and mental composite score (MCS) at three and six weeks after operation. The scale score (higher better) is norm‐based scoring which is referred to as "50/10" scoring because the score has been standardised so that the general US population has a mean of 50 and SD of 10. We used the data at six weeks. There may be little or no difference between the two groups in the SF‐12 v2 PCS (MD 0.00, 95% CI ‐6.05 to 6.05; Analysis 2.11; 38 participants) and MCS (MD 1.00, 95% CI ‐5.05 to 7.05; Analysis 2.12; 38 participants). The certainty of the evidence is low, downgraded twice for imprecision.
Wound closure time in the operation (minutes)
Six studies reported the time taken for closure (Khan 2006; Kobayashi 2015; Ranaboldo 1992; Selvadurai 1997; Steele 1983; Subramanian 2005). Two studies reported wound closure time as speed (rate) seconds/cm (Eggers 2011; Zwart 1989). However, as the data of four trials were insufficient (missing statistics or data) (Eggers 2011; Khan 2006; Ranaboldo 1992; Zwart 1989) (see also Characteristics of included studies), only data from the remaining four trials contributed to the meta‐analysis. As there was extreme heterogeneity (I2= 99%; 1384 participants; Analysis 2.13), we presented the results narratively. The mean wound closure time in the skin staple group ranged from 0.9 to 4.5 minutes. Mean differences ranged between 0.30 and 5.50 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. The certainty of the evidence is low, downgraded twice for inconsistency.
Cost
Four studies (342 participants) reported cost (Chughtai 2000; Eggers 2011; Murphy 1995; Ranaboldo 1992). However as the data of all studies were insufficiently reported because of missing statistics, we presented them narratively. Three of these trials favoured subcuticular sutures (Chughtai 2000; Murphy 1995; Ranaboldo 1992). Chughtai 2000 reported the average cost per case was USD 4.5 for sutures and USD 15 for staples. Murphy 1995 reported the mean cost per patient was USD 4 for sutures and USD 12 for staples. Ranaboldo 1992 reported the mean cost per patient was GBP 1.41 pounds for sutures and GBP 13.99 for staples (including removal cost). Eggers 2011 reported that the total cost associated with surgery for each of the closure groups (including all aspects of surgery associated with materials, labour, and operating room expenses) was USD 1056.3 for sutures and USD 802.8 for staples. They concluded that the staples were to be favoured because most of the cost differential was attributed to procedure times.
It is uncertain whether subcuticular sutures reduce the cost compared with skin staples. The certainty of the evidence is very low downgraded once for high risk of bias (detection and other bias), once for imprecision and once for inconsistency (some studies favoured subcuticular sutures and the other favoured skin staples).
Summary of comparison
Moderate‐certainty evidence shows that subcuticular sutures probably have little effect on SSI; but probably decrease the incidence of wound complications and hypertrophic scars when compared with skin staples. Low‐certainty evidence also shows that subcuticular sutures may reduce the risk of wound dehiscence compared with skin staples. In addition, subcuticular sutures are associated with slightly higher patient satisfaction (high‐certainty evidence). However, wound closure time may be longer compared with staples. See summary of findings Table 2.
Comparison 3. Subcuticular sutures compared with tissue adhesives (17 studies, 1419 participants)
None of the trials reported the incidence of hypertrophic or keloid scar.
Primary outcome: SSI
Ten trials compared the use of subcuticular sutures with tissue adhesives for SSI. However, as four trials had no cases of infection (Gennari 2004; Keng 1989; Ong 2002; Teoh 2018), only data from the remaining six trials contributed to the meta‐analysis (Eggers 2011; Khan 2006; Maartense 2002; Martin 2017; Sebesta 2004; Van den Ende 2004). Overall, 4.0% (35/869 participants) developed an SSI. There is moderate‐certainty evidence (downgraded once for imprecision) showing no clear difference in the incidence of SSI between participants treated with subcuticular sutures and those treated with tissue adhesives (RR 0.77, 95% CI 0.41 to 1.45; Analysis 3.1). There is no evidence of heterogeneity (Tau2= 0, I2= 0%). Confidence intervals are wide, spanning both appreciable benefits and harms, so clear differences between treatments are not apparent. There are no trials that are considered to be at a low risk of selection bias.
In the sensitivity analysis, there was no change in the interpretation of results by adopting the fixed‐effects model. One study was considered to have potential unit of analysis issues (Keng 1989). We conducted a post hoc sensitivity analysis removing this study from the meta‐analysis. Again, there was no evidence of a difference in the proportion of participants developing infection between the groups (RR 0.77, 95% CI 0.41 to 1.45; Analysis 3.2). ITT sensitivity analyses showed that the results did not change depending on the methods of imputation of the missing data (Analysis 3.2).
Secondary outcomes
Wound complications
Eleven trials compared the use of subcuticular sutures with tissue adhesives for wound complications. However, as one trial had no cases of complications (Jallali 2004), only data from the remaining 10 trials contributed to the meta‐analysis (Ademuyiwa 2009; Brown 2009; Jan 2013; Keng 1989; Khan 2006; Krishnamoorthy 2009; Sebesta 2004; Soni 2013; Swtizer 2003; Van den Ende 2004). A total of 152 participants (152/1058 (14.4%)) developed wound complications. There is no clear difference in the incidence of wound complications between participants treated with subcuticular sutures and those treated with tissue adhesives (RR 0.62, 95% CI 0.35 to 1.11; Tau2= 0.32, I2= 46%; Analysis 3.3) (low‐certainty evidence, downgraded once due to high risk of bias across varying domains (detection, attrition and other bias affecting 54% of the analysis weight) and once due to imprecision). Although the point estimate is on the side of a possible benefit, the 95% confidence intervals includes the possibility of both benefit and harm so clear differences between treatments are not apparent. No publication bias was evident in the funnel plot (Figure 6) and the small study size effect was not apparent.
Funnel plot of comparison: 3 Subcuticular sutures compared with tissue adhesives, outcome: 3.3 Wound complications.
Wound dehiscence
Eleven trials reported superficial wound dehiscence. However as five trials had no cases of dehiscence (Brown 2009; Gennari 2004; Martin 2017; Ong 2002; Teoh 2018), only data from the remaining six trials contributed to the meta‐analysis (Eggers 2011; Jan 2013; Sebesta 2004; Soni 2013; Swtizer 2003; Van den Ende 2004). Overall, 2.1% (24/1155 participants) developed wound dehiscence. Subcuticular sutures may decrease wound dehiscence in comparison with tissue adhesives (RR 0.23, 95% CI 0.07 to 0.74; Tau2= 0, I2= 0%; Analysis 3.4). The evidence is low‐certainty, downgraded once due to high risk of bias across varying domains (detection, attrition and other bias affecting 45% of the analysis weight) and once due to imprecision (low numbers of events).
Proportion of re‐closure of the skin incision required
Van den Ende 2004 reported the proportion of re‐closure of the skin incision. There is low‐certainty evidence (downgraded twice due to imprecision) showing no clear difference in the proportion of re‐closure between participants treated with subcuticular sutures and those treated with tissue adhesives (RR 0.33; 95% CI 0.01 to 7.99; Analysis 3.5; 100 participants. This comparison was underpowered, with only one event in total.
Wound pain intensity
Eggers 2011 reported this outcome using a VAS scale that ran from 0 to 10, where 10 represented maximal pain. Eggers 2011 reported pain intensity at three and six weeks. Thus, we included the data at six weeks. There may be little or no difference between the groups in the pain intensity after 30 days (MD 0.55, 95% CI ‐0.35 to 1.45; Analysis 3.6; 56 participants) (low‐certainty evidence, downgraded twice due to imprecision).
Length of hospital stay
Three studies reported this outcome (Eggers 2011; Gennari 2004; Khan 2006). However, as the data of one trial was considered to be skewed (Khan 2006), only data from the remaining two trials contributed to the meta‐analysis. Khan 2006 reported there was no significant difference in this outcome. In the meta‐analysis, there may also be little or no difference between the groups in the length of hospital stay (MD 0.22 days, 95% CI ‐0.39 to 0.84; Tau2= 0, I2= 0%; Analysis 3.7; 189 participants) (low‐certainty evidence, downgraded once due to risk of bias (attrition and other bias) and once due to imprecision).
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months).
Gennari 2004 reported cosmetic outcomes by blinded surgeon and patient at six months and one year after operation. Because we could not include the data because of missing statistics, we presented it narratively. Gennari 2004 used a score system that ran from 1 to 10, where 10 represented maximal cosmesis. One‐year follow‐up was performed in 97 patients (73%). There were similar outcomes between groups on the wound cosmetic evaluation by both plastic surgeons (tissue adhesive 8.8 versus suture 8.8) and patients (tissue adhesive 8.6 versus suture 8.9) at the 1‐year follow‐up visit. There may be little or no difference in the cosmesis of scar between the two groups. The evidence was downgraded to low certainty due to risk of bias (attrition and other bias) and imprecision. Many studies did not contribute to this outcome because, although they evaluated cosmetic appearance, they did so less than six months after surgery (Ademuyiwa 2009; Brown 2009; Eggers 2011; Jallali 2004; Jan 2013; Keng 1989; Khan 2006; Krishnamoorthy 2009; Maartense 2002; Martin 2017; Sebesta 2004; Soni 2013; Swtizer 2003; Teoh 2018; Van den Ende 2004).
