Introduction
Septoplasty is a commonly performed surgical procedure for the treatment of nasal airway obstruction caused by septal deviation.Reference Carrie, Fouweather, Homer, O’Hara, Rousseau and Rooshenas1, Reference Shin and Jang2 Although generally considered safe and effective, post-operative complications, including bleeding, septal haematoma, synechia, septal perforation and patient discomfort, continue to affect surgical outcomes and patient satisfaction.Reference Shin and Jang2, Reference Brescia, Franz, Frasconi, Marciani, Soldati and Frigo3
Various methods have been introduced to stabilise the nasal septum after surgical procedure and facilitate the appropriate healing of the mucoperichondrial flaps. Among these methods, intranasal splints are widely used because they provide structural support, reduce dead space and maintain separation between opposing mucosal surfaces. These effects are thought to decrease the risk of haematoma formation and post-operative adhesions.Reference Lau, Elhassan and Singh4, Reference Quinn, Bonaparte and Kilty5 However, until their removal, intranasal splints may also contribute to nasal obstruction, local irritation, sleep disturbance, epiphora and patient discomfort.Reference Karatas, Pehlivanoglu, Salviz, Kuvat, Cebi and Dikmen6, Reference Amer, Tomoum, Askar and Saafan7
Trans-septal suturing, commonly referred to as quilting suturing, has emerged as an alternative stabilisation technique that eliminates the need for intranasal foreign material. By securing the mucoperichondrial flaps and obliterating dead space, trans-septal sutures are intended to provide stabilisation comparable to that achieved using intranasal splints.Reference Certal, Silva, Santos, Correia and Carvalho8 Several studies have suggested potential advantages of this approach, including reduced post-operative pain and improved patient comfort; however, concerns regarding increased operative time and technical complexity persist.Reference Alhejaily, Alhejaily, Alosaimi, Bogari and Aldosari9, Reference Sowerby and Wright10
Previous systematic reviews have primarily focused on comparisons between trans-septal suturing and nasal packing, or have evaluated stabilisation techniques indirectly. Furthermore, even those reviews that included intranasal splinting as a comparator did not fully address patient-centred outcomes such as post-operative pain trajectories, sleep disturbance and overall recovery experience, nor did they systematically evaluate operative time as a measure of surgical burden. Consequently, evidence directly and comprehensively comparing trans-septal suturing and intranasal splinting – particularly with respect to patient-centred recovery outcomes – remains limited.Reference Certal, Silva, Santos, Correia and Carvalho8, Reference Alhejaily, Alhejaily, Alosaimi, Bogari and Aldosari9, Reference AlBilasi, Alghamdi, AlDhawi, Albalawi and Alzahrani11, Reference Aldhwaihy, Alhazzani and Alkarzae12
To address this gap, we conducted a systematic review and meta-analysis comparing trans-septal suturing and intranasal splinting following septoplasty. The objective was to evaluate differences in post-operative complications, operative burden and patient-centred recovery outcomes between these two stabilisation strategies.
Materials and Methods
Study design and registration
This study was designed as a systematic review and meta-analysis, and was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann and Mulrow13 The protocol was prospectively registered with the Open Science Framework (https://osf.io/dygp2/).
Literature search and eligibility criteria
A systematic search of PubMed, Embase, Scopus, Web of Science and the Cochrane Library was conducted in June 2026. Search terms were developed using the PICO [Patient/Population, Intervention, Comparator/Control, and Outcome(s)] framework and incorporated both controlled vocabulary and free-text keywords related to septoplasty, trans-septal suturing, quilting sutures, and intranasal splints (Supplementary Table 1).
Studies were considered eligible if they directly compared trans-septal suturing with intranasal splinting following septoplasty and reported at least one post-operative clinical outcome. Intranasal splinting was operationally defined as any biocompatible sheet or device placed within the nasal cavity to support the septum without traversing it, including conventional silicone, silastic and Doyle-type splints.
Functionally equivalent devices – including the nasal septal retainer and synthetic polyurethane foam (SPF/Nasopore) – were also eligible under this definition, provided they served the same structural function as a conventional splint. Studies in which an intranasal splint (INS) was combined with concurrent anterior nasal packing were eligible only when the packing–splint combination was reported and analysed as a single comparator arm; studies using packing alone, without any splinting component, were not eligible as an INS arm.
Randomised controlled trials and comparative observational studies were included. Studies were excluded if they were non-comparative, lacked a direct comparison between the two interventions or did not provide sufficient quantitative data for analysis. Review articles, case reports, conference abstracts, editorials and animal studies were also excluded. Two investigators independently screened records and assessed full-text articles for eligibility. Any disagreements were resolved through consensus or consultation with a third reviewer. The study selection process is presented in Figure 1.
