Observational study of type of surgical training and outcome of definitive surgery for primary malignant melanoma

BMJ 2002; 325 doi: (Published 30 November 2002) Cite this as: BMJ 2002;325:1276
  1. Rona M MacKie, Leverhulme professorial research fellow (R.M.Mackie{at},
  2. Caroline A Bray, medical statistician,
  3. David J Hole, professor of epidemiology and biostatistics
  1. Department of Public Health, University of Glasgow, Glasgow G12 8RZ
  1. Correspondence to R M MacKie
  • Accepted 17 April 2002

The incidence of primary cutaneous malignant melanoma continues to rise,1 coinciding with narrower excision margins of normal skin being recommended around primary melanomas. 2 3 The bulk of surgery for primary melanoma is now done on an outpatient basis under local anaesthesia. This change has occurred at a time when training in dermatological surgery has developed, leading to a much higher proportion of excisions of primary melanoma being done by dermatologists. In 1979 in the west of Scotland only 3% of all primary melanomas were removed by dermatologists. By 1998 this figure had risen to 40%. Plastic surgeons now excise 26% of primary melanomas compared with 65% 20 years ago, and general surgeons excise 34% compared with 32%.

Details of melanomas treated, by surgical groups adjusted for type of surgical training. Values are numbers (percentages) unless stated otherwise

View this table:

We aimed to establish if the change in type of surgeon removing primary cutaneous malignant melanoma has affected the prognosis, and also whether any evidence exists for a specialist treatment effect such as has been observed for breast cancers, with better outcomes for surgeons carrying out breast cancer surgery regularly.4

Participants, methods, and results

We identified 4159 melanoma patients from the files of the Scottish melanoma group. All patients had had their primary melanoma removed between 1979 and 1998. We divided the surgeons performing the definitive excision of the primary melanoma into dermatological, plastic surgery, or general surgery training. We recorded age, sex, tumour thickness, presence of ulceration, and maximum diameter of the primary tumour and noted mortality and cause of death up to 1998. We also looked at the effect within the three surgical groups of treating up to six or more than six primary melanomas annually.

An average of 10 years' follow up information was available for all patients. To test for an association between tumour thickness and type of surgical experience we used the χ2 statistic for trend, aggregated over the period of diagnosis. We used the Cox proportional hazards model to compare the survival of patients in relation to surgical experience,5 with adjustment for thickness, ulceration, and maximum diameter of tumour and sex, age, and deprivation category of patients.

The table shows the division of patients by tumour thickness, ulceration, maximum diameter of primary melanoma, and outcome by surgical training. Dermatologists treated a significantly higher proportion of thin melanomas (P<0.001). The proportion of ulcerated melanomas was higher in the plastic surgery group than in the dermatological group (P<0.001) and higher in the general surgical group than the plastic surgery group (P<0.001).

After adjustment for thickness, the best outcome was in the dermatological surgeon treatment group (P=0.008). Statistical significance was lost when adjustment was made for ulceration and then for maximum diameter. We found no evidence that surgeons in any of the three categories who performed more than six primary melanoma excisions annually had better outcomes than those who performed fewer excisions.


Survival of melanoma patients does not depend on the surgical background of the person removing the primary tumour. The object of this study was to provide an evidence base for primary care guidelines on appropriate specialist referral. The data show that the growth in dermatological surgeons excising primary melanomas has had no adverse affect on patient outcome. We found no evidence that any type of surgeon performing excisions of primary melanomas regularly had a better outcome than those who carried out fewer excisions, possibly because wide local excision is a relatively simple procedure. We therefore provide an evidence base to recommend referral of suspected primary melanomas to the dermatological, plastic surgery, or general surgical service with the shortest surgical waiting time.


Contributors: RMM collected the data, wrote the paper, and is the guarantor for the study. CAB and DJH analysed the data.


  • Funding The Scottish Melanoma Group is funded by NHS Scotland. RMM receives financial support from the Leverhulme Trust.

  • Conflict of interest None declared.


View Abstract