The study was conducted in the Reinier de Graaf Gasthuis (RdGG), a district teaching hospital with expertise in parastomal hernia repair surgery. All PSH repairs performed in our centre are registered in a prospectively collected database. Patients undergoing robotic PSH repair from March 2022 to January 2026 were included in this study. In all cases informed consent for use of information was obtained in the electronic patient file. All operations were performed using a Da Vinci Xi system (Intuïtive Surg, Sunnyvale, USA). The following data was collected: age at time of operation, gender, medical and surgical history, type of stoma, parastomal hernia classification according to the European Hernia Society [10, 12], American Society of Anesthesiologists (ASA) classification of physical status, Body Mass Index, surgical time, length of stay, post-operative 90-day complications, recurrence status and reoperation.

The decision to perform robot assisted PSH repair was multi-factorial and based on patient characteristics, surgical history and previous repairs, patient condition and patient preference and expectations. The option of reversal of the stoma was always considered first. Patients who were not suited for an extensive repair, e.g., due to frailty, were counselled for local procedures or a primary repair. Emergency cases were always performed by local repair also. Patients were counselled on the use of a non-synthetic mesh.

Whenever possible and necessary, patients were prehabilitated. Three important reasons for prehabilitation were pulmonary function, cardiac comorbidity and obesity. Patients were referred to a pulmonologist for pulmonary prehabilitation and underwent cardiac review to estimate operative risk and if possible, to adjust medication to improve cardiac function. A BMI of <33 was strived for and patients with higher BMI were referred for bariatric review or offered outpatient dietician of physiotherapy coaching to lose weight.

Patients were counselled that the primary choice of operation was a robotic Pauli repair with chance of conversion to robotic Sugarbaker or robotic sandwich repair if necessary. Conversion to an open repair was also mentioned for consent.

The robotic Pauli technique was described in detail previously [9, 11]. In summary, all procedures were performed under general anaesthesia after intrathecal morphine was administered (necessitating an indwelling urinary catheter). Antibiotic prophylaxis was administered half an hour before the operation. Patients were placed in supine position, slightly tilted to the side contralateral to the stoma. Insufflation to 12mmHg was achieved by Veress needle to the left upper abdomen (Palmer’s point). Three robotic 8 mm trocars were inserted into the contralateral flank (opposing stoma). Adhesiolysis was performed as necessary. The stoma conduit was mobilized as much as possible to achieve a “free hanging” conduit with limited bulk in the subcutis. The retromuscular plane was developed around the stoma and a limited posterior component separation (PCS/TAR) was performed to further develop the plane deep into the flank. Approximation of the anterior fascia (around the conduit) was performed with 0 v-loc [Medtronic, Fridley USA]. An Ovitex mesh [TelaBio, Malvern USA] was placed and fixed to the lateral abdominal wall with v-loc 3/0 sutures [Medtronic, Fridley USA]. In most cases an Ovitex 1s permanent mesh (with 12 g/m² polypropylene suture reinforcement) was used, which can be introduced down a 15 mm trocar, placed through the rectus muscle in the upper abdomen (either left or right). Mesh size was adjusted as necessary. Peritoneum/posterior layer was closed using v-loc 3/0 sutures [Medtronic, Fridley USA].

The surgical procedure has been altered several times in order to improve quality (or speed) of the operation. We describe a number of these technical adjustments here (some already previously described):

Additional suture fixation of the stoma conduit to the lateral abdominal wall, to mesh and peritoneum to reduce the chance of a sliding hernia recurrence

In a number of cases a robotic Pauli repair was not possible, either due to limited possibility to lateralize the conduit, or due to the absence of a (usable) posterior layer. In these cases, either a robotic Sugarbaker repair or a robotic sandwich repair was performed.

The robotic Sugarbaker is essentially the same procedure as the described robotic Pauli, but without closure of a posterior layer. An Ovitex 2S mesh was used in these cases to ensure a safe mesh surface for both stoma conduit and the other viscera.

The robotic sandwich repair was performed either intraperitoneally (in the absence of a posterior layer and the impossibility of adequate lateralization) or in retromuscular position. In both cases, a keyhole mesh was placed to reinforce the abdominal wall. A second mesh was placed in Sugarbaker configuration, but with limited lateralization.