PRISMA flow diagram of study selection.
Figure 1 Long description
Flowchart sequence: Identification of studies via databases and registers, Records identified from PubMed, the Cochrane Central Register of Controlled Trials, Embase, Web of Science, SCOPUS and Google Scholar, n equals 1,708, Records removed before screening, Duplicate records removed, n equals 1,485, Records screened, n equals 223, Records excluded after screening of title or abstract, n equals 196, Reports sought for retrieval, n equals 27, Reports not retrieved, n equals 2, Reports assessed for eligibility, n equals 25, Reports excluded, No quantifiable data or no relevant data, n equals 8, Reports of included studies, n equals 17.
Data collection and outcomes
Data extraction was performed independently by two reviewers using a predefined data extraction form. Information collected included study characteristics, patient demographics, sample size, surgical techniques, operative duration, post-operative complications and patient-reported outcomes.Reference Kim, Jang and Hwang14–Reference Kim, Jang and Hwang16 For studies with more than two treatment arms (Özkırış et al., 2013Reference Özkırış, Kapusuz and Saydam32; Cayonu et al., 2014Reference Cayonu, Acar, Horasanlı, Altundag and Salihoglu21; Karataş et al., 2014Reference Karataş, Yüksel, Doğan and Sentürk22; Baklaci et al., 2019Reference Baklaci, Güler, Kuzucu, Kum and Özcan34; Arora et al., 2020Reference Arora, Munjal, Khurana, Munjal and Saggar26; Soni 2020Reference Soni27 and Sari 2022Reference Sari, Atar, Kumral, Uygan, Karaketir and ŞG30), only the trans-septal suturing and INS arms, as defined above, were extracted for analysis; additional comparator arms consisting of packing alone (e.g. gauze or Merocel packing) were excluded from data extraction.
The primary endpoints were post-operative adhesion, bleeding, septal haematoma, septal perforation, crust formation and infection. Secondary endpoints included operative time, post-operative pain, sleep disturbance, epiphora and other recovery-related outcomes when reported. For analytical purposes, outcomes were categorised into operative outcomes, bleeding-related events, healing-related complications and patient-reported recovery measures.
Quality assessment
Methodological quality was assessed independently by two reviewers. Randomised controlled trials were evaluated using the Cochrane Risk-of-Bias tool, whereas non-randomised comparative studies were appraised using the Newcastle-Ottawa Scale.
Statistical analysis
All analyses were performed using R software (version 4.6.0; R Foundation for Statistical Computing, Vienna, Austria). For binary outcomes, pooled odds ratios (ORs) with corresponding 95 per cent confidence intervals (CIs) were calculated. Continuous outcomes were synthesised using standardised mean differences (SMDs) and 95 per cent CIs to account for variations in measurement scales across studies. A random-effects model was applied for all meta-analyses. Between-study heterogeneity was assessed using Cochran’s Q statistic and quantified with the I2 statistic, with values greater than 50 per cent considered indicative of substantial heterogeneity.Reference Higgins, Thompson, Deeks and Altman17
Pre-specified subgroup analyses examined differences according to the timing of post-operative assessment for patient-reported outcomes and according to suturing technique (conventional quilting vs device-assisted suturing) for operative time. Statistical interaction tests were performed to compare subgroup effects. Potential publication bias was explored through visual inspection of funnel plots and Egger’s regression test when at least 10 studies were available for a given outcome.Reference Sterne, Sutton, Ioannidis, Terrin, Jones and Lau18 Formal assessment was not performed for outcomes with fewer studies because of limited statistical reliability.
Results
In total, 17 studies comprising 1502 participants were included in the meta-analysis.Reference Ardehali and Bastaninejad19–Reference Ghasembaglou, Shahidi and Bahman35 Study characteristics are presented in Table 1, and the risk-of-bias assessments are presented in Supplementary Tables 2 and 3.
Characteristics of included studies
Table 1 Long description
Seventeen included septoplasty studies are listed with columns for design, sample size, participant age and sex, country, treatment groups, and measured outcomes. Most studies are randomised controlled trials, with several cohort studies; sample sizes range from 23 to 228 participants. Countries are mainly Turkey, with additional studies from Iran, Iraq, Japan, India, Saudi Arabia, Jordan, and China. Across studies, the most common comparison is trans-septal suturing versus intranasal splints and or nasal packing materials such as Merocel, gauze, paraffin gauze, synthetic polyurethane foam, anterior nasal packing, or a septal retainer. Outcomes repeatedly assessed include adhesion formation, septal haematoma, local infection, septal perforation, and bleeding; many studies also report post-operative pain, and some include crusting, epiphora, operative time, bleeding severity, and sleep disturbance. Age reporting varies by study, using either means with standard deviations or ranges, and sex is missing in some cohorts. Because study characteristics rather than results are summarized, it supports understanding of what was compared and what outcomes were tracked, but it does not indicate which treatment performed better.
