Family Practice Advance Access originally published online on June 23, 2008
Family Practice 2008 25(4):221-227; doi:10.1093/fampra/cmn036
Excision of malignant melanomas in North Wales: effect of location and surgeon on time to diagnosis and quality of excision
a Department of Primary Care and Public Health, North Wales Clinical School/School of Medicine, Cardiff University, Wrexham Technology Park, Wrexham LL13 7YP
b Department of Primary Care and Public Health
c South East Wales Trials Unit, School of Medicine, Cardiff University, Neuadd Meirionydd, Heath Park, Cardiff CF14 4YS
d Department of Dermatology, Wrexham Maelor Hospital, Croesnewydd Road, Wrexham LL13 7TD
e Cancer Services, Wrexham Maelor Hospital, Croesnewydd Road, Wrexham LL13 7TD
f National Public Health Service for Wales, Preswylfa, Hendy Road, Mold, Flintshire CH7 1PZ
g Department of Audit, Research and Effectiveness, Wrexham Maelor Hospital, Croesnewydd Road, Wrexham LL13 7TD, UK
Correspondence to Richard D Neal, Department of Primary Care and Public Health, North Wales Clinical School/School of Medicine, Cardiff University, Wrexham Technology Park, Wrexham LL13 7YP, UK; Email: nealrd{at}cf.ac.uk
Received 31 January 2008; Accepted 26 May 2008.
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Background. The epidemiology of melanoma is changing and its current management is variable, with some lesions being removed in general practice. We aimed to determine the quality of excision and time to diagnosis relating to the excising surgeon and the place of excision.
Method. Analysis of data from the North Wales Melanoma Database.
Results. In total, 578 cases were diagnosed 1993–2001. There was a gender difference with anatomical location, with 107 (65%) males with lesions on their trunk compared to 57 (35%) females. Median Breslow thickness was 1.10 mm (range 0.05–16.0 mm). Ninety-five (16%) lesions were removed in general practice, of which 49 (52%) were referred on to hospital. In total, 266 (61%) lesions were excised with ‘adequate’ margins and 170 (39%) excised with margins narrower than the guidelines. General practice excisions were from a younger group than hospital excisions. There were no differences in quality of excision between general practice and hospital excisions. Time to diagnosis was shorter overall for general practice excisions than hospital excisions (median 12 versus 41 days, P < 0.001).
Conclusion. These findings are of policy importance in that there is no evidence from this study that general practice excisions are managed poorly or have a worse prognosis.
Keywords. Breslow thickness, diagnostic delays, general practice, melanoma, quality of excision.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Cutaneous melanoma accounts for 2% of all cancers and 1% of all cancer deaths. Survival rates for early disease are high compared to many other cancers, although significant mortality exists despite treatment.1 Management of lesions suspicious of melanoma is by excision biopsy followed by a wide local excision (WLE) if the diagnosis is confirmed, with margins dictated by the Breslow thickness.2 The standard practice for the diagnosis and treatment of melanomas in North Wales is outlined in the North and South Wales Melanoma Group Guidelines.3 UK guidelines state that ‘suspected’ melanoma should be referred urgently by GPs to a dermatologist or surgeon/plastic surgeon and be seen within 2 weeks of referral and that the excision of all pigmented lesions suspected to be skin cancer is discouraged in primary care.4,5 Patients who have had melanomas removed by their GP should be referred immediately to specialists for further assessment.6 A small percentage of melanomas are removed by GPs, either deliberately (i.e. with diagnosis suspected) or accidentally (with diagnosis unsuspected). There is a relative paucity of evidence from small studies regarding the extent and quality of general practice management of melanoma and outcomes associated with this.7–13 Significantly longer times to diagnosis are usually patient rather than physician related,14 but there is no convincing evidence that an increased time to diagnosis worsens clinical outcomes in terms of Breslow thickness or survival.15–20 Reported associations between time to diagnosis and other factors are often conflicting, for example with respect to age and gender.16–18,20–22
This paper aims to determine the quality of excision and the time to diagnosis relating to the excising surgeon and the place of excision of melanomas in North Wales. A secondary aim was to describe the primary care management of lesions. We are able to do this through analysis of data from a large comprehensive cohort of patients providing findings of both policy and clinical importance.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
The North Wales Melanoma Database records details of all patients in North Wales with a diagnosis of primary cutaneous malignant melanoma. These patients are identified from histology reports that are sent directly from the pathology departments of the three North Wales Trusts (Wrexham Maelor Hospital, Ysbyty Glan Clwyd and Ysbyty Gwynedd) to their cancer services offices. Privately removed lesions are also sent to these pathology laboratories. Pathology laboratories in Chester, Shrewsbury and Liverpool (Whiston Hospital) are asked for data relating to lesions removed from patients resident in North Wales. Ocular, non-cutaneous and in situ melanomas are not included (in line with the majority of epidemiological studies which consider melanoma and in situ melanoma separately). North Wales has a population of 623 000 and is served by about 250 GPs with six local health boards.