Patient satisfaction
Only two trials (255 participants) reported this outcome (Gennari 2004; Khan 2006). However, as the data of one trial were considered to be skewed (Khan 2006), only the Gennari 2004 trial contributed to the meta‐analysis. Khan 2006 assessed this outcome with a VAS between 0 and 100, where 100 represented maximal satisfaction at 8 to 12 weeks after operation, and reported there was no significant difference in patient satisfaction. Gennari 2004 reported this outcome using a satisfaction score rated by the patients (scale from 1 to 10, where 10 represented best score) in the first three weeks after surgery. We calculated the SD from the reported P value. Tissue adhesives may improve patient satisfaction compared with subcuticular sutures (MD ‐2.05, 95% CI ‐3.05 to ‐1.05; 131 participants; Analysis 3.8). The evidence was downgraded to low certainty due to risk of bias (attrition and other bias) and imprecision.
Quality of Life
Eggers 2011 (56 participants) reported general health as judged by responses to the SF‐12 v2 (QualityMetric Inc., Lincoln, RI) survey of the physical composite score (PCS) and mental composite score (MCS) at three and six weeks after operation. The scale score (higher better) is norm‐based scoring which is referred to as "50/10" scoring because the score has been standardised so that the general US population has a mean of 50 and SD of 10. We used the data at six weeks. There may be little or no difference between the two groups in the SF‐12 v2 PCS (MD 2.00, 95% CI ‐2.98 to 6.98; Analysis 3.9) and MCS (MD ‐1.50, 95% CI ‐6.78 to 3.78; Analysis 3.10). The certainty of the evidence is low, downgraded twice for imprecision.
Wound closure time in the operation (minutes)
Sixteen studies reported the time taken for closure. However, as the data of two trials were considered skewed (Khan 2006; Krishnamoorthy 2009) and the data of three studies (Eggers 2011; Keng 1989; Teoh 2018) were insufficient (missing data or statistics), only data from the remaining 11 trials (895 participants) contributed to the meta‐analysis. Keng 1989, Khan 2006 and Krishnamoorthy 2009 suggested that subcuticular sutures took significantly more time (about 0.5 to 4 minutes) than tissue adhesives.
In five studies (Jallali 2004; Jan 2013; Maartense 2002; Martin 2017; Van den Ende 2004), we calculated the SD from the reported P values. We also calculated the SD from the reported 95% CI or the SEM (standard error of the mean) in two trials (Gennari 2004; Swtizer 2003).
As there was extreme heterogeneity (I2= 97%; Analysis 3.11), we presented the results narratively. The mean wound closure time in the tissue adhesives group ranged from 0.3 to 3.7 minutes. Mean differences ranged between ‐0.34 and 10.39 minutes across 11 studies. Further analyses were not undertaken due to statistical heterogeneity in the results. The certainty of the evidence is very low downgraded once due to high risk of bias across varying domains (detection, attrition and other bias) and twice for inconsistency (I2= 97% which was mostly quantitative (i.e. most studies agreed in the direction of effect but differed in its magnitude only).
Cost
Seven studies reported the cost. However, as the data of three trials were insufficient (missing data or statistics) (Eggers 2011; Martin 2017; Van den Ende 2004), only data from the remaining four trials contributed to the meta‐analysis. We calculated SDs from the reported P value (Maartense 2002) and from the SEM (Gennari 2004). Two studies (Brown 2009; Sebesta 2004) reported the cost (USD) and the others (Gennari 2004; Maartense 2002) reported the cost (EUR). As there was extreme heterogeneity (I2= 96%; Analysis 3.12), we presented the results narratively. The mean cost in the tissue adhesives group ranged from 31.96 to 65.1 USD and from 20.3 to 34.01 EUR. Mean differences ranged between ‐57.36 and ‐4.26 USD (‐16.19 and ‐10.30 EUR). Further analyses were not undertaken due to statistical heterogeneity in the results. The certainty of the evidence is very low downgraded once due to high risk of bias across varying domains (attrition and other bias) and twice for inconsistency (I2= 96% which was mostly quantitative (i.e. all studies agreed in the direction of effect but differed in its magnitude only).
Summary of comparison
There is no clear difference in the incidence of SSI (moderate‐certainty evidence) and wound complications (low‐certainty evidence) between participants treated with subcuticular sutures and those treated with tissue adhesives. Low‐certainty evidence suggests that subcuticular sutures may reduce the risk of wound dehiscence compared with tissue adhesives; but may also achieve lower patient satisfaction ratings. See summary of findings Table 3.
Comparison 4. Subcuticular sutures compared with surgical tapes (9 studies, 638 participants)
None of the trials reported the proportion of re‐closure, incidence of keloid scar, patient satisfaction or QOL.
Primary outcome: SSI
Data from six trials compared the use of subcuticular sutures with surgical tapes for SSI. However, as two trials had no cases of infection (Barker 1984; Grottkau 2010), only data from the remaining four trials contributed to the meta‐analysis (Lazar 2011; Maartense 2002; O'Leary 2013; Pitcher 1983). Overall, 2.8% (10/354 participants) developed an SSI. There is no clear difference in the incidence of SSI between the two groups (RR 1.31, 95% CI 0.40 to 4.27; Analysis 4.1) with no evidence of heterogeneity (Tau2= 0, I2= 0%). Using risk difference, there were 10 more SSI per 1000 with subcuticular sutures compared with surgical tapes (30 fewer to 40 more) (RD 0.01, 95% CI ‐0.03 to 0.04; Tau2= 0, I2= 0%). It is uncertain whether subcuticular sutures increase or decrease the risk of SSI. The certainty of the evidence is very low, downgraded once for high risk of bias (attrition bias) and twice for imprecision. There are no trials that are considered to be at a low risk of selection bias. In the sensitivity analysis, there was no change in the interpretation of results by adopting the fixed‐effects model. ITT sensitivity analyses showed that the results did not change depending on the methods of the missing data (Analysis 4.2).
Secondary outcomes
Wound complications
Five trials compared the use of subcuticular sutures with surgical tapes for wound complications and contributed to the meta‐analysis (Anatol 1997; Barker 1984; O'Leary 2013; Pitcher 1983; Rosen 1997). A total of 128 participants (128/492 (26%)) developed wound complications. There may be little or no difference in the incidence of wound complications between the two groups (RR 0.90, 95% CI 0.61 to 1.34 ; Tau2= 0.07, I2= 42%; Analysis 4.3). The certainty of the evidence is low, downgraded once due to high risk of bias (attrition, reporting and other bias in trials accounting for 65% of the analysis weight) and once due to imprecision (the confidence intervals overlapped 1 and both 0.75 and 1.25).
Wound dehiscence
Four trials reported wound dehiscence. However, as three trials had no cases of dehiscence (Grottkau 2010; Lazar 2011; Rebello 2009), only Anatol 1997 data contributed to the meta‐analysis. All events occurred in the surgical tape group. It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical tapes because the certainty of the evidence is very low (RR 0.07, 95% CI 0.00 to 1.47; Tau2= 0, I2= 0%; Analysis 4.4) (downgraded once due to high risk of bias (attrition and other bias) and twice due to imprecision).
Hypertrophic scar
Barker 1984 reported this outcome at one year after operation. It is uncertain whether subcuticular sutures increase or reduce the risk of hypertrophic scar compared with surgical tapes because the certainty of the evidence is very low (RR 1.68; 95% CI 0.07 to 40.14; Analysis 4.5). The evidence was downgraded to very low certainty, once due to high risk of bias (attrition bias) and twice due to imprecision.
Wound pain intensity within seven days
Only three studies reported the wound pain intensity (Lazar 2011; O'Leary 2013; Rosen 1997). However, as the data of one trial were insufficient because of missing statistics (Rosen 1997), only data from the remaining two trials contributed to the meta‐analysis. Both studies assessed this outcome using a satisfaction score rated by the patients (scale, 1‐10, where 10 represented worst score). In one study, we calculated the SD from the reported P value (O'Leary 2013). There may be little or no difference in the pain intensity between the two groups (MD 0.41, 95% CI ‐0.02 to 0.83; Tau2= 0.03, I2= 28%; Analysis 4.6). The evidence was downgraded to low certainty, once due to high risk of bias (attrition bias affecting 56% of the analysis weight) and once due to imprecision (low numbers of participants).
Length of hospital stay
Lazar 2011 reported this outcome. There may be little or no difference between the groups in the length of hospital stay (MD 0.20 days, 95%CI ‐1.25 to 1.65; Analysis 4.7). The certainty of the evidence is low, downgraded twice for imprecision (very few numbers of participants).
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months)
Barker 1984 reported this outcome, but we did not use the data because the dropout numbers were not acceptable (> 30%) and not evenly distributed across the arms (see also Characteristics of included studies). Many studies did not contribute to this outcome because, although they evaluated cosmetic appearance, they did so less than six months after surgery (Grottkau 2010; Lazar 2011; Maartense 2002; O'Leary 2013; Rebello 2009).