ANP = anterior nasal packing; INS = intranasal splint; NR; not reported; NSP = nasal septal retainer; RCT = randomised controlled trial; SPF = synthetic polyurethane foam (Nasopore); TSS = trans-septal suturing.
Post-operative complications
Post-operative adhesion was reported in nine studies involving 841 patients. Meta-analysis demonstrated no significant difference between trans-septal suturing and INS (OR, 1.35; 95 per cent CI, 0.59–3.13; p = 0.478), with no evidence of heterogeneity (I2 = 0 per cent). Similarly, pooled data from five studies comprising 327 patients showed comparable rates of crust formation between the groups (OR, 1.23; 95 per cent CI, 0.44–3.44; p = 0.693; I2 = 24 per cent). Septal haematoma was evaluated in four studies involving 974 patients and occurred at similar frequencies in both groups (OR, 1.14; 95 per cent CI, 0.23–5.79; p = 0.872; I2 = 0 per cent). Analysis of five studies including 1009 patients similarly demonstrated no significant difference in septal perforation rates (OR, 0.57; 95 per cent CI, 0.18–1.80; p = 0.339; I2 = 0 per cent) (Figure 2).
Forest plots comparing post-operative adhesion (a), crust formation (b), septal haematoma (c), and septal perforation (d) between trans-septal suturing and INS. Odds ratios (ORs) with 95 per cent confidence intervals (CIs) are shown for individual studies and pooled estimates.
Figure 2 Long description
A) Forest plot with columns labeled, Study, Experimental Events Total, Control Events Total, Odds Ratio, OR, 95 percent CI, Weight fixed and Weight random. The forest axis shows odds ratio values on a logarithmic scale with tick labels 0.01, 0.1, 1, 10, 100. Studies listed are Adelheid 2009, Duka 2010, Duka 2013, Caglar 2014, Karatas 2015, Akin 2016, Bulut 2018, Karaca 2019, Arora 2020, Son 2020, Almutairi 2021, Alfouh 2022 and Wang 2026. The common effect model line reports OR equals 1.35, 95 percent CI 0.59 to 3.13, Weight fixed 100.0 percent. The random effect model line reports OR equals 1.35, 95 percent CI 0.59 to 3.13, Weight random 100.0 percent. Heterogeneity line reads, I squared equals 0.0 percent, tau squared equals 0, p equals 0.8379. B) Forest plot with the same column headers and the same forest axis tick labels 0.01, 0.1, 1, 10, 100. Studies listed are Amin 2015, Bulut 2018, Arora 2020, Son 2020 and Almutairi 2021. The common effect model line reports OR equals 1.23, 95 percent CI 0.44 to 3.44, Weight fixed 100.0 percent. The random effect model line reports OR equals 1.23, 95 percent CI 0.44 to 3.44, Weight random 100.0 percent. Heterogeneity line reads, I squared equals 24.0 percent, tau squared equals 0.5464, p equals 0.2612. C) Forest plot with the same column headers and a forest axis with tick labels 0.1, 0.5, 1, 2, 10. Studies listed are Orkinos 2013, Caglar 2014, Karatas 2015, Amin 2015, Son 2020, Arora 2020, Bulut 2018, Almutairi 2021, Alfouh 2022 and Wang 2026. The common effect model line reports OR equals 1.14, 95 percent CI 0.23 to 5.79, Weight fixed 100.0 percent. The random effect model line reports OR equals 1.14, 95 percent CI 0.23 to 5.79, Weight random 100.0 percent. Heterogeneity line reads, I squared equals 0.0 percent, tau squared equals 0, p equals 0.6977. D) Forest plot with the same column headers and the same forest axis tick labels 0.01, 0.1, 1, 10, 100. Studies listed are Adelheid 2009, Duka 2010, Duka 2013, Caglar 2014, Karatas 2015, Amin 2015, Son 2020, Arora 2020, Bulut 2018, Karaca 2019, Gok 2019, Ghasemzadeh 2020, Almutairi 2021, Alfouh 2022, Sarar 2022 and Wang 2026. The common effect model line reports OR equals 0.57, 95 percent CI 0.18 to 1.80, Weight fixed 100.0 percent. The random effect model line reports OR equals 0.57, 95 percent CI 0.18 to 1.80, Weight random 100.0 percent. Heterogeneity line reads, I squared equals 0.0 percent, tau squared equals 0, p equals 0.8507.
