Treatment options for patients with pilonidal sinus disease: PITSTOP, a mixed-methods evaluation

Pattern of disease and management options

Patients often present with acute infection and abscess formation. Abscesses usually require hospitalisation with incision and drainage of the abscess cavity. One in five patients present with recurrent symptoms following emergency surgery. 5 This picture of relapsing and remitting infections is typical of chronic PSD.

Treatment of chronic PSD is surgical, usually using two essential components: excision and closure. The exceptions are phenol injection and fistuloscope/diathermy as stand-alone treatments (which aim to induce fibrosis and obliterate the tracks) and seton insertion which may induce fibrosis, allowing the possibility of a simpler subsequent surgery. There is no clear consensus as to which approach for each component is superior. For those procedures that involve excision, the tissue removed may be minimal (e.g. curettage or excision of the ‘pit’) or there may be substantial excision of the affected area and surrounding tissue to ensure complete removal of disease. The resultant wound may be left open to heal slowly by secondary intention, or it is closed with glue6–8 or sutures. The skin closure technique may be midline or off-midline. In the off-midline technique, the wound is positioned adjacent to the natal cleft, rather than in the natal cleft itself, in order to theoretically aid healing. 8,9 Examples include: the Karydakis flap, Bascom cleft closure (Bascom II), rhomboid and Limberg flaps.

Monetary and humanistic burden

Pilonidal disease is relatively common and represents a significant burden to primary and secondary care in the NHS. The 2012 hospital episode statistics (HES) data reported 13,239 hospital admissions for PSD. 10 At present, both emergency and the most common elective excisional surgical treatments leave large open wounds that may take months to heal. 6,7 Patients consequently require prolonged wound care from community healthcare services. As the disease tends to affect young otherwise healthy adults, this prolonged need for dressings and general wound care impacts on education, work, intimacy and social life, pain, recurrent infection and fear of wound deterioration, all severely affecting QoL. 11,12 Alternative techniques including minimally invasive interventions that aim to close the wound away from the midline may reduce the burden to the patient, but their efficacy outside the care of dedicated enthusiasts is not clear.

Current evidence base

The optimum treatment that is both easy to perform and results in rapid healing and minimal complications is not clear. This is reflected in varied practice throughout the UK with a perceived random selection of the procedure techniques detailed above. Some of these procedures result in lengthy healing times and long periods of incapacity. The literature on PSD is large but mainly consists of single-centre cohort studies looking at individually favoured techniques. Many of these have reported very low recurrence and infection rates for almost all procedures. 13 It has proven difficult to replicate these results in ‘real life’. In addition to the literature being mainly from single-centre cohorts, most studies make no attempt to stratify patients or detail the extent of disease or the adjuvant management (antibiotics, anaesthetic, postoperative care). There have been numerous randomised controlled trials (RCTs), and nearly 40 systematic reviews that focus on management – including two Cochrane Reviews. Most of these systematic reviews include meta-analyses of cohort studies only or analyse comparative RCTs and non-RCTs (often combining these data) for numerous interventions with varied controls. The methodological flaws of many individual studies and systematic reviews, coupled with the uncertainty of front-running interventions and an absence of a universally accepted control, make the value and interpretation of the data difficult. 14

The first Cochrane Review demonstrated that healing through secondary intention had lower overall recurrence rates compared to primary closure but at the expense of longer healing times. 14 Another systematic review reached the same conclusion but also compared two types of closure, suggesting off-midline to be preferable to midline. 15 The authors also concluded that outcome measures, such as time to healing, were poorly analysed, and health economic data were lacking. They proposed that future trials should be adequately powered, multicentric and include valid methods of assessing surgical outcomes. A systematic review of wound care after excision found no best practice guidelines and only one clinical pathway. 16

The second Cochrane Review focused on fibrin glue (FG) in the treatment of PSD. 17 The authors concluded this was a promising and appealing option as monotherapy given the non-invasive nature and that it could be performed as a day-case procedure, under local anaesthesia. These conclusions echo the conclusions of a previous meta-analysis, both suggesting a need for further research. 18 Nevertheless, the research to date has largely considered FG as an adjunct to surgery and although small, single-centre observational studies6,7,19–21 have been published, there is no RCT of FG as monotherapy in treatment of PSD.

Survey design and development

A survey was developed to identify the most frequently used interventions for specific clinical scenarios in current PSD practice. As this was a novel survey, it was designed by study collaborators and followed the CHERRIES statement checklist of recommendations. 22 The survey included questions on the following: the mean number of primary elective procedures performed annually, factors affecting choice of procedure, treatment choice for recurrent disease presentation and the factors affecting treatment choice for recurrent disease treatment. The survey was piloted to determine the clinical sensibility.

Data collection

To maximise completion rates, the survey could be completed online or on paper. The online survey was hosted on the Research Electronic Data Capture^TM (REDCap) system managed centrally by the University of Sheffield Clinical Trials Research Unit (CTRU). REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external sources. 23,24 Paper surveys were returned by post or via e-mail. The questionnaires were anonymised at the respondent level.

Sampling

The survey was disseminated via the UK surgical trainee research collaboratives, led jointly by the South Yorkshire Surgical Research Group and the North-West Research Collaborative. Collaborators were asked to deliver the questionnaire to consultant colorectal surgeons in their units. The first point of contact was made through the National Research Collaborative e-mail lists, and electronic contact to local collaborative leads was cascaded locally. The collaborators were asked to circulate the survey locally to three consultants and thereafter return the completed questionnaires to the REDCap system.

Data analysis

All aspects of data management were provided by the CTRU in accordance with their standard operating procedures. The data emanating from this survey were captured and stored in the REDCap software.

Respondent practice overview

Respondents reported a median caseload of 15 patients per year [interquartile range (IQR) 10–20 patients] and indicated that recurrent disease accounted for 20% of overall workload (IQR 10–30%). Of those estimating their recurrence rates (n = 97), 19 (19.5%) were unaware of their recurrence rate, 14 (14.4%) estimated their rate to be <5%, 36 (37.1%) to be in the 6–15% range, and 28 (28.8%) in the 16–30% range.

With regards to hair management, depilation was recommended by 54 (49.5%), laser hair removal by 32 (29.4%), salt baths by 14 (12.8%), shaving by 52 (47.7%) and waxing by 32 (29.4%).

Operative strategies employed

A wide range of treatment strategies were employed by responding surgeons, summarised in Table 1. Excision of disease with wound left open was the most frequently used strategy (71 responses; 65.1%), followed by Karydakis flap (62 responses; 58.1%). Curettage with phenol injection (1 response; 0.9%) and endoscopic pilonidal sinus treatment (EPSiT) (2 responses; 1.8%) were the least frequently performed interventions.

Participants were asked to provide a first-, second, and third-choice preference for their interventions. Karydakis was the first-preference treatment for 24/96 respondents (25.0%), followed by Bascom’s II (n = 18; 18.7%), and curettage and glue (n = 15; 15.5%). For second-preference treatments, local excision with wound left open was the most popular with 21/85 participants (24.7%), followed by local excision with midline closure for 15 (17.6%) and Karydakis procedure for 14 (16.4%). The most popular third-preference treatment was local excision with wound left open (27/32; 84.4%), followed by local excision with midline closure for 12 (37.5%), and Bascom’s II for 7 (21.9%) respondents (see Appendix 3, Figure 12).

