A template for future pilonidal sinus research

Striving for an evidence-based approach is laudable but not always easy in surgery. This is particularly the case for pilonidal sinus disease. The literature from the last 20 years is vast but consists of mainly single-centre studies [1]. Most are of a retrospective cohort design, extolling exceptional outcomes with that centre's technique and management and therefore not reflective of the real world. Even the randomized controlled studies are almost exclusively single-centre based with other significant methodological flaws. A brief but by no means exhaustive trawl of the literature from the last 5 years has shown no change in this trend. Hardly any of the ‘randomized controlled trials’ identified were multicentre, and those that were ‘multicentre’ mainly involved two hospitals. For the comparative studies the number of participants was as low as six and most were below 90. Whilst 90 may be enough there is a dearth of sample size calculations to justify the number. A possible explanation for this obvious and, at first glance, easy to remedy lack of large multicentre trials relates to the vast number of interventions that exist with no obvious front-running procedure. Many experts have their preferred technique which, if the literature is to be believed, works very well for them. When it comes to involving surgeons from other centres, they may have their own favoured technique and lack equipoise. Even if they can be persuaded to trial an alternative technique, they may require training. A robust quality assurance element of the trial is necessary to ensure that everyone is doing the alternative procedure(s) correctly. There is also a geographical issue that makes international studies difficult, with some countries having popular operations that are not popular or even carried out in other countries. For instance, one of the most popular interventions in Belgium is sinus laser closure (SiLaC), which is uncommon in the UK. In the UK pit picking with glue is a frequently performed minimally invasive procedure but is very rarely carried out in Australia (Peter Wysocki, Logan Hospital, Australia Neils Komen, University of Antwerp, Belgium, personal communication). There are of course ways around this when designing a trial. Keeping the trial within one country and choosing an intervention that is easy to learn and/or commonly utilized within that country is clearly preferable. Also keeping the comparator broad whilst maintaining clinical relevance. Groupings of interventions that include minimally invasive procedures [e.g. pit picking, glue, endoscopic pilonidal sinus treatment (EPSiT), SiLaC] and more major skin excisional procedures (e.g. asymmetric closure, flap procedures) is an option that keeps comparators broad and is consistent with the recent consensus on research priorities [2]. Some may argue such a grouping is too broad. The lack of stratification of disease is a major shortcoming of the current pilonidal sinus literature. One operation is unlikely to suit all disease presentations. A person with minimal disease (e.g. pits in the midline in the absence of any lateral disease) does not necessarily require a major skin excisional procedure. There is a danger of making the surgery worse than the disease itself. Conversely, poor outcomes may be expected with the use of a minimally invasive procedure on a patient with complex disease. A validated classification system is now available, and it or an equally validated system should be incorporated into all future comparative research [3]. Perhaps the most problematic issue with the current literature relates to outcome measures. Most published research incorporates ‘recurrence’ as the primary outcome. And yet recurrence is rarely defined. A review of all the literature cited in the recent Italian guidelines reveals that 85% of the references fail to define ‘recurrence’ at all [4]. In the true sense of the word recurrence should be described as disease healing and then further disease arising at that site or at a de novo site. Such a phenomenon is probably more a function of follow-up, and extensive follow-up is required to give a realistic estimate—perhaps as much as 5 years [5]. What is more relevant and more feasible when comparing techniques is treatment failure by a specific timepoint. Any effective intervention should theoretically result in complete healing within the space of a few weeks. Data from the PITSTOP study suggest that up to 15% of patients had persistent symptoms at 6 months [6]. A 6-month assessment is therefore one potential assessment timepoint. Patients involved in a recent consensus ranked time away from work (or usual activities) as the most important outcome [2]. Such an outcome measure is a surrogate for treatment failure as well as speed of recovery, and could be considered an alternative primary outcome to treatment failure alone. If treatment failure is used as an outcome, it is essential that it should be patient reported. Pilonidal disease is a condition that exerts impact through symptoms. Whilst treatments are aimed at cure, they all commonly target symptomatic relief. The best person to assess symptoms is the patient, using a patient-reported outcome. This is consistent with the patient and surgeon consensus and tackles the current criticism of lack of data of patient-reported outcome measures in the surgical literature in general [2, 7]. For other outcomes, a core outcome set is preferable. No such tool exists currently and there is a need for development [8]. Given the recent interest in techniques that require what may be expensive equipment (EPSiT and SiLaC) health economics should also be included as an outcome. Such an outcome is rarely, if ever, present in the current literature. The above gives a template for future methodologically robust and clinically meaningful research (see Figure 1): studies that are clinically meaningful, of appropriate size and that are likely to recruit from multiple centres, not just specialist hospitals or even individual surgeons; a classification of disease so that equivalent disease severity is being compared; and well-defined and patient-relevant outcomes including heath economics. Steven R. Brown: Conceptualization; writing – original draft; writing – review and editing. None. I have no conflict of interest. None. No patients were involved. The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.