British Journal of Plastic Surgery (2002), 55, 616-622

9 2002 The BritishAssociation of Plastic Surgeons.Published by Elsevier Science Ltd. All rights reserved,
doi: 10.1054/bjps.2002.3948

BRITISH JOURNAL OF [ ~ I PLASTIC SURGERY

Incomplete excision of basal cell carcinoma: a prospective multicentre audit

E Kumar*?$, S. Watson*, A. N. Brain*-~, R J. Davenport*5;, L. J. McWilliam*, S. S. Banerjee? and D. L. Bisset5;

* University Hospital of South Manchester; "~Christie Hospital Withington, Manchester," and
5;Royal Bolton Hospital, Bolton, UK

SUMMARY. The audit of incomplete excision of basal cell carcinoma can be used as a parameter for clinical gover-
nance in plastic surgery units. However, there are no national standards, and all the previous reports from the UK have
been retrospective and from regional units only. This 1 year prospective audit was undertaken simultaneously in the
plastic surgery units of three different categories of hospital: a regional plastic surgery unit (University Hospital of
South Manchester), a supraregional cancer hospital (Christie Hospital) and a district general hospital (Royal Bolton
Hospital). A total of 757 lesions were excised from 600 patients, with 34 lesions (4.5%) being incompletely excised.
The incidences of incomplete excision were similar in the regional unit (3.2%) and the district general hospital (3.1%),
but higher in the supraregional cancer hospital (7.5%). The commonest site for incomplete excision was the eyebrow,
followed by the postauricular area, the nose and the temple. There were no statistically significant differences in the
distributions of the age and sex of the patients, the site and size of the lesions or the methods of repair between the three
hospitals. However, there were significant differences in the distribution of syndromal, multiple and recurrent lesions,
the grade of surgeon, and the clinical and histological subtypes. When the various confounding factors were adjusted
by logistic regression, the variables most likely to have affected the incidence of incomplete excision were found
to be grade of surgeon, minimal excision margin and histological subtype. 9 2002 The British Association of Plastic
Surgeons. Published by Elsevier Science Ltd. All rights reserved.

Keywords:basal cell carcinoma, incomplete excision, grade of surgeon, excision margin, histological subtype.

Accountability and continuous improvement of quality are Patientsand methods

the main features of clinical govemance.I There is a need to
develop processes to monitor continuously and to improve
the quality of care provided by clinical departments. In
plastic surgery it is difficult to identify clinical performance
indicators that are easily quantifiable. Excision of basal cell
carcinoma (BCC) is one of the commonest operations per-
fomled in plastic surgery departments. All grades of sur-
geon perform this operation. The outcome of this procedure
is strongly related to completeness of excision. If the lesion
is completely excised, the recurrence rate is about 1%,2,3
whereas if it is incompletely excised, the average recur-
rence rate is 31%. 4 As the incidence of incomplete excision
can easily be monitored, it could be used as tool for clinical
governance.5 However, all the data published from the UK
have been retrospective surveys conducted in regional plas-
tic surgery departments. Prospective audits are needed to
collect data for benchmarking and to set national standards.
This prospective study was designed to quantify the
incidence of incomplete excision of BCC over a 1 year
period in three different categories of hospital: a regional
plastic surgery unit, a supraregional cancer hospital and a
district general hospital. The data were further analysed to
identify variables affecting the incidence of incomplete
excision.
This paper was presented at the Summer Meeting of the British
Association of Plastic Surgeons, Stirling, UK, 4-6 July 2001.
Three hospitals participated in this prospective audit: the
University Hospital of South Manchester, which is a
regional plastic surgery unit, the Christie Hospital, which
is a supraregional cancer hospital, and the Royal Bolton
Hospital, which is a district general hospital. All patients
who underwent excision of BCC in the plastic surgery
departments at these hospitals between March 1999 and
February 2000 were included in the audit.
Three forms were designed for data collection. The
first was completed by the operating surgeon at the time
of surgery, and included data on the age and sex of the
patient, the site and size of the lesion, whether it was a
recurrent lesion or part of a syndromal lesion, clinical
type, method of repair and grade of surgeon. The
second form was completed by the histopathologist, and
included data on the histological type of the lesion
and the margin of clearance. Incomplete excision was
defined as the presence of tumour at any excision margin.
The third form was completed by the surgeon for incom-
pletely excised lesions, and included data on the further
management.
Statistical analyses were performed using the two-
sample /-test, Fisher's exact test, the X2 test, one-way
analysis of variance and logistic regression where appro-
priate. A P value of <0.05 was taken as significant. For
statistical analysis, each lesion was considered to be a
separate entity.