Case vignettes

Case vignettes demonstrated heterogeneity across respondents. Case one (recurrent disease) showed a preference for rhomboid flap or ‘other’ procedures (22.6% and 25.5%, respectively). For case two (female with primary disease and cosmesis concerns), preferences turned to favour conservative management (21.6%), followed by excision and primary closure (16.0%) and cleaning/curettage of tracts (14.1%). Case three assessed recurrent disease and requirement for minimal time off work. For this scenario, most respondents opted for conservative management with hair removal (25.4%), followed by curettage of tracts (16.0%). Of note, 15.1% would offer a Karydakis procedure in this setting. Responses are summarised in Table 2.

Training

Surgical training programmes were the key training setting for commonly offered procedures. These included training in wide local excision with wound left open or closed for 59/71 (83.1%) and 36/48 (75.0%) of those offering the procedures, respectively. Similar numbers were seen for Bascom’s I (21/27; 77.7%) and Karydakis procedure (49/62; 79.0%). For some procedures, no formal training was reported by 5–10% of respondents. Courses, observation of colleagues and reference material such as text or video was also variably used. A summary of training experiences is presented in Table 3.

Overview

The key finding of this survey is the heterogeneity and variation in practice of consultant colorectal surgeons who treat PSD. It demonstrates a relatively low annual volume of operative procedures (around 15) when compared to other conditions such as colorectal cancer surgery. This number is slightly higher than the median of four cases per year identified through HES.

One in four surgeons perceived they had treatment failure rates approaching 30% in this study. This is somewhat at odds with the published literature, which often claims cure in > 90% of cases. 13 This demonstrates dissonance between published reports and real-world experience of clinicians. Conversely, 1 in 10 respondents reported recurrence rates of < 5%, which is concordant with the literature. This gap in outcomes may arise from issues with the quality of the literature, where the often surprisingly high quality of outcomes has been challenged. 14 Alternatively, it may reflect a small group of clinicians with a high volume of practice and associated good outcomes. 25 This poses three key questions. First, should complex pilonidal disease, or even all pilonidal disease, be managed by a group of high-volume surgeons? This is an approach that is advocated in other aspects of surgery such as rectal cancer and inflammatory bowel disease. 26,27 The second question to ask is whether we should improve training opportunities (highlighted here as limited) for colorectal surgeons to improve their skill set. Finally, we should ask whether there is a need for better monitoring of outcomes in PSD, as with registries established for other conditions. The findings of this study largely match those from a previous survey conducted in 2011,28 suggesting little has changed in a decade, making these questions more important to improve care.

Surgeons expressed preferences for some treatments such as excise and leave open. The literature suggests these should be considered outdated as they are associated with significant wound morbidity. 15 Sixty-five per cent of surgeons used the leave open technique with healing occurring by secondary intention, and 44% used a midline closure technique. Surgeons expressed a stronger preference for asymmetric closure than when previously assessed,28 in keeping with global trends. 29 In contrast, minimally invasive techniques such as EPSiT and pit picking appear to be relatively unpopular treatments. This suggests that surgeons are focused on cure rather than symptomatic relief. This may contrast with stated patient preferences where they are willing to trade a less major procedure in exchange for a higher risk of recurrence. 30

The survey does have limitations. Surveys of this nature can present artificial choices and do not permit qualification of answers. The use of vignettes does, however, allow some direct comparison of preferences. The survey may have drawn in experts and enthusiastic practitioners in pilonidal disease. However, the heterogeneity of responses does not reflect unity of thought, and responses are in keeping with published surveys. The response rate of 54.5% should reassure us as to representativeness of the survey.

Implications for policy-makers

This study presents three key actions for policy-makers. First, there is a need to agree a general framework for interventions to standardise pathways of care. This should be supported by best available evidence. Where such evidence is not available, funding should be secured to inform such guidance. Secondly, this is an area with no clear registry or oversight. Policy-makers should consider whether collection of granular data on practice and outcomes might aid initiatives to improve care, or even justify specialisation or centralisation of practice. Finally, surgeons’ practice is driven by their postgraduate training and persists into their independence. Therefore, it is important to offer opportunities for further training in new techniques. This would allow implementation of new techniques and may support the decommissioning of outdated procedures.

Implications for researchers

The level of heterogeneity likely speaks in part to uncertainty. Researchers should consider whether the findings presented here would support the delivery of specific procedures in a head-to-head RCT, or whether a ‘bucket’ approach would be more pragmatic.

Aims

A prospective, multicentre observational cohort study was conducted to:

• describe the disease characteristics of participants undergoing treatment for PSD

• describe procedures currently in use and quantify clinical outcomes and PROs associated with each

• identify patient-specific disease features that might predict poor outcome in each treatment group

• derive a case mix-adjusted estimate of the risks associated with common treatment options.

Participants

Eligible patients were undergoing definitive elective treatment for PSD at study recruiting centres.

Study procedures

Statistical methods

Patient and public involvement

Two patient and public involvement (PPI) representatives joined the study. The PPI representatives reviewed all patient-facing documents to ensure readability, understanding and format. One PPI representative sat on the steering committee panel and provided an instrumental patient voice in the management of the study. This PPI representative suggested the inclusion of a supplementary participant information sheet to be available for patients who were waiting for their clinic appointment. The patient representatives were also consulted when writing the plain English summaries and dissemination materials.

Recruitment and participant flow

Thirty-one UK sites recruited participants over a 46-month period from May 2019 to March 2022 (see Appendix 2). Figure 1 shows the flow of participants through the cohort: in total, 729 participants consented to be part of the cohort study. Participants were excluded from analyses if they did not have a procedure during the study (n = 45), if they were ineligible due to an incorrect diagnosis (n = 7), or if there was not enough information provided in order to categorise their procedure (n = 10). A total of 667 participants were included in the analysis cohort, of whom 476 (71%) provided follow-up data at 6 months. The number of participants who consented was lower than our anticipated sample size of 800, with 100 patients in each of the front-running management strategies.

FIGURE 1.

Participant flow for the PITSTOP study. PIS, patient information sheet.

In response to COVID-19, participants were able to be followed up at the end of the study; this was completed for 574 participants and used to update complications and recurrence data.

Baseline characteristics

The characteristics of the cohort participants are included in Tables 5 and 6. There were more males (73%), and 85% of participants were white; the average age of participants was 29 years. Just over half (54%) of participants reported no previous procedures, and 22% reported a previous elective procedure for PSD.

Data completion

Data completion rates for the cohort outcomes are presented in Appendix 3, Table 25. Return to normal activities, recurrence and wound healing were considered complete if a patient contributed those data at either clinic visit, 6-month follow-up or study close (or in the case of recurrence, it was apparent from AE reporting). Complication data were considered complete if the participant contributed data to at least one follow-up time point. Data for recurrence (94%), complication (96%) and return to normal activities (94%) were collected for most participants. The characteristics of participants that attended 6-month follow-up were compared to those that did not attend 6-month follow-up (see Appendix 3, Table 37), and the distribution of characteristics was similar between the groups; there were marginally more participants that had lateral openings in the attenders (50%) compared to the non-attenders (44%). All analyses were conducted on available data.

Treatment decisions

The breakdown of treatments received is presented in Table 7. Recorded procedure details were categorised into four categories, which were further combined into major or minor procedure categories. Over half (60%) of the participants received a major treatment, most commonly asymmetric closure (41%). Of the participants that received minimal excision, the most common treatment options were glue (n = 106, 16% of the cohort) and pit picking (n = 60, 9% of the cohort).