Post-operative haemorrhage
Eight studies involving 597 patients reported post-operative bleeding events. The pooled analysis showed no significant difference between trans-septal suturing and INS (OR, 1.41; 95 per cent CI, 0.48–4.15; p = 0.537), with moderate heterogeneity across studies (I2 = 48.9 per cent). Clinically significant haemorrhage was analysed in five studies comprising 540 patients. Event rates were comparable between the two interventions (OR, 1.22; 95 per cent CI, 0.40–3.73; p = 0.730; I2 = 10.6 per cent). Bleeding severity scores were reported in seven studies including 290 patients. Although higher scores tended to be observed in the trans-septal suturing group, the pooled effect did not reach statistical significance (SMD, 0.47; 95 per cent CI, −0.07 to 1.00; p = 0.088). Substantial heterogeneity was present (I2 = 74.1 per cent). Stratification according to assessment timing (post-operative day 1 vs day 2) revealed no significant subgroup effect (p = 0.627) (Figure 3).
Forest plots comparing overall post-operative bleeding (a), clinically significant post-operative haemorrhage (b), and bleeding severity scores (c) between trans-septal suturing and intranasal splinting. Subgroup analysis was performed according to the timing of post-operative assessment (post-operative day 1 vs post-operative day 2) for bleeding severity scores. Odds ratios (ORs), standardised mean differences (SMDs), and 95 per cent confidence intervals (CIs) are presented for individual studies and pooled estimates.
Figure 3 Long description
A) Forest plot with a table and a log scale axis. Column headers read, Study, Experimental Events Total, Control Events Total, Odds Ratio, OR, 95 percent CI, Weight common, Weight random. Study rows list Cayon 2014, Amin 2015, Sari 2016, Baklaci 2019, Arora 2020, Soni 2020, Ghasemboglou 2020, Al Muslem 2021, Sar 2022, Wang 2026. Odds ratio values and 95 percent confidence interval values shown at right are: Cayon 2014, 5.44 0.61, 48.40; Amin 2015, 0.48 0.08, 2.86; Sari 2016, 9.47 0.47, 191.57; Baklaci 2019, 0.07 0.00, 1.34; Arora 2020, 0.75 0.48, 1.17; Soni 2020, 4.75 0.48, 46.91; Ghasemboglou 2020, 0.64 0.10, 4.20; Al Muslem 2021, 0.40 0.04, 3.82; Sar 2022, 6.00 0.93, 42.73; Wang 2026, 0.30 0.03, 2.73. The axis tick labels are 0.01, 0.1, 1, 10, 100. The common effect model row shows Experimental Total 286 and Control Total 311, with 1.39 0.65, 2.97. The random effects model row shows 1.41 0.48, 4.15. Text below reads, Heterogeneity: I squared equals 48.9 percent, tau squared equals 1.1367, p equals 0.0570. B) Forest plot with the same column headers as A. Study rows list Cayon 2014, Kayahan 2017, Baklaci 2018, Kuboki 2019, Arora 2020, Soni 2020, Ghasemboglou 2020, Al Muslem 2021, Sar 2022, Wang 2026. Odds ratio values and 95 percent confidence interval values shown at right are: Cayon 2014, 1.00 0.06, 16.44; Kayahan 2017, not legible; Baklaci 2018, not legible; Kuboki 2019, 0.18 0.01, 4.01; Arora 2020, 0.00 0.00, 0.00; Soni 2020, 3.35 0.32, 35.36; Ghasemboglou 2020, 0.64 0.10, 4.20; Al Muslem 2021, not legible; Sar 2022, not legible; Wang 2026, 10.29 0.47, 225.93. The axis tick labels are 0.01, 0.1, 1, 10, 100. The common effect model row shows Experimental Total 261 and Control Total 279, with 1.22 0.40, 3.73. The random effects model row shows 1.22 0.40, 3.73. Text below reads, Heterogeneity: I squared equals 10.6 percent, tau squared less than 0.0001, p equals 0.3456. C) Forest plot with a table and a linear axis. Column headers read, Study, Experimental Total Mean SD, Control Total Mean SD, Standardised Mean Difference, SMD, 95 percent CI, Weight common, Weight random. The axis tick labels are 2, 1, 0, 1, 2. A subgroup label reads, g, POD1. Study rows and standardised mean difference values with 95 percent confidence interval values shown at right are: Amin 2015, minus 0.11 minus 0.62, 0.40; Kiboki 2018, 1.18 0.28, 2.08; Wang 2026, 1.06 0.18, 1.95. The random effects model row for this subgroup shows 0.64 0.23, 1.51. A second subgroup label reads, g, POD2. Study rows and standardised mean difference values with 95 percent confidence interval values shown at right are: Amin 2015, minus 0.41 minus 0.93, 0.10; Kiboki 2018, 1.48 0.54, 2.43; Baklaci 2019, 0.52 minus 0.11, 1.15; Soni 2020, 0.04 minus 0.78, 0.86; Wang 2026, 0.17 minus 0.17, 0.50. The random effects model row for this subgroup shows 0.35 minus 0.41, 1.12. The overall common effect model row shows Experimental Total 142 and Control Total 148, with 0.26 0.00, 0.61. The overall random effects model row shows 0.47 minus 0.07, 1.00. Text below includes, Heterogeneity: I squared equals 74.1 percent, tau squared equals 0.3769, p equals 0.0007. Additional lines include, Test for subgroup differences, common effect: chi squared equals 0.62, df equals 1, p equals 0.4311 and Test for subgroup differences, random effects: chi squared equals 0.24, df equals 1, p equals 0.6271.
