In de behandeling van sinus pilonidalis heeft primaire wondgenezing voordelen ten aanzien van secundaire wondgenezing, met name voor de duur van de wondgenezing en de tijd tot hervatting van dagelijkse activiteiten. Het sluiten van een wond in de middenlijn moet achterwege gelaten worden omdat dit in veel gevallen leidt tot infectie of dehiscentie van de wond en een hoog recidief percentage (Stauffer, 2018).

Er zijn diverse verschuivingsplastieken ontwikkeld om de wond primair te sluiten buiten de middenlijn met als doel de genezing van de wond na een operatie voor sinus pilonidalis te optimaliseren. Binnen de beschreven verschuivingsplastieken wordt er onderscheid gemaakt tussen technieken waarbij de incisie het midden van de bilnaad doorkruist (“crossing the midline”) en technieken waarbij de incisie paramediaan van de bilnaad blijft (“off midline”). Het is de vraag of primaire wondgenezingstechnieken waarbij de incisie door de middenlijn kruist ook meer risico op wondgenezingsstoornissen en recidieven hebben dan technieken die helemaal uit de middenlijn blijven.

Recurrence

Return to daily activities

Quality of life

Surgical complications

Wound infection

Wound dehiscence

Seroma

Complications that required re-operation

Postoperative pain

Description of studies

Systematic review

Ray (2020) performed a systematic review and meta-analysis of RCTs and included studies comparing Limberg flap with excision with closure, Karydakis procedure or Bascom procedure among patients with pilonidal sinus. The inclusion criteria were: randomized controlled trials, comparing Limberg flap versus excision of sinus pilonidalis ± closure/Karydakis procedure/Bascom procedure. No exclusion criteria were applied. The electronic databases of Embase, Medline through PubMed, the Cochrane Central Register of Controlled Trials (CENTRAL, provider Wiley Online Library) and Open Grey were searched from their inception until November 2019. Eighteen studies were included in the systematic review, of which we used ten studies in our analyses (studies with closure in the midline were excluded in our analyses). They included eighteen studies, of which ten

 studies compared Limberg flap with Karydakis or Bascom procedures and met our PICO. These ten studies comprised 1601 patients of which 865 were randomized to the (modified) Limberg flap group, and 736 to the Karydakis flap or Bascom procedure group. Eight studies were performed in Turkey, and two in Egypt. Five studies used the Limberg flap technique, four studies used the modified Limberg flap technique, and one study used both Limberg flap and modified Limberg flap techniques in the intervention group. Seven studies used the Karydakis flap procedure, one study the modified Karydakis flap procedure and two studies used the Bascom cleft lift procedure in the control group. The mean follow-up duration was 21 months and ranged from 1 month to a mean of 33 months in the individual studies. The following relevant outcome was reported: recurrence.

RCTs

Alvandipour (2019) performed a single blind RCT, comparing the Limberg flap technique with the Karydakis flap technique. In total, 65 patients were included, of which 28 were randomized to the Limberg flap group, and 37 to the Karydakis flap group. The inclusion criteria were patients with intergluteal pilonidal disease, whose age ranged between 15 and 65 years. This study was performed in Iran. The follow-up duration was 6 months. They reported the following relevant outcomes: recurrence (6 months), return to daily activities, surgical complications (wound infection and wound dehiscence), and postoperative pain.

Riaz (2019) performed an RCT, comparing the Limberg flap technique with the Karydakis flap technique. In total, 204 patients were included, of which 102 were randomized to the Limberg flap group and 102 to the Karydakis flap group. They included patients presenting with sacrococcygeal pilonidal sinus, with an age ranging between 15 to 60 years, and an ASA grade I to III. The study was performed in Pakistan. The follow-up duration was 6 months. The following relevant outcomes were reported: recurrence (6 months), surgical complications (wound infection, wound dehiscence, seroma) and postoperative pain.

Results

Recurrence (crucial)

Ten studies reported on the outcome recurrence. Eight studies reported the recurrence rate with a follow-up period of at least 12 months, of which six compared the Limberg flap with the Karydakis flap (Arslan, 2014; Ates, 2011; Bali, 2015; Bessa, 2013; Can, 2010; Sevinç, 2016), and two with Bascom cleft lift (Guner, 2013; Sarhan, 2016). In total, 1399 patients were analysed, with 763 in the Limberg flap group, and 636 in the Karydakis/Bascom procedure group. In the Limberg flap group, 32/763 (4.2%) had a recurrence, compared to 24/636 (3.8%) in the Karydakis/Bascom group. The pooled risk difference is 0.01 (95% CI: -0.01 to 0.03), favouring the Karydakis/Bascom procedure, which is not clinically relevant (Figure 1).