616
Incomplete excision of BCC 617

Results Table 1 Site distribution of basal cell carcinoma

Altogether, 600 patients underwent excision of BCC during Hospital Head and neck Trunk and limb
the study period. Some patients had more than one lesion (number of incomplete (number of incomplete
excised. The total number of lesions excised was 757, of excisions) excisions)
which 34 were incompletely excised (25 at the lateral University Hospital of 369 (14) 66 (0)
margin, 5 at the deep margin and 4 at both margins). South Manchester
The overall incidence of incomplete excision was Christie Hospital 194 (16) 32 (1)
4.5%. This incidence was similar at the University Royal Bolton Hospital 81 (3) 15 (0)
Hospital of South Manchester and the Royal Bolton total 644 (33) 113 (1)
Hospital (3.2% and 3.1%, respectively), whereas it was
7.5% at the Christie Hospital.
Table 2 Site distribution of incompletely excised lesions

Distribution of multiple and syndromal lesions Site Numberof Numberof incompletely

lesions excised lesions(%)
The percentage of patients who had more than one lesion
eyebrow 10 2 (20.0)
was 13.7% at the University Hospital of South postauricular 20 3 (15.0)
Manchester, 13.6% at the Royal BoRon Hospital and nose 95 9 (9.5)
26.9% at the Christie Hospital. In the total sample, the temple 88 7 (8.0)
incidence of incomplete excision in patients with multi- outer canthus 15 1 (6.7)
ple lesions was 4.3%. Similarly, there was a higher inci- lower eyelid 34 2 (5.9)
pinna 21 I (4.8)
dence of syndromal lesions at the Christie Hospital (8.8%) preauricular 24 1 (4.2)
than at the University Hospital of South Manchester scalp 25 1 (4.0)
(0.2%) and the Royal BoRon Hospital (0%). At the upper lip 26 1 (3.8)
Christie Hospital the incidence of incomplete excision in lower limb 31 1 (3.2)
inner canthus 38 1 (2.6)
patients with syndromal lesions was 10%. cheek 84 2 (2.4)
forehead 102 2 (2.0)

Age and sex distribution

The overall age of the patients ranged from 17 years to slightly higher in larger lesions (6.18% versus 4.5%), but
100 years. The patients with completely excised lesions this was not statistically significant (P = 0.45).
had a mean age of 70.6 years, while those with incom-
pletely excised lesions had a mean age of 72.1 years.
This difference was not statistically significant. There Distribution of recurrent lesions
was no difference in the age distribution of the patients
Overall, 61 lesions were recurrent and the rest were new
between the three hospitals.
lesions. The incidence of incomplete excision was 3.7%
Altogether, 50.2% of lesions were in male patients and
in the new lesions, but it was 13.1% in the recurrent
49.8% were in female patients. There was a slightly
lesions (Fisher's exact test: P = 0.004).
higher incidence of females in the group of patients with
The proportion of recurrent lesions was almost the
incompletely excised lesions (55.9%), but this was not
same at the University Hospital of South Manchester
statistically significant (Fisher's exact test: P = 0.488).
(7.1%) and the Royal Bolton Hospital (8.1%), but it was
There was no statistical difference in the sex distribu-
slightly higher at the Christie Hospital (10.6%), although
tion between the three hospitals.
statistically this difference was not significant.

