Volume-outcome analysis of colorectal cancer-related outcomes

Abstract

Significant associations between caseload and surgical outcomes highlight the conflict between local cancer care and the need for centralization. This study examined the effect of hospital volume on short-term outcomes and survival, adjusting for the effect of surgeon caseload.
Between 1998 and 2002, 8219 patients with colorectal cancer were identified in a regional population-based audit. Outcomes were assessed using univariable and multivariable analysis to allow case mix adjustment. Surgeons were categorized as low (26 or fewer operations annually), medium (27-40) or high (more than 40) volume. Hospitals were categorized as low (86 or fewer), medium (87-109) or high (more than 109) volume.
Some 7411 (90.2 per cent) of 8219 patients underwent surgery with an anastomotic leak rate of 2.9 per cent (162 of 5581), perioperative mortality rate of 8.0 per cent (591 of 7411) and 5-year survival rate of 46.8 per cent. Medium- and high-volume surgeons were associated with significantly better operative mortality (odds ratio (OR) 0.74, P = 0.010 and OR 0.66, P = 0.002 respectively) and survival (hazard ratio (HR) 0.88, P = 0.003 and HR 0.93, P = 0.090 respectively) than low-volume surgeons. Rectal cancer survival was significantly better in high-volume versus low-volume hospitals (HR 0.85, P = 0.036), with no difference between medium- and low-volume hospitals (HR 0.96, P = 0.505).
This study has confirmed the relevance of minimum volume standards for individual surgeons. Organization of services in high-volume units may improve survival in patients with rectal cancer.

Full text

Original article
Volume–outcome analysis of colorectal cancer-related
outcomes
D. W. Borowski1,4,D.M.Bradburn
2, S. J. Mills2,4,B.Bharathan
1,4,R.G.Wilson
3,A.A.Ratcliffe
5
andS.B.Kelly
1on behalf of the members of the Northern Region Colorectal Cancer Audit Group
(NORCCAG)
Departments of Surgery, 1North Tyneside General Hospital, North Shields, 2Wansbeck General Hospital, Ashington, and 3James Cook University
Hospital, Middlesbrough, 4Northern Region Colorectal Cancer Audit Group, Hexham General Hospital, Hexham, UK, and 5Impact Planning and
Improvement, Bill and Melinda Gates Foundation, Seattle, Washington, USA
Correspondence to: Mr S. B. Kelly, Department of Surgery, North Tyneside General Hospital, Rake Lane, North Shields NE29 8NH, UK
(e-mail: s.b.kelly@ncl.ac.uk)
Background: Significant associations between caseload and surgical outcomes highlight the conflict
between local cancer care and the need for centralization. This study examined the effect of hospital
volume on short-term outcomes and survival, adjusting for the effect of surgeon caseload.
Methods: Between 1998 and 2002, 8219 patients with colorectal cancer were identified in a regional
population-based audit. Outcomes were assessed using univariable and multivariable analysis to allow
case mix adjustment. Surgeons were categorized as low (26 or fewer operations annually), medium
(27–40) or high (more than 40) volume. Hospitals were categorized as low (86 or fewer), medium
(87–109) or high (more than 109) volume.
Results: Some 7411 (90·2 per cent) of 8219 patients underwent surgery with an anastomotic leak rate of
2·9 per cent (162 of 5581), perioperative mortality rate of 8·0 per cent (591 of 7411) and 5-year survival
rate of 46·8 per cent. Medium- and high-volume surgeons were associated with significantly better
operative mortality (odds ratio (OR) 0·74, P=0·010 and OR 0·66, P=0·002 respectively) and survival
(hazard ratio (HR) 0·88, P=0·003 and HR 0·93, P=0·090 respectively) than low-volume surgeons.
Rectal cancer survival was significantly better in high-volume versus low-volume hospitals (HR 0·85,
P=0·036), with no difference between medium- and low-volume hospitals (HR 0·96, P=0·505).
