Onlay mesh versus suture repair for smaller umbilical hernias in adults—early results from SUMMER trial: randomized clinical trial Authors: Mathias Bergström, Björn Widhe, Gabriel Granåsen, Anna Löf Granström, Johan Ohlsson, Simone Schult, Ursula Dahlstrand, Johanna Österberg, Peter Loogna, Sven Bringman, Maria Melkemichel Published: 2025 ABSTRACT AbstractBackgroundMesh repair is recommended for umbilical hernias larger than 1 cm to reduce recurrence rates, yet current evidence remains limited for smaller umbilical hernias. Important questions concern optimal mesh positioning and wound complications/surgical-site occurrences. The aim of this study was to report the preliminary results of a trial investigating surgical-site occurrences in suture versus mesh repair for umbilical hernias less than or equal to 2 cm.MethodsA randomized, controlled, parallel-group, double-blind, multicentre trial across six Swedish surgical units is comparing 4 × 4 cm macroporous lightweight onlay mesh repair with conventional suture repair for primary elective umbilical hernias less than or equal to 2 cm. Intraoperative centralized web-based randomization ensured allocation concealment. The primary outcome of the trial is recurrence at 3 years, whereas secondary outcomes (the focus of this study) include surgical-site occurrences and pain intensity at 30 days post-surgery.ResultsFrom February 2020 to January 2024, 290 participants were randomly assigned to either suture or mesh repair. After exclusion and loss to follow-up, the remaining population for analysis was 144 participants for suture repair and 135 participants for mesh repair. Surgical-site occurrences (Clavien–Dindo grade greater than or equal to I) affected 32 mesh repair participants (23.7%) compared with 26 suture repair participants (18.1%), without any significant increase in surgical-site occurrences for mesh repair (OR 1.39 (95% c.i. 0.78 to 2.51)). Clinically relevant surgical-site occurrences (Clavien–Dindo grade greater than or equal to II) were less common in the mesh group (2 participants; 1.5%) compared with the suture group (4 participants; 2.8%). The median duration of surgery was 32 min for suture repair and 45 min for mesh repair (P < 0.001). Assessment of pain intensity revealed that 82.0% of suture repair participants and 73.0% of mesh repair participants reported no pain (P = 0.061).ConclusionThis randomized clinical trial provides high-level evidence for mesh repair for umbilical hernias less than or equal to 2 cm. With regard to early postoperative outcomes, such as surgical-site occurrences, onlay mesh repair can be considered comparable to suture repair and is safe to use for smaller umbilical hernias.Registration numberNCT04231071 (http://www.clinicaltrials.gov). Introduction Umbilical hernia (UH) repair is a widely performed surgical procedure globally, yet an optimal technique for smaller UHs remains undefined. With UHs accounting for 10% of all abdominal wall hernias1 and approximately 4000 repairs performed annually in Sweden2, identifying the optimal repair technique is crucial for improving patient outcomes. Traditionally, mesh repair has been reserved for larger UHs, whereas smaller UHs have been repaired using a simple suture or Mayo technique3; however, suture repair has shown disappointing recurrence rates, with some studies reporting rates of up to 21%4,5. According to the European classification from 2009, a small UH is defined as having a defect size of less than 2 cm1. Previous research suggests that even smaller UHs may benefit from mesh repair, with lower recurrence rates compared with suturing. Studies, including observational cohort studies4–7, retrospective studies8,9, older randomized clinical trials (RCTs) with small sample sizes10,11, and heterogeneous meta-analyses12–17, have reported recurrence rates of 4–15% for suture repair compared with 0–5% for mesh repair. Notably, a large RCT for UHs sized 1–4 cm found significantly lower recurrence rates with mesh repair (4%) versus suturing (12%)18. However, a non-significant decrease was observed for UHs sized 1–2 cm, possibly due to an insufficient sample size for that subgroup18. That trial used a preperitoneal sublay mesh technique, which is technically challenging to perform in smaller UHs without enlarging the defect and may involve a higher risk of complications. Despite the first guidelines from 2020 recommending preperitoneal mesh for all UHs as small as 1 cm19, many surgeons are hesitant to adopt this technique for