Site distribution
Overall, 644 lesions were in the head and neck area, and
113 were located on the trunk and limbs. There was no
difference in the distribution between the three hospitals.
Overall, 33 of the incompletely excised lesions were in
the head and neck area, with only one in the rest of the
body (Table 1). The highest incidence of incomplete
excision was at the eyebrows (20%), followed by the
postauricular area (15%), the nose (9.5%) and the temple
(8%) (Table 2).
Distribution of margin of excision
The surgeons decided the excision margin on the basis of
clinical need in each case. The margin was noted by the
surgeons in 666 lesions. The distribution varied signifi-
cantly between the three hospitals (X2(4)=27.05;
P<0.001).
The incidence of incomplete excision was 4.7% in the
1-2.5mm margin group, 4.1% in the 3-4ram margin
group and 2.9% in the 5mm or more margin group
(Table 3). Statistically, this variation was not significant.

Size distribution
In total, 86.2% of the lesions were small (up to 2 cm
maximum diameter), the rest were larger. The distribu-
tion was similar at the three hospitals (X2(2)--0.37;
P - 0 . 8 3 ) . The incidence of incomplete excision was
Method of repair
Direct closure was the most common method. There was
no significant difference in the distribution of the differ-
ent methods of repair between the three hospitals,
618 British Journal of Plastic Surgery

although the Christie Hospital used slightly more split-
skin grafts at the expense of full-thickness grafts.
There was one incomplete excision out of the four
lesions allowed to heal by secondary intention. Ignoring
this group because of its very small size, the incidence of
incomplete excision was lowest (3.1%) in lesions
repaired by direct closure and highest in those repaired
by split-skin grafting (10.3%) (Table 4). The incidence of
incomplete excision was significantly lower in the direct-
closure group than the other methods when considered
together (Fisher's exact test: P - - 0.025).
Grade of surgeon
The surgeons were divided into four groups: junior trainee
(senior house officer), senior trainee (specialist registrar),
non-consultant career grade (associate specialist and clinical
assistant) and consultant. All grades of surgeon were
involved in the excision of BCC at the University Hospital
of South Manchester and at the Christie Hospital, whereas at
the Royal Bolton Hospital all excisions were done by either
the consultant or the associate specialist. The incidence of
incomplete excision was highest for the senior house offi-
cers, and lowest for the non-consultant career grades
(X2(3) = 8.41; P = 0.038) (Table 5).
Table 6 shows the ratio of wounds closed by other
methods to those closed directly (complexity ratio) for
the individual grades o f surgeon.
Distribution of clinicai types
Clinical type was arbitrarily classified into five groups
based on the predominant clinical appearance. It was
recorded in 709 lesions. At all three hospitals the most
common lesions were nodular. There was a higher inci-
dence of morphoeic lesions at the Christie Hospital
(15.7%) than at the other hospitals (8.6% at the
University Hospital of South Manchester and 8.7 at
the Royal Bolton Hospital) (Table 7). The difference in the
distribution of clinical types between the three hospitals
was statistically significant (X2(8)= 19.74; P = 0.011).
The incidence of incomplete excision was highest in mor-
phoeic lesions (Table 7). Though clinically significant, the
numbers in each group were too small for statistical analysis.

Table 3 Distribution of excision margin

Hospital 1-2.5 mm (%) 3--4 mm (%) 5 mm or more (%)

University Hospital of South Manchester 135 (35.1) 183 (47.5) 67 (17.4)

Christie Hospital 63 (33.5) 64 (34) 61 (32.4)
Royal Bolton Hospital 37 (39.8) 47 (50.5) 9 (9.7)
total 235 (35.3) 294 (44.1) 137 (20.6)
incomplete excisions (percentageof total excisions) 11 (4.7) 12 (4.1) 4 (2.9)

Table 4 Distribution of method of repair

Hospital Direct closure Flap Full-thickness skin Split-skin Secondary healing
graft graft

University Hospital of South Manchester 274 78 61 19 3

Christie Hospital 149 35 23 17 2
Royal Bolton Hospital 63 15 15 3 0
incomplete excisions (%) 15 (3.1) 9 (7.0) 5 (5.1) 4 (10.3) 1 (20.0)