Wound closure time in the operation (minutes)
Six studies reported the time taken for closure. However, as the data of two trials were insufficient (missing statistics) (Barker 1984; Pitcher 1983), only data from the remaining four trials contributed to the meta‐analysis (Grottkau 2010; Lazar 2011; Maartense 2002; Rebello 2009). In three studies (Grottkau 2010; Maartense 2002; Rebello 2009), we calculated the SD from the reported P values. As there was considerable heterogeneity (I2= 90%; Analysis 4.8), we presented the results narratively. The mean wound closure time in the surgical tape group ranged from 1.33 to 5.33 minutes. Mean differences ranged between 0.74 and 6.36 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. The certainty of the evidence is very low, downgraded once due to imprecision (low numbers of participants (169)) and twice for inconsistency (I2= 90% which was mostly quantitative (i.e. all studies agreed in the direction of effect but differed in its magnitude only)).
Cost
Only three trials reported this outcome (Anatol 1997; Lazar 2011; Maartense 2002). However, as the data of these trials were insufficient (missing statistics), we presented them narratively. Lazar 2011 (36 participants) reported the mean cost per patient was USD 4.36 for sutures and USD 32.91 for surgical tapes using Steri‐Strip S. The other trials favoured surgical tapes (Anatol 1997; Maartense 2002). Anatol 1997 (190 participants) reported the cost per closure was USD 12 to 14 for subcuticular sutures and USD 4.72 for tapes. Maartense 2002 (92 participants) reported the median cost per patient was 17.82€ for sutures and 8.68€ for surgical tapes.
It is uncertain whether subcuticular sutures increase the cost compared with surgical tapes. The certainty of the evidence is very low, downgraded one level for high risk of bias (attrition and other bias), one level for imprecision (narrative synthesis), and one level for inconsistency.
Summary of comparison
Very low‐certainty evidence suggests that it is uncertain whether subcuticular sutures increase or decrease the risk of SSI and wound dehiscence compared with surgical tapes. Low‐certainty evidence shows that there may be little or no difference between subcuticular sutures and surgical tape groups in the incidence of wound complications. See summary of findings Table 4.
Comparison 5. Subcuticular sutures compared with surgical zippers (3 studies, 447 participants)
None of the studies reported the proportion of re‐closure, incidence of hypertrophic scar or keloid, wound pain intensity, length of hospital stay, patient satisfaction or QOL. Two studies had one or two domains rated at high risk of bias (Tanaka 2016; Xu 2014).
Primary outcome: SSI
Three trials compared the use of subcuticular sutures with surgical zippers for SSI. However, as one trial had no cases of infection (Xu 2014), only data from the remaining two trials contributed to the meta‐analysis (Roolker 2002; Tanaka 2016). Overall, 1.2% (5/424 participants) developed an SSI. It is uncertain whether subcuticular sutures reduce or increase the risk of SSI compared with surgical zippers because the certainty of the evidence is very low (RR 0.80, 95% CI 0.08 to 8.48, Tau2= 1.04, I2= 35%; Analysis 5.1; 424 participants) (very low‐certainty evidence, downgraded once for risk of bias (attrition and reporting bias in trials accounting for 58% of the analysis weight) and twice for imprecision). There were no trials that were considered to be at a low risk of selection bias. In the sensitivity analysis, there was no change in the interpretation of results by adopting the fixed‐effects model and by ITT analyses based on the worst‐worst and the worst‐best scenarios (Analysis 5.8).
Secondary outcomes
Wound complications
All three studies reported this outcome. It is uncertain whether subcuticular sutures have an effect on wound complications compared with surgical zippers because the certainty of the evidence is very low (RR 0.55, 95% CI 0.15 to 2.04; Tau2= 0.63, I2= 47%; Analysis 5.2; 424 participants) (very low‐certainty evidence, downgraded once for risk of bias (detection, attrition and reporting bias) in trials accounting for 62% of the analysis weight), twice for imprecision).
Wound dehiscence
All three trials reported wound dehiscence. However as one trial had no cases of dehiscence (Xu 2014), only data from the remaining two trials contributed to the meta‐analysis (Roolker 2002; Tanaka 2016). It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical zippers because the certainty of the evidence is very low (RR 0.78, 95% CI 0.19 to 3.16; Tau2= 0.29, I2= 28%; Analysis 5.3; 424 participants) (very low‐certainty evidence, downgraded once for risk of bias (attrition and reporting bias in trials accounting for 46% of the analysis weight) and twice for imprecision).
Cosmesis of scar (as defined by the authors for a minimum follow‐up of six months).
Xu 2014 reported this outcome. Xu 2014 assessed cosmesis by patients using a VAS scale that ran from 0 to 10, where 10 represented the best score, and by one surgeon using the Hollander Incision Evaluation Score (Hollander 1995) at seven days, two weeks, six months, and one year after surgery. We used the data at one year. There maybe little or no difference between the two group in cosmesis using the VAS scale (MD ‐0.3, 95% CI ‐0.72 to 0.12; Analysis 5.4) and using the Hollander Score (MD ‐0.1, 95% CI ‐0.25 to 0.05; Analysis 5.5). The study is considered to be at a high risk of detection bias (Figure 3). The evidence was downgraded to low certainty, once for high risk of bias in blinding of outcome assessment in the single trial and once for imprecision (low numbers of participants; 90 participants).
Two studies did not contribute to this outcome because, although they evaluated cosmetic appearance, they did so less than six months after surgery (Roolker 2002; Tanaka 2016).
Wound closure time
All three trials reported this outcome. As there was extreme heterogeneity (I2= 100%; Analysis 5.6), we presented the results narratively. The mean wound closure time in surgical zippers group was 0.76 to 2.1 minutes. All trials suggested that subcuticular sutures took significantly more time than surgical zippers. Mean differences ranged between 4.38 and 8.25 minutes across three studies. Further analyses were not undertaken due to statistical heterogeneity in the results. The evidence was downgraded to very low certainty, once for high risk of bias (detection, attrition and reporting bias) and twice for inconsistency (I2= 100%).
Cost
Roolker 2002 reported this outcome. We calculated the SD from the reported P value. Subcuticular sutures may reduce the cost compared with surgical zippers. The mean cost was 8 USD in the subcuticular sutures group compared with 12 USD in the surgical zippers group with a difference in means of ‐5.00 USD (95% CI ‐8.76 to ‐1.26; 120 participants; Analysis 5.7). The evidence was downgraded to low certainty, due to risk of bias and imprecision.
Summary of comparison
Very low‐certainty evidence suggests that it is uncertain whether subcuticular sutures increase or decrease the risk of SSI and wound complications and dehiscence compared with surgical zippers. Low‐certainty evidence shows that subcuticular sutures may reduce costs compared to surgical zippers. See summary of findings Table 5.
Subgroup analyses
The following analyses were aimed at exploring possible sources of heterogeneity for the primary outcome, that is, SSI. We performed subgroup analyses for comparisons which included a sufficient number of studies (comparison 1: subcuticular sutures versus transdermal sutures and comparison 2: subcuticular sutures versus skin staples). We did not have sufficient studies for comparisons 3 through 5. Following the protocol, we divided the included studies into subgroups.
Subgroup analyses for comparison 1 (subcuticular sutures versus transdermal sutures)
-
1. absorbable versus non‐absorbable subcuticular sutures
There was no heterogeneity between the two subgroups ( I2 = 0%), and there was no change in the interpretation of results (Analysis 1.12).
-
2. location of surgery on the body (trunk versus extremities)
Substantial heterogeneity between the two subgroups ( I2 = 74%) was noted and the transdermal sutures were preferable in the extremities subgroup (RR 2.46, 95% CI 1.06 to 5.70; Analysis 1.13). However, this result in the extremities subgroup comprised only one study that was rated as being at high risk of bias in one domain (Corder 1991).
-
3. CDC class 1 (clean) versus class 2 (clean‐contaminated)
We did not include in the analysis studies that were unclear about contamination level (Clough 1975; Taube 1983; Zwart 1989). There was no change in the interpretation of results (Analysis 1.14) and there was moderate heterogeneity between the two subgroups ( I2 = 50.1%).
-
4. continuous versus interrupted skin suture
Substantial heterogeneity between the two subgroups ( I2 = 67.8%) was noted and the subcuticular sutures were preferable in the continuous sutures subgroup (RR 0.47, 95% CI 0.26 to 0.84; Analysis 1.15).
-
5. endoscopic versus open surgery.
There were insufficient studies reporting these data to undertake these analyses.
Subgroup analyses for comparison 2 (subcuticular sutures versus skin staples)
-
1. absorbable versus non‐absorbable subcuticular sutures
There was no heterogeneity between the two subgroups ( I2 = 0%), and there was no change in the interpretation of results (Analysis 2.14).
-
2. location of surgery on the body (trunk versus extremities)
There was no heterogeneity between the two subgroups ( I2 = 0%), and there was no change in the interpretation of results (Analysis 2.15).
-
3. CDC class 1 (clean) versus class 2 (clean‐contaminated)
We did not include in the analysis studies that were unclear about contamination level (Imamura 2016; Ranaboldo 1992; Zwart 1989). There was no heterogeneity between the two subgroups ( I2 = 0%), and there was no change in the interpretation of results (Analysis 2.16).