Further treatment details are presented in Appendix 3, Table 26. Median length of surgery was 30 minutes, and 95% were performed as day cases. Procedures were typically performed by consultant surgeons (68%).

Figure 2 shows the treatment received by disease characteristics. Participants with recurrent disease (defined as reporting any previous procedure, including acute drainage) were more likely to be given asymmetric closure; participants that were not recurrent were more likely to receive minimal excision. Over half (56%) of Wysocki type 1 (only midline pit or sinuses) participants underwent minimal excision, whereas over half (53%) of Wysocki type 4 (disease after treatment with definitive intent) were given asymmetric closure. The extent of overlap is noteworthy; for all disease characteristic categories there were a number of participants that received each treatment type, suggesting there is variety in the types of procedures considered appropriate for patients with different disease characteristics. The distribution of the number of pits was similar across procedure types (see Appendix 3, Figure 13). The proportion of patients with a minor procedure varied substantially across the sites (see Appendix 3, Figure 19), although this may be due to differing case mix across centres.

FIGURE 2.

Treatment choice by disease characteristic. Note: ‘recurrent’ defined as any reported previous procedure; type 1: only midline pit or sinuses; type 2: any midline disease with secondary sinus/es or abscess scar/s; type 3: any midline or secondary disease extending below tip of coccyx; type 4, any disease after treatment with definitive intent.

Shared decision-making and decision regret

Participant ratings of their pre-op consultation were high (Table 8), with the median (IQR) of the CollaboRATE mean score response being 3 (3–4), where 3 represents ‘a lot of effort was made’. The CollaboRATE top score was given in 36% of cases, reflecting that ‘every effort’ was made to help the patient understand their health issue, listen to the things that matter most and include what matters most to the patient in choosing what to do next. The decision regret (DR) scale, completed at month 6 follow-up, was low (median 8, IQR 0–20), and was broadly similar across the treatment categories.

The relationship between CollaboRATE mean score and DR is shown in Appendix 3, Figure 14. There is little clear evidence of a correlation between shared decision-making (SDM) and DR, with the majority of patients being in the top left corner of the graph (representing participants that were happy with their collaboration and had few regrets about their procedure).

Decision regret was low among patients [mean (SD) 14.5 (16.7)] and was broadly similar across the procedure types. The majority of patients reported being either satisfied or very satisfied (83%) with their procedure. Seven (21%) of the participants that received a leave-open procedure reported being either dissatisfied or very dissatisfied. The majority of patients returned to normal activity by the end of follow-up (n = 550, 88%) and 75% reported the wound as having healed during the study follow-up. Almost half (47%) of participants experienced a complication during follow-up.

Outcome summaries

The continuous outcome measures, recorded over time, are presented in Appendix 3, Table 27. Self-reported pain related to pilonidal sinus was at its highest on average on day 1 after procedure compared to baseline and reduced to its lowest at the 6-month visit. The highest average pain was reported by participants that received asymmetric closure. Patient-reported EQ-5D-5L health utility reduced from baseline to day 7 (overall means 0.80 and 0.69, respectively) but had recovered at both clinic and 6-month visits (overall means 0.83 and 0.89, respectively). Participants that received minimal excision reported the least change at day 7 and the highest health utility and QoL satisfaction at clinic visit and 6 months, although the scores were more similar among treatment groups at 6 months.

Other outcome measures are presented in Appendix 3, Tables 28–30. Repacking procedures were reported by 68 participants (12%) by day 7, while 87 participants (17%) reported repacking at clinic visit. At day 7, 226 (39%) patients reported a re-dressing procedure; re-dressing by day 7 was most common in asymmetric closure (49%) and midline closure (52%). Nearly half of participants experienced a complication during follow-up (n = 301, 45%), the most common of which were infection (26%) and discharge (18%). The numbers of complications were broadly similar across the three major surgery groups, and were lower for patients who received minimal excision, particularly for bleeding, dehiscence and infection.

Risk-adjusted treatment comparisons

The primary comparison between treatments was made between major procedures (asymmetric closure, leave open, midline closure, n = 396) and minor procedures (minimal excision, n = 270). No factors were found to be collinear and so all were included in the risk adjustment. Non-linearity of continuous features (BMI, natal cleft depth, pit spread and pit distance) was investigated and all features were deemed to be sufficiently modelled using linear terms. For the propensity score modelling, sufficient overlap in risk score was observed for all outcomes, and thus risk-adjusted analysis was deemed appropriate for major versus minor procedures.

Pain

The propensity score model identified sex, presence of pus and Wysocki classification as the most important features in treatment choice. Patients were more likely to undergo major procedure if they were female, had pus, or were classified as Wysocki type 4 (disease after treatment with definitive intent) and least likely to have major procedure if type 1 (only midline pit or sinuses). These factors were used in the propensity-adjusted models for all outcome comparisons.

Pain on day 1, adjusted for factors predictive of treatment choice and outcome via augmented IPW, was higher for patients who received major procedures compared to minor procedures by 1.58 points (95% CI 1.14 to 2.01) (see Appendix 3, Table 31). This was very similar to the unadjusted difference (mean difference 1.62, 95% CI 1.23 to 2.02). The number of participants included in each analysis varied according to the factors included to adjust the models. A similar difference in pain at day 7 was observed (augmented IPW-adjusted mean difference 1.53, 95% CI 1.12 to 1.95). The difference in pain between treatment groups was similar regardless of the risk adjustment method. A post hoc sensitivity analysis that included baseline pain as a covariate yielded mean differences and 95% CIs that were within 0.1 points of these estimates.

Pain over time for major and minor procedures is shown in Figure 3 (pain at clinic visit and month 6 were not prespecified as outcomes with formal comparisons); the raw difference in means was closer at clinic visit than at day 7, and there was no difference in mean pain reported at 6 months between the procedure types.

FIGURE 3.

Pain with major and minor procedures, baseline to 6 months after surgery.

Return to normal activities

Time to return to normal activity was compared between major and minor procedures (Table 11 and Figure 4). While nearly all participants had returned to normal activity by the end of follow-up, the time taken was far quicker among those undergoing minimal excision (median 7 days) than those who had major procedures (median 32 days). At 6 months, 12% of participants who had major procedures and 4% of participants who had minor procedures were yet to return to normal activity (see Figure 4). Several participants dropped out, providing either a censored time or no data at all; in the best-case scenario where these were assumed to have recovered, the proportion of participants who returned to normal activity at 6 months would be 96% for major procedures and 98% for minor procedures. Participants who received major procedures took on average 21 days longer to return to normal activity than those receiving minor procedures, and the difference was greater using risk adjustment models. The mean difference as estimated by augmented IPW was 25.9 days (95% CI 18.4 to 33.4 days), with regression adjustment approaches providing estimates closer to the unadjusted difference. Similar findings were present when comparing asymmetric closure procedures to the minimally invasive approaches (see Appendix 3, Figure 19 and Table 40).

FIGURE 4.

Time to return to normal activities by major or minor procedure type.

TABLE 11 - Comparison of time to return to normal activity between major and minor procedures

Reference group: minor procedures risk-adjusted difference estimated using parametric accelerated survival time with adjustment for covariates as listed, full model includes adjustment for sex, BMI, natal cleft depth, gluteal hair, smoking status, Wysocki classification, pit density, lateral distance, lateral distribution and presence of pus. Propensity-adjusted for sex, Wysocki classification and presence of pus. IPW + regression adjusted for sex, Wysocki classification, natal cleft depth, lateral distribution and presence of pus.