Table 5 Distribution of grades of surgeon

Hospital Consultant Non-consultant Specialist registrar Senior house officer

career grade

University Hospital of South Manchester 43 179 95 118

Christie Hospital 22 49 111 44
Royal Bolton Hospital 43 53 0 0
total 108 281 206 162
incomplete excisions (percentage of total for grade) 5 (4.6) 5 (1.8) 13 (6.3) 11 (6.8)

Table 6 Incidence of incomplete excision and complexity ratio by grade of surgeon

Surgeon Incidence of incomplete Direct closure (n) Other methods Other methods/
excision ( %) of closure (n ) direct closure

consultant 4.6 62 46 0.7

associate specialist 1.8 146 99 0.7
specialist registrar 6.3 122 84 0.7
senior house officer 6.8 125 37 0.3
clinical assistant 5.5 31 5 0.2
Incomplete excision of BCC 619

Table 7 Distribution of clinical types

Hospital Nodular Ulcerated Superficial Morphoeic Other
University Hospital of South Manchester 175 129 69 36 11
Christie Hospital 84 39 39 31 4
Royal Bolton Hospital 34 30 14 8 6
total 293 198 122 75 21
incomplete excisions (percentage of clinical type) 13 (4.4) 9 (4.5) 4 (3.3) 5 (6.7) 0

Table 8 Distribution of histological subtypes

Hospital Nodular Morphoeic Micronodular Superficial Mixed Others
University Hospital of South Manchester 219 13 1 32 21 0
Christie Hospital 23 41 3 26 62 8
Royal Bolton Hospital 45 9 9 10 23 0
total 287 63 13 68 106 8
incomplete excisions (percentage of 11 (3.8) 6 (9.5) 1 (7.7) 2 (2.9) 5 (4.7) 0 (0)
histological subtype)

Table 9 Adjustment of potential confounding factors by logistic
regression
Potential confounding factor P for hospital outcome,
i.e. clearance
age
sex
syndromal lesion
recurrent lesion
overall site
size
method of repair
grade of surgeon
excision margin
clinical type
histological subtype
Distribution of histological subtypes
For this study, BCC was classified into five histological
subtypes: nodular, micronodular, superficial, morphoeic
and infiltrative/mixed, based on the classification pro-
posed by Rippey.6 Histological subtypes were reported in
537 lesions. The distribution of subtypes varied markedly
between the three hospitals (X2(10) = 213.67; P < 0.001).
Overall, the incidence of incomplete excision was
lowest in superficial BCC and highest in morphoeic BCC
(Table 8). It was not possible to perform a valid statistical
analysis because of the small numbers in each group.
Adjustment for confounding factors
Logistic regression was carded out to see whether there
was a significant relationship between hospital and com-
plete excision rate after adjusting for potential confound-
ing factors. Histological subtype, excision margin and
grade of surgeon account for a significant proportion of
the variation in the incidence of incomplete excision
between the three hospitals (Table 9).
Management of incompletely excised lesions
Two patients with incompletely excised lesions died dur-
ing the early postoperative period as a result of unrelated
conditions. Of the remaining incompletely excised lesions,
50% were managed by regular follow-up and the rest were
treated with further intervention (Table 10). The average
age of the patients managed by regular follow-up was
77 years (range: 44-95 years; median: 77.5 years) and that
0.036 of the patients who underwent further intervention was
0.042 68 years (range: 33-83 years; median: 70.5 years).
0.057
0.062
0.042
0.034 Discussion
0.044
0.14 Clinical governance has been defined as a system
0.13 through which health-service organisations are account-
0.048 able for continuously improving the quality of their ser-
0.46 vice and safeguarding high standards of care. 1 In order to
assess the quality of care or performance, reliable quanti-
tative assessments are needed. These will help to identify
poor performance and to set in place corrective measures.
In plastic surgery, it is difficult to identify clinical perfor-
mance indicators that are easily quantifiable. The inci-
dence of incomcolete excision of BCC, however, is one
such parameter.J
The incidence of incomplete excision of BCC varies
markedly in the world literature from the 0.7% reported
by Emmett and Broadbent in Australia 7 to the 50%
reported by Rakofsky from the USA. 8 In the UK,
the reported incidence varies from 4.7% to 19%
(Table 11).2'5'9-12 The 19% reported by Taylor and Barisoni
can be excluded from the comparison because in more
than half their lesions there was no comment on the com-
pleteness of excision. 9 All these reports are retrospective.
Our overall incidence of 4.5% compares well with the
published data. As in other published series, the excision
was incomplete at the lateral margin in the majority of
incomplete excisions. 2"5'1~,12The incidences of incomplete
excision were similar at the district general hospital and
the regional plastic surgery unit (3.1% and 3.2%, respec-
tively). The incidence was higher (7.5%) at the suprare-
gional cancer hospital. The data were analysed further to
see whether we could identify the reasons for this.
The percentage of patients with multiple lesions was
similar at the University Hospital of South Manchester
and the Royal Bolton Hospital, but was significantly
620 British Journal of Plastic Surgery