smaller UHs due to its challenges. These guidelines also highlight the need for further higher-quality data to refine treatment recommendations, particularly for small UHs defined today as less than 1 cm. Also, a meta-analysis has underscored the uncertain role of mesh in very small UHs under 1 cm15. Alternative approaches, such as ventral mesh patches, have been suggested, but are associated with an increased risk of wound complications/surgical-site occurrences (SSOs)20 and morbidity4,7,21. Although mesh repair reduces recurrence rates, optimal mesh positioning and the increased risk of SSOs are key concerns. The appropriate mesh placement for smaller UHs remains uncertain. One alternative to enhance mesh use in smaller UHs is to use small onlay mesh over the sutured defect. This approach may offer acceptable strength in preventing hernia recurrence to a sublay placement, while reducing the risk of complications. Additionally, it can be easier to perform, regardless of the hernia size or the surgeon’s level of expertise. No RCTs have yet compared suture repair with small onlay mesh repair for smaller UHs. The Suture UMbilical MEsh Repair (SUMMER) trial is designed to provide high-level evidence for mesh repair of primary elective UHs less than or equal to 2 cm. This randomized, controlled, parallel-group, double-blind, multicentre trial aims to compare recurrence rates and SSOs after suture versus onlay mesh repair. The hypothesis is that mesh repair could reduce recurrence rates without increasing SSOs compared with suture repair. This paper presents early results from the trial, namely SSOs and pain intensity at 30 days post-surgery. Methods Trial design and participants The SUMMER trial is a prospective, randomized, controlled, parallel-group, double-blind, multicentre study for primary elective UHs less than or equal to 2 cm. Conducted across six Swedish surgical units—four district hospitals (Södertälje Hospital, Enköping Hospital, Mora Lasarett, and Danderyds Hospital) and two elective care providers (Frölunda Hospital, Sophiahemmet Capio)—the trial randomized participants 1 : 1 to either suture repair or onlay mesh repair between February 2020 and January 2024. Eligible patients were recruited from surgical outpatient clinics based on specified inclusion and exclusion criteria (Table S1). All participating surgeons received standardized training, including detailed video instructions on the surgical technique to ensure uniformity. Participants had a 30-day follow-up visit after surgery and will continue to be monitored at 1 and 3 years post-surgery. The study followed CONSORT guidelines and statements22 and was approved by the Regional Ethics Review Board in Stockholm, Sweden (DNR 2018/22-65 and DNR 2019/05-608). The trial is registered with ClinicalTrials.gov under the identifier NCT04231071. The trial study protocol was published before data analysis23. Adults over 18 years old referred for a UH at participating surgical units were considered eligible to participate in the trial. Patients with a primary UH with a defect less than or equal to 2 cm, as assessed during the outpatient clinical examination, chosen for elective open repair, and who fulfilled the inclusion criteria without any exclusion criteria were given oral and written information about the trial. Written consent was obtained before surgery and randomization. A UH in this study was defined according to the European Hernia Society definition from 2009 as a midline abdominal wall defect from 3 cm above to 3 cm below the umbilicus1. Patients screened for inclusion who met exclusion criteria were registered for their exclusion reasons and are presented in Fig. 1. Trial participants included but not operated on and randomized were also registered and are presented in Fig. 1. CONSORT study flow chart BMI, body mass index. Surgical techniques All surgical procedures were conducted under general anaesthesia and without any preoperative antibiotics being given to the participants. No drains were inserted. In the suture repair group, an open incision was made in the umbilical area. The hernia sac was dissected and reduced. Resection of the hernia sac was permissible but discouraged. Measurement of the defect size was done before randomization intraoperatively, ensuring it to be less than or equal to 2 cm. When a trial participant fulfilled all the inclusion criteria, randomization to an allocation took place perioperatively. When allocated to suture repair, the defect in the aponeurosis was