-
4. continuous versus interrupted skin suture
We did not include studies in the analysis that were unclear about the methods of suturing (continuous or interrupted). There was no heterogeneity between the two subgroups ( I2 = 0%), and there was no change in the interpretation of results (Analysis 2.17).
-
5. endoscopic versus open surgery.
There were insufficient studies reporting these data to undertake these analyses.
Discussion
Summary of main results
This review compared subcuticular sutures with any other skin closure methods for participants undergoing non‐obstetric operations. Sixty‐six trials with a total of 7487 participants were included: 25 studies compared subcuticular sutures with transdermal sutures; 18 studies compared subcuticular sutures with skin staples; 17 studies compared subcuticular sutures with tissue adhesives; nine studies compared subcuticular sutures with surgical tapes and three studies compared subcuticular sutures with surgical zippers. Most participants appeared to be adults, although several studies enrolled only children or both adults and children. The majority of studies included only CDC class 1 (clean) surgeries, however, two‐thirds of participants were enrolled in the studies which included CDC class 2 to 4 surgeries. The results are summarised in 'Summary of findings' tables.
Primary outcome: surgical site infection (SSI)
SSI is the most frequent healthcare‐associated infection, accounting for 31% of all healthcare‐associated infections among hospitalised patients (Mangram 1999; Magill 2012). It increases medical costs, prolongs hospital stay, and occasionally leads to mortality. There was no clear difference in the risk reduction of SSI between wounds closed with subcuticular sutures and any of the other skin closure methods reported here because of the small sample sizes or low event rates. The effect estimates are imprecise, which is reflected in the results of our GRADE assessments. The certainty of evidence varied from moderate to very low. Compared with skin staples (data from 4163 participants), subcuticular sutures probably have little or no effect on SSI because the certainty of evidence is moderate. There is moderate‐certainty evidence showing no clear difference in the incidence of SSI between participants treated with subcuticular sutures and those treated with tissue adhesives (data from 869 participants); confidence intervals were wide, spanning both appreciable benefits and harms so clear differences between treatments are not apparent. Compared with transdermal sutures (data from 3107 participants), there may be little difference in the incidence of SSI because the certainty of evidence is low. It is uncertain whether subcuticular sutures reduce or increase the risk of SSI compared with surgical tapes (data from 354 participants) or surgical zippers (data from 424 participants) because the certainty of the evidence is very low. The studies in the comparisons with transdermal sutures, surgical tapes or surgical zippers have some issues with high risk of bias in key domains. This is also reflected in the results of our GRADE assessments.
Compared with transdermal sutures, an ITT sensitivity analysis based on the worst‐best scenario showed that the results changed in favour of the transdermal sutures group. It indicated a high risk of bias due to missing outcome data. The reason may be that the majority of included studies that compared subcuticular sutures with transdermal sutures are old studies that did not report the detail of dropouts.
Secondary outcomes
The incidence of wound complications was reported for all comparisons, although the definition of wound complications varied across the studies. Subcuticular sutures for skin closure probably, on average, decrease the incidence of wound complications compared with skin staples (moderate‐certainty evidence based on 2973 participants). Although we downgraded the GRADE assessment due to imprecision (low number of events), the difference was probably more than the minimal clinically important difference. No other comparison showed evidence for a difference in risk of wound complications. We added a secondary outcome 'wound dehiscence' as a post hoc decision because, contrary to our expectations, many studies reported this outcome separately and we considered it clinically relevant. One notable finding in this review was that a clear difference in risk of wound dehiscence became apparent, with a higher risk of dehiscence associated with tissue adhesives. The difference was greater than the minimal clinically important difference. The certainty of the evidence is low, based on 1155 participants; subcuticular sutures may reduce the risk of wound dehiscence compared with adhesives. In addition, subcuticular sutures may decrease the incidence of wound dehiscence compared with skin staples. The difference was more than the minimal clinically important difference. The certainty of the evidence is low, based on 1984 participants. Many studies were also underpowered, and dehiscence was not used as a basis for sample size calculation. This is reflected in the results of our GRADE assessments. We are uncertain whether subcuticular sutures reduce wound dehiscence compared with the other closure methods (as the certainty of evidence has been assessed as very low), although most trials had lower numbers of dehiscence events in the subcuticular sutures group.
Cosmetic outcome is an important long‐term outcome for the patient and we considered appearance at or beyond six months after surgery to be meaningful. Many studies assessed cosmetic outcomes at less than six months. Therefore, we could not include limited numbers of studies. There may be little or no difference between subcuticular sutures and any of the other skin closure methods in most comparisons; this is low‐certainty evidence.
Patient satisfaction is also important when comparing closure methods provided that SSI, wound complications, dehiscence, hypertrophic/keloid scar and cosmetic appearance are satisfactory. Patient satisfaction may include ratings for cosmesis, overall comfort, ability to shower, dressing changes, tension on wounds, hygiene problems, allergic reactions and overall satisfaction. An important benefit of absorbable subcuticular sutures and tissue adhesives is that they do not need to be removed. Subcuticular sutures probably result in greater patient satisfaction than transdermal sutures (moderate‐certainty evidence based on 290 participants in a single trial) or skin staples (high‐certainty evidence based on 1232 participants in a single trial). The difference between subcuticular sutures and staples was very small (the mean patient satisfaction score (scale from 1 to 5) with subcuticular sutures was 0.20 higher), therefore this might be a less clinically important benefit than other comparisons. In contrast, there may be greater satisfaction with tissue adhesives than with subcuticular sutures (low‐certainty evidence based on 131 participants in a single trial).
The results of wound closure time in the pooled analyses were associated with considerable heterogeneity. It may depend on factors such as wound type and length. Thus, we presented the results narratively in most comparisons except for a comparison with transdermal sutures. There is moderate‐certainty evidence that wound closure time is probably longer in the subcuticular sutures group compared with transdermal sutures. Low‐certainty evidence suggested that it may take longer time to close a wound with subcuticular sutures than with skin staples. However, the difference between groups was only a few minutes (mean difference ranged from 0.3 to 5.81 minutes). The clinical impact of the difference depends on the total operation time of each procedure. The effect estimates for wound closure time suggest that this takes longer when subcuticular sutures are compared with tissue adhesives, surgical tapes or surgical zippers, however, the certainty of this evidence is very low.
Costs were not adequately reported in many studies with a few exceptions. There may be some differences in the cost when subcuticular sutures are compared with transdermal sutures or surgical zippers in favour of subcuticular sutures, but this is low‐certainty evidence. The differences were only USD 5 to USD 10, therefore, they are unlikely to exceed the minimal clinically important difference. It depends on the setting. In addition, the difference may be offset by increases in costs necessary for longer operations when subcuticular sutures are compared with transdermal sutures. We are uncertain whether subcuticular sutures increase or reduce the cost compared with skin staples or surgical tapes (very low‐certainty evidence), although most trials suggest lower costs in the subcuticular sutures groups compared with skin staples.
The proportion of re‐closure of the skin incision, incidence of hypertrophic or keloid scars and QOL were reported for a limited number of comparisons. There is moderate‐certainty evidence based on 1195 participants that skin closure by subcuticular sutures probably, on average, decreases the proportion of hypertrophic scars compared with skin staples. The proportion of re‐closure and the incidence of keloid scar was very low where reported and the effect of subcuticular sutures is very uncertain (low or very low‐certainty evidence). Only one trial with a small number of participants reported QOL. There was no clinically important difference in QOL (low‐certainty evidence).
Wound pain intensity and length of hospital stay were reported for many comparisons. We had planned to include any readmissions for wound‐related complications in the length of hospital stay, but none of the studies reported readmissions. There may be little or no clinically important difference between subcuticular sutures and any other methods of skin closure in most comparisons; this is low‐certainty evidence.
Overall completeness and applicability of evidence
There was no apparent exclusion of particular patient groups from the included studies. Therefore, the results of this review would apply to a wide range of people undergoing non‐obstetric surgery. However, due to the variety of the surgical procedures, the overall effects demonstrated in this review should be interpreted with caution. For example, incisions in areas of high tension were excluded when subcuticular sutures were compared with tissue adhesives and surgical tapes. Tissue adhesives and surgical tapes have not been evaluated in this population. In addition the surgical zipper is a new device for skin closure, and we were able to include only three studies (two: orthopaedic surgery, one: cardiac operation; two of three studies for paediatric populations). Note that a subgroup analysis comparing SSI between the study groups based on degree of contamination was carried out in an attempt to evaluate the effect of wound contamination on infection. However, wound classification was not consistently reported in some of the included studies. Other clinical factors that could affect wound outcomes such as endoscopic versus open surgery were not adequately addressed in the included studies. We encourage readers to also examine if surgical characteristics of the included populations were consistent with their particular interests.
It is also noteworthy that there was notable variation among studies with regard to the detailed suture methods such as nature of materials or continuous/interrupted suture methods. We examined their influences, where possible, with regards to the primary outcome only (for the reasons explained in Subgroup analysis and investigation of heterogeneity). However, the methods of continuous or interrupted suture were not consistently reported in many of the included studies.