Table I0 Management of incompletely excised lesions

University Hospital of Christie Hospital Royal Bolton Hospital Total
South Manchester

incomplete excisions 14 17 3 34
further surgical excision 4 7 1 12
Mohs' surgery 1 1 1 3
radiotherapy 1 0 0 1
regular follow-up 6 9 1 16
died 2 0 0 2

Table 11 Incidence of incomplete excision of basal cell carcinoma (UK literature)

Author(s) Year Incidence of incomplete excision (%) Total number of lesions

Taylor and Barisoni9 1973 (19) 388

Richmond and Davie12 1987 67 (7.9) 850
Park and Watson2 1994 23 (10.7) 215
Griffithsn 1999 99 (7.1) 1392
Kumar et al5 2000 41 (4.7) 879
Schreuder and Powell1~ 1999 7 (13.7) 51

higher at the Christie Hospital. As the incidence of
incomplete excision in multiple lesions was almost the
same as that in single lesions, this does not seem to be an
important variable. However, the situation is different
with syndromal lesions. There was a higher percentage of
syndromal cases at the Christie Hospital, and the inci-
dence of incomplete excision was higher in these lesions.
The age and sex distributions of the patients at each of
the three hospitals were similar, probably because they
came from the same geographical region. Similar to the
findings of Hauben et all3 and Fleischer et al, 14 there was
no statistically significant difference in the age or sex
distribution between the complete and the incomplete
excision groups.
BCC occurs predominantly in the sun-exposed areas
of the body: 85% of the lesions in our study were in the
head and neck area. The distribution was similar in all
three hospitals (Table 1). As in previous reports, the inci-
dence of incomplete excision was very low in the trunk
and extremities. The commonest site for incomplete exci-
sion was the eyebrow, followed by the postauricular area,
the nose and the temple. The majority of previous publi-
cations quote the nose and the midface as the sites of
highest incomplete excision, but, as discussed by Kumar
et al, 5 they do not compare the overall distribution of
BCC at each site separately. Griffiths, who calculated the
incidence of incompletely excised BCC at each site sepa-
rately, reported that the highest incidence occurred at the
eyebrow, followed by the inner canthus, the eyelids and
the nose. xl In a previous retrospective study at the
University Hospital of South Manchester the highest inci-
dence was found in the scalp, followed by the ear, canthi,
eyebrows and nose. 5 The eyebrow is a difficult area to
reconstruct, and there may be a natural reluctance to use
wider margins. In the postauricular area the lesions are
hidden, and late presentation and larger size may con-
tribute to a higher incomplete excision rate.
The overall size distribution was similar in all three
hospitals. The incidence of incomplete excision was
slightly higher in lesions larger than 2 cm. This was not
statistically significant. Kumar et al reported a similar
finding. 5 Fleischer et al also found that size was not an
independent risk factor, 14 although greater subclinical
growth was noted in larger lesions by Burg et a115 and
Epstein. 16
Burg et al also noted more extensive subclinical exten-
sion in recurrent lesions. 15 In our study, the rate of
incomplete excision was higher in recurrent lesions
(13.1%). Emmett and Broadbent also reported a higher
incidence of incomplete excision in recurrent lesions. 7
The proportion of recurrent lesions was only slightly
higher than average at the Christie Hospital (10.6%).
The prevalence o f recurrent lesions varies between
reports. 7,9,17 When comparing incomplete excision rates
in different hospitals, this variable may become
important.
The minimum margin advised for excision of BCC
varies widely from 2 to 10mm. 4'18'19 In small well-
defined nodulocystic lesions a margin of 2 - 4 m m is ade-