Conclusion: This study has confirmed the relevance of minimum volume standards for individual
surgeons. Organization of services in high-volume units may improve survival in patients with rectal
cancer.
Presented to the Annual Scientific Meeting of the Association of Surgeons of Great Britain and Ireland, Manchester,
UK, April 2007, and published in abstract form as Br J Surg 2007; 94(Suppl 2): 157
Paper accepted 8 March 2010
Published online in Wiley InterScience (www.bjs.co.uk). DOI: 10.1002/bjs.7111
Introduction
Higher surgeon volume, hospital caseload and subspe-
cialization have all been associated with better outcomes
following major surgery1–3. Although the evidence is
largely consistent in favour of a volume –outcome rela-
tionship for some operations, such as pancreatic cancer
resections4, such an association is more controversial even
within one healthcare system for other types of surgery,
such as resections for oesophagogastric cancer5–8. In colo-
rectal cancer, volume–outcome relationships have been
reported for both hospital volume9–11 and individual sur-
geon caseload12–16, but few studies have examined the
effects of both simultaneously17– 23. There remains debate
whether colorectal cancer surgery should be carried out
in high-volume centres to realize the benefits of cumula-
tive experience within the institution, or to focus on the
individual surgeon’s workload and specialist skills.
Within a national health service in any country, the
outcome of this debate has substantial implications for
healthcare delivery. In the UK there is a conflict between
Copyright 2010 British Journal of Surgery Society Ltd British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
retention of colorectal cancer services in district or small
local hospitals, with geographic advantages for many
patients, and concentration of cancer services in large-
volume regional units24. Previous UK data did not support
the centralization of colorectal cancer surgery in high-
volume centres17,18 or teaching hospitals25–27, and a model
of care for these patients in district general hospitals
working within a regional cancer network is advised28,29.
A recent meta-analysis suggested that institutional vol-
ume and surgeon’s specialist training played a more
important role than individual surgeon caseload, partic-
ularly for long-term and cancer-related outcomes30,31.
Most studies, however, originated from countries where
healthcare provision may depend on market forces
and competition10,11,21,23,32– 38. The relevance of a vol-
ume–outcome relationship for every healthcare system is
questionable and each country should rationalize accord-
ing to local data. In the absence of conclusive evidence
from UK data, the aim of this study was to interro-
gate a large, population-based data set of patients with
colorectal cancer for the individual effects of hospital vol-
ume and surgeon caseload on short-term outcomes and
5-year survival.
Methods
The Northern Region Colorectal Cancer Audit Group
(NORCCAG) was established in 1997 across the 17
acute hospitals in the Northern Region of England to
audit the process of care and outcome for patients with
colorectal cancer. Ethical approval was obtained from the
Northern and Yorkshire Multi-Centre Research Ethics
Committee. The audit was funded by the participating
hospitals and cancer networks, and has latterly been funded
by the strategic health authority. The management of the
audit is overseen by a multidisciplinary steering group
consisting of clinicians from each of the participating
hospitals. The chairperson of the steering group and
the audit coordinators are responsible for data security
and confidentiality. Identifiable data relating to patients,
surgeons and institutions are coded and known only to the
audit facilitators to maintain confidentiality.
NORCCAG carries out a population-based audit of
all patients who undergo surgery for colorectal cancer
within the National Health Service (NHS) in the Northern
Region of England. As private practice is comparatively
uncommon, and major surgery is frequently performed
within NHS hospitals in this region even for patients with
private healthcare provision, it is thought that the audit
includes the vast majority of colorectal cancers within the
region. During the reporting period of this study, polyp
cancers removed at endoscopy were excluded. Very few
colorectal operations were carried out laparoscopically and
were therefore not identified specifically. For each patient
only the main operation for colorectal cancer was included,
and the figures do not reflect reoperations for the same
cancer. However, some patients had metachronous disease,
which was included as a separate event.