Two studies reported the recurrence rate with 6 months follow-up (Alvandipour, 2019; Riaz, 2019). Alvandipour (2019) reported 0/27 (0%) recurrences in the Limberg flap group and 1/37 (3%) in the Karydakis flap group (risk difference: -0.03, 95% CI: -0.11 to 0.05), favouring Limberg flap procedure. Riaz (2019) reported 9/102 (9%) recurrences in the Limberg flap group, and 5/102 (5%) in the Karydakis procedure group (risk difference: 0.04, 95% CI: -0.03 to 0.11), favouring the Karydakis procedure. These studies were not taken into account in the pooled analyses, because of the short follow-up period.

Figure 1 Outcome recurrence rate (>12 months follow-up) with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Return to daily activities (important)

Eight studies reported on the outcome return to daily activities, of which six reported the mean and standard deviation (Alvandipour, 2019; Arslan, 2014; Ates, 2011; Can, 2010; Sarhan, 2016; Tokac, 2015). In total, 1074 patients were analysed, with 593 patients in the Limberg flap group, and 481 in the Karydakis/Bascom procedure group. The pooled mean difference in days until return to daily activities was 0.24 days (95% CI: -1.05 to 1.52), favouring the Karydakis/Bascom procedure, which is not clinically relevant (figure 2).

Two studies reported the median time to return to daily activities and were therefore not taken into account in the pooled analyses, but only described here. Bali (2015) reported a median of 8 days (6 – 12, not specified whether this was a range or interquartile range) in the Limberg flap group and 17 (14 – 20) in the Karydakis procedure group, and Ersoy (2009) reported a median of 14 days (range: 5 – 46) in the Limberg flap and 15 days (range: 5 – 45) until return to daily activities in the Karydakis procedure group.

Figure 2 Outcome return to daily activities with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Quality of life (important)

One study reported the outcome quality of life, measured with the SF-36, on day ten and day 30 (Guner, 2013), see table 1. They reported: “the SF-36 survey showed statistically significant differences between the two groups on postoperative day ten in regard to bodily pain and role limitations due to physical problems. The results of these variables were higher in the Bascom cleft lift patients. A statistically significant difference was not observed between the other parameters.

Table 1 – Quality of life scores, measured with the SF-36

Data are expressed as the mean ± SD. Higher scores indicate better health status. Surgical complications (important)

Wound infection

Eleven studies reported on the outcome wound infection (Alvandipour, 2019; Arslan, 2014; Ates, 2011; Bali, 2015; Bessa, 2013; Can, 2010; Ersoy, 2009; Guner, 2013; Riaz, 2019; Sevinç, 2016; Tokac, 2015). In total, 1575 patients were analysed, with 845 in the Limberg flap group and 730 in the Karydakis/Bascom procedure group. In the Limberg flap group, 51/845 (6.0%) had a wound infection, and in the Karydakis/Bascom procedure group, 55/730 (7.5%) had a wound infection. The pooled risk ratio is 0.84 (95% CI: 0.57 to 1.24), favouring the Limberg flap, which is not clinically relevant (figure 3).

Figure 3 Outcome wound infection with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Wound dehiscence

Nine studies reported on the outcome wound dehiscence infection (Alvandipour, 2019; Arslan, 2014; Ates, 2011; Bali, 2015; Can, 2010; Guner, 2013; Riaz, 2019; Sarhan, 2016; Sevinç, 2016). In total, 1470 patients were analysed, with 792 in the Limberg flap group and 678 in the Karydakis/Bascom procedure group. In the Limberg flap group, 47/792 (5.9%) had wound dehiscence, and in the Karydakis/Bascom procedure group 40/678 (5.9%) had wound dehiscence. The pooled risk difference is 0.00 (95% CI: -0.03 to 0.03), favouring neither intervention, which is not clinically relevant (figure 4).

Figure 4 Outcome wound dehiscence with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Seroma

Six studies reported on the outcome seroma (Arslan, 2014; Ates, 2011; Guner, 2013; Riaz, 2019; Sarhan, 2016; Sevinç, 2016). In total, 1190 patients were analysed, with 651 in the Limberg flap group and 539 in the Karydakis/Bascom procedure group. All patient were treated with a suction drain postoperatively. In the Limberg flap group, 31/651 (5%) had seroma, and in the Karydakis/Bascom procedure group 38/539 (7%) had seroma. The pooled risk difference is -0.02 (95% CI: -0.07 to 0.02), favouring Limberg flap, which is not clinically relevant (figure 5).

Figure 5 Outcome seroma with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Complications that required re-operation

Two studies reported on the outcome complications that required a re-operation (Ates, 2011; Can, 2010). In total, 414 patients were analysed, with 211 in the Limberg flap and 203 in the Karydakis/Bascom procedure group. In the Limberg flap group, 6/211 (3%) had complications that required a re-operation, and in the Karydakis/Bascom procedure group 5/203 (2%) had complications that required a re-operation. Complications that required reoperation were wound dehiscence or wound infection that needed removal of the stitches and secondary wound healing. Because there were only two studies that reported this outcome, the results were not pooled, and the clinical relevance could not be determined.