Infection and epiphora
Eleven studies involving 756 patients assessed post-operative infection, with events observed in four studies. Although a lower infection rate was observed in the trans-septal suturing group, the difference was not statistically significant (OR, 0.27; 95 per cent CI, 0.06–1.24; p = 0.092; I2 = 0 per cent). Post-operative epiphora was evaluated in three studies comprising 147 patients. While the pooled estimate favoured trans-septal suturing, no significant difference was identified between the groups (OR, 0.46; 95 per cent CI, 0.15–1.36; p = 0.161; I2 = 0 per cent) (Figure 4).
Forest plots comparing post-operative infection (a) and epiphora (b) between trans-septal suturing and intranasal splinting. Odds ratios (ORs) with 95 per cent confidence intervals (CIs) are presented for individual studies and pooled estimates.
Figure 4 Long description
A) Forest plot with a study table and a logarithmic odds ratio scale. Table headers read, Study, Experimental Events, Experimental Total, Control Events, Control Total, Odds Ratio, OR, 95 percent CI, Weight common, Weight random. Study rows list, Arderali 2009, Kula 2010, Ozkiris 2013, Cagiran 2014, Kubik 2018, Ghasembaglou 2019, Arora 2020, Al Muslem 2021, Alrooh 2022, Sari 2022, Wang 2026. The odds ratio axis shows tick labels 0.01, 0.1, 1, 10, 100. Four study effect entries are printed: 0.12 with 95 percent CI 0.01 to 2.34, 0.10 with 95 percent CI 0.00 to 2.04, 3.06 with 95 percent CI 0.12 to 76.95, 0.19 with 95 percent CI 0.01 to 4.10. Summary rows read, Common effect model, 0.27 with 95 percent CI 0.06 to 1.24 and Random effects model, 0.27 with 95 percent CI 0.06 to 1.24. Totals printed are Experimental total 365 and Control total 391. The heterogeneity line reads, Heterogeneity: I squared equals 0.0 percent, tau squared equals 0.0001, p equals 0.4012. B) Forest plot with the same table headers and odds ratio scale. Study rows list, Amin 2015, Ghasembaglou 2019, Soni 2020. The odds ratio axis shows tick labels 0.01, 0.1, 1, 10, 100. Study effect entries printed are 0.58 with 95 percent CI 0.12 to 2.67, 0.45 with 95 percent CI 0.07 to 2.76, 0.18 with 95 percent CI 0.01 to 4.01. Summary rows read, Common effect model, 0.46 with 95 percent CI 0.15 to 1.36 and Random effects model, 0.46 with 95 percent CI 0.15 to 1.36. Totals printed are Experimental total 73 and Control total 74. The heterogeneity line reads, Heterogeneity: I squared equals 0.0 percent, tau squared equals 0, p equals 0.8051.
Operative time
Operative duration was reported in four studies including 224 patients. Compared with INS, trans-septal suturing was associated with a significantly longer operative time (SMD, 0.65; 95 per cent CI, 0.19–1.10; p = 0.005), with moderate heterogeneity across studies (I2 = 60.2 per cent). Subgroup analysis according to suturing technique demonstrated that conventional quilting sutures were associated with a significant increase in operative time (SMD, 0.77; 95 per cent CI, 0.28–1.27), whereas device-assisted suturing using the Maniceps Septum Stitch system resulted in operative times comparable to those of splint placement (SMD, 0.07; 95 per cent CI, −0.76 to 0.91). However, the interaction between subgroups was not statistically significant (p = 0.160) (Figure 5).
Forest plot comparing operative time between trans-septal suturing and intranasal splinting. Subgroup analysis was conducted according to suturing technique (conventional quilting sutures vs device-assisted suturing). Standardised mean differences (SMDs) and 95 per cent confidence intervals (CIs) are shown for individual studies and pooled estimates.