Most patients had two procedures, a primary diagnostic excision biopsy followed by a wide excision, as directed by the Breslow thickness on the biopsy. The margin analysed in this study refers to the WLE. In cases where there was no clinical doubt about the diagnosis, some patients had just one excision with a margin deemed appropriate by the excising surgeon. In these cases, the margin analysed in this study refers to this definitive excision.
A primary treatment pro forma is sent to the relevant GP or consultant and initial non-returns are followed up. Where the surgeon removing the lesion has not returned the primary treatment pro forma, a dermatologist reviews the notes to extract the necessary data. Annual follow-up forms for each patient are sent to the clinician for 10 years. Disease-free patients are discharged from consultant care to GP care. The data collected are shown in Box 1.
| BOX 1 Data collected for the North Wales Melanoma Database Data from consultant or GP.
Pathology reports.
Follow-up data.
|
Prior to the research, the data set was anonymized. Data collection has been ongoing since 1993 and for the analysis in this paper is reported up until 2001. Data were transferred from an Excel spreadsheet to SPSS and checked for errors, such as coding errors, multiple entries for the same original lesion and date errors. We classified the quality of the excision as ‘adequate’ when excision margins were within the guidelines6 and ‘narrow’ when they were less than stipulated. Components of the time journey to diagnosis were calculated in days in line with a previous definition23: GP delay (days from first presentation in general practice to referral or GP biopsy); referral delay (days from referral to first being seen in clinic); secondary care delay (days from first being seen in clinic to biopsy or WLE for GP excisions); and total delay (days from first presentation to biopsy or WLE for GP excisions).
Statistical analysis
Logistic regression was used to determine predictors of quality of the excision. Associations between certain factors (age, gender and overall deprivation, lesion location, lesion thickness, place of excision and excising surgeon) and the quality of the excision were assessed using univariate statistics (such as t-test, 
2 and Mann–Whitney). Factors that reached a 10% level of significance were entered into a logistic regression model, which was fitted using a forward stepwise approach.
Appropriate univariate statistics again were used to explore demographic differences between lesions excised in general practice and secondary care and whether patients with lesions excised in general practice differed according to whether they were referred to hospital or not.
The difference in patients’ delay (GP, secondary care and overall delay) between where the lesion was excised (GP or hospital) was examined by using the Mann–Whitney non-parametric test, due to the skewed nature of the delay data.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Patient demographics and lesion characteristics
There were 578 cases diagnosed 1993–2001 inclusive. In total, 239 (41%) of the sample were male with an average age of 58.3 years (SD = 16.9 years) compared to an average age of 59.6 years (SD = 19.0 years) in females. By location, 164 (28%) lesions were on the trunk, 162 (28%) were on the lower limb, 123 (21%) were on the upper limb and 115 (20%) were on the head or neck. Data on 14 (2%) patients were missing. There was a significant difference in the location of lesions between gender, with 107 (65%) of males with lesions on their trunk compared to 57 (35%) in females (
2 = 71.58, P < 0.001). Lesions of the lower limb were primarily found in females: 130 (80%) compared to 32 (20%) in males. At diagnosis, 508 (88%) patients had a primary lesion only, 14 (2%) had local node involvement and three (1%) patients had distant spread. Clinical stage data were missing for 53 (9%) patients. Median Breslow thickness for all lesions was 1.10 mm (range 0.05–16.0 mm). For the 14 patients with node involvement at diagnosis, the median Breslow thickness was 1.90 mm (range 0.50–5.70). Data were missing for 70 lesions; the cause for this, in part, due to incomplete data capture at one particular hospital in the early stages of data collection that was rectified once it was identified.