Figure 6 Outcome complications that required re-operation with Limberg flap versus Karydakis/Bascom procedure

Z: p-value of pooled effect; df: degrees of freedom, I²: statistical heterogeneity, CI: confidence interval

Postoperative pain (important)

Six studies reported on the outcome postoperative pain (Alvandipour, 2019; Ates, 2011; Bali, 2015; Ersoy, 2009; Guner, 2013; Sarhan, 2016). Due to the different scales used for the Visual Analogous Scale (VAS), timepoints at which pain was assessed, and different ways of data presentation (mean and standard deviation, or median and range) it was not possible to pool the data. Therefore, the data is presented in table 2.

Table 2 – Pain scores measured with VAS

Level of evidence of the literature

Recurrence (crucial)

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome measure recurrence was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), and one level for imprecision (low number of events and confidence interval includes ‘no effect’). The level of evidence was therefore graded as low.

Return to daily activities (important)

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome return to daily activities was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), one level for inconsistency (conflicting results across studies), one level for imprecision (low number of patients and confidence interval includes ‘no effect’). The level of evidence was therefore graded as very low.

Quality of life (important)

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome quality of life was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), two levels for imprecision (very low number of patients). The level of evidence was therefore graded as very low.

Surgical complications (important)

Wound infection

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome wound infection was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), one level for inconsistency (inconsistent results) and one level for imprecision (low number of patients). The level of evidence was therefore graded as low.

Wound dehiscence

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome wound dehiscence was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), one level for inconsistency (inconsistent results) and one level for imprecision (low number of patients). The level of evidence was therefore graded as low.

Seroma

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome seroma was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), one level for inconsistency (conflicting results) and one level for imprecision (low number of patients). The level of evidence was therefore graded as low.

Complications that required re-operation

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome complications that required re-operation was downgraded by one level for risk of bias (no intention-to-treat analyses and blinding), one level for inconsistency (conflicting results) and two levels for imprecision (very low number of patients). The level of evidence was therefore graded as very low.

Postoperative pain (important)

The certainty of the evidence started high, as the evidence originated from RCTs. The level of evidence regarding the outcome postoperative pain was downgraded by one level for risk of bias (mostly no intention-to-treat analyses and blinding), two levels for imprecision (very low number of patients). The level of evidence was therefore graded as very low.

A systematic review of the literature was performed to answer the following question:

P: Patients with symptomatic pilonidal sinus

I: Excision with closure by flap repair (primary healing), with techniques that cross the midline  ((modified) Limberg flap, Dufourmentel, V-Y plasty, Z-plasty, Rhomboid flap, advancement flap, rotation flap, transposition flap)

C: Excision with closure by flap repair (primary healing), with off-midline closure (Karydakis, Bascom cleft lift)

O: Recurrence, return to daily activities, quality of life, pain, surgical complications (wound infection, wound dehiscence, seroma, complication that required re-operation)

Relevant outcome measures

The guideline development group considered recurrence as a crucial outcome measure for decision making; and surgical complications, postoperative pain, return to daily activities, and quality of life as an important outcome measure for decision making.

The working group defined the outcome measures as follows:

• Recurrence: a healed surgical site with de novo midline pits/sinus/secondary sinus opening after a symptom free period. Total recurrence rate of the study should be measured preferably with a study follow-up of at least 12 months.
• Return to daily activities: time from surgery to first day of return to daily activity (including work, sports, study or other activities).
• Quality of life: quality of life measured with a validated questionnaire.
• Surgical complications: wound infection, wound dehiscence, seroma or any complication that required re-operation.
• Postoperative pain: postoperative pain measured with a Visual Analog Scale (VAS) or the reported use of analgetics.

The guideline development group used the GRADE standard limits of 25% for dichotomous outcome measures as a minimal clinically (patient) important difference. Risk ratios were described where possible. In case studies that reported no events in one of the groups, risk differences were used. A risk difference of 10% was considered clinically relevant.

The following limits were used for minimal clinically (patient) important differences per outcome:

• Recurrence: GRADE standard limits of 25% for dichotomous outcome measures (RR <0.80 or RR >1.25)
• Return to daily activities: 1 working week (5 days)
• Quality of life: 0.5 sd
• Surgical complications: GRADE standard limits of 25% for dichotomous outcome measures (RR <0.80 or RR >1.25)
• Postoperative pain: 2 points on a VAS-scale.

Search and select (Methods)

The databases Medline (via OVID) and Embase (via Embase.com) were searched with relevant search terms until April 22^nd, 2021. The detailed search strategy is depicted under the tab Methods. The systematic literature search resulted in 52 hits. Studies were selected based on the following criteria: included patients with symptomatic pilonidal disease, compared midline flap repair techniques with off-midline flap repair techniques, reported at least one of the outcomes of interest, the study design is a systematic review (SR) of randomised controlled trials (RCTs) or RCT, and was written in English language.

Thirty-nine studies (14 SRs and 25 RCTs) were initially selected based on title and abstract screening. After reading the full text, 35 studies were excluded (see the table with reasons for exclusion under the tab Methods), and four studies (1 SR and 3 RCTs) were included.