Summary of results from this paper
The procedures studied in this project are those commonly undertaken by GPs anywhere [2, 13, 14, 18, 19, 26,27,28,29,30,31,32]. We selected participants who had sufficient experience to demonstrate the capacity of GPs in Ireland to perform minor surgery and to have a sufficient established scope of practice to test an accreditation procedure for community surgery [17]. This was a purposefully heterogenous group of GPs with a seven fold difference between the numbers of cases managed by the busiest and least busy practices in terms of minor surgery activity. Inter-practice referrals were common.
We report on key quality indicators of safety: submission of tissue for histological examination, accuracy of clinical diagnosis, completeness of excision of malignancy, avoidance of unnecessary excisions and surgical complications. On each of these measures, the participants had results in line with available comparable data [7, 13].
Interpretation of results in light of literature
There has been a debate as to the scope of the GP’s practice and it is especially evident in the area of technical procedures [33]. Policies tending to devolve hospital surgical work to primary care, such as that underpinning the present project, have been seen in other jurisdictions [34].
There has been little argument that minor surgical services provided in the community are both convenient and cost effective [24, 27] and encouraging community provision does not invariably divert work from hospitals as much as it generates new activity, albeit to address an unmet need [24, 34].
The role of the GP in managing skin cancers has been resisted by specialist providers out of concern about safety [22, 23, 25, 35] or on health economic grounds [36,37,38].
The data we present represents the largest report of surgical work by GPs in Ireland.
Previous Irish surveys have established the range of procedures that are commonly or sometimes provided in general practice [18, 19]. These were the procedures commissioned in the current project and reflect the scope of primary care surgery internationally [2, 24, 28, 29, 32].
The variation between respondents in the numbers of cases reflects practice size and location.
O’Kelly reported that 65% of GPs in Ireland in 2015 stated that they provide some minor surgical services [14]. Earlier Irish surveys by Clarke [18] and White [19] indicate that a third of GPs frequently undertake the more common procedures we have studied. Barriers to wider provision in the community reported internationally, and likely at play in Ireland, include limited training, lack of time, equipment and insufficient compensation [39,40,41]. In establishing the present project, we were limited by budget rather than any difficulty attracting volunteers.
Overcoming the barriers that do exist to the provision of procedures in primary care would seem to require incentivization of the GP. In the course of this project, participants were compensated by way of enhanced income for their surgical clinics and by the achievement of accreditation. An economic assessment of whether this is the best way to supply patient demand for minor surgery was beyond our remit. On the other hand, the project arose from a concern as to the inordinate unit cost of provision in Irish hospitals [15, 16].
Economics aside, any commissioner of a devolved surgical service will have a concern as to safety. Quality assurance in minor surgery has many dimensions, including the doctor’s qualification, maintenance of professional competence, the quality of their premises, equipment and policies, as well as their clinical and technical ability.
Qualifications, infrastructure and governance are domains of high quality community surgical practice. These domains underpinned the accreditation standards, which have been reported separately [17].
In this study, we had the opportunity to measure the safety of the surgical service from the perspective of clinical and technical proficiency in relation to the cases undertaken as part of that accreditation procedure.
The key indicators of quality which we studied are established measures in minor surgical practice [6, 7, 13, 14, 22–25].
The diagnostic accuracy of the GP is important because better clinical diagnosis avoids unnecessary excisions as well as missed diagnoses of skin malignancy. This is measured by the clinicopathological correlation. Two aspects are of interest. The overall proportion of cases where the GP’s clinical diagnosis is confirmed on histology and the occurrence of an unrecognized malignancy are simple measures which can be audited.
It is well established and entirely expected that skin specialists (dermatologists and plastic surgeons) perform better at pre-operative diagnosis than either GPs or general surgeons [22, 25, 35, 42]. Early British reports from single pathology laboratories following the contractual encouragement of GP skin surgery in 1991 indicated alarmingly low rates of correct clinical diagnosis in GP specimens [25, 42]. In a British randomised trial of GP minor surgery, 45% of clinical diagnoses were confirmed at histology [6]. Overall clinicopathological concordance for all lesions among our cases was reassuring by comparison.