Time to return to normal activity	n	Difference (days) (95% CI)a
Raw difference	607	21.0 (16.3 to 25.7)
Risk-adjusted – Wysocki	600	20.3 (15.6 to 24.9)
Risk-adjusted – chosen model (natal cleft depth, Wysocki, lateral distribution)	502	19.8 (14.7 to 24.9)
Risk-adjusted – full model	403	19.8 (14.0 to 25.6)
Propensity-adjusted – inverse weighting	589	35.6 (19.8 to 51.4)
IPW + regression	502	25.9 (18.4 to 33.4)

Wound healing

Participants having major procedures also took longer to heal than those who had a minor procedure. The median time to healing was 30 days among people undergoing minimal procedures, compared to 70 days among those undergoing a major procedure. However, as highlighted in Figure 5, around 25% of participants in both groups had wounds that had not healed. Some of the individuals lost to follow-up (LTFU) may have healed, but a best-case scenario where those censored prior to 6 months were assumed to have healed would still mean 10% of participants considered their wound unhealed at 6 months. Unadjusted and regression-based risk adjustments both estimated the difference in wound healing to be over a month greater following a major procedure, while propensity score methods estimated larger differences but with wider CIs (augmented IPW estimate 53.5 days, 95% CI 28.8 to 78.2 days; Table 12).

FIGURE 5.

Time to wound healing by major or minor procedure type.

TABLE 12 - Comparison of time to healing between major and minor procedures

Reference group: minor procedures risk-adjusted difference estimated using parametric accelerated survival time with adjustment for covariates as listed, full model includes adjustment for sex, BMI, natal cleft depth, gluteal hair, smoking status, Wysocki classification, pit density, lateral distance, lateral distribution and presence of pus. Propensity-adjusted for sex, Wysocki classification and presence of pus. IPW + regression adjusted for sex, Wysocki classification and presence of pus.

	Major procedure		Minor procedure
Model	n	Median (IQR) days	n	Median (IQR) days	N	Mean difference (95% CI)a
Raw difference	336	70 (31–52)	217	30 (14–54)	553	39.7 (27.0 to 52.4)
Risk-adjusted – Wysocki					546	36.7 (23.8 to 49.6)
Risk-adjusted – chosen model (Wysocki, BMI, smoking status, pus)					452	34.8 (19.9 to 49.6)
Risk-adjusted – full model					368	38.2 (22.3 to 54.1)
Propensity-adjusted –IPW					536	111.3 (−10.9 to 233.4)
Augmented IPW					452	53.5 (28.8 to 78.2)

Pairwise comparisons of asymmetric closure and minimal excision (removing participants that had the major procedures – leave open and midline closure) produced results in keeping with the comparison between major and minor procedures (see Appendix 3, Tables 37–43).

Surgeon variation

A post hoc analysis looked at recurrence rates among surgeons who operated on at least 10 participants. In total 13 surgeons undertook at least 10 procedures (range 10–55 procedures, median 14). In total 282/667 participants underwent procedure by one of these surgeons. While recurrence and treatment failure rates varied among the 13 surgeons, outcomes were more favourable among participants whose surgeons treated 10 or more cases (see Appendix 3, Table 32). Overall recurrence at 6 months was lower among participants treated by these surgeons (40%) compared with surgeons that treated fewer cases (60%), with similar differences seen for recurrence at any time (42 vs. 48%) and treatment failure (40 vs. 60%).

Adverse events

Adverse events and serious adverse events (SAEs) are presented in Table 13; 107 (16%) patients experienced at least one AE during follow-up, and the most common category of AE was wound infection (59%). Eleven participants experienced an SAE, nine of which were inpatient hospitalisation.

Design

A multiple-case design was employed to compare more than one data type between and within more than one person. 59 The case study was nested in the observational cohort, with two embedded units of analysis at baseline and 6 months: qualitative longitudinal semistructured interviews and quantitative cohort data.

Participants

Sampling was purposive and sought to recruit PITSTOP cohort participants with symptomatic PSD referred for elective surgical treatment. This sampling method aimed for maximum variation based on the following: Wysocki classification (an indicator of disease severity), surgical procedure (excision and closure techniques) and NHS Foundation Trust. Initial contact was made by telephone. All potential participants were e-mailed the participant information sheet and provided verbal consent prior to participation in the interview.

Data collection

At baseline, participants completed a patient-reported experience measure (PREM) of SDM. SDM is scored 0 to 9 (0 indicating poor SDM; 9 indicating good SDM). 60 At 6 months, participants completed a PREM of DR. Using a 5-point scale, healthcare DR is scored 0 (low DR) to 100 (high DR). 61 In addition, the following outcome measures were collected: pain, length of time to healing and post-surgery complications. Semistructured telephone interviews were conducted between June 2019 and September 2020. A minimum of 20 interviews was considered adequate to understand common perceptions and experiences of treatment choices, thereby achieving data saturation. 62,63 A topic guide was designed using the coping in deliberation (CODE) and Sekhon’s Acceptability framework. 64,65 Baseline interviews adapted key ‘choice’ (e.g. ‘did you let the surgeon choose your treatment?’) and ‘options’ (e.g. ‘did the surgeon talk you through the risks and benefits?’) questions from the CODE framework. 64 At 6 months, the interview guide asked CODE questions related to decision ‘consolidation’ (e.g. ‘was this the right decision?’). Throughout, probing questions covered dimensions of Sekhon’s acceptability framework,65 as well as intervention attributes, to inform the discrete choice experiment (DCE). The interviews were recorded using an encrypted digital recorder and transcribed verbatim.

Patient and public involvement

One patient representative participated in a pilot interview. The aim of the pilot interview was to assess the apprehension of the topic guide questions and review interview cohesion. No amendments were made to the topic guide.

Data analysis

Transcripts were analysed using the National Centre for Social Research ‘Framework’ analysis approach. This approach was chosen because it allows for coding of a priori and de novo themes. 66 After familiarising ourselves with the data set, we independently coded a sample of transcripts using the CODE and Acceptability frameworks64,65 on NVivo (QSR International, Warrington, UK) version 11 before conferring. During the analysis and interpretation, integration of qualitative and quantitative data occurred to understand: (1) how disease characteristics and surgeon preferences interacted with patient values in treatment choices; and (2) how participants appraised treatments given particular outcomes. We used joint display tables to look for convergences and divergences between cohort data (disease features/treatment choices/outcomes) with experiences, views and values. 67,68 Patient experts were invited to provide feedback on a lay summary of triangulated results to assess acceptability.

Sample

An initial expression of interest was made by 266 cohort participants. The final sample comprised 20 patients (median age 28; range 20–64) from 13 NHS Foundation Trusts in the UK (Table 14). Only 13 participants completed baseline (median 16 minutes, range 6–47 minutes) and follow-up interviews (median 18 minutes, range 11–37 minutes).

Coping in deliberation framework: health threat

Newly diagnosed participants were unfamiliar with, and expressed confusion about, the cause and prognosis of PSD. They detailed their experience of soreness, inflammation, discharge and odour which disrupted employment, physical activity and relationships. Participants discussed the impact this had on psychosocial well-being.

I was told initially, 'Oh that could be it, and then it might go away' … but once you get it once, that’s it: it’s coming back … If I was a bit more aware of that I would have probably started to look into the surgeries quicker.

18: no previous pilonidal disease

Some participants were reluctant to address their condition, deciding to tolerate distress and delay treatment. Often, an exacerbation of symptoms would cause participants to present to emergency services for treatment.