quate, whereas a larger margin is recommended for other
lesions. In this audit, the incidence of incomplete exci-
sion was lowest when the excision margin was 5 m m or
more. Paradoxically, more lesions at the Christie Hospital
were excised with a margin of 5 m m or more. The impor-
tance of data on excision margin is difficult to interpret.
In all three hospitals the margin was decided by the sur-
geon without a formal protocol. Larger margins were
excised in morphoeic, recurrent, large or ill-defined
lesions. These are precisely the lesions that are likely to
be incompletely excised. If all lesions were excised with
margins of 5 m m or greater, the rate of incomplete exci-
sion would be lower, but considerably more cosmetic and
functional problems would result, which, in the majority,
would be unnecessary.
Various plastic surgical methods were used to repair
the wounds following BCC excision. There was no sig-
nificant difference in the distribution of different methods
of repair between the three hospitals, although at the
Christie Hospital there were slightly more split-skin
grafts at the expense o f full-thickness grafts. Method of
Incomplete excision of BCC
repair gives an indication of the size and depth of the
lesion and the complexity of the surgery. Direct closure
was the most common method. The direct closure group
had the lowest incidence of incomplete excision, as these
lesions are usually small and in areas of relative skin sur-
plus. The highest rate of incomplete excision was in the
group treated with split-skin grafts; this method is mainly
used for larger lesions in areas where no surplus skin is
available. Kumar et al also reported their lowest incom-
plete excision rate in the direct closure group. 5 A high
incidence of incomplete excision in the direct closure
group is indicative of an inappropriate excision margins.
Allen suggested that incomplete excision is less fre-
quent when the surgery is performed by trained plastic
surgeons. 2~ In the series with the lowest incomplete exci-
sion rate (0.7%), reported by Emmett and Broadbent, all
surgery was performed in a private consultant plastic
surgery practice by two experienced surgeons. 7 However,
no publication compares the incomplete excision rate for
different grades of surgeon or for surgeons of differing
experience. In our audit, the incidence of incomplete
excision was lowest in cases treated by the associate spe-
cialist, who excised 245 lesions or 32.4% of all lesions.
In all the other grades there were multiple persons, and it
was not possible to analyse the outcome as a function of
the volume/experience of each surgeon. A positive rela-
tionship between operator volume and outcome has been
shown in many other clinical situations. 21-23 The lowest
incomplete excision rate (3.1%) was at the Royal Bolton
Hospital, where all surgery was performed by either the
consultant or the associate specialist. The incomplete
excision rate at the University Hospital of South
Manchester was almost the same at 3.2%, and here all
grades were involved. The highest incidence of incom-
plete excision was in the junior trainee group, but it was
only slightly higher than in the senior trainee group. Both
these rates are still towards the lower end of the spectrum
reported in the literature. These low rates were achieved
by appropriate delegation and supervision. The junior
trainees were given smaller lesions in areas of relative
skin surplus, and the defects were mostly closed directly,
as evidenced by the low complexity ratio in this group
(Table 6). A very high incomplete excision rate in the
trainee group may be due to inappropriate delegation
and/or inadequate supervision.
In deciding an appropriate margin, clinical subtyping
will generally be more valuable than histological subtyp-
ing, as a prior incision biopsy is rarely performed. We