In relation to the clinical suspicion of malignancy in particular, our results compare well with recent audit data from the United Kingdom [13], the small number of malignant lesions included in the only extant randomized trial of GP minor surgery, and Corwin’s results from New Zealand [22]. The anxiety that the GP surgeon is missing skin cancer has been a long-standing concern of specialist surgeons and dermatologists [6, 25, 42]. However, this concern could apply equally to GPs and non-specialist skin surgeons who refer rather than operate.
Some unanticipated histological results are inevitable whatever the surgical setting. For this reason, adherence with the requirement that all excised material be studied by the pathologist is a further key indicator of safe practice. The majority of lesions eligible for histology were referred appropriately by the GPs in our study and accord with other reports [7, 13]. Though the ideal is that all excised tissue be submitted, in practice, GPs frequently do not submit small lesions such as skin tags and warts or grossly recognizable lesions such as sebaceous cysts and seborrheic keratoses [6]. These are commonly treated by curettage or shave excision. Our results appear to reflect that practice.
Where skin cancers are excised by the GP, the adequacy of resection margins is an important indicator of quality and a proxy for cure. Comparisons are difficult as studies often report a categorical assessment rather than one based on measurement or do not state how margins were assessed at all. We have only reported data relating to peripheral margins as deep margins are not reliably reported in our sample or generally in Ireland [43]. In adopting the RCPath guidance [20, 21], which regards a margin of 1 mm or less as being involved, we provide some indication of the completeness of excision. Internationally, much of the evidence indicates a greater likelihood of adequate excision of malignancy by specialist surgeons [7, 23, 44]. However, GPs may perform as well as non-skin specialist surgeons in excising BCC [43] and in one Scottish study, GPs of varying levels of experience had rates of complete excision as good as those of skin specialists [10].
However, prognosis is based on the clinical margin rather than the histological outcome and we did not study this [45]. Furthermore, a close margin in a BCC at a low risk site is a different prospect than the same margin for an aggressive subtype of SCC at a high risk site. Because of these issues, both our study and results reported elsewhere from similar surveys are inevitably of limited value in determining safety. Ideally, a randomised comparison of tumour recurrence rate over at least 5 years would be required to compare the GP and skin specialist outcomes.
An additional indicator of quality which relates to clinical diagnostic ability and surgical judgement is the rate at which benign lesions are excised as full-thickness skin resections. It is generally inappropriate to undertake surgical excisions of warts, skin tags, seborrheic keratoses other than by curettage, electrocautery or cryoablation. Our results for possibly inappropriate excisions compare favourably with those of a recent review of pathology requests at a regional laboratory in one Irish region [36].
A final and important measure of safety is the rate of surgical complication. Due to the limited length of follow-up, which extended to 1 month after the last recruitment, late complications such as poor scar and tumour recurrence are not assessed in our study. Our results are similar to those in Botting’s audit of over 6000 cases in Britain [13].
Limitations
Besides the limitations relating to quality indicators already referred to, several general limitations pertain to these descriptive data.
The study was not primarily designed as a test of the quality and safety of Irish community surgery but as a description of work undertaken in the course of the design and piloting of a community surgery accreditation process, and this limits the strength of conclusions that can be drawn from our findings.
The requirement of the GPs to undertake double data entry is less than ideal when seeking to collect comprehensive details of surgical work, and structured clinical records linked to automated data extraction may be a better solution. This limitation may have resulted in missing or miscoded data.
The participating GPs were chosen on the basis of their surgical experience from a self-selected group of volunteers. Their performance cannot be taken to represent all Irish GPs. Nevertheless, minor surgery is provided by volunteers within general practice and our cohort may be deemed an appropriate representation of what the more experienced among them can achieve. However, the number of cases may be higher among this group than among the average general practitioner.