closed transversely with a running non-absorbable monofilament suture (2/0). The umbilical skin was adapted to the aponeurosis using an absorbable monofilament suture (3/0 or 4/0). The skin closure was completed with an intracutaneous running suture using the same suture material. In the mesh repair group, the procedure followed the steps outlined above for the suture repair. Additionally, after closing the defect, dissection of the subcutaneous tissue from the aponeurosis was performed to facilitate the placement of a 4 × 4 cm macroporous lightweight partially resorbable mesh (Ultrapro Advanced, © Ethicon US, LLC, 2020). The same mesh size was used in all trial participants, regardless of the size of the UH defect. The mesh was secured using five non-absorbable monofilament sutures (2/0)—one in the centre and one in each corner. To reduce the risk of nerve entrapment, the sutures were placed in a transverse direction. The subcutaneous tissue and the umbilical skin were adapted to the mesh using an absorbable monofilament suture (3/0 or 4/0). Skin closure was completed with an intracutaneous running suture using the same suture material. Any trial participants in whom an unintentional opening in the umbilical skin was created during the procedure were excluded from the study. Outcomes of interest The primary outcome of the trial is hernia recurrence at the site of the previous UH repair, evaluated at the 3-year follow-up. These data are not presented in this paper. Early recurrences were assessed at the 30-day follow-up through a physical abdominal wall examination by a surgeon. In cases of uncertainty concerning recurrence, CT was used for confirmation, as it better maintains allocation concealment compared with ultrasonography. The secondary outcomes include SSOs (defined as seromas (clear fluid accumulations), haematomas (blood accumulations), and wound infections (surgical-site infections necessitating any enactment)) and pain intensity at 30 days post-surgery. Findings regarding chronic pain, assessed via the Ventral Hernia Pain Questionnaire (VHPQ) at the 1-year follow-up visit, are not presented in this paper. SSOs were evaluated at the 30-day postoperative follow-up visit and were classified using the Clavien–Dindo grading system24, with a grade greater than or equal to I indicating the presence of a postoperative complication. Any bedside wound opening was classified as Clavien–Dindo grade IIIa in this trial, which, according to the Clavien–Dindo classification, includes procedures requiring surgical intervention not performed under general anaesthesia. Participants were instructed to return to their surgical unit for any SSOs occurring before their 30-day visit. These events were documented in the case report forms (CRFs) for the 30-day follow-up. Systematic complications, such as pulmonary, cardiovascular, or urinary tract issues, were also monitored. All SSO events were monitored using the same approach as recurrence assessments. Complications were managed following local medical standards. Pain intensity was measured using the Numerical Rating Scale (NRS) during the 30-day postoperative follow-up, where participants verbally reported their pain level on a scale from zero to ten during their outpatient clinic visit. Statistics The sample size was calculated with a focus on the primary outcome of recurrence at 3 years post-surgery and not for the secondary outcomes of SSOs. Recurrence rates in previous studies have varied widely, ranging from a few percent to as high as 20%. Based on previous reports, recurrence rates of 12% in the suture group and 3% in the mesh group were assumed. The study was powered to detect a risk difference of 9% in recurrence and an OR of approximately 4.4, with 80% power at a significance level of 0.050, with a 95% confidence interval. The sample size was adjusted to address potential patients lost to follow-up of 10%, resulting in a need of 144 trial participants in each arm. All CRFs were collected and managed using REDCap (Research Electronic Data Capture), an online data capture tool hosted by the Karolinska Institutet in Sweden. Baseline data were entered before surgery, with all other data registered prospectively. Intraoperative randomization was conducted through REDCap, ensuring complete allocation concealment. Each surgical unit implemented a 1 : 1 randomization in random sequence blocks of four or six, stratified by surgical unit and hernia size (less than 10 mm or 10–20 mm). Decisions to exclude participants