Patients were initially identified from histopathology
records, multidisciplinary team (MDT) meetings and hos-
pital waiting lists. Data were collected on a standard three-
part form according to prospectively agreed protocols. The
management form, completed by specifically trained audit
facilitators, contained the patient’s demographic, referral
and admission details, as well as information on investiga-
tions, clinical management, treatment and outcome. The
surgical form was completed by the operating surgeon
and the pathology form, based on the national minimum
data set for colorectal cancer histopathology reporting,
by pathologists or audit staff from standardized hospital
pathology reports.
All audit data were stored in a MicrosoftAccess
2000 database (Microsoft, Redmond, Washington, USA),
developed and maintained by an external medical software
company (Xentec, North Tyneside, UK). The data were
regularly validated by random checks on 10 per cent of the
electronic records against the data collection pro forma,
and by targeted cross-referencing. For a small number of
patients, further interrogation of hospital case notes was
required to rectify missing and non-coherent data. Details
of all patients were submitted to the Office for National
Statistics (ONS) to obtain the date and cause of death, and
establish operative mortality and survival rates.
Surgeon and hospital volume
During the study period, a total of 140 surgeons
performed surgery for colorectal cancer in 17 hospitals. To
determine the caseload of individual surgeons, allowing
for retirement or uptake of practice, the mean annual
number of operations performed for colorectal cancer was
considered across the years in which a surgeon’s firm
had recorded activity. All patients were stratified into
three groups of nearly equal size according to the annual
individual surgeon caseload: 112 low-volume surgeons
(LVS) who performed 26 or fewer colorectal cancer
operations per year (range 1–26), 18 medium-volume
surgeons (MVS) who carried out 27–40 operations per
year (range 27–40) and ten high-volume surgeons (HVS)
who performed more than 40 operations per year (range
43–63).
Using a similar approach, patients were stratified into
three groups according to the mean annual colorectal
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
cancer caseload of the hospital where surgery was carried
out: eight low-volume hospitals (LVH) with 86 or fewer
operations per year (range 36–84), five medium-volume
hospitals (MVH) with 87– 109 operations per year (range
87–95) and four high-volume hospitals (HVH) with more
than 109 operations per year (range 109– 148).
Surgeon and hospital volume for colonic
and rectal cancer
To allow separate analysis of the effects of provider volume
on the outcomes of surgery for colonic and rectal cancer,
the mean annual operative caseload was calculated for
each individual surgeon and each hospital based on their
respective numbers of colonic and rectal cancer cases.
Patients in each subgroup were stratified into three near-
equal size categories according to the respective surgeon
and hospital caseload.
For patients with cancer of the colon, 107 LVS
performed 11·3 or fewer colonic cancer operations per year
(range 0·8–11·3), 22 MVS carried out 11·7– 23 operations
per year and 11 HVS surgeons performed more than
23 operations per year (range 24–37). Hospital volume
groups for colonic cancer were: eight LVH with 53 or
fewer operations per year (range 22–53), five MVH with
54–59 operations per year, and four HVH with more than
59 operations per year (range 69–90).
For patients with cancer of the rectum, 82 LVS
performed 13·5 or fewer rectal cancer operations per
year (range 0·2–13·5), 15 MVS carried out 13·8–21·4
operations per year and seven HVS performed at least 22·3
operations per year (range 22·3–29·2). Thirty-six surgeons
performed no rectal cancer surgery in the period covered
by this study. Hospital volume groups for rectal cancer
were eight LVH with 33 or fewer rectal cancer operations
per year (range 14–33), five MVH with 34 –39 operations
per year and four HVH with more than 39 operations per
year (range 40–71).
Definitions
Rectal cancer was defined as a cancer located within 15 cm
from the anal verge. Colonic tumours were defined as right-
sided cancer if they were located proximal to the splenic
flexure and left-sided cancer if located distal to the splenic
flexure. A splenic flexure cancer was defined as right or left
sided according to the surgical procedure carried out.