Excision of the lesions
Location of excisions and excising surgeon..
In total, 95 (16.4%) lesions were removed in general practice. For hospital-excised lesions, dermatologists excised 316 (54.7%), surgeons (unspecified) 107 (18.5%), other hospital doctors 44 (7.6%), maxillofacial surgeons 6 (1.0%), plastic surgeons 5 (0.9%), hospital-based GPs 2 (0.3%) and 3 (0.5%) unknown. Of the lesions excised in general practice, 56 GPs (60%) excised one lesion only, with the remainder excising two or more lesions. The proportion of GP-excised lesions fluctuated between 1993 and 1997 (Figure 1). However, from 1997 onwards there was a decline in the number of lesions excised in general practice.
|
Margins and quality. The cross tabulation of margin of excision and Breslow thickness is shown in Table 1. Patients with excisions with adequate margins were younger [54.8 years (SD = 16.5)] than those with margins narrower than the guidelines [60.4 years (SD = 19.0)] (t = –3.17, P = 0.002) (Table 2). There was also a difference between anatomical location (
2 = 21.15, P < 0.001) with the majority of lesions on the trunk, upper and lower limb excised with adequate margins, but only 38% of head and neck lesions excised with adequate margins. A difference in quality was found between those who excised the lesion, with dermatologists excising more lesions with adequate margins than GPs and surgeons (includes surgeons, plastic surgeons, maxillofacial surgeons, other hospital doctors and unknown doctors) (2 = 16.94, P < 0.001). There were some differences in quality between where it was excised (hospital or general practice) with statistical significance reaching the 10% level (2 = 3.82, P = 0.051). There was no difference in terms of Breslow thickness, deprivation score or gender.
|
|
|
Age, anatomical location, place of excision and the excising surgeon were entered into the logistic regression model. Only anatomical location and the excising surgeon were independently associated with quality of the lesion, indicating that head and neck lesions and those excised by a surgeon were more likely to be excised inadequately (Table 3).
General practice excisions. Melanomas excised in general practice were on average from a younger group than those excised in hospital (t = 2.43, P = 0.015) (Table 4). Gender, deprivation score, anatomical location and Breslow thickness did not differ between the two groups. Of the 95 excisions in general practice, there was only evidence of 49 (52%) being referred on to hospital. There were no differences between patients being referred to hospital or not with respect to gender, deprivation or Breslow thickness. However, referred patients were younger [50.8 years (SD = 17.2)] than those not referred [59.3 years (SD = 19.8) (t = 2.25, P = 0.027)].
|
Time to diagnosis
The number of patients for whom time to diagnosis was calculated was 340 (59%) for GP delay, 402 (70%) for referral delay, 439 (76%) for secondary care delay and 358 (62%) for total delay (Table 5). The number of patients for whom we were able to calculate total delay was greater than for GP delay since more patients had data on both GP date and biopsy date than for GP date and referral date. Time from first presentation to biopsy (total delay) was considerably shorter overall for excisions in general practice (median 12 days) than for secondary care excisions (median 41 days) (P < 0.001). Patients with missing time to diagnosis data were on average slightly older than with data.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Findings within the context of the literature
Gender and age and anatomical location. In terms of gender, age and anatomical location, our sample was similar to other published data.24–28 Our findings diverge slightly from most of the reported literature in that men developed melanoma at slightly earlier age than women.24,25
GP excisions. These data show that one-sixth of lesions were removed in general practice. This figure is in the middle of the range of proportions in previously reported case series.7–13 This is despite the fact that most national guidelines state that suspected melanoma should be removed by specialists in secondary care. Although we did not assess preoperative diagnosis (probably only a minority were suspected to be melanomas prior to excision7,10,11), most of the GPs in this study excised only one melanoma, showing that melanoma remains an infrequent challenge for GPs performing surgery. However, we do not know from these data whether GPs were ignoring guidelines or getting the diagnosis wrong. The fact that the proportion of lesions removed in general practice in latter years of the study decreased somewhat may suggest greater adherence to guidelines. Our findings corroborate the findings of previous studies, but on a much larger sample, in that patients having GP excisions were younger8,11 and had no difference in Breslow thickness.7,10,11 The unexpected finding that only 52% of patients were referred after excision in primary care is worrying; all should be referred for further assessment.