adopted a simple classification: nodular, ulcerated, mor-
phoeic, superficial and others. Some lesions had more
than one feature but were classified on the basis of their
predominant clinical appearance. Though there are many
papers reporting the incidence of incomplete excision or
the incidence of recurrence in different histological sub-
types, there are no reports of these according to clinical
subtype. In this study, the most common clinical type was
nodular. The highest incomplete excision rate was in mor-
phoeic lesions (6.7%). This supports the prevalent clinical
view that morphoeic lesions require a wider margin.
In spite of various papers describing histological sub-
types of BCC, there is no agreed classification, and
histopathologists do not routinely report it. We adopted
621
the simplified classification described by Rippey.6 The
distribution of these histological subtypes was markedly
different across the three hospitals. The incidences of
morphoeic lesions and mixed lesions at the Christie
Hospital were very high, and the Royal Bolton Hospital
had a remarkably high incidence of micronodular lesions.
Was there a true difference or was this due to individual
histopathologists' subjective bias? This question could
only be resolved if all the specimens were reviewed by
one histopathologist, which was not possible in our study.
The nodular type was the most common type in the sam-
ple as a whole. Superficial BCC had the lowest incidence
of incomplete excision (2.9%). The incomplete excision
rate was also low in nodular lesions (3.8%). As in
o t h e r papers, 7'13'16'24 the incidence was highest in
morphoeic/infiltrative lesions. A higher percentage of
morphoeic/infiltrative lesions may result in a higher
incomplete excision rate. If an incision biopsy shows a
morphoeic/infiltrative or a micronodular subtype, a wider
excision margin is warranted. Similarly, a wider margin
should be excised in recurrent BCC if the original lesion
was one of these aggressive subtypes.
When adjustment was made for various confounding
factors by logistic regression (Table 9), the most likely
variables to affect the incidence of incomplete excision
were grade of surgeon, excision margin and histological
subtype. These three factors account for a significant pro-
portion of the variation in clearance between the three
hospitals. These variables should be critically scrutinised
if the incidence of incomplete excision is high.
The majority of authors recommend further wider
excision for incompletely excised lesions. 2'7'11"19 Some
advocate a policy of 'wait and s e e ' . 3'13'17 In our survey,
two patients died of unrelated causes and one patient
opted for radiotherapy; approximately half the remaining
lesions were managed with regular follow-up, and the
other half by further excision. The average age was
higher in follow-up group. This is similar to previous
observations by Kumar et al.s
Incomplete excision of BCC is an inferior result,
needing further treatment, and can easily be identified
by the histopathologist. As excision of BCC is a high-
volume procedure in all plastic surgery units, and is
performed by all grades of surgeon, regular ongoing audit
of incomplete excision rates will be a very useful perfor-
mance indicator for continuous quality assessment,
which is an integral part of clinical governance.
References
1. Scally G, Donaldson LJ. Clinical governance and the drive for
quality improvement in the new NHS England. BMJ 1998; 317:
61-5.
2. Park AJ, Strick M, Watson JD. Basal cell carcinomas: do they need
to be followed up? J R Coil Surg Edinb 1994; 39: 109-11.
3. Pascal RR, Hobby LW, Lattes R, Crikelair GE Prognosis of
'incompletely excised' versus 'completely excised' basal cell
carcinoma. Plast Reconstr Surg 1968; 41: 328-31.
4. Goldberg DP. Assessment and surgical treatment of basal cell skin