from the trial post-randomization were made by surgeons during operations when exclusion criteria were met; reasons were documented in the CRFs and are presented in Fig. 1. Trial participants are to be kept unaware of the allocated surgical method of repair throughout the entire 3-year follow-up interval. No details regarding the allocation were recorded in the hospitals’ medical records for the patients. Instead, the allocation was documented separately on paper and securely concealed by the clinical secretary. This prevented trial participants from unblinding their allocation via online access to their hospital records. Also, the follow-up surgeons who assessed the outcomes were not part of the surgery and remained uninformed about the allocation. A statistical analysis plan was written before participant inclusion. The analysis included participants who adhered to the inclusion criteria, were free from exclusionary factors, and completed the 30-day follow-up post-surgery. Details of excluded eligible patients are shown in Fig. 1. Differences in baseline characteristics and post-surgical outcomes between the surgical methods were tested using Fisher’s exact test for categorical variables and Wilcoxon’s test for continuous variables. Regarding pain intensity, an NRS score of zero indicated no pain and an NRS score of greater than or equal to one indicated pain. Logistic regression analysis yielding an estimated OR, adjusted for strong predictors and stratified factors (BMI and hernia defect size), was performed to compare SSO risk and pain intensity risk. Sensitivity analysis addressed the potential impact of participants lost to follow-up by treating dropouts as either cases (having SSOs) or non-cases. All statistical analyses were performed in RStudio (version 2023.12.1+402). Results Between February 2020 and January 2024, 415 of 1040 eligible patients were included in the trial, with 290 participants randomized—146 participants underwent suture repair and 144 participants underwent onlay mesh repair (Fig. 1). Follow-up analyses at 30 days included 144 participants for suture repair and 135 participants for onlay mesh repair. Exclusions occurred due to participants being lost to follow-up, defects being larger than 2 cm, an incorrect initial diagnosis, or multiple defects (Fig. 1). Reasons for participants being lost to follow-up were being unable to reach three participants and one participant declining further participation for the mesh group and being unable to reach one participant for the suture group. Repairs were performed by 86 surgeons across the six surgical units, with a balanced distribution regarding the level of surgical expertise in both groups (Table 1). Table 1Characteristics of the trial participants with a completed follow-up visit at 30 days after surgeryCharacteristicTotal (n = 279)Suture repair (n = 144)Mesh repair (n = 135) P* Sex 0.789 Male202 (72.4)103 (71.5)99 (73.3) Female77 (27.6)41 (28.5)36 (26.7) Age (years), median (interquartile range) 51 (42–59)51 (42–58)51 (42–61)0.213 BMI (kg/m2), median (interquartile range) 27.8 (24.8–30.3)27.5 (24.6–30.4)28.4 (25.6–30.2)0.350 ASA fitness grade 0.624 I171 (61.3)92 (63.9)79 (58.5) II102 (36.6)49 (34.0)53 (39.3) III6 (2.2)3 (2.1)3 (2.2) IV0 (0)0 (0)0 (0) Risk factor Diabetes16 (5.7)7 (4.9)9 (6.7)0.610 Chronic obstructive pulmonary disease13 (4.7)7 (4.9)6 (4.4)1.000 Ischaemic heart disease0 (0)0 (0)0 (0)1.000 Smoking 0.305 Never smoker200 (71.7)109 (75.7)91 (67.4) Previous smoker56 (20.1)25 (17.4)31 (23.0) Current smoker23 (8.2)10 (6.9)13 (9.6) Preoperative imaging 0.546 None211 (75.6)108 (75.0)103 (76.3) CT40 (14.3)19 (13.2)21 (15.6) Ultrasonography28 (10.0)17 (11.8)11 (8.1) Size of hernia defect (mm) 1.000 <1034 (12.2)18 (12.5)16 (11.9) 10–20245 (87.8)126 (87.5)119 (88.1) Duration of surgery (min), median (interquartile range) 37 (29–47)32 (25–40)45 (36–55)<0.001 Level of surgical expertise 0.803 Consultant179 (64.2)91 (63.2)88 (65.2) Resident100 (35.8)53 (36.8)47 (34.8)Values are n (%) unless otherwise indicated. Continuous variables are presented as median (interquartile range). *Continuous variables were tested using Wilcoxon’s test and categorical variables were tested using Fisher’s exact test. Characteristics of the trial participants with a completed follow-up visit at 30 days after surgery Values are n (%) unless otherwise indicated. Continuous variables are presented as median (interquartile range). *Continuous variables were tested using Wilcoxon’s test and categorical