Margins. A number of patients had excision margins narrower than defined in the guidelines. The guidelines are based on large, randomized controlled studies showing disease-free or survival benefits from defined margins. The guidelines accept that surgical considerations can be taken into account when defining margins, and in certain circumstances narrower margins may be necessary. With head and neck lesions, narrower margins may be necessary because of surgical and cosmetic considerations, and these margins are appropriate in the clinical circumstances; these are lesions which really should not be excised in primary care. With some older patients, medical status could have influenced surgical margins because major surgical procedures were inappropriate. We found no evidence of poorer quality excisions in general practice, despite several previous reports of less than satisfactory margins in this group.11–13 We can conclude that dermatologists are better at adequately excising lesions and this is not necessarily attributable to them excising thinner lesions than GPs and surgeons. While we acknowledge that the satisfactory nature of excision involves more than just width, for example the correct direction of incision is important, we have no data about this.
Time to diagnosis. This is the first large-scale reported UK study of time to diagnosis in melanoma. It is the first to look at time to diagnosis following initial presentation using medical records (as reported by clinicians) as a source of data rather than historical memory.29 The techniques for sampling and analysis overcome previous methodological difficulties with the literature on delays and time to diagnosis.16–20,21,30 Our data suggest that, predictably, GP excisions are diagnosed sooner after first presentation of the lesion, although this may be ‘accidental’. Of more importance, the findings put the time to diagnosis for patients diagnosed through different pathways into perspective, showing that the slightly longer median times for secondary care diagnoses are unlikely to have a difference in outcome; indeed, primary care excisions were slightly thicker. What is more concerning is the minority of patients with very long time to diagnosis. These are the patients who potentially have most to gain from being diagnosed earlier.
Strengths and limitations of the study
The findings reported in this paper are based on a comprehensive cohort of data from a large geographical area over a long time period, using comprehensive data collection. The analysis is based on a large number of cases, ensuring generalizability of data and sufficient statistical power. However, there are limitations to this study:
- Definition of lesions in the data set. Our data set did not contain data relating to lesions of the eye or of the vulva, other non-cutaneous lesions or in situ lesions. While this is in keeping with much of the current literature, comparisons with other data sets that may have contained these lesions need caution. Most studies look at invasive melanoma only because it is a disease that can be easily defined. Additionally, as part of this work, we were considering the effects of treatment on mortality (data analysis not presented in this paper), therefore including in situ melanoma in the data would have confused the picture because treatment of in situ melanoma has a 100% cure rate.
- Missing data. There were some missing data throughout this data set. While effort was put into trying to achieve completeness of the data, it is inevitable with data sets of this nature that some data are missing. We cannot be certain that the missing data introduces bias into the findings; hence, all must be interpreted with some caution.
- Data about doctors. We did not have any data on the surgical or dermatological expertise of the GPs (e.g. whether they were a GP with a special interest); we are aware that they do vary in their surgical skills and this may be reflected in their activity described in this paper.
- The design of this study does not allow general conclusions about who performs better, as it follows daily practice referral patterns that may be subject to all kinds of unknown biases, for example, GPs are likely to refer obvious melanomas and biopsy the less suspicious ones.
- It may be difficult to extrapolate the findings from North Wales to other settings. There is wide local variation in services; for example, in North Wales very few lesions are excised by plastic surgeons.