To allow case mix-adjusted analysis, a patient’s health
and co-morbidity was determined using the American Soci-
ety of Anesthesiologists (ASA) grading system39 , based on
the preoperative assessment by the anaesthetist. Urgency of
surgery was defined as elective or emergency by the surgeon
at the time of operation; the latter included both emergency
and urgent cases as defined by the National Confidential
Enquiry into Peri-Operative Deaths40. Tumour stage was
defined according to the modified Dukes’ classification41;
Dukes’ ‘D’ included patients in whom distant metastases
were confirmed by preoperative or perioperative radiolog-
ical imaging, operative findings or histology, and whose
operation was deemed palliative by the surgeon because
of unresectable or macroscopic residual disease. Com-
plete tumour resection (R0) was defined as macroscopic
clearance as indicated by the surgeon at the time of oper-
ation, with histological confirmation of complete tumour
excision. A resection was defined as R1 when there was
macroscopic clearance but histological evidence of mar-
gin positivity, and R2 when macroscopically incomplete as
indicated by the surgeon. Operative mortality was defined
as death within 30 days of surgery, or as an inpatient during
the same admission before discharge.
Statistical analysis
Relationships between variables of interest were considered
in univariable analysis using χ2test, χ2test for trend,
ANOVA and Student’s ttest as appropriate, assuming
statistical significance at the 0·01 level. Correlation
between colonic and rectal cancer caseload as non-
parametric variables was examined using Spearman’s rank
test. Univariable overall survival analysis was carried
out using the Kaplan–Meier method for incomplete
observations42, considering the time elapsed between the
day of surgery and death from any cause, or censored at
31 October 2006 to allow for delay in the reporting of this
event by the ONS. The log rank test was used to compare
survival with significance assumed at the 0·05 level.
To allow case mix-adjusted analysis, variables signifi-
cantly associated with anastomotic leakage, operative death
and 5-year survival in univariable analysis were introduced
into a set of forward-selective logistic regression models
under assessment of collinearity and effect modification.
The basic model was initially developed on a 70 per cent
random sample of patients and its predictive value tested
on the remaining 30 per cent before application to the
entire cohort. Surgeon and hospital characteristics were
initially introduced separately into each best regression
model, and later used simultaneously to estimate their
individual effects.
Factors showing a significant association with survival
were examined using a forward-selective Cox proportional
hazard model43 built by employing a similar methodology
to the strategy used in logistic regression; volume and
specialization characteristics were included in the final
predictive model from each hierarchy.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
All statistical analysis was performed with SPSS
version 11 (SPSS, Chicago, Illinois, USA). Additional
survival analysis was carried out using Stataversion 8.0
(StataCorp, College Station, Texas, USA).
Results
Between January 1998 and December 2002, 7411
(90·2 per cent) of the 8219 patients with colorectal cancer
included in the audit underwent surgery. Their median age
was 71 (range 14–98) years. Patients treated in HVH were
more frequently treated under the care of HVS (Table 1).
Most surgeons performed surgery for both colonic and
rectal cancer, except the 36 who did not perform any
operations for rectal cancer during the study period. There
was good correlation between individual surgeon’s caseload
for colonic and rectal cancer (rs=0·852, P<0·001).
Surgery for both colonic and rectal cancer was performed
in all hospitals, with good correlation between colonic and
rectal cancer caseload (rs=0·895, P<0·001).
Tab le 1 Patient population in relation to hospital volume and
surgeon caseload (140 surgeons, 17 hospitals)
Hospital volume
LVH MVH HVH P*
Surgeon volume <0·001
LVS (≤26) 1225 (48·1) 650 (25·5) 670 (26·4)
MVS (27– 40) 1139 (47·4) 1109 (46·2) 154 (6·4)
HVS (>40) 176 (7·1) 541 (22·0) 1747 (70·9)
Total 2540 (34·3) 2300 (31·0) 2571 (34·7)
Values in parentheses are percentages. LVH, low-volume hospitals;
MVH, medium-volume hospitals; HVH, high-volume hospitals; LVS,
low-volume surgeons; MVS, medium-volume surgeons; HVS,
high-volume surgeons. *χ2test for trend.