variables were tested using Fisher’s exact test. Characteristics of the trial participants are presented in Table 1; both groups had a similar sex ratio, with approximately 70% men and 30% women (Table 1). The median age was 51 years and the median BMI was 27.5 kg/m2 for suture repair and 28.4 kg/m2 for mesh repair. Nearly 98% of participants had an ASA fitness grade of I–II. There were more smokers and diabetic participants in the mesh repair group. The hernia size was predominately 10–20 mm in both groups (greater than 87%). The median duration of surgery was 32 min for suture repair and 45 min for onlay mesh repair (P < 0.001). Postoperative surgical-site occurrences Overall SSO (Clavien–Dindo grade greater than or equal to I) rates were 18.1% (26 participants) for suture repair and 23.7% (32 participants) for mesh repair (Table 2). Clinically relevant SSO (Clavien–Dindo grade greater than or equal to II) rates were 2.8% (4 participants) for suture repair and 1.5% (2 participants) for mesh repair (Table 2). Of these clinically relevant SSOs, one mesh repair participant required an intervention under local anaesthesia (Clavien–Dindo grade IIIa) and two suture repair participants were treated with antibiotics for potential wound infections. Logistic regression analysis showed no significant increase in the risk of SSOs for mesh repair (OR 1.39 (95% c.i. 0.78 to 2.51)) (Table 3). This pattern persisted across the different categories of SSOs (Table 3). Table 2Characteristics of the postoperative complications of the trial participants with a completed follow-up visit at 30 days after surgeryCharacteristicTotal (n = 279)Suture repair (n = 144)Mesh repair (n = 135) P*Overall surgical-site occurrences†58 (20.8)26 (18.1)32 (23.7)0.302Seromas22 (7.9)9 (6.3)13 (9.6)0.375 Seromas (Clavien–Dindo Scale‡) No seroma257 (92.1)135 (93.8)122 (90.4) Grade I21 (7.5)8 (5.6)13 (9.6) Grade II1 (0.4)1 (0.7)0 (0) Grade IIIa0 (0)0 (0)0 (0) Grade IIIb0 (0)0 (0)0 (0) Grade IV0 (0)0 (0)0 (0) Grade V0 (0)0 (0)0 (0)Wound infections16 (5.7)7 (4.9)9 (6.7)0.610 Wound infections (Clavien–Dindo Scale‡) No wound infection263 (94.3)137 (95.1)126 (93.3) Grade I12 (4.3)5 (3.5)7 (5.2) Grade II3 (1.1)2 (1.4)1 (0.7) Grade IIIa1 (0.4)0 (0)1 (0.7) Grade IIIb0 (0)0 (0)0 (0) Grade IV0 (0)0 (0)0 (0) Grade V0 (0)0 (0)0 (0)Haematomas40 (14.3)18 (12.5)22 (16.3)0.396 Haematomas (Clavien–Dindo Scale‡) No haematoma239 (85.7)126 (87.5)113 (83.7) Grade I39 (14.0)17 (11.8)22 (16.3) Grade II1 (0.4)1 (0.7)0 (0) Grade IIIa0 (0)0 (0)0 (0) Grade IIIb0 (0)0 (0)0 (0) Grade IV0 (0)0 (0)0 (0) Grade V0 (0)0 (0)0 (0) Systematic complications 4 (1.4)2 (1.4)2 (1.5)1.000 Pulmonary0 (0)0 (0)0 (0) Cardiovascular0 (0)0 (0)0 (0) Urinary tract2 (0.7)1 (0.7)1 (0.7) Others2 (0.7)1 (0.7)1 (0.7)Hernia recurrences0 (0)0 (0)0 (0)Values are n (%). *Fisher’s exact test was used for statistical tests. †Overall SSOs include seromas, wound infections, and haematomas with Clavien–Dindo grade greater than or equal to I. The trial participants could have a combination of these complications. ‡Complications presented according to the international Clavien–Dindo Scale. Any bedside wound opening was classified as Clavien–Dindo grade IIIa, which, according to Clavien–Dindo classification, includes procedures requiring surgical intervention not performed under general anaesthesia. Characteristics of the postoperative complications of the trial participants with a completed follow-up visit at 30 days after surgery Values are n (%). *Fisher’s exact test was used for statistical tests. †Overall SSOs include seromas, wound infections, and haematomas with Clavien–Dindo grade greater than or equal to I. The trial participants could have a combination of these complications. ‡Complications presented according to the international Clavien–Dindo Scale. Any bedside wound opening was classified as Clavien–Dindo grade IIIa, which, according to Clavien–Dindo classification, includes procedures requiring surgical intervention not performed under general anaesthesia. Table 3Logistic regression analysis* of postoperative complications at 30 days after surgery Overall surgical-site occurrencesSeromasWound infectionsHaematomasOR† (95% c.i.) P OR† (95% c.i.) P OR† (95% c.i.) P OR† (95% c.i.) P Suture repair1 (reference)111Mesh repair1.39 (0.78,2.51)0.2651.58 (0.66,3.95)0.3131.35 (0.49,3.91)0.5621.33 (0.68,2.64)0.407*Adjusted for body mass index and size of hernia defect in the analysis. †Adjusted OR. Logistic regression analysis* of postoperative complications at 30 