Issues for policy, practice and research
Our data show that there is no evidence that lesions excised in general practice have a worse prognosis given Breslow thickness at diagnosis than those excised in secondary care; however, given some of the limitations described above, we do not feel that we can advocate any change to the referral guidance as a result. We cannot, for sure, conclude that excisions by GPs are ‘safe’, especially since we may have missed some lesions not sent for histology. We therefore recommend vigilance on the part of GPs in recognizing suspected cases and referring them appropriately, to continue to send excised pathological specimens to pathology, and ensure that all confirmed primary care melanoma excisions are referred to secondary care. However, most GPs are not exposed to enough melanomas in daily practice to be confident in diagnosis, and there may therefore be a need for more training in the recognition and management of skin cancer. We also found no difference between primary care (with shorter times to diagnosis) and secondary care excisions and Breslow thickness, suggesting that efforts to minimize times to diagnosis after initial presentation may, at best, be only influencing psychosocial outcomes. The effectiveness of the urgent skin cancer referral guidance needs fuller evaluation.
|
Declaration |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
Funding: No specific project funding was received. Kerry Hood is funded by the Wales Office for Research and Development.
Ethical approval: The study was approved by the chair of the North East Wales National Health Service Trust research ethics committee.
Conflicts of interest: None declared.
|
Acknowledgments |
|---|
We would like to acknowledge the members of the North Wales Melanoma Group: Dr Richard Williams and Dr Diane Williamson (Ysbyty Glan Clwyd), Dr Andrew Macfarlane (Ysbyty Gwynedd) and Dr Robert Lister and Dr Sue Lewis-Jones (Wrexham Maelor Hospital). We would like to thanks all clinicians who have continued to provide data for the study. We thank Dr Nefyn Williams for constructive comments on drafts of this paper.
|
Notes |
|---|
Neal RD, Cannings-John R, Hood K, Sowden J, Lawrence H, Jones C and Jones J. Excision of malignant melanomas in North Wales: effect of location and surgeon on time to diagnosis and quality of excision. Family Practice 2008; 25: 221–227.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionDeclarationReferences |
|---|
1 Thomas JM, Giblin V. Cure of cutaneous melanoma. BMJ (2006) 332:987–988.
2 Breslow A. Thickness, cross-sectional areas and depth of invasion in the prognosis of cutaneous melanoma. Ann Surg (1970) 172:902–908.[Web of Science][Medline]
3 North and South Wales Melanoma Group. Management of Primary Malignant Melanoma of the Skin (1998) Swansea: Guidelines for Wales.
4 Department of Health. Referral Guidelines for Suspected Cancer (2000) London: Department of Health.
5 NICE. Referral Guidelines for Suspected Cancer (2005) London: NICE.
6 Roberts DLL, Anstey AV, Barlow RJ, et al. UK guidelines for the management of cutaneous melanoma. Br J Dermatol (2002) 146:7–17.[Web of Science][Medline]
7 Khorshid SM, Pinney E, Newton Bishop JA. Melanoma excision by general practitioners in North-East Thames region, England. Br J Dermatol (1998) 138:412–417.[CrossRef][Web of Science][Medline]
8 McKenna DB, Marioni JC, Lee RJ, Prescott RJ, Doherty VR. A comparison of dermatologists’, surgeons’, and general practitioners’ surgical management of cutaneous melanoma. Br J Dermatol (2004) 151:636–644.[CrossRef][Web of Science][Medline]
9 Williams RB, Burdge AH, Lewis Jones S. Skin biopsy in general practice. BMJ (1991) 303:1179–1180.
10 Jackson AM, Morgan DR, Ellison R. Diagnosis of malignant melanoma by general practitioners and hospital specialists. Postgrad Med J (2000) 76:295–298.
11 Herd RM, Hunter JAA, McLaren KM, Chetty U, Watson AC, Gollock JM. Excision biopsy of malignant melanoma by general practitioners in south east Scotland 1982-91. BMJ (1992) 305:1476–1478.
12 McWilliam L, Knox F, Wilkinson N, Oogarah P. Performance of skin biopsies by general practitioners. BMJ (1991) 303:1177–1179.
13 Schofield JK, O'Neill E, Tatnall FM. Dermatological surgery in general practice: management of malignant skin tumours. J Dermatolog Treat (1993) 4:153–155.[CrossRef]
14 Doherty VR, MacKie RM. Reasons for poor prognosis in British patients with cutaneous malignant melanoma. BMJ (1986) 292:987–989.