I said to [my girlfriend], 'Look, I can’t really see it properly. Is it still getting bigger?' And, she said, 'Oh bloody hell … get in the car.' So, we went straight to [hospital].

03: one previous episode of PSD

Other participants discussed barriers to secondary care treatment referrals due to their GP not taking their condition seriously.

… just gave me some antibiotics … it just kept getting more painful and worse … I went back three times … then she put me on sort of the path to go back to surgery, but she didn’t send me [as] an urgent patient … so I had to wait for maybe like five months.

14: no previous PS

Another saw their GP numerous times over 25 years and was repeatedly dissuaded from surgery.

he basically sort of said to me that it’s a very precarious operation … that the success rate wasn’t very high … that it was something that if I could live with.

15: no previous pilonidal disease

Coping in deliberation framework: choice

Once referred to secondary care, 9/20 participants were offered a choice of treatment. Absence of choice was rarely viewed negatively. Due to their own limited knowledge of the condition, some participants viewed healthcare professionals as best placed to make decisions about their care, especially where emergency surgery was concerned. If offered a choice, participant preference was based on one or more of the following factors: previous experience of surgery (n = 3); surgeon’s guidance (n = 3); invasiveness of the treatment (n = 3); or anticipated recovery time (n = 2). One participant rejected their consultant’s treatment advice to undergo a ‘leave open’ procedure due to the employment opportunity costs – they would require additional time off work (Table 15). Other participants utilised social networks (significant others, friends and relatives) or the internet to support decision-making, with some acquiring a sense of control from independently researching the condition and treatment options.

Coping in deliberation framework: key outcomes at the time of decision-making

Not all participants specified a single important outcome when undergoing surgery. The following were discussed: avoiding recurrence (n = 8), return to normal activities (n = 6) and/or the elimination of symptoms (n = 7). Six participants were not aware of procedural risks; others expressed awareness of risks presented by anaesthesia (n = 2), infection or bleeding (n = 4), the wound not healing (n = 5) and recurrence (n = 8).

Coping in deliberation framework: consolidation

After surgery, most participants were anxious of aggravating the wound and/or delaying wound healing. Many implemented adaptations, including physical (altering seating positions) and behavioural (reducing exercise duration and changing the type of exercise) changes, which negatively impacted their well-being.

It has made me reticent to engage in some activities … exercise and things like that … through the pain and discomfort, and also the chance of sort of popping the cyst...

01: no previous pilonidal disease

Many participants required daily or weekly primary care wound management support via GP appointments or district nurse visits. Due to its location, most experienced difficulties managing the wound independently. They used mirrors for physical inspection or were dependent on others to help examine and manage the wound (including cleaning, dressing and packing). Some participants cited the reliance on others as a loss of independence, while others acknowledged how emotional support alleviated distress.

I think the worst part of it is that you always have to rely on someone else to do, like, a dressing for you … you can’t drive cos you can’t sit down … you basically you can’t do anything.

06: sinus excised and left open

Acceptability framework: key outcomes at 6-month follow-up

Six months post surgery, some participants reported hoping surgery would address: pain (n = 3); recurrence (n = 5); wound healing (n = 1); the smell (n = 1); the inconvenience (n = 1); impaired ability to perform ADL (n = 1). In five cases (01, 05, 11, 14, 17; see Appendix 3, Table 33), these priorities had changed since baseline. During the recovery period, some participants accepted the recurrent nature of the condition. Others managed their own treatment expectations by considering any symptomatic improvement as an indicator of effectiveness.

I’ve still got some kind of stuff going down on there that is just a recurring thing … if I’ve had four operations, it probably won’t get rid of [it].

06

Some participants specified a single important outcome 6 months after surgery. The following were discussed: the wound healing in the expected time (n = 3); avoidance of recurrence (n = 4); return to ADL (n = 1). Three participants who had undergone PSD surgery for the first time found their treatment to be effective. When asked, they did not feel they would have done anything differently (see Appendix 3, Table 33).

Design and theoretical/conceptual framework

Discrete choice experiments are an attribute-based measure of benefit; they assume that healthcare interventions, services or polices can be described by their attributes. In a DCE, respondents are required to make trade-off decisions regarding the quantity or quality of a good or service. The resulting choices are analysed to estimate the overall utility (value) and willingness to trade between services. The use of DCEs to identify patient preferences in health and healthcare has increased. Where no clear treatment decision exists, accurate quantification of patient preferences for risks and benefit is crucial.

Development of survey

The survey contained four separate sections. These were:

1. patient characteristics and disease history

2. a treatment ranking exercise

3. DCE-specific tasks

4. survey feedback questions.

The ranking exercise and DCE tasks were developed by conducting qualitative interviews with 20 patients. The patients were asked to identify the key attributes and levels that they considered when choosing a treatment (see Chapter 4 for further information). A thematic analysis of the interview data was conducted and identified a list of factors that patients considered important when making treatment decisions. PPI members and clinicians reviewed an initial list of themes and selected attributes considered the most important for inclusion in the DCE and ranking task.

Nine attributes were included. These were:

1. type of excision and closure

2. type of anaesthetic

3. length of hospital stay

4. wound care

5. pain medication requirement

6. infection risk

7. healing time

8. risk of recurrence

9. scarring.

Currently, there are 18 theoretically possible surgical options for PSD. 14,110 Clinicians developed a treatment classification of five treatment groups. This was considered important due to the potential cognitive burden of ranking 18 treatment categories. The treatment categories, related descriptions and attributes were informed by the literature, clinical input and PPI piloting (see Appendix 3, Box 1).

Two attributes were included in the DCE: risk of infection/persistence and recovery time. This selection was based on two reasons. Firstly, these two attributes were assessed as most important by patients and clinicians. Secondly, DCE attributes must be independent to avoid presenting implausible combinations of attribute and level profiles. The levels for the DCE attributes were selected based on plausible values published in the literature and additional input from clinicians.

The choice tasks were constructed based on an orthogonal design using a design catalogue. 111 The survey contained 16 hypothetical DCE tasks. Participants were asked to choose between two combinations of outcomes with varying levels (see Appendix 3, Box 1). Forced unlabelled choices were presented – ‘treatment A’ or ‘treatment B’. An ‘opt-out’ alternative was not provided for the purposes of realism. A dominant task was included – where one treatment option is logistically superior – to test participant understanding of the task (see Appendix 3, Box 2).

Patient and public involvement

Patient and public involvement representatives were heavily involved in the design and implementation of the DCE. They assisted with the following tasks:

• Reviewed the initial list of themes to finalise the attributes.

• Reviewed the acceptability of the treatment category descriptions.

• Prior to the dissemination of the survey, three PPI representatives reviewed the survey to assess comprehensibility, interpretation and complexity of tasks.

Sampling

Discrete choice experiment sample sizes can vary from 100 to 1000 plus. 112 An Orme113 rule of thumb formula was adopted to estimate the minimum sample size. Using this formula – {N = 500 × [4 (maximum number of levels)]/[2 (number of alternatives) × 16 (number of tasks)]} – the estimated sample size was 63.

Recruitment

All participants aged 16 years and above with symptomatic PSD, referred for elective surgery and participating in the PITSTOP cohort study, were invited to take part in the survey. Interested participants were e-mailed a Qualtrics^® (Qualtrics, Provo, UT, USA) link which included a participant information sheet, consent form and the questionnaire. Participants with symptomatic PSD and not participating in the PITSTOP cohort could also take part in the DCE by accessing a digital QR code advertised on a study leaflet. The study leaflet was displayed in NHS Foundation Trust colorectal outpatient clinics and disseminated via the PITSTOP Twitter website.