In addition, we could not explore the influence of studies sponsored by companies, because funding was not reported in many of the included studies. It should be considered in the future studies.
Quality of the evidence
Five comparisons are presented in the 'Summary of findings' tables (summary of findings Table for the main comparison: subcuticular sutures versus transdermal sutures, summary of findings Table 2: skin staples, summary of findings Table 3: tissue adhesives, summary of findings Table 4: surgical tapes and summary of findings Table 5: surgical zippers). Following a GRADE assessment, the certainty of evidence was high to very low across the outcomes assessed. The quality of the evidence for comparison 4 (surgical tapes) and comparison 5 (zippers) was low to very low, depending on the outcome.
One of the main factors affecting the quality of evidence was the presence of an unclear or high risk of bias, for many included studies, in more than one important domain. As shown in Figure 2 and Figure 3, the majority of included studies were at unclear risk of selection bias; we found 16 studies to be at high risk of attrition bias and nine studies to be at high risk of other potential bias (unit of analysis issues, baseline imbalance, etc); we also found selective outcome reporting to be present. It may be difficult for both the surgeons and assessors of postoperative short‐term outcomes to be blinded to the intervention. The influence of the post‐randomisation exclusion of participants was tested by imputing outcomes for the missing participants under different scenarios. Compared with transdermal sutures, this showed that the worst‐best scenario resulted in different conclusions, and indicated a high risk of bias due to missing outcome data. We reflected this in our assessment of attrition bias in these individual studies, and also in our overall GRADE criteria. Fourteen studies were identified as having potential unit of analysis issues as it did not appear that paired or clustered data were accounted for in the analysis. We also reflected this in our assessment of other potential sources of bias in these individual studies, and also in our overall GRADE criteria.
The main factor affecting the quality of evidence was the lack of precision of results. Almost all outcome results were imprecise due to small sample size and limited number of outcome events, leading to wide confidence intervals. We downgraded evidence by one level when the confidence interval of the overall effect of an outcome crossed the line of no effect (or 1), in addition to the confidence intervals crossing either 0.75 (appreciable benefit) or 1.25 (appreciable harm) and by two levels when there were very few events and CIs around effects included both appreciable benefit and appreciable harm.
We downgraded for inconsistency when the meta‐analyses had I2 values which suggested substantial or considerable heterogeneity, or the included studies reported qualitatively different results. Such downgrading was necessary for some outcomes.
There was no downgrading for indirectness as the included studies were in agreement with the review question.
The outcomes assessed for publication bias were SSI and wound complications. A visual inspection of the funnel plot revealed no evidence of publication bias, therefore, no results were downgraded for publication bias.
Potential biases in the review process
Although this is the first, large and comprehensive systematic review for subcuticular sutures for skin closure and the funnel plots did not show any apparent asymmetry, there is reason to suspect that the available literature may still be affected by publication bias. Even when a study was published, lack of unified outcome measures across studies, especially in the older trials, led to suspicion of outcome reporting bias. We included only trials in which it was clear that the comparison involved subcuticular sutures versus other methods of skin closure. As mentioned in Excluded studies, 15 studies were excluded because it was not clear whether they involved this comparison. Attempts to contact the authors were unproductive. While the majority of these studies were unlikely to be included in this review, some of the studies may have met the inclusion criteria for this review. Many of the included studies were not ideal in terms of risk of bias as individual studies and did not provide sufficient information on long‐term patient outcomes. Fourteen studies were identified as having potential unit of analysis issues. Of these studies, only four studies were identified as it did not appear that paired data were accounted for in the analysis and the remaining studies were identified as it did not appear that clustered data were accounted for in the analysis. We adopted a pragmatic but conservative post hoc approach to analyses including clustered and paired data because a very small number of participants have more than one wound in almost all of these studies. We included such studies in meta‐analyses where possible (where unadjusted clustered data would produce too narrow CIs and unadjusted paired data too wide CIs). We undertook a post hoc sensitivity analysis to explore the impact of including data that had been inappropriately unadjusted. In all cases, it had little effect on the estimate of effect or the confidence intervals. We are therefore confident that our post hoc approach to data from these trials is unlikely to have affected the findings of the review, and that fully including the data increases the comprehensiveness of the review.
Agreements and disagreements with other studies or reviews
This is the first systematic review on this topic.
One related Cochrane systematic review entitled 'Continuous versus interrupted skin sutures for non‐obstetric surgery' showed that there was no significant difference between the groups in the proportion of participants who developed superficial SSI, but superficial wound dehiscence may be reduced by using continuous subcuticular sutures (Gurusamy 2014). In this review, it must be noted that in most of the included studies the continuous skin suture groups received subcuticular sutures, while the interrupted skin suture groups received non‐absorbable transdermal sutures. The results of our review can be said to be in agreement with Gurusamy 2014 in finding no difference in the incidence of SSI. There was no clear difference in the wound dehiscence in our study possibly due to a low number of events.
In obstetric surgery, four systematic reviews and meta‐analyses that evaluated subcuticular sutures have been published. A Cochrane systematic review did not find conclusive evidence about how the skin should be closed (Mackeen 2012). Staples are associated with similar outcomes in terms of wound infection, pain and cosmesis compared with subcuticular sutures. Mackeen 2012 showed staples are associated with an increased risk of skin separation compared with subcuticular sutures. The results of Mackeen 2012 are generally in agreement with our review. We also show subcuticular sutures may reduce the risk of skin separation (wound dehiscence) compared with skin staples.
The others concluded that there was a possible benefit with subcuticular sutures compared with skin staples, because of a lower incidence of wound complications and faster methods of skin closure (Clay 2011; Mackeen 2015; Tuuli 2011). Clay 2011 and Mackeen 2015 also reported wound separation (dehiscence) was significantly more frequent in the group that received staples. These results are in agreement with our review. Although Mackeen 2015 reported there were no significant differences in pain, patient satisfaction, or cosmesis, we found subcuticular sutures were associated with higher scores of patient satisfaction.
Study flow diagram.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Funnel plot of comparison: 1 Subcuticular sutures compared with transdermal sutures, outcome: 1.1 Surgical site infection.
Funnel plot of comparison: 2 Subcuticular sutures compared with skin staples, outcome: 2.1 Surgical site infection.
Funnel plot of comparison: 3 Subcuticular sutures compared with tissue adhesives, outcome: 3.3 Wound complications.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 1 Surgical site infection.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 2 Surgical site infection (sensitivity analyses).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 3 Wound complications.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 4 Wound dehiscence.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 5 Re‐closure.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 6 Hypertrophic scar.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 7 Length of hospital stay.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 8 Patient satisfaction (within 30 days).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 9 Patient satisfaction (after 60 days).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 10 Wound closure time.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 11 Cost.
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 12 Surgical site infection (subgroup analysis 1).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 13 Surgical site infection (subgroup analysis 2).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 14 Surgical site infection (subgroup analysis 3).
Comparison 1 Subcuticular sutures compared with transdermal sutures, Outcome 15 Surgical site infection (subgroup analysis 4).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 1 Surgical site infection.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 2 Surgical site infection (sensitivity analyses).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 3 Wound complications.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 4 Wound dehiscence.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 5 Hypertrophic scar.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 6 Pain intensity within seven days.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 7 Pain intensity after 30 days.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 8 Length of hospital stay.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 9 Cosmesis of scar.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 10 Patient satisfaction (within 30 days).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 11 SF‐12v2 PCS.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 12 SF‐12v2 MCS.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 13 Wound closure time.
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 14 Surgical site infection (subgroup analysis 1).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 15 Surgical site infection (subgroup analysis 2).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 16 Surgical site infection (subgroup analysis 3).
Comparison 2 Subcuticular sutures compared with skin staples, Outcome 17 Surgical site infection (subgroup analysis 4).
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 1 Surgical site infection.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 2 Surgical site infection (sensitivity analyses).
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 3 Wound complications.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 4 Wound dehiscence.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 5 Re‐closure.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 6 Pain intensity after 30 days.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 7 Length of hospital stay.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 8 Patient satisfaction (within 30 days).
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 9 SF‐12v2 PCS.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 10 SF‐12v2 MCS.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 11 Wound closure time.
Comparison 3 Subcuticular sutures compared with tissue adhesives, Outcome 12 Cost.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 1 Surgical site infection.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 2 Surgical site infection (sensitivity analyses).
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 3 Wound complications.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 4 Wound dehiscence.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 5 Hypertrophic scar.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 6 Pain intensity within seven days.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 7 Length of hospital stay.
Comparison 4 Subcuticular sutures compared with surgical tapes, Outcome 8 Wound closure time.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 1 Surgical site infection.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 2 Wound complications.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 3 Wound dehiscence.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 4 Cosmesis of scar (VAS).
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 5 Cosmesis of scar (HWES).
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 6 Wound closure time.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 7 Cost.