Figure 5 Long description
Study table and forest plot with two subgroups and an overall summary. Table headers read, Study, Total under Experimental, Total under Control, Standardised Mean Difference, SMD, 95 percent CI, Weight under Random and Weight under Fixed. Subgroup label reads, Conventional suture technique. Study rows list: Ozkinas 2013 with SMD 0.36 and 95 percent CI 0.04 to 0.75. Sari 2014 with SMD 1.09 and 95 percent CI 0.33 to 1.65. Son 2020 with SMD 0.99 and 95 percent CI 0.33 to 1.65. A subgroup summary row labeled, Random effects model, shows SMD 0.77 and 95 percent CI 0.28 to 1.27. Subgroup label reads, Maniceps septum stitch device. Study row lists: Kuboki 2018 with SMD 0.07 and 95 percent CI minus 0.76 to 0.91. A subgroup summary row labeled, Random effects model, shows SMD 0.62 and 95 percent CI 0.35 to 0.89. Overall summary row labeled, Random effects model, shows SMD 0.65 and 95 percent CI 0.19 to 1.10. Forest plot x-axis tick labels run from minus 1.5 to 1.5, with ticks at minus 1.5, minus 1.0, minus 0.5, 0, 0.5, 1.0 and 1.5. The y-axis consists of the study and model rows listed in the table. Study estimates are drawn as squares with horizontal confidence interval lines and subgroup and overall pooled estimates are drawn as diamonds.
Patient-centred outcomes
Fifteen studies involving 1041 patients reported post-operative pain scores. Trans-septal suturing was associated with significantly lower post-operative pain compared with INS (SMD, −1.09; 95 per cent CI, −1.68 to −0.49; p = 0.0003). However, substantial heterogeneity was observed (I2 = 93.3 per cent), indicating considerable variability in the magnitude and direction of treatment effects across studies. Although the pooled estimate favoured trans-septal suturing, this finding should therefore be interpreted with caution. Notably, Kuboki et al. (2018)Reference Kuboki, Kikuchi, Asaka, Onda, Nakayama and Kojima25 reported a significant effect in the opposite direction (SMD, 1.23), indicating that the direction of effect was not consistent across all studies. No significant difference was observed between post-operative day-1 and day-2 assessments (p = 0.939).
Given the subjective nature of post-operative pain and the difficulty of blinding patients to splint placement, the potential for reporting and performance bias should also be considered. Sleep disturbance was evaluated in five studies including 204 patients. Patients treated with trans-septal suturing experienced significantly less sleep disturbance than those receiving INS (SMD, −0.32; 95 per cent CI, −0.60 to −0.05; p = 0.022), with no evidence of heterogeneity (I2 = 0 per cent). Findings were consistent across assessment time points (p = 0.748) (Figure 6).
Forest plots comparing post-operative pain (a) and sleep disturbance (b) between trans-septal suturing and intranasal splinting. Subgroup analyses were performed according to the timing of outcome assessment (post-operative day 1 vs post-operative day 2). Standardised mean differences (SMDs) and 95 per cent confidence intervals (CIs) are presented for individual studies and pooled estimates.
Figure 6 Long description
A) Forest plot with subgroup headings, g equals POD1 and g equals POD2. Table columns read, Study; Experimental Total, Mean, SD; Control Total, Mean, SD; Standardised Mean Difference; SMD; 95 percent CI; Weight common; Weight random. The x axis is labeled, Standardised Mean Difference, with tick labels minus 3, minus 2, minus 1, 0, 1, 2, 3 and a vertical reference line at 0. Under g equals POD1, SMD and 95 percent CI entries are: Ardalan 2009, minus 0.51, 95 percent CI minus 0.90 to minus 0.12; Ozkiris 2013, minus 2.62, 95 percent CI minus 3.16 to minus 2.08; Karatas 2014, minus 2.50, 95 percent CI minus 2.99 to minus 2.01; Amin 2015, minus 0.93, 95 percent CI minus 1.47 to minus 0.40; Kubilay 2018, 1.23, 95 percent CI 0.33 to 2.14; Sonu 2020, minus 1.77, 95 percent CI minus 2.52 to minus 1.03; Alfano 2022, minus 0.53, 95 percent CI minus 0.94 to minus 0.13. Subgroup pooled rows list: Common effect model, minus 1.15, 95 percent CI minus 1.29 to minus 1.01; Random effects model, minus 1.11, 95 percent CI minus 2.08 to minus 0.14. Heterogeneity line reads, I squared equals 94.4 percent, tau squared equals 