cancer. Clin Plast Surg 1997; 24: 673-86.
5. Kumar P, Orton CI, McWilliam LJ, Watson S. Incidence of
incomplete excision in surgically treated basal cell carcinoma:
a retrospective clinical audit. Br J Plast Surg 2000; 53: 563~.
622
6. Rippey JJ. Why classify basal cell carcinomas? Histopathology
1998; 32: 393-8.
7. Emmett AJJ, Broadbent GG. Basal cell carcinoma in Queensland.
Aust N Z J Surg 1981; 5l: 576-90.
8. Rakofsky SI. The adequacy of the surgical excision of basal cell
carcinoma. Ann Ophthalmol 1973; 5: 596-600.
9. Taylor A, Barisoni D. Ten years" experience in the surgical treat-
ment of basal cell carcinoma. Br J Surg 1973; 60: 522-5.
10. Schreuder F, Powell BW. Incomplete excision of basal cell carcino-
mas: an audit. Clin Perform Qual Health Care 1999; 7:119-20.
11. Griffiths RW. Audit of histologically incompletely excised basal
cell carcinomas: recommendations for management by re-
excision. Br J Plast Surg 1999; 52: 24-8.
12. Richmond JD, Davie RM. The significance of incomplete excision in
patients with basal cell carcinoma. Br J Plast Surg 1987; 40: 63-7.
13. Hauben DJ, Zirkin H, Mahler D, Sacks M. The biologic behavior of
basal cell carcinoma: part I. Plast Reconstr Surg 1982; 69: 103-9.
14. Fleischer AB, Feldman SR, Barlow JO, et al. The speciality of the
treating physician affects the likelihood of tumour free resection
margins for basal cell carcinoma: results from a multi-institutional
retrospective study. J Am mcad Dermato12001; 44: 224-30.
15. Burg G, Hirsch RD, Konz B, Braun-Falco O. Histographic surgery:
accuracy of visual assessment of the margins of basal cell epithe-
lioma. J Dermatol Surg 1975; 1: 21-4.
16. Epstein E. How accurate is the visual assessment of basal carci-
noma margins? Br J Dermatol 1973; 89: 37--43.
17. Griffith BH, McKinney P. An appraisal of the treatment of basal cell
carcinoma of the skin. Plast Reconstr Surg 1973; 51: 565-71.
18. Preston DS, Stern RS. Nonmelanoma cancers of the skin. N Engl J
Med 1992; 327: 1649~2.
19. Binstock JH. Recurrent basal cell carcinoma. Plast Reconstr Surg
1981; 67: 697-9.
20. Allen C. Incompletely excised basal cell carcinomas of the skin.
Med J Aust 1997; 167: 557.
21. Ellis SG, Weintraub W, Holmes D, Shaw R, Block PC, King SB HI.
Relation of operator volume and experience to procedural out-
come of percutaneous coronary revascularization at hospitals
with high interventional volumes. Circulation 1997; 95:
2479-84.
British Journal o f Plastic Surgery
22. Sosa JA, Bowman HM, Teilsch JM, Powe NR, Gordon TA,
Udelsman R. The importance of surgeon experience for clinical
and economic outcomes from thyroidectomy. Ann Surg 1998;
228: 320-30.
23. Witt PD, Wahlen JC, Marsh JL, Grames LM, Pilgram TK. The
effect of surgeon experience on velopharyngeal functional out-
come following palatoplasty: is there a learning curve? Plast
Reconstr Surg 1998; 102: 1375-84.
24. Rippey JJ, Rippey E. Characteristics of incompletely excised basal
cell carcinoma of the skin. Med J Aust 1997; 166: 581-3.
The Authors
Prem Kumar MS, FRCS, Associate Specialist in Plastic Surgery
Stewart Watson FRCS, MRCP, Consultant Plastic Surgeon
Department of Bums, Plastic and Reconstructive Surgery, University
Hospital of South Manchester, Southmoor Road, Wythenshawe,
Manchester M23 9LT, UK.
Anne N. Brain FRCS, Consultant Plastic Surgeon
Samnitra S. Banerjee MD, FRCPath, Consultant Histopathologist
Christie Hospital, Manchester M20 4BX, UK.
Peter J. Davenport LRCP, FRCS, Consultant Plastic Surgeon
David L. Bisset BSc, MB, ChB, FRCSPath, Consultant
Histopathologist
Royal Bolton Hospital, Bolton BL4 0JR, UK.
Lorna J. McWilliam MB, ChB, FRCPath, Consultant
Histopathologist
Manchester Royal Infirmary, Manchester M13 9WL, UK.
Correspondence to Mr Prem Kumar.
Paper received 24 June 2002.
Accepted 9 August 2002, after revision.