Comparison 5 Subcuticular sutures compared with surgical zippers, Outcome 8 Surgical site infection (ITT sensitivity analyses).
| Subcuticular sutures compared with transdermal sutures for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with transdermal sutures | Risk with subcuticular sutures | |||||
| Surgical site infection (SSI) Incidence of wound infection follow‐up: 7 to 42 days | 71 per 1,000 | 78 per 1,000 | RR 1.10 | 3107 | ⊕⊕⊝⊝ | There may be little difference between subcuticular and transdermal sutures groups in the incidence of SSI. |
| Wound complications Incidence of wound complications follow‐up: 5 to 42 days | 102 per 1,000 | 85 per 1,000 | RR 0.83 | 1489 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures have an effect on wound complications compared with transdermal sutures. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 7 to 42 days | 61 per 1,000 | 21 per 1,000 | RR 0.35 | 866 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with transdermal sutures (as the certainty of the evidence has been assessed as very low) . |
| Cosmesis of scar (cosmesis) follow‐up: 6 months to 12 months | Insufficient data reported. We were unable to carry out further analyses. | ‐ | 950 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures improve the cosmesis of scar compared with transdermal sutures. | |
| Patient satisfaction (at 30 days) | The mean patient satisfaction score (at 30 days) was 7.4 | The mean patient satisfaction score with subcuticular sutures was 1.6 higher (1.32 to 1.88 higher). | MD 1.60 (1.32 to 1.88) | 290 | ⊕⊕⊕⊝ | Patient satisfaction at 30 days is probably higher in subcuticular sutures group compared with transdermal sutures group. |
| Wound closure time (minutes) | The mean wound closure time was 5.40 minutes | The mean wound closure time with subcuticular sutures was 5.81 minutes longer (5.13 to 6.49 minutes longer) | MD 5.81 (5.13 to 6.49) | 585 | ⊕⊕⊕⊝ | Wound closure time is probably longer in subcuticular sutures group compared with transdermal sutures group. |
| Cost | The mean cost was 16 Naira | The mean cost with subcuticular sutures was 8 Naira lower (13.05 lower to 2.95 lower). | MD ‐8.00 (‐13.05 to ‐2.95) | 100 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the cost compared with transdermal sutures. In the study, participants used non‐absorbable (Nylon) subcuticular sutures. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded two levels: one level due to several trials at high risk of bias in at least one domain (attrition, selection, reporting and other bias); one level for imprecision as the confidence intervals overlapped 1 and 1.25. 2 Downgraded three levels: one level due to high risks of bias across varying domains (attrition, selection, reporting and other bias); one level for imprecision as the confidence intervals overlapped 1 and both 0.75 and 1.25); one level for inconsistency. 3 Downgraded three levels: one level due to risk of bias (attrition and selection bias) and two levels due to imprecision (study 95% CIs are wide). 4 Downgraded three levels: one level for high risk of attrition bias; one level for imprecision (narrative synthesis); one level for inconsistency (two reaching significance and two not). 5 Downgraded one level: one level for imprecision (low numbers of participants). 6 Downgraded one level for inconsistency. 7 Downgraded two levels: one level for risk of bias (the risk of bias in the included single study was unclear in almost every domain); one level for imprecision (low numbers of participants). | ||||||
| Subcuticular sutures compared with skin staples for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with skin staples | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 10 to 42 days | 90 per 1,000 | 73 per 1,000 | RR 0.81 (0.64 to 1.01) | 4163 | ⊕⊕⊕⊝ | There is probably little or no difference between subcuticular sutures and skin staples groups in the incidence of SSI. |
| Wound complications Incidence of wound complications follow‐up: 10 to 42 days | 110 per 1,000 | 87 per 1,000 | RR 0.79 | 2973 | ⊕⊕⊕⊝ | Subcuticular sutures probably on average decrease wound complications compared with skin staples. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 10 to 42 days | 59 per 1,000 | 37 per 1,000 | RR 0.63 | 1984 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the risk of wound dehiscence compared with skin staples. |
| Cosmesis of scar follow‐up: 6 months to 1 year | The cosmetic score in the subcuticular sutures group was on average 0.12 SDs (95% CI: 0.11 lower to 0.36 higher) higher in the patients treated with subcuticular sutures than in the patients treated with skin staples. | SMD 0.12 (‐0.11 to 0.35) | 291 | ⊕⊕⊝⊝ | As a rule of thumb, 0.2 SD represents a small effect, 0.5 a moderate effect, and 0.8 a large effect. There may be little or no difference between subcuticular sutures and skin staples groups in cosmesis of scar. | |
| Patient satisfaction (at 30 days) | The mean patient satisfaction score (at 30 days) was 4.2 | The mean patient satisfaction score with subcuticular sutures was 0.20 higher (0.10 to 0.30 higher). | MD 0.20 (0.10 to 0.30) | 1232 | ⊕⊕⊕⊕ | Patient satisfaction at 30 days after surgery is slightly higher in subcuticular sutures group compared with skin staples group. |
| Wound closure time (minutes) | The mean wound closure time ranged from 0.9 to 4.5 minutes | Mean differences ranged between 0.30 and 5.50 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 1384 | ⊕⊕⊝⊝ | Wound closure time may be a few minutes longer in subcuticular sutures group compared with skin staples group. |
| Cost | Three trials favoured subcuticular sutures. It cost almost 5 to 15 USD lower per participant than staples. Another one favoured staples because most of the cost differential was attributed to procedure times. We were unable to carry out further analyses because of insufficient data. | ‐ | 342 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the cost compared with skin staples. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level: one level for imprecision (the confidence intervals overlapped 1 and 0.75). 2 Downgraded one level: one level for imprecision (low numbers of events). 3 Downgraded two levels: one level due to high risks of bias (detection and other bias) in one trial accounting for 24% of the analysis weight; one level for imprecision (low numbers of events). 4 Downgraded two levels: one level due to high risk of bias (attrition and reporting bias); one level for imprecision (the confidence intervals overlapped 0 and minimal clinically important difference). 5 Downgraded two levels: two levels for inconsistency (I2 = 99%). 6 Downgraded three levels: one level for high risk of bias (detection and other bias); one level for imprecision (narrative synthesis); one level for inconsistency. | ||||||
| Subcuticular sutures compared with tissue adhesives for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with tissue adhesives | Risk with Subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 7 to 42 days | 50 per 1,000 | 39 per 1,000 | RR 0.77 | 869 | ⊕⊕⊕⊝ | There is no clear difference in the incidence of SSI between participants treated with subcuticular sutures and those treated with tissue adhesives. Confidence intervals are wide, spanning both appreciable benefits and harms so clear differences between treatments are not apparent. |
| Wound complications Incidence of wound complications follow‐up: 10 to 42 days | 170 per 1,000 | 106 per 1,000 | RR 0.62 | 1058 | ⊕⊕⊝⊝ | There is no clear difference in the incidence of wound complications between participants treated with subcuticular sutures and those treated with tissue adhesives. Although the point estimate on the side of a possible benefit, the 95% confidence intervals includes the possibility of both benefit and harm so clear differences between treatments are not apparent. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 10 to 42 days | 43 per 1,000 | 10 per 1,000 | RR 0.23 | 1155 | ⊕⊕⊝⊝ | Subcuticular sutures may decrease wound dehiscence in comparison with tissue adhesives. |
| Cosmesis of scar assessed with: score system scale from: 1 to 10 (best score) | The study reported there were similar outcomes between the two groups (mean score: subcuticular sutures 8.8 vs tissue adhesives 8.8). We were unable to carry out further analyses because of insufficient data. | ‐ | 99 | ⊕⊕⊝⊝ | There may be little or no difference in cosmesis of scar between subcuticular and tissue adhesives groups. | |
| Patient satisfaction (within 30 days) scale from: 1 to 10 follow up: 14 to 21 days | The mean patient satisfaction score (within 30days) was 9.5 | The mean patient satisfaction score with subcuticular sutures was 2.05 lower (3.05 to 1.05 lower). | MD ‐2.05 (‐3.05 to ‐1.05) | 131 | ⊕⊕⊝⊝ | Patient satisfaction within 30 days after surgery may be lower in subcuticular sutures group compared with tissue adhesives. |