days after surgery *Adjusted for body mass index and size of hernia defect in the analysis. †Adjusted OR. Pain intensity—Numerical Rating Scale Measurement of pain intensity using the NRS showed that 82.0% of suture repair participants and 73.0% of mesh repair participants reported no pain, with NRS scores of zero, and that 18.0% of suture repair participants and 27.0% of mesh repair participants had NRS scores of greater than or equal to one (Fig. 2). Analysis showed an adjusted, increased, non-significant OR of 1.72 (c.i. 0.98 to 3.07; P = 0.061) for the risk of pain with an NRS score of greater than or equal to one for mesh repair. Additionally, 95.3% of the suture group and 96.5% of the mesh group had NRS scores less than or equal to three. No pain with an NRS score of greater than or equal to eight was reported. Postoperative pain intensity 30 days after surgery according to the Numerical Rating Scale Sensitivity analysis No pattern of imbalance in the baseline characteristics of those lost to follow-up was observed (Table S2). The sensitivity analysis revealed similar effects as the main analysis (Table 3), both when treating all lost to follow-up as cases and non-cases of SSOs (Table S3). Discussion This large-scale RCT provides evidence for mesh repair for UHs less than or equal to 2 cm. Although the sample size for this trial was not primarily calculated to investigate SSOs, the early results suggest that onlay mesh repair seems to be comparable to suture repair for UHs less than or equal to 2 cm with regard to SSOs. This highlights the safety of onlay mesh repair for smaller UHs during the early postoperative interval. However, a 3-year follow-up after surgery is necessary to investigate differences in the trial’s primary outcome of recurrences. The findings revealed a higher SSO rate in the mesh repair group (23.7%) compared with the suture repair group (18.1%) across all observed SSOs, when SSOs were defined as Clavien–Dindo grade greater than or equal to I. However, for clinically relevant SSOs requiring any enactment (Clavien–Dindo grade greater than or equal to II), the rate was comparably low between the groups. Specifically, SSO rates were lower in the mesh repair group (1.5%) compared with the suture repair group (2.8%). Previous randomized trials have reported higher SSO rates with mesh repair. Kaufmann et al.18 found SSO rates of 3% for suture repair versus 8% for mesh repair in 300 trial participants, defining SSOs as requiring interventions. Others used mesh plugs in 68 patients for repairs of hernias greater than 3 cm and standard flat preperitoneal mesh in 32 patients for hernias less than 3 cm in diameter, finding SSO rates of 11% for both suture and mesh repair10. Similarly, another study investigated two different mesh repairs against Mayo suture repair for UHs less than 4 cm and found SSO rates of 12% for the 17 patients repaired using the Prolene Hernia System (Ethicon) and 20% for the 15 patients repaired using standard flat onlay mesh repair11. The results of the present study point to lower SSO rates when applying the same SSO definition used in these studies, indicating that a small onlay mesh technique with minimal subcutaneous dissection does not increase the risk of complications requiring intervention. A seroma after onlay mesh repair may be the SSO that is most frequently reported. In the present study, 9.6% of mesh repair participants developed seromas, and none required an intervention involving aspiration. Fonseca et al.25 reported an SSO rate of 33% for onlay mesh repair, mainly due to seromas (23%) and surgical-site infections (10%), linked to larger microporous meshes. Others found a higher infection rate with small-pore meshes, 6.7% for mesh repair and 4.9% for suture repair26, yet no mesh removal was necessary, affirming the safety of the macroporous lightweight onlay mesh. In addition, McCombie et al.27 used a lightweight onlay mesh for UHs and reported an even lower SSO rate than the present study; the seroma rate was 7.8% and all non-infected seromas were treated with aspiration. Larger onlay mesh repairs, involving greater subcutaneous dissection, tend to have a higher seroma rate when compared with other mesh techniques12,13,16,17,25,28. However, using small 4 × 4 cm macroporous lightweight onlay mesh can minimize dissection and reduce the risk of seroma formation. This approach was supported by a retrospective study conducted by some of the authors of this trial, who used the same small 4 × 4 cm onlay