15 Baade PD, English DR, Youl PH, McPherson M, Elwood JM, Aitken JF. The relationship between melanoma thickness and time to diagnosis in a large population-based study. Arch Dermatol (2006) 142:1422–1427.
16 Brochez L, Verhaeghe E, Bleyen L, Naeyaert J-M. Time delays and related factors in the diagnosis of cutaneous melanoma. Eur J Cancer (2001) 37:843–848.[CrossRef][Web of Science][Medline]
17 Blum A, Brand CU, Ellwanger U, et al. Awareness and early detection of cutaneous melanoma: an analysis of factors related to delay in treatment. Br J Dermatol (1999) 141:783–787.[CrossRef][Web of Science][Medline]
18 Oliviera SA, Christos PJ, Halpern AC, Fine JA, Barnhill RL, Berwick M. Patient knowledge, awareness, and delay in seeking medical attention for malignant melanoma. J Clin Epidemiol (1999) 52:1111–1116.[CrossRef][Web of Science][Medline]
19 Krige JE, Isaacs S, Hudson DA, et al. Delay in the diagnosis of cutaneous malignant melanoma. Cancer (1991) 68:2064–2068.[CrossRef][Web of Science][Medline]
20 Betti R, Vergami R, Tolomio E, Santambrogio R, Crosti C. Factors of delay in the diagnosis of melanoma. Eur J Dermatol (2003) 13:183–188.[Web of Science][Medline]
21 Rampen FHJ, Rumke Ph, Hart AAM. Patients’ and doctors’ delay in the diagnosis and treatment of cutaneous melanoma. Eur J Surg Oncol (1989) 15:143–148.[Web of Science][Medline]
22 Richard MA, Grob JJ, Avril MF, et al. Delays in diagnosis and melanoma prognosis (I): the role of patients. Int J Cancer (2000) 89:271–279.[CrossRef][Web of Science][Medline]
23 Allgar VL, Neal RD. Delays in the diagnosis of six cancers: analysis of data from the National Survey of NHS Patients: cancer. Br J Cancer (2005) 92:1959–1970.[CrossRef][Web of Science][Medline]
24 Jones WO, Harman CR, Ng AK, Shaw JH. Incidence of malignant melanoma in Auckland, New Zealand: highest rates in the world. World J Surg (1999) 23:732–735.[CrossRef][Web of Science][Medline]
25 Saxe N, Hoffman M, Krige JE, Sayed R, King HS, Hounsell K. Malignant melanoma in Cape Town, South Africa. Br J Dermatol (1998) 138:998–1002.[CrossRef][Web of Science][Medline]
26 Katalinic A, Kunze U, Schafer Y. Epidemiology of cutaneous melanoma and non-melanoma skin cancer in Schleswig-Holstein, Germany: incidence, clinical subtypes, tumour stages and localization. Br J Dermatol (2003) 149:1200–1206.[CrossRef][Web of Science][Medline]
27 Marrett LD, Nguyen HL, Armstrong BK. Trends in the incidence of cutaneous malignant melanoma in New South Wales, 1983–1996. Int J Cancer (2001) 92:457–462.[CrossRef][Web of Science][Medline]
28 Cancer Research UK. CancerStats Key Facts on Skin Cancer. http://info.cancerresearchuk.org/cancerstats/types/skin/incidence/#age (last accessed on 14 May 2008).
29 Roberts RO, Bergstrahl EJ, Schmidt L, Jacobsen SJ. Comparison of self-reported and medical record health care utilization measures. J Clin Epidemiol (1996) 49:989–995.[CrossRef][Web of Science][Medline]
30 Schmidt-Wendtner MH, Baumert J, et al. Delay in the diagnosis of cutaneous melanoma: an analysis of 233 patients. Melanoma Res (2002) 12:389–394.[CrossRef][Web of Science][Medline]
31 National Assembly for Wales. Welsh Index of Multiple Deprivation (2002) Cardiff: National Assembly for Wales.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||