| Wound closure time (minutes) | The mean wound closure time ranged from 0.3 to 3.7 minutes | Mean differences ranged between ‐0.34 and 10.39 across 11 studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 895 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with tissue adhesives (as the certainty of the evidence has been assessed as very low). |
| Cost | The mean cost ranged from 31.96 to 65.1 USD and from 20.3 to 34.01 EUR | Mean differences ranged between ‐57.36 and ‐4.26 USD (‐16.19 and ‐10.30 EUR). Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 422 | ⊕⊝⊝⊝ | Two studies reported the cost by using USD, the others reported the cost by using EUR. It is uncertain whether subcuticular sutures reduce the cost compared with tissue adhesives (as the certainty of the evidence has been assessed as very low). |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded one level: one level for imprecision (the confidence intervals overlapped 1 and both 0.75 and 1.25). 2 Downgraded two levels: one level due to high risks of bias across varying domains (detection, attrition and other bias) accounting for 54% of the analysis weight; one level for imprecision (the confidence intervals overlapped 1 and 0.75). 3 Downgraded two levels: downgraded one level due to high risks of bias across varying domains (detection, attrition and other bias) accounting for 45% of the analysis weight and one level for imprecision (low numbers of events). 4 Downgraded two levels: one level due to high risks of bias (attrition and other bias); one level for imprecision (narrative synthesis). 5 Downgraded two levels: one level due to high risks of bias (attrition and other bias); one level for imprecision (low numbers of participants). 6 Downgraded three levels: one level due to high risks of bias across varying domains (detection, attrition and other bias); two levels for inconsistency (I2= 97%). 7 Downgraded three levels: one level due to high risks of bias across varying domains (attrition and other bias); two levels for inconsistency (I2= 96%). | ||||||
| Subcuticular sutures compared with surgical tapes for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with surgical tapes | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 7 to 30 days | 25 per 1,000 | 33 per 1,000 | RR 1.31 | 354 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of SSI compared with surgical tapes. |
| Wound complications Incidence of wound complications follow‐up: 5 to 42 days | 293 per 1,000 | 263 per 1,000 | RR 0.90 | 492 | ⊕⊕⊝⊝ | There may be little or no difference between subcuticular sutures and surgical tape groups in the incidence of wound complications. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 7 to 42 days | 23 per 1,000 | 2 per 1,000 | RR 0.07 | 264 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical tapes (as the certainty of the evidence has been assessed as very low). |
| Cosmesis of scar | One study reported this outcome, but the data of this trial could not be included as it was insufficient. | |||||
| Patient satisfaction | Not reported in any of the studies. | |||||
| Wound closure time (minutes) | The mean wound closure time ranged from 1.33 to 5.33 minutes | Mean differences ranged between 0.74 and 6.36 minutes across four studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 169 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with surgical tapes (as the certainty of the evidence has been assessed as very low). |
| Cost | Two studies reported the cost per participant was 10‐15 USD higher in subcuticular sutures. The other reported the cost was about 30 USD higher in surgical tapes. We were unable to carry out further analyses because of insufficient data. | ‐ | 315 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures increase the cost with surgical tapes. | |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1Downgraded two levels: one level for high risk of bias (attrition bias) and two levels for imprecision (very few events and wide 95% confidence intervals). 2 Downgraded two levels: one level due to high risks of bias across varying domains (attrition, reporting and other bias) accounting for 65% of the analysis weight; one level for imprecision (the confidence intervals overlapped 1 and both 0.75 and 1.25). 3 Downgraded three levels: one level due to high risks of bias across varying domains (attrition and other bias); two levels for imprecision (very few events and wide 95% confidence intervals). 4 Downgraded three levels: one level for imprecision (low numbers of participants); two levels for inconsistency (I2= 90%). 5 Downgraded three levels: one level for high risk of bias (attrition and other bias); one level for imprecision (narrative synthesis); one level for inconsistency (the included studies reported the opposite results, leading to qualitative heterogeneity). | ||||||
| Subcuticular sutures compared with surgical zippers for skin closure in non‐obstetric surgery | ||||||
| Patient or population: skin closure in non‐obstetric surgery | ||||||
| Outcomes | Anticipated absolute effects* (95% CI) | Relative effect | № of participants | Quality of the evidence | Comments | |
| Risk with surgical zippers | Risk with subcuticular sutures | |||||
| Surgical site infection Incidence of wound infection follow‐up: 14 to 42 days | 14 per 1,000 | 11 per 1,000 | RR 0.80 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of SSI compared with surgical zippers. |
| Wound complications Incidence of wound complications follow‐up: 14 to 42 days | 83 per 1,000 | 45 per 1,000 | RR 0.55 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound complications compared with surgical zippers. |
| Wound dehiscence Incidence of wound dehiscence follow‐up: 14 to 42 days | 32 per 1,000 | 25 per 1,000 | RR 0.78 | 424 | ⊕⊝⊝⊝ | It is uncertain whether subcuticular sutures reduce the risk of wound dehiscence compared with surgical zippers. |
| Cosmesis of scar | The mean cosmesis of scar (VAS) score was 7.7 | The mean cosmetic VAS score with subcuticular sutures was 0.3 lower (0.72 lower to 0.12 higher). | MD ‐0.3 | 90 | ⊕⊕⊝⊝ | There may be little or no difference between subcuticular sutures and surgical zippers groups in the cosmesis of scar. |
| Patient satisfaction | Not reported in any of the studies. | |||||
| Wound closure time (minutes) | The mean wound closure time was 0.76 to 2.1 minutes | Mean differences ranged between 4.38 and 8.25 minutes across three studies. Further analyses were not undertaken due to statistical heterogeneity in the results. | ‐ | 424 | ⊕⊝⊝⊝ | It is uncertain whether it takes longer time to close a wound with subcuticular sutures than with surgical zippers. |
| Cost | The mean cost was 13 USD | The mean cost with subcuticular sutures was 5 USD lower (8.76 lower to 1.26 lower). | MD ‐5.00 (‐8.76 to ‐1.26) | 120 | ⊕⊕⊝⊝ | Subcuticular sutures may reduce the cost compared with surgical zippers. |
| *The risk in the intervention group (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 Downgraded three levels: one level due to high risks of bias (attrition and reporting bias) accounting for 58% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 2 Downgraded three levels: one level due to high risks of bias (detection, attrition and reporting bias) accounting for 62% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 3 Downgraded three levels: one level due to high risks of bias (attrition and reporting bias) accounting for 46% of the analysis weight; two levels for imprecision (very low numbers of events and wide 95% confidence intervals). 4 Downgraded two levels: one level due to high risk of bias in blinding of outcome assessment in the single trial; one level for imprecision (low numbers of participants). 5 Downgraded three levels: one level for high risks of bias (detection, attrition and reporting bias); two levels for inconsistency (I2= 100%). 6 Downgraded two levels: one level for risk of bias (unclear in almost every domain in included single study); one level for imprecision (low numbers of participants). | ||||||
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Surgical site infection Show forest plot | 20 | 3107 | Risk Ratio (M‐H, Random, 95% CI) | 1.10 [0.80, 1.52] |
| 2 Surgical site infection (sensitivity analyses) Show forest plot | 20 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Low risk of selection bias | 1 | 290 | Risk Ratio (M‐H, Random, 95% CI) | 0.93 [0.57, 1.50] |
| 2.2 Excluding unit of analysis issues | 17 | 2768 | Risk Ratio (M‐H, Random, 95% CI) | 1.04 [0.75, 1.42] |
| 2.3 Worst‐worst scenario | 20 | 3186 | Risk Ratio (M‐H, Random, 95% CI) | 1.23 [0.97, 1.55] |
| 2.4 Worst‐best scenario | 20 | 3186 | Risk Ratio (M‐H, Random, 95% CI) | 1.59 [1.05, 2.42] |
| 3 Wound complications Show forest plot | 9 | 1489 | Risk Ratio (M‐H, Random, 95% CI) | 0.83 [0.40, 1.71] |
| 4 Wound dehiscence Show forest plot | 6 | 866 | Risk Ratio (M‐H, Random, 95% CI) | 0.35 [0.08, 1.54] |