mesh technique for UHs less than or equal to 2 cm and reported a seroma rate of 5%29. Thus, one explanation for the low rate (1.5%) of clinically relevant SSOs in the mesh repair group in the present study may be attributed to the minimal subcutaneous dissection for the small mesh insertion. The most prevalent complication in this present study was a haematoma, with rates of 16.3% for mesh repair and 12.5% for suture repair, with only one participant (in the suture group) having a Clavien–Dindo grade II complication and requiring some intervention. Though these rates are relatively high compared with previous studies10,11,18, the low number of cases that required intervention (only one) suggest benignity, with expectant management as the primary approach. Moreover, pain management after surgery remains a critical aspect of patient recovery. The present study showed that those participants who underwent suture repair reported less pain on the pain intensity scale using NRS 30 days post-surgery compared with those participants who underwent mesh repair, although the difference was not significant between the groups. Assessment of potential chronic pain occurrence between the groups will be reported 1-year post-surgery. This large, multicentre RCT aims to determine whether using small onlay mesh for repair of UHs less than or equal to 2 cm can reduce recurrence rates compared with suture repair. The early SSO results support the safety of using small onlay mesh for repair of smaller UHs. The involvement of 86 participating surgeons across six surgical units can be considered a strength. Efforts such as extensive training and standardized procedural documentation were employed to minimize variability in surgical technique. The multicentre design reduces biases related to individual surgeon experience and ensures the generalizability of the results to routine clinical settings. Consequently, the surgical method of repair is considered straightforward to endorse in clinical practice. Another strength was the focus on only smaller elective primary UHs with a stringent definition, unlike other studies that have included various types of ventral hernias with a broader size range4,5,9–12,17,18. Also, the use of online data registration ensured accurate data entry and a totally concealed allocation before randomization. Moreover, the absence of crossover allocation likely resulted from randomization occurring close to treatment allocation, minimizing post-randomization cases. Additional strengths were the high follow-up rates—94% for mesh repair and 99% for suture repair at 30-day follow-up after surgery. Participants lost to follow-up were primarily due to the inability to reach participants (4 of 5), considered to be due to participants lacking interest in returning after a successful minor hernia operation. Furthermore, the double-blind design ensured unbiased follow-up assessments by both participants and surgeons. Additionally, the frequent documentation of reasons for not including eligible patients helped detect potential selection bias, although this documentation was not always fully complete. The trial faced limitations, including challenges in objectively diagnosing SSOs across the six surgical units, as no standard requirements for aspiration or radiology were specified in the study protocol for confirming seromas or haematomas. Consequently, this lack of a systematic approach may have led to variability in SSO classification and management, although likely to be random across both groups and followed by a clinically relevant approach. Additionally, the trial’s definition of a UH was based on the European guidelines from 2009, which differ from the 2020 guidelines. Also, discrepancies may have arisen from defining surgical-site wound infections as complications requiring enactment in the study protocol against the number of wound infections classified as only Clavien–Dindo grade I, with 3.5% of suture repair participants and 5.2% of mesh repair participants having wound infections but not treated with antibiotics. Furthermore, the inclusion interval was extended due to the COVID-19 pandemic. In conclusion, the early results of secondary outcomes show acceptable SSO rates and support the safety of using small onlay mesh repair for smaller hernias. This trial is likely to impact future standards of practice regarding the repair of UHs. Small onlay mesh repair is recommended as an alternative, safe, and easy mesh technique for surgeons to choose to treat smaller