Standard follow up consisted of an in-office appointment at 8–12 weeks post-operatively, a telephone appointment after 6 months and again a yearly physical appointment after that. Additional imaging was performed only if indicated and therefor all recurrences were clinically diagnosed, followed by imaging if recurrence was suspected.

If patients presented with complications in their referring hospital, we were contacted by their local surgeons. Complications described are 90-day postoperative complications.

Data were analysed using IBM SPSS Software, PASW Statistics version 29.0.2.0. Descriptive statistics were used to present all data. Median values were presented with interquartile range 9IQR). Only factors that were expected to be relevant risk factors for adverse outcomes were provided with odds ratio’s and 95% confidence intervals (CI), using binary logistics. Categorical variables were analysed using Pearson Chi², Fisher-Freeman-Halton Exact test. Continuous data were analysed for normalcy with Shapiro-Wilk and Q-Q plots to decide on (non)parametric tests: Mann Whitney U test was used for non-normal data. Median values were presented with interquartile range (IQR).

All analyses performed should be deemed explorative due to small number of event.

A total of 90 patients was operated for PSH between March 2022 and February 2026, of which 74 were treated robotically (baseline characteristics in Table 1). Sixteen patients underwent another type of repair (Figure 1) and are left out of the further description. In the robotic group, the Pauli technique was performed in 64 (86,4%) cases. Conversion to robotic Sugarbaker repair was necessary in 6 cases (8.1%), either due to the necessity of explantation of a synthetic (Sugarbaker) mesh, leading to obliteration of the posterior layer, or (in two cases) no utilisable posterior layer was present at all. Conversion to a robotic sandwich repair was performed in 4 cases (5.4%): 1 end ileostomy and 3 urinary conduit PSH repairs. In these 4 cases lateralization of the conduit was severely hampered by a short mesentery/ureter.

In 22 cases (30%) a concomitant (midline) incisional hernia was present (EHS classification II & IV). In these cases, the midline hernia was addressed by a retromuscular repair and a synthetic mesh (Progrip; Medtronic, Fridley, USA) was placed separate to the Ovitex mesh used for the PSH. In cases where the synthetic mesh would overlap the Ovitex mesh for adequate overlap, the synthetic mesh was placed onto the Ovitex mesh, to avoid any contact between the synthetic mesh and the stoma conduit.

Complications arose in 23 patients (31,1%), of which 7 (9.5%) were severe complications (Clavien Dindo ≥ 3) and are summarized in Table 2. Further descriptive analytics are given in Table 3.

Complications arose in 53.3% in end ileostomies compared to 22.2% in colostomies and 35.7% in urinary conduits. In patients with ≥2 previous repairs, complications occurred in 53.8% compared to 28.1% in patient who underwent no or 1 previous repair.

A univariate analysis of complications related to type of stoma, number of previous repairs and whether previous mesh was still in situ during our operation is shown in Table 4.

One patient died after a robotic Pauli repair. His body mass index was 41 kg/m² and he had a longstanding PSH with significant loss of domain necessitating an open reduction of the hernia contents through a peristomal incision. There was a bowel injury, which was not recognized until day 3 after surgery. He was reoperated and a short segment of small bowel was removed. Thereafter, the patient rapidly developed multiorgan failure and died within 24 h.

One patient developed a hematoma in the subcutaneous cavity after robotic Pauli, which subsequently infected. The extra-abdominal conduit became necrotic and perforated on day 5. A laparotomy was performed, where no important intraabdominal complications were seen. Notably, the intraabdominal conduit was unaffected, and the mesh was not visibly contaminated. Nevertheless, the mesh was explanted, and a new stoma was formed on the contralateral side of the abdomen. The patient developed an incisional hernia in the old stoma port site and a parastomal hernia at the new location at 4 months follow-up.