| 5 Re‐closure Show forest plot | 2 | 246 | Risk Ratio (M‐H, Random, 95% CI) | 1.16 [0.09, 14.57] |
| 6 Hypertrophic scar Show forest plot | 2 | 233 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.25, 3.39] |
| 7 Length of hospital stay Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8 Patient satisfaction (within 30 days) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 9 Patient satisfaction (after 60 days) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 10 Wound closure time Show forest plot | 2 | 585 | Mean Difference (IV, Random, 95% CI) | 5.81 [5.13, 6.49] |
| 11 Cost Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 12 Surgical site infection (subgroup analysis 1) Show forest plot | 19 | 2966 | Risk Ratio (M‐H, Random, 95% CI) | 1.09 [0.78, 1.52] |
| 12.1 absorbable subcuticular sutures group | 15 | 2351 | Risk Ratio (M‐H, Random, 95% CI) | 1.13 [0.72, 1.76] |
| 12.2 non‐absorbable subcuticular sutures group | 4 | 615 | Risk Ratio (M‐H, Random, 95% CI) | 0.98 [0.56, 1.71] |
| 13 Surgical site infection (subgroup analysis 2) Show forest plot | 19 | 2966 | Risk Ratio (M‐H, Random, 95% CI) | 1.09 [0.78, 1.52] |
| 13.1 Trunk | 17 | 2751 | Risk Ratio (M‐H, Random, 95% CI) | 1.01 [0.73, 1.39] |
| 13.2 Extremities | 2 | 215 | Risk Ratio (M‐H, Random, 95% CI) | 2.46 [1.06, 5.70] |
| 14 Surgical site infection (subgroup analysis 3) Show forest plot | 16 | 2493 | Risk Ratio (M‐H, Random, 95% CI) | 1.09 [0.74, 1.61] |
| 14.1 CDC class 1 | 5 | 731 | Risk Ratio (M‐H, Random, 95% CI) | 1.98 [0.77, 5.11] |
| 14.2 CDC class 2 | 11 | 1762 | Risk Ratio (M‐H, Random, 95% CI) | 0.95 [0.64, 1.40] |
| 15 Surgical site infection (subgroup analysis 4) Show forest plot | 7 | 1193 | Risk Ratio (M‐H, Random, 95% CI) | 0.71 [0.49, 1.03] |
| 15.1 subcuticular continuous sutures group | 6 | 903 | Risk Ratio (M‐H, Random, 95% CI) | 0.47 [0.26, 0.84] |
| 15.2 subcuticular interrupted sutures group | 1 | 290 | Risk Ratio (M‐H, Random, 95% CI) | 0.93 [0.57, 1.50] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Surgical site infection Show forest plot | 15 | 4163 | Risk Ratio (M‐H, Random, 95% CI) | 0.81 [0.64, 1.01] |
| 2 Surgical site infection (sensitivity analyses) Show forest plot | 15 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Low risk of selection bias | 4 | 2865 | Risk Ratio (M‐H, Random, 95% CI) | 0.91 [0.72, 1.16] |
| 2.2 Excluding unit of analysis issues | 12 | 3875 | Risk Ratio (M‐H, Random, 95% CI) | 0.81 [0.61, 1.09] |
| 2.3 Worst‐worst scenario | 15 | 4219 | Risk Ratio (M‐H, Random, 95% CI) | 0.87 [0.67, 1.13] |
| 2.4 Worst‐best scenario | 15 | 4219 | Risk Ratio (M‐H, Random, 95% CI) | 0.94 [0.65, 1.37] |
| 3 Wound complications Show forest plot | 9 | 2973 | Risk Ratio (M‐H, Random, 95% CI) | 0.79 [0.64, 0.98] |
| 4 Wound dehiscence Show forest plot | 7 | 1984 | Risk Ratio (M‐H, Random, 95% CI) | 0.63 [0.43, 0.94] |
| 5 Hypertrophic scar Show forest plot | 3 | 1195 | Risk Ratio (M‐H, Random, 95% CI) | 0.77 [0.60, 0.98] |
| 6 Pain intensity within seven days Show forest plot | 3 | 218 | Mean Difference (IV, Random, 95% CI) | ‐1.86 [‐10.37, 6.65] |
| 7 Pain intensity after 30 days Show forest plot | 3 | 196 | Std. Mean Difference (IV, Random, 95% CI) | 0.18 [‐0.30, 0.66] |
| 8 Length of hospital stay Show forest plot | 5 | 2794 | Mean Difference (IV, Random, 95% CI) | ‐0.58 [‐1.57, 0.42] |
| 9 Cosmesis of scar Show forest plot | 3 | 291 | Std. Mean Difference (IV, Random, 95% CI) | 0.12 [‐0.11, 0.35] |
| 10 Patient satisfaction (within 30 days) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 11 SF‐12v2 PCS Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 12 SF‐12v2 MCS Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 13 Wound closure time Show forest plot | 4 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 14 Surgical site infection (subgroup analysis 1) Show forest plot | 15 | 4163 | Risk Ratio (M‐H, Random, 95% CI) | 0.81 [0.64, 1.01] |
| 14.1 absorbable subcuticular sutures group | 13 | 4051 | Risk Ratio (M‐H, Random, 95% CI) | 0.80 [0.62, 1.02] |
| 14.2 non‐absorbable subcuticular sutures group | 2 | 112 | Risk Ratio (M‐H, Random, 95% CI) | 1.0 [0.07, 14.64] |
| 15 Surgical site infection (subgroup analysis 2) Show forest plot | 15 | 4163 | Risk Ratio (M‐H, Random, 95% CI) | 0.81 [0.64, 1.01] |
| 15.1 Trunk | 10 | 3657 | Risk Ratio (M‐H, Random, 95% CI) | 0.84 [0.67, 1.05] |
| 15.2 Extremities | 5 | 506 | Risk Ratio (M‐H, Random, 95% CI) | 0.70 [0.30, 1.60] |
| 16 Surgical site infection (subgroup analysis 3) Show forest plot | 12 | 3554 | Risk Ratio (M‐H, Random, 95% CI) | 0.75 [0.55, 1.02] |
| 16.1 CDC class 1 | 8 | 712 | Risk Ratio (M‐H, Random, 95% CI) | 0.68 [0.39, 1.17] |
| 16.2 CDC class 2 | 4 | 2842 | Risk Ratio (M‐H, Random, 95% CI) | 0.75 [0.49, 1.15] |
| 17 Surgical site infection (subgroup analysis 4) Show forest plot | 10 | 3628 | Risk Ratio (M‐H, Random, 95% CI) | 0.80 [0.60, 1.06] |
| 17.1 subcuticular continuous sutures group | 6 | 549 | Risk Ratio (M‐H, Random, 95% CI) | 0.77 [0.42, 1.44] |
| 17.2 subcuticular interrupted sutures group | 4 | 3079 | Risk Ratio (M‐H, Random, 95% CI) | 0.78 [0.54, 1.12] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Surgical site infection Show forest plot | 10 | 869 | Risk Ratio (M‐H, Random, 95% CI) | 0.77 [0.41, 1.45] |
| 2 Surgical site infection (sensitivity analyses) Show forest plot | 10 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Excluding unit of analysis issues | 9 | 823 | Risk Ratio (M‐H, Random, 95% CI) | 0.77 [0.41, 1.45] |
| 2.2 Worst‐worst scenario | 10 | 901 | Risk Ratio (M‐H, Random, 95% CI) | 0.94 [0.61, 1.46] |
| 2.3 Worst‐best scenario | 10 | 901 | Risk Ratio (M‐H, Random, 95% CI) | 1.08 [0.43, 2.73] |
| 3 Wound complications Show forest plot | 11 | 1058 | Risk Ratio (M‐H, Random, 95% CI) | 0.62 [0.35, 1.11] |
| 4 Wound dehiscence Show forest plot | 11 | 1155 | Risk Ratio (M‐H, Random, 95% CI) | 0.23 [0.07, 0.74] |
| 5 Re‐closure Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6 Pain intensity after 30 days Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 7 Length of hospital stay Show forest plot | 2 | 189 | Mean Difference (IV, Random, 95% CI) | 0.22 [‐0.39, 0.84] |
| 8 Patient satisfaction (within 30 days) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 9 SF‐12v2 PCS Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 10 SF‐12v2 MCS Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 11 Wound closure time Show forest plot | 11 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 12 Cost Show forest plot | 4 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 12.1 Cost (USD) | 2 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| 12.2 Cost (EUR) | 2 | Mean Difference (IV, Random, 95% CI) | 0.0 [0.0, 0.0] | |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Surgical site infection Show forest plot | 6 | 354 | Risk Ratio (M‐H, Random, 95% CI) | 1.31 [0.40, 4.27] |
| 2 Surgical site infection (sensitivity analyses) Show forest plot | 6 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 2.1 Excluding unit of analysis issues | 5 | 267 | Risk Ratio (M‐H, Random, 95% CI) | 1.31 [0.40, 4.27] |
| 2.2 Worst‐worst scenario | 6 | 373 | Risk Ratio (M‐H, Random, 95% CI) | 1.27 [0.67, 2.42] |
| 2.3 Worst‐best scenario | 6 | 373 | Risk Ratio (M‐H, Random, 95% CI) | 2.30 [0.52, 10.29] |
| 3 Wound complications Show forest plot | 5 | 492 | Risk Ratio (M‐H, Random, 95% CI) | 0.90 [0.61, 1.34] |
| 4 Wound dehiscence Show forest plot | 4 | 264 | Risk Ratio (M‐H, Random, 95% CI) | 0.07 [0.00, 1.47] |
| 5 Hypertrophic scar Show forest plot | 1 | Risk Ratio (M‐H, Fixed, 95% CI) | Totals not selected | |
| 6 Pain intensity within seven days Show forest plot | 2 | 118 | Mean Difference (IV, Random, 95% CI) | 0.41 [‐0.02, 0.83] |
| 7 Length of hospital stay Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8 Wound closure time Show forest plot | 4 | 169 | Mean Difference (IV, Random, 95% CI) | 2.63 [0.67, 4.60] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Surgical site infection Show forest plot | 3 | 424 | Risk Ratio (M‐H, Random, 95% CI) | 0.80 [0.08, 8.48] |
| 2 Wound complications Show forest plot | 3 | 424 | Risk Ratio (M‐H, Random, 95% CI) | 0.55 [0.15, 2.04] |
| 3 Wound dehiscence Show forest plot | 3 | 424 | Risk Ratio (M‐H, Random, 95% CI) | 0.78 [0.19, 3.16] |
| 4 Cosmesis of scar (VAS) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 5 Cosmesis of scar (HWES) Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 6 Wound closure time Show forest plot | 3 | Mean Difference (IV, Random, 95% CI) | Totals not selected | |
| 7 Cost Show forest plot | 1 | Mean Difference (IV, Fixed, 95% CI) | Totals not selected | |
| 8 Surgical site infection (ITT sensitivity analyses) Show forest plot | 3 | Risk Ratio (M‐H, Random, 95% CI) | Subtotals only | |
| 8.1 Worst‐worst scenario | 3 | 447 | Risk Ratio (M‐H, Random, 95% CI) | 1.11 [0.10, 13.05] |
| 8.2 Worst‐best scenario | 3 | 447 | Risk Ratio (M‐H, Random, 95% CI) | 2.26 [0.03, 198.64] |