UHs. Future follow-up of the trial participants will determine whether this method significantly reduces recurrence rates compared with suture repair. Supplementary Material Acknowledgements The authors are grateful for the enthusiastic commitment of all participating surgical units in the trial. The authors are also thankful to Joshua Juvrud and Filippa Widman for a language review of the paper. Contributor Information Mathias Bergström, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; Department of Surgery, Södertälje Hospital, Södertalje, Sweden. Björn Widhe, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; Department of Surgery, Södertälje Hospital, Södertalje, Sweden. Gabriel Granåsen, Department of Public Health and Clinical Medicine, Umeå University, Umeå, Sweden. Anna Löf Granström, Department of Women’s and Children’s Health, Karolinska Institutet, Stockholm, Sweden; Department of Clinical Sciences, Danderyds Hospital, Stockholm, Sweden; Department of Paediatric Surgery, Karolinska University Hospital, Stockholm, Sweden. Johan Ohlsson, Department of Clinical Sciences, Danderyds Hospital, Stockholm, Sweden; Department of Surgery, Danderyds Hospital, Stockholm, Sweden. Simone Schult, Department of Surgery, Högsbo Hospital HN, Västra Frölunda, Gothenburg, Sweden. Ursula Dahlstrand, Department of Clinical Science, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden; Department of Surgery, Enköping Hospital, Enköping, Sweden. Johanna Österberg, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; Department of Surgery, Mora Hospital, Mora, Sweden. Peter Loogna, Department of Surgery, Capio Surgery, Stockholm, Sweden. Sven Bringman, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; Department of Surgery, Södertälje Hospital, Södertalje, Sweden. Maria Melkemichel, Department of Clinical Science and Education, Södersjukhuset, Karolinska Institutet, Stockholm, Sweden; Department of Breast, Endocrine Tumours and Sarcoma, Karolinska University Hospital, Stockholm, Sweden. Funding The study was funded by research grants from Region Stockholm, Karolinska Institutet, Stockholm, Sweden and a Medical Resident Research funding grant from Region Stockholm, Karolinska Institutet, Stockholm, Sweden. The study was also funded by grants from Södertälje Hospital, Södertalje, Sweden. The funders had no involvement in approving the study, conducting the study, analysing the data, or writing the final manuscript. There was no commercial sponsorship. Author contributions Mathias Bergström (Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Supervision, Validation, Visualization, Writing—original draft, Writing—review & editing), Björn Widhe (Conceptualization, Investigation, Methodology, Project administration, Resources, Software, Supervision, Validation, Visualization, Writing—review & editing), Gabriel Granåsen (Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Resources, Software, Supervision, Validation, Writing—review & editing), Anna Löf Granström (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Johan Ohlsson (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Simone Schult (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Ursula Dahlstrand (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Johanna Österberg (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Peter Loogna (Data curation, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), Sven Bringman (Conceptualization, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—review & editing), and Maria Melkemichel (Conceptualization, Data curation, Funding acquisition, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing—original draft, Writing—review & editing) Disclosure Maria Melkemichel is a board member of the Swedish Hernia Register, but declares no conflict of interest related to the submitted work. Johanna Österberg is also a board member of the Swedish Hernia Register and has given lectures and workshops for Ethicon, Baxter, and Mölnlycke, but declares no conflict of interest related to the submitted work. Ursula Dahlstrand is the president of the Swedish Hernia Register, but declares no conflict of interest related to the submitted work. All other authors declare no conflict of interest. Supplementary material Supplementary material is available at BJS Open online. Data availability All data can be made available upon request. References Associated Data Supplementary Materials Data Availability Statement All data can be made available upon request. References