1.6218, p less than 0.0001. Under g equals POD2, SMD and 95 percent CI entries are: Ozkiris 2013, minus 2.62, 95 percent CI minus 3.16 to minus 2.08; Karatas 2014, minus 2.50, 95 percent CI minus 2.99 to minus 2.01; Amin 2015, 0.39, 95 percent CI minus 0.90 to 1.67; Kubilay 2018, 0.45, 95 percent CI 0.18 to 0.71; Bakbak 2019, 0.47, 95 percent CI 1.10 to 1.24; Alfano 2022, minus 0.53, 95 percent CI minus 1.03 to minus 0.03; Wang 2020, minus 0.35, 95 percent CI minus 1.77 to 1.07. Subgroup pooled rows list: Common effect model, minus 1.15, 95 percent CI minus 1.35 to minus 0.95; Random effects model, minus 1.06, 95 percent CI minus 1.85 to minus 0.27. Heterogeneity line reads, I squared equals 93.1 percent, tau squared equals 1.2013, p less than 0.0001. Overall pooled rows list: Common effect model, minus 1.15, 95 percent CI minus 1.29 to minus 1.01; Random effects model, minus 1.09, 95 percent CI minus 1.68 to minus 0.49. Test lines list, chi squared equals 1.2824, p equals 0.2573 and chi squared equals 0.00, p equals 0.9899. B) Forest plot with subgroup headings, g equals POD1 and g equals POD2. Table columns read, Study; Experimental Total, Mean, SD; Control Total, Mean, SD; Standardised Mean Difference; SMD; 95 percent CI; Weight common; Weight random. The x axis is labeled, Standardised Mean Difference, with tick labels minus 1.0, minus 0.5, 0, 0.5, 1.0 and a vertical reference line at 0. Under g equals POD1, SMD and 95 percent CI entries are: Amin 2015, minus 0.27, 95 percent CI minus 0.78 to 0.24; Kubilay 2018, minus 0.27, 95 percent CI minus 1.09 to 0.55. Subgroup pooled rows list: Common effect model, minus 0.27, 95 percent CI minus 0.70 to 0.17; Random effects model, minus 0.27, 95 percent CI minus 0.70 to 0.17. Heterogeneity line reads, I squared equals 0 percent, tau squared equals 0, p equals 0.9990. Under g equals POD2, SMD and 95 percent CI entries are: Amin 2015, minus 0.16, 95 percent CI minus 0.67 to 0.35; Sonu 2020, 0.62, 95 percent CI 1.25 to 0.02; Wang 2020, minus 0.45, 95 percent CI minus 1.28 to 0.38. Subgroup pooled rows list: Common effect model, minus 0.36, 95 percent CI minus 0.72 to minus 0.00; Random effects model, minus 0.36, 95 percent CI minus 0.72 to minus 0.00. Heterogeneity line reads, I squared equals 0 percent, tau squared equals 0, p equals 0.6371. Overall pooled rows list: Common effect model, minus 0.32, 95 percent CI minus 0.60 to minus 0.05; Random effects model, minus 0.32, 95 percent CI minus 0.60 to minus 0.05. Test lines list, chi squared equals 0.00, p equals 0.8535 and chi squared equals 0.10, p equals 0.7480.
GRADE assessment
The certainty of evidence for the major clinical outcomes was assessed using the GRADE framework (Supplementary Table 4). The evidence supporting lower post-operative pain with trans-septal suturing was rated as low certainty because of serious risk of bias related to the subjective nature of pain assessment and the inability to adequately blind participants, together with substantial between-study heterogeneity (I2 = 93.3 per cent). Nevertheless, no additional downgrade for publication bias was applied because funnel plot inspection and Egger’s regression test (p = 0.841) showed no evidence of small-study effects. Evidence for reduced sleep disturbance was also rated as low certainty, primarily because of the subjective outcome assessment and imprecision. The certainty of evidence for rare complications, including clinically significant haemorrhage, septal hematoma and septal perforation, was very low because of the small number of events and wide CIs.
Publication bias
Assessment of publication bias was feasible only for post-operative pain because this outcome was reported in at least 10 studies. Egger’s regression test showed no evidence of small-study effects (p = 0.841).
Discussion
This systematic review and meta-analysis demonstrated that trans-septal suturing and INS provide comparable outcomes with respect to major post-operative complications following septoplasty. Rates of adhesion formation, crusting, septal haematoma, septal perforation, post-operative bleeding, and clinically significant haemorrhage were similar between the two approaches, indicating that both techniques offered effective post-operative septal stabilisation.