Data analysis

Descriptive statistics were calculated for patient characteristics, disease history and survey feedback variables. DCE responses were modelled using conditional logistic regression where the dependent variable was the preferred treatment choice, and the independent variables were risk of infection/persistence and recovery time. Linearity of the attributes was assessed before deciding to treat risk as a linear variable (Figure 6). Regression coefficients were used to estimate the relative importance of attributes. Maximum acceptable risk (MAR) is the rate at which patients are willing to sacrifice a benefit of one attribute in exchange for an improvement in another. MARs were calculated by dividing the ratio of recovery time coefficients by the infection/persistence coefficient and 95% CIs calculated using the delta method. 114 Latent class models were used to analyse individual heterogeneity and to identify subsets of patients with varying preferences. The optimal number of classes was selected using the Bayesian information criterion (BIC) and consistent Akaike information criterion (CAIC) and model parsimony. Data were analysed using Stata^® 17 (StataCorp LP, College Station, TX, USA).

FIGURE 6.

Regression model 1 results reproduced in a diagram.

Participants

One hundred and eleven participants were included in the DCE survey. The completion rate was 74% (423 unique visitors entered the survey, 150 participants consented to take part and 3 participants declined). Table 16 reports the characteristics of the 111 included participants. Of these, 75 (68%) respondents were male and 73 (66%) were between the ages of 17 and 29 years; 89 respondents (80%) were employed and 97 (87%) were white. Except for six respondents, the rest of the sample reported having had at least one surgery for PSD. The respondents had various types of surgeries, including excision of the skin and closure of the wound with stitches (26%); excision of the sinuses only and leave the wound open to heal (23%); excision of skin and leave the wound open (23%); excision of the skin and closure of the wound with a skin flap and stitches (9%); and excision of the sinuses and closure of the wound with glue (19%).

Patient preferences

Appendix 3, Table 34 presents the regression modelling results of the DCE. Patients preferred treatments with lower risk of infection/persistence and this attribute was modelled linearly (see model 2 in Appendix 3, Table 34). Risk of infection/persistence was the most important attribute when patients were choosing a treatment, with an attribute importance score of 70%. Patients also preferred treatments with shorter recovery time; for example, compared to a treatment that takes 12 weeks to recover, a treatment with a 1-, 2- or 6-week recovery period was preferred. However, their preferences were not linear, so this attribute was modelled as a categorical variable. Treatments with shorter recovery time had an attribute importance score of 30%.

Maximum acceptable risk

When choosing a treatment, patients were willing to make trade-offs between risk of infection/persistence and recovery time (Table 17). These trade-offs were measured using MAR, which is the maximum risk of infection/persistence participants are willing to accept to have a treatment with faster recovery times. The highest-risk patients were willing to accept was a 17.08 risk of infection/persistence in return for a treatment with 2-week recovery period compared to a treatment with 12 weeks recovery period. Patients were willing to accept a 10.49 increase in risk of infection/persistence to have a treatment with 6-week recovery period compared to 12-week recovery period. Patients were willing to accept a 6.59 increase in risk of infection/persistence to have a treatment with a faster recovery period (2 weeks compared to 6 weeks).

Preference heterogeneity

Differences in preferences between patients were explored using the latent class modelling approach. This identified two groups of respondents with different preferences (Table 18). The first subgroup of respondents (class 1) were risk-averse and so they were only willing to accept a small risk (1.51–2.15) in exchange for a treatment with faster recovery time (Table 19). The second subgroup of respondents (class 2) showed stronger preferences for treatments with shorter recovery time: they were willing to accept higher risks of infection/persistence (22.35–34.67) to receive treatments with quicker recovery time. Of all the demographic variables used to predict whether a respondent belonged to class 1 or 2, only age of the respondents was statistically significant (see Table 18). The results show that respondents in the 17–29 age group were more likely to belong to class 1 and respondents above the age of 30 were more likely to belong to class 2.

Ranking of treatments

Patients ranked the treatments they preferred in order of importance, and Figure 7 shows the results of this ranking task. The best preferred treatment was complex flap (e.g. Limberg, Dufourmental) procedures (27%), followed by excision of the sinuses only (22%), glue (19%), excision of the skin and closure of the wound with stitches (18%) and lastly leave open (14%). The least preferred treatment was leave open (35%), followed by glue (23%), complex flap procedures (18%), excision of the sinuses only (17%) and excision of the skin and closure of the wound with stitches (7%).

FIGURE 7.

Ranking of treatments.

Participants’ understanding of and engagement with the survey

Table 20 reports the results of the questions used to test the internal validity of the DCE.

Most of the respondents said that they understood the DCE tasks (91%) and ranking task (86%). Ninety-three (84%) of the respondents correctly answered the DCE question with a logically correct answer (dominance test). None of the respondents were always choosing the same side (left or right) profiles of the DCE tasks. Fewer than 25% said that the DCE task was confusing and that they needed more information.

Stakeholders

Two stakeholder groups were defined: patients with previous experience of PSD and clinicians with an interest in PSD. Clinicians included those with certificates of completion of training in general surgery and nurse specialists with wound care practice. Participants were recruited via e-mail, national organisations and social media. Snowball sampling was also used for clinician recruitment. Participants were invited to participate in one, two or all three phases of the consensus exercise.

Patient and public involvement

Fifteen patient representatives with relevant experience were recruited to the patient stakeholder group following substudy conception, and contributed to the delivery and analysis. Of these, 6/15 PPI representatives attended the initial workshop and supported the generation of the longlist of recommendations; 4/15 PPI representatives attended the virtual consensus meeting and highlighted the importance of ensuring the final set of recommendation statements were conceivable to a patient audience.

Generation of longlist

In accordance with Delphi methodology, the study consisted of three phases. In phase 1, an online workshop was conducted. This was based on Rolfe’s critical reflection model, ‘What? So What? Now What?’. 118 In the ‘What?’ phase, data or information is presented. In this case, researchers presented findings from the cohort study, mixed-methods study and survey-based work. In the ‘So What?’ phase, participants are encouraged to reflect and discuss the information. Participants were asked to consider how the presented data reflected their experiences, and how this matched wider experiences. In the ‘Now What?’ phase, participants discuss how the information should be used to influence the next stage. Participants were asked to frame their ideas as statements related to policy or research ideas. In the workshop, the following data were presented: preliminary PITSTOP cohort data, a systematic review of classification systems,110 a mapping reviewing of PSD,119 the PITSTOP DCE survey109 and the PITSTOP mixed-methods study. 30 Participants were asked to consider two questions: ‘How can we use this data to improve and/or inform clinical practice?’ and ‘What are the key research questions generated by this data?’. A longlist of potential practice and research recommendation statements was generated. The steering committee assessed the readability of these statements. Prior to attending the workshop, all participants received an information sheet and completed an online electronic consent form.

e-Delphi consensus

In phase 2, a three-round e-Delphi consensus was conducted. The Delphi surveys were delivered using Qualtrics. In round 1, all participants were e-mailed a participant information sheet and a link to the survey. Upon accessing the survey, participants were asked to complete an online consent form. The following information was captured: age, gender, demographics, ethnicity, e-mail address and stakeholder respondent group (patient, surgeon and specialist nurse). The longlist of recommendation statements was presented in a random order, and each statement was supplemented with a written summary to aid understanding. At the end of the survey, respondents were encouraged to propose any additional statements. Additional items were reviewed at the end of round 1.