One patient underwent PSH repair for a previously relocated urinary conduit. The initial formation of the conduit was complicated by an infection, leading to resection of the conduit and left nephrectomy. A new conduit was formed in the left abdomen. The single (right) ureter now transversed the abdomen over the bowels from right to left severely hampering lateralisation of the conduit. After robotic sandwich repair, urine production declined over the first postoperative day and on day 2 a CT scan was performed, showing a urinoma proximal to the kidney. We assume that, in attempting to achieve some lateralization of the conduit, a traction lesion of the proximal ureter arose, which was unrecognized during the operation. The patient received a nephrostomy catheter, and a stent was placed in the ureter. Patient recovered without lasting complications. The stent and nephrostomy were removed 6 weeks later and no recurrence of PSH has arisen.

One patient, in which we removed the intraperitoneal mesh (after laparoscopic Sugarbaker repair) and performed a robotic Sugarbaker with reinforced tissue matrix developed an infected hematoma, which was drained percutaneously. Over time the infection did not completely subside, leaving the patient with a chronic abscess between mesh and abdominal wall. The reason for stoma formation in the first place had been a severe slow-transit colon. We performed a laparotomy with subtotal colectomy with ileo-sigmoidal anastomosis, with an uneventful course.

One patient, who previously underwent a Hartmann’s procedure for diverticular perforation, developed a perforation in the subcutaneous conduit (which still contained many diverticula) necessitating a relaparotomy, resection of all affected colon and relocation of the stoma.

One patient developed ischemia of the stoma conduit, with no other complications. A relaparoscopy was initiated but converted to a laparotomy. The conduit was resected and a new colostomy was placed through the (unaffected) mesh, which was left in situ in keyhole configuration. The patient developed a wound infection in de midline laparotomy which was treated conservatively.

Nine recurrences arose in our robotic PSH repair cohort (12%) at median 15-month follow-up (Table 5). Eight recurrences developed after robotic Pauli repair (12.5%), one after robotic Sugarbaker repair (20%) and none after a sandwich repair. Median time to recurrence was 4 months (range 2–12 months; IQR 2 months). All recurrences arose within the first year after surgery.

Three recurrences arose in the 40 patients who were not previously operated for PSH (7.5%), of which two were sliding of the conduit. In cases with only one previous attempt to repair (n = 13) no recurrence was seen. Conversely, looking at multiply recurrent cases (two or more recurrences) the number of further recurrences after our surgery increased to 6 out of 21 cases (28.6%).

Twenty-four patients had previously placed mesh (for PSH repair) still in situ (intraperitoneal or retromuscular) of which 6 (25%) developed a re-recurrence after our attempt for recurrent PSH repair. In the intraperitoneal mesh group 4 out of 13 developed a recurrent PSH (30.8%), while in the retromuscular mesh group 2 out of 11 patients developed a recurrence (18.1%).

A univariate analysis of recurrence related to type of stoma, number of previous repairs and whether previous mesh was still in situ during our operation is shown in Table 6.

Of the 9 recurrences, 4 developed a sliding (hernia) of the stoma conduit, of which three underwent a local repair to shorten the conduit. So far, these patients have had no new complaints. One patient was not significantly troubled by the recurrent bulge and has not undergone a second intervention.

Five patients developed a new true parastomal hernia necessitating further operations. Two of these patients underwent a local repair with reduction of the herniating abdominal content and refashioning of the mesh, which had been dislocated from the abdominal wall. Both these patients developed a further recurrence. One underwent an open operation with relocation of the stoma and placement of a Sugarbaker configuration mesh (reinforced tissue matrix). The other awaits further (open) surgery. One patient underwent an open bilateral PCS/TAR (posterior component separation/transversus abdominis release) with a large synthetic mesh and placement of a new reinforced tissue matrix to separate the stoma conduit from the synthetic mesh. This reintervention has been successful to date. One patient developed an early incarcerated recurrence (6 weeks after robotic Pauli repair) and first underwent a local repair to treat the incarceration, rapidly followed by a non-incarcerated recurrence, after which we performed an open relocation of the stoma. She has developed a new PSH at the new site but currently has no wish for repair. The last patient has undergone a new attempt at a robotic (sandwich) repair of his (now 8th) recurrence of PSH.

No mesh infections were seen in our cohort. In 3 cases the mesh was removed during a reoperation for a severe complication. In one case this removal was due to a chronic abscess, which may have been related to mesh. During this operation the mesh seemed not to be affected, but this is of course not evincible. No mesh erosions were seen.