Although the two techniques differed mechanistically, their clinical performance appeared largely equivalent. Intranasal splints maintained separation between opposing mucosal surfaces while providing structural support, whereas trans-septal suturing stabilised the septum through approximation of the mucoperichondrial flaps and elimination of dead space.Reference Lau, Elhassan and Singh4, Reference Certal, Silva, Santos, Correia and Carvalho8 The comparable incidences of haematoma, perforation, and adhesion formation observed in the present analysis suggested that either strategy could reliably preserve septal integrity when appropriately performed.
In contrast, patients treated with trans-septal suturing experienced significantly less post-operative pain and sleep disturbance than those receiving intranasal splints. These findings are likely to reflect the avoidance of intranasal foreign material, which may contribute to mucosal irritation, nasal obstruction, pressure-related discomfort and impaired sleep during the early post-operative period.Reference Karatas, Pehlivanoglu, Salviz, Kuvat, Cebi and Dikmen6, Reference Alhejaily, Alhejaily, Alosaimi, Bogari and Aldosari9 However, the certainty of evidence supporting these patient-reported benefits was limited. Participant blinding is inherently difficult when comparing trans-septal suturing with INS, and awareness of treatment allocation may influence subjective symptom reporting. Moreover, substantial heterogeneity was observed for post-operative pain. Accordingly, these findings suggested that trans-septal suturing may improve early post-operative comfort rather than establishing definitive superiority over INS.
Furthermore, although statistical significance was not achieved, trans-septal suturing was associated with lower rates of post-operative infection and epiphora. These trends may be attributable to the temporary presence of splints within the nasal cavity, which could promote secretion retention, local irritation or bacterial colonisation.Reference Ritter, Alkan, Yahav, Soudry and Reifen36 However, the limited number of available studies and low event rates precluded definitive conclusions and therefore further investigation is warranted.
The principal drawback of trans-septal suturing was the increase in operative time. Conventional quilting sutures required a longer surgical duration than splint placement, reflecting the additional technical steps involved.Reference Sowerby and Wright10 Nevertheless, subgroup analysis suggested that this disadvantage may be mitigated by the adoption of device-assisted suturing systems. Although the interaction test was not statistically significant, the smaller effect estimate observed in the device-assisted subgroup suggested that technological advances may improve procedural efficiency without compromising clinical outcomes.
These findings have important practical implications for post-operative management following septoplasty. Given the comparable efficacy of the two techniques in preventing major complications, treatment selection may be guided primarily by patient comfort, surgeon preference and operative considerations. Trans-septal suturing may be particularly advantageous when early post-operative comfort is a priority, whereas INS remains a reasonable option when procedural simplicity or operative efficiency is preferred.
Several limitations should be considered when interpreting these findings. Firstly, significant heterogeneity was observed for certain continuous outcomes, particularly post-operative pain and bleeding severity. Variability in assessment methods and timing of outcome measurements may have contributed to this inconsistency. In addition, the lack of participant blinding may have influenced subjective patient-reported outcomes, particularly pain and sleep disturbance. Secondly, some patient-centred outcomes, including infection and epiphora, were reported in relatively few studies, thereby limiting statistical power. Thirdly, the INS comparator itself was heterogeneous: in addition to conventional silicone or silastic splints, several included studies used functionally analogous devices, such as the nasal septal retainer and synthetic polyurethane foam (SPF/Nasopore); differences in splint design, duration of placement and suturing technique may therefore have introduced additional clinical heterogeneity across studies. Despite these limitations, this study represents one of the most comprehensive direct comparisons of trans-septal suturing and INS currently available. Overall, the two techniques provided comparable protection against post-operative complications following septoplasty. Trans-septal suturing may improve early post-operative comfort, with lower pain and sleep disturbance, at the expense of a modest increase in operative time; however, the limited certainty of evidence precludes definitive conclusions regarding its superiority over INS.
Summary
What is already known on the subject
What this paper adds to our understanding
• Trans-septal suturing and intranasal splinting showed comparable rates of major post-operative complications.
• Trans-septal suturing significantly reduced post-operative pain and sleep disturbance.
• Trans-septal suturing required a modest increase in operative time.
• Device-assisted suturing achieved operative times comparable to intranasal splinting.
• Trans-septal suturing is a safe, patient-centred alternative to intranasal splinting after septoplasty.
Conclusion
Trans-septal suturing and INS provide comparable protection against major post-operative complications following septoplasty. However, trans-septal suturing is associated with reduced post-operative pain and sleep disturbance, albeit with a modest increase in operative time. These findings support trans-septal suturing as an effective alternative to INS, particularly when post-operative patient comfort is prioritised.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S0022215126106185.
Data availability statement
The English in this document has been checked by at least two professional editors, both native speakers of English. For a certificate, please see: http://www.textcheck.com/certificate/ORAPlH.
Funding statement
None.
Competing interests
The authors declare no conflict of interest.