In rounds 2 and 3, the remaining longlisted items were presented in random order. Ratings of items were reviewed after the close of each round. Respondents received an e-mail copy of results which included their vote and how that compared to each stakeholder group’s votes.

During each round, participants voted on the importance of each recommendation using a 9-point Likert scale (one being ‘not important’ and nine being ‘very important’). Recommendations were shortlisted if the following was satisfied: (1) > 70% participants within both stakeholder groups rate the recommendation as 7–9; or (2) 90% participants within a single stakeholder group rate the recommendation as 7–9. Recommendations that reached consensus after three rounds were considered at the consensus meeting. All items had to be rated to complete the surveys. Only those who completed a survey round were eligible to participate in the subsequent round. At the end of round 3, all participants were asked if they were interested in participating in the virtual consensus meeting.

Virtual consensus meeting

In phase 3, an online consensus meeting was held to finalise the set of recommendations. Invitations were issued to interested participants, with efforts made to encourage participation from members of the public/patients. A target of > 15 participants with at least three patient representatives was felt to be reasonable for this prioritisation exercise as it reflected proportions recruited to round 1 of the consensus. The meeting was held using the Google Meet™ videoconferencing platform (Google Inc., Mountain View, CA, USA). Electronic consent was taken prior to the meeting. Participants were presented with a total of 34 statements: 15 policy and 19 research. After the presentation, participants were instructed to complete two separate Qualtrics surveys. A constant sum question type was used to calculate the sum of scores for each statement. In the first survey, participants were asked to distribute 100 points between the 15 policy statements dependent on priority. This could be all points to a single item, an even division, or however the respondent felt appropriate, as long as 100 points were distributed. The total points allocated to each item were then calculated, and the five highest-scoring items were considered as top priorities. In the second survey, participants distributed 100 points between the 19 policy statements dependent on priority. The same approach to summing points was conducted.

Longlisting of potential outcomes

Following the initial ‘So What, Now What’ workshop, 33 items were generated for the longlist by clinicians and patients. The flow of items is presented in Figure 8.

FIGURE 8.

Flow of items through the e-Delphi.

Delphi round 1

Consent forms were completed by 57 potential candidates from both stakeholder groups, and 57 completed round 1. Characteristics of respondents’ participation among stakeholder groups are presented in Table 21. This included 15 patients, 40 surgeons and 1 nurse specialist. A total of 33 items were considered for level of priority; 15 items reached a priori consensus for inclusion for the final consensus meeting. Respondents proposed a further 12 items for review. Outcomes voted on in each round, along with the proportion of each panel rating them 7–9, are presented in Appendix 3, Tables 35 and 36.

Delphi round 2

In round 2, 53 participants completed the survey. This included 14 patients, 38 surgeons and 1 nurse specialist, 95% of those who completed round 1. Respondents were sent a summary of voting patterns from the first round and were asked to vote on 30 items: 18 for reconsideration and an additional 12 statements proposed. Of these, 18 items met the a priori criteria for inclusion in the consensus meeting.

Delphi round 3

In round 3, 51 participants completed the survey. This included 14 patients, 36 surgeons and 1 nurse specialist. This reflected 96% of those who completed round 2 and 91% who completed round 1. One further item was carried to the consensus meeting.

Consensus meeting

The consensus meeting was attended by three patient representatives and 15 clinicians. One clinician withdrew during the meeting due to work commitments. The top five policy statements and research recommendations are presented in Table 22.

Overview

Research in surgery has been much maligned over the years,120 and pilonidal disease is no exception. 14 This is reflected in relatively weak or vague guidance to support practice. The Delphi we conducted has identified high-priority practice and research topics to guide the further development of the field. Consensus statements on practice topics strongly reflect the findings from previous sections. The top practice recommendation reflects the need to avoid harm related to interventions. This is supported by findings from the mixed-methods study which showed high levels of regret associated with poor wound healing. The cohort study supports the top two policy recommendations as it demonstrates the ongoing morbidity from poor wound healing after major procedures. In contrast, the third policy recommendation does not relate to clinical outcomes, but to the need to train surgeons in new techniques as highlighted by the clinician survey. This is particularly important if the top two priorities are to be achieved, as upskilling of surgeons will be required to facilitate techniques beyond ‘excise and leave open’. The need for a classification tool with clinical reference is clear. In other areas of colorectal surgery, classification tools and systems facilitate decision-making around treatment pathways. 121,122 This inconsistency in PSD means that it can be difficult to compare outcomes between patients and surgeons due to an inconsistent baseline description and treatment selection. Finally, mixed-methods and cohort studies highlighted the importance of return to work in this typically young and economically active patient group.

The recommended research priorities provide direction on next steps. Priorities one and three discuss the potential design of a future RCT. These suggest that a pragmatic approach to design, using an umbrella-type approach with interventions grouped into severity or grade of procedure, would be a appropriate. 123 We have seen similar approaches in other proctology studies. 124,125 The fifth priority also demonstrates the need to understand interventions which work across the treatment pathway. These are not limited to surgical interventions, as adjuvant treatment such as hair removal, use of antibiotics and wound dressings may also play a role in this pathway. 119 With this in mind, it may be more appropriate and efficient for a funder to commission a multiarm, multistage trial, with rerandomisation of patients who develop recurrence of PSD. The importance of measuring relevant outcomes in a consistent manner is emphasised, with the need for a core outcome set identified as a priority. 126 In addition, participants highlighted the need for the development of patient-reported outcome measures (PROMs). PROMs are important in any core outcome set. 127 A multidimensional PROM may include an assessment of return to work, wound healing and recurrent symptoms, all of which have been identified as essential in prior work packages.

Patient and public involvement

The consensus was based on the engagement of patients and members of the public. Initial findings were discussed with patient representatives, and these ideas were used when delivering the initial workshop. This was also attended by patients who were able to express their priorities. This engagement continued at all stages. During the final consensus meeting, patient representatives were invited to comment regularly to ensure their voice was considered in the final ranking.

Impact for policy-makers

This exercise sets out five clear policy statements which could form the basis for the development of future guidelines by informing PICO (population, intervention, control/comparison, outcome) development. The findings of limited training opportunities following qualification as a consultant surgeon may not be limited to this area, and policy-makers should be cognisant of this when commissioning services.

Impact for researchers

Researchers can use the findings of this study to direct future research. The directives here are for pragmatic trials established to address questions along the treatment pathway. Studies should also have a clear patient focus.

Conclusion

This consensus exercise has involved patients and clinicians to identify five key policy and five key research priorities. These should form the foundation of future work in the field.

Rationale

There is no commonly used classification tool to characterise PSD. The absence of such a tool represents an important knowledge gap, since surgeons faced with PSD typically have little exposure to the disease and there exists little guidance on how to treat it.

At the inaugural meeting of the Pilonidal Society in Berlin in 2017, a panel of 13 surgeons gathered to establish a simple classification system for PSD. 136 Each member of the panel had a special interest in PSD and had either written other classifications132,134 or had published widely on PSD. Components of a classification system were longlisted and refined by online consensus involving a large group of PSD surgeons137 to form a final configuration that was felt to be easy to use, clinically meaningful and had potential statistical validity.

The tool classifies PSD into one of four categories:

• type 1: only midline pit or sinuses

• type 2: any midline disease with secondary sinus/es or abscess scar/s

• type 3: any midline or secondary disease extending below tip of coccyx

• type 4: any disease after treatment with definitive intent.