Patient characteristics in relation to surgeon and hospital
volume groups are shown in Table 2. Patient sex was
not associated with surgeon volume, but surgeons with
a higher caseload operated on a significantly younger,
fitter population with more rectal cancers and fewer
Tab le 2 Patient demographics and case mix in relation to surgeon and hospital caseload (7411 patients)
All
(n=7411)
LVS
(n=2545)
MVS
(n=2402)
HVS
(n=2464) P*
LVH
(n=2540)
MVH
(n=2300)
HVH
(n=2571) P*
Sex 0·808 0·647
M 4256 (57·4) 1454 (57·1) 1386 (57·7) 1416 (57·5) 1458 (57·4) 1306 (56·8) 1492 (58·0)
F 3155 (42·6) 1091 (42·9) 1016 (42·3) 1048 (42·5) 1082 (42·6) 994 (43·2) 1079 (42·0)
Age (years) <0·001 0·002
<65 2133 (28·8) 672 (26·4) 710 (29·6) 751 (30·5) 721 (28·4) 649 (28·2) 763 (29·7)
65– 74 2647 (35·7) 912 (35·8) 842 (35·1) 893 (36·2) 865 (34·1) 824 (35·8) 958 (37·3)
75– 84 2166 (29·2) 794 (31·2) 678 (28·2) 694 (28·2) 784 (30·9) 655 (28·5) 727 (28·3)
≥85 465 (6·3) 167 (6·6) 172 (7·2) 126 (5·1) 170 (6·7) 172 (7·5) 123 (4·8)
ASA grade <0·001 0·029
I 924 (12·5) 285 (11·2) 360 (15·0) 279 (11·3) 347 (13·7) 303 (13·2) 274 (10·7)
II 3210 (43·3) 970 (38·1) 1062 (44·2) 1178 (47·8) 1080 (42·5) 1008 (43·8) 1122 (43·6)
III 2245 (30·3) 832 (32·7) 690 (28·7) 723 (29·3) 744 (29·3) 726 (31·6) 775 (30·1)
IV/V 449 (6·1) 206 (8·1) 144 (6·0) 99 (4·0) 143 (5·6) 165 (7·2) 141 (5·5)
Unknown 583 (7·9) 252 (9·9) 146 (6·1) 185 (7·5) 226 (8·9) 98 (4·3) 259 (10·1)
Urgency <0·001 <0·001
Elective 5955 (80·4) 1712 (67·3) 2095 (87·2) 2148 (87·2) 2099 (82·6) 1840 (80·0) 2016 (78·4)
Emergency 1456 (19·6) 833 (32·7) 307 (12·8) 316 (12·8) 441 (17·4) 460 (20·0) 555 (21·6)
Tumour site <0·001 0·355
Right colon 2344 (31·6) 967 (38·0) 705 (29·4) 672 (27·3) 782 (30·8) 762 (33·1) 800 (31·1)
Left colon 2137 (28·8) 805 (31·6) 657 (27·4) 675 (27·4) 790 (31·1) 628 (27·3) 719 (28·0)
Rectum 2883 (38·9) 745 (29·3) 1028 (42·8) 1110 (45·0) 954 (37·6) 892 (38·8) 1037 (40·3)
Unknown 47 (0·6) 28 (1·1) 12 (0·5) 7 (0·3) 14 (0·6) 18 (0·8) 15 (0·6)
Tumour stage 0·002 0·878
Dukes’ A 1058 (14·3) 289 (11·4) 366 (15·2) 403 (16·4) 357 (14·1) 333 (14·5) 368 (14·3)
Dukes’ B 2324 (31·4) 785 (30·8) 768 (32·0) 771 (31·3) 811 (31·9) 715 (31·1) 798 (31·0)
Dukes’ C 1988 (26·8) 700 (27·5) 630 (26·2) 658 (26·7) 684 (26·9) 581 (25·3) 723 (28·1)
Dukes’ ‘D’ 1417 (19·1) 551 (21·7) 439 (18·3) 427 (17·3) 477 (18·8) 477 (20·7) 463 (18·0)
No resection 461 (6·2) 167 (6·6) 149 (6·2) 145 (5·9) 157 (6·2) 143 (6·2) 161 (6·3)
Unknown 163 (2·2) 53 (2·1) 50 (2·1) 60 (2·4) 54 (2·1) 51 (2·2) 58 (2·3)
Values in parentheses are percentages. LVS, low-volume surgeons; MVS, medium-volume surgeons; HVS, high-volume surgeons; LVH, low-volume
hospitals; MVH, medium-volume hospitals; HVH, high-volume hospitals. ASA, American Society of Anesthesiologists. *χ2test and χ2test for trend, as
appropriate.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
emergencies. In contrast, higher hospital caseload was
only associated with a greater proportion of emergency
operations and younger patients.