While there may be agreement among the panel for the eventual classification (referred to as the ‘Wysocki classification’), none of the parameters of ease of use, reliability and validity have been tested. We aimed to do so by assessing the level of agreement between surgeons when used in clinical practice.

Aims

The aims of this substudy were to:

• Quantify how well different assessors agree in their classification.

• Quantify how classification relates to surgeon’s experience.

Additional exploratory aims were to:

• Identify specific patients with low agreement, which may in turn help clarify or even modify the classification system.

• To present, within each subtype, the frequency with which each treatment option is chosen.

Participating surgeons

Sampling was purposive and sought to recruit 15 colorectal surgeons. This sampling method aimed for maximum variation based on experience: five expert surgeons who registered an interest in pilonidal disease and who offered a specialised service, five surgeons who carried out pilonidal sinus surgery as part of a general surgical service and five final-year colorectal trainees. Initial contact with surgeons was made by e-mail, and interested surgeons were e-mailed the participant information sheet.

Participating patients

Stimuli required for the validation exercise were obtained from the main cohort study.

All participants referred for elective surgery and participating in the PITSTOP cohort study were asked if they agreed to the surgical site being photographed before surgery (an optional item on the consent form).

Participant photographs of the PSD surgical site were usable for the exercise if they satisfied the following criteria:

• The participant was eligible for the main cohort study (aged 16 years and above with symptomatic PSD, referred for elective surgery).

• The participant consented to a preoperative photograph to be taken of the surgical site.

• The photograph was usable (i.e. unblurred).

• The participant associated with the photograph had been classified using the Wysocki classification system at the time of procedure.

Assessment schedule

Each participant was asked to independently rate 36 cases using the Wysocki classification, with allocation of surgeons to cases selected to ensure overlap with other assessors (Figure 9). A total of 90 photographs were each assessed by two specialist surgeons, two general surgeons and two trainee surgeons. Surgeons were sent photographs accompanied by the medical history (previous PSD history, including number of elective procedures and emergency drains) electronically. Surgeons were not told the classification as recorded by the original surgeon at the time of procedure.

FIGURE 9.

Allocation of patients with PSD to surgical assessors.

The substudy was also used to provide an exploratory assessment of surgical opinion. Participating surgeons were asked to record their preferred treatment for each patient they assessed, with the aim being to quantify variation in practice among practitioners. Surgeons recorded their assessments in Microsoft Excel^® (Microsoft Corporation, Redmond, WA, USA) spreadsheets which were returned to CTRU and combined into an analysis data set.

Statistical methods

Agreement was quantified as both raw and chance-corrected agreement. Raw agreement is the percentage of patients for whom the assessments agree, while chance-corrected agreement is the ratio of observed to expected agreement. Both raw and chance-corrected agreement are essentially proportions in which one means complete agreement while zero reflects complete disagreement. As the four categories are not ordinal, agreement is a simple yes/no construct in which any difference is considered ‘disagreement’.

Raw agreement was defined as (100 × number in which all raters agree / number rated) and was accompanied by a 95% Wilson score interval. Chance-corrected agreement was defined using the unweighted kappa statistic and the unweighted Gwet AC1 statistics. 138 Agreement among surgeons was reported overall and within for expert, general and trainee surgeons. Finally, the agreement was calculated for each surgeon in relation to the original assessment made at the time of procedure. Analyses were conducted using Stata version 17. 139

Sample size

The sample size was based on: (1) a hypothesis test to rule out a minimal kappa statistic;140 (2) the standard error of raw agreement; and (3) the number of patients expected to consent and provide usable photographs. For (1), an internal pilot was undertaken in which study surgeons were asked to assess photographs obtained either online or via published articles. A total of 41 pictures were assessed by seven surgeons (five specialists and two trainees) and yielded an overall kappa statistic of 0.42 (0.55 if trainees were excluded). Since the assessments were based on pictures alone and did not include prior history, these may be an underestimate. Based on this, a target of κ = 0.45 was used. The expected lowest limit for chance-corrected agreement was set at κ = 0.3, which represents the lowest acceptable agreement if this classification were to be introduced into practice. The sample size calculation also depends on the expected prevalence of each class, which was estimated from the cohort study (approximately 25% type 1, 50% type 2, 10% type 3 and 15% type 4 at the point of data review). A sample size of 90 was adequate to rule out differences of 15% between expected and minimum kappa with 90% power and 5% significance (1); to estimate raw agreement to within a CI half-width of ±10% (2); and to be accommodated by the number of photographs available (expected around 150) (3).

Participants

Fifteen surgeons participated in the classification exercise as described in Chapter 7, Assessment schedule. A total of 166 patients consented to and provided a photograph of the diseased area, of which three were too unclear to use and were removed. Ninety participants were randomly selected from these for assessment, all of whom had been assigned a classification of 1–4 by the original treating clinician. The majority were classified as type 2 disease at the time of procedure, and 60% underwent a major procedure, with the most common treatments being a Karydakis asymmetric closure (n = 42) or glue (n = 29) (Table 23).

Agreement among participating surgeons

The agreement among surgeons is summarised in Table 24. Of the 540 assessments (90 patient photographs and case histories each having 6 assessments), 14 (3%) of assessments were classified ‘none of the above’, affecting 12 (13%) of the patients.

Overall, the six assessors all reached the same consensus in 38% of participants, with a chance-corrected kappa statistic of 0.52 (95% CI 0.42 to 0.61) and a Gwet AC1 statistic of 0.63 (95% CI 0.56 to 0.69). Agreement between pairs of surgeons was higher, with specialist surgeons agreeing in 72% of patients, general surgeons agreeing in 70% of patients and trainee surgeons agreeing in 71% of patients. The overall agreement is lower since this measure required all six assessors to agree. All six surgeons agreed in 34 (38%) of patients, and five of the six surgeons agreed for 19 (21%). The chance-corrected kappa agreement was above 0.5 (conventionally considered ‘moderate’) and the Gwet AC1 measure over 0.6 for all subgroups of surgical expertise.

Agreement between participating surgeons and original classification

Each assessor’s agreement with the original classification is described in Figure 10 and 11. Surgeons were less likely to agree with the original classification than with other surgeons given the same photograph. Raw agreement ranged between 47% (17/36) and 75% (27/36) with chance-corrected kappa statistics ranging from 0.11 to 0.59, and chance-corrected Gwet AC1 agreement statistics from 0.35 to 0.71.

FIGURE 10.

Raw agreement between assessment and original.

FIGURE 11.

Kappa chance-corrected agreement between assessment and original.

Treatment choice

Surgeons differed markedly when asked how they would treat the individual participants. Overall, surgeons in the substudy recommended a minimally invasive procedure in 46% of cases, but the figure ranged from 0% to 94% (34/36) of patients. There was substantial variation in practice among all levels of expertise, and the variation in recommendation did not appear related to the actual patients assessed (Appendix 3, Figure 16).

The surgeons surveyed were slightly more likely to recommend minimally invasive surgery (46%) than those that actually received this approach (40%). Although the percentage favouring asymmetric closure (45%) was similar to the treatment actually received (49%), the specific procedure types differed: the surgeons surveyed were more likely to use a Bascom’s cleft lift (22%) than a Karydakis (16%). The use of midline closure (5%) and leave open (4%) approaches were uncommon, most notably among the specialist surgeons surveyed.