Complete tumour resection
Tumour resection was performed in 6949 (93·8 per cent)
of 7411 patients undergoing surgery; of these, 5086
(73·2 per cent) had complete margin-negative tumour
resection (R0). R0 resection rates increased with higher
surgeon caseload (LVS: 1673 of 2378, 70·4 per cent;
MVS: 1653 of 2253, 73·4 per cent; HVS: 1760 of 2318,
75·9 per cent; P<0·001), but were not associated with
hospital volume (LVH: 1758 of 2383, 73·8 per cent;
MVH: 1521 of 2156, 70·5 per cent; HVH: 1807 of 2410,
75·0 per cent; P=0·340).
Restoration of bowel continuity
Excluding operations that included only the formation of a
defunctioning stoma, transanal resection of rectal tumours
and non-curative R2 resections, elective resection of at
Tab le 3 Univariable and multivariable analysis of anastomotic dehiscence and operative death
Anastomotic leak (n=5581) Operative death (n=7411)
Unadjusted Adjusted Unadjusted Adjusted
n*P‡Odds ratio†P§n*P‡Odds ratio†P§
Sex <0·001 0·033
F48(1·9) 1·00 227 (7·2) 1·00
M 114 (3·7) 1·81 (1·28, 2·57) 0·001 364 (8·6) 1·43 (1·18, 1·73) <0·001
Age (years) 0·929 <0·001
<65 51 (3·3) 1·00 62 (2·9) 1·00
65– 74 47 (2·3) 0·57 (0·39, 0·90) 0·014 173 (6·5) 2·07 (1·52, 2·83) <0·001
75– 84 56 (3·4) 0·93 (0·62, 1·40) 0·718 263 (12·1) 3·62 (2·67, 4·91) <0·001
≥85 8 (2·4) 0·69 (0·32, 1·50) 0·348 93 (20·0) 5·81 (4·00, 8·44) <0·001
ASA grade 0·062 <0·001
I6(0·8) 1·00 15 (1·6) 1·00
II 73 (2·9) 4·26 (1·83, 9·93) 0·001 117 (3·6) 1·57 (0·90, 2·72) 0·111
III 60 (3·7) 5·84 (2·45, 13·90) <0·001 291 (13·0) 4·37 (2·55, 7·49) <0·001
IV/V 9 (3·2) 4·86 (1·66, 14·30) 0·004 126 (28·1) 8·39 (4·74, 14·90) <0·001
Unknown 14 (3·4) 5·25 (1·97, 14·00) 0·001 42 (7·2) 2·44 (1·31, 4·52) 0·005
Urgency 0·043 <0·001
Elective 122 (2·7) 1·00 319 (5·4) 1·00
Emergency 40 (3·9) 2·18 (1·42, 3·35) <0·001 272 (18·7) 2·93 (2·38, 3·60) <0·001
Tumour site <0·001 0·025
Right colon 39 (1·8) 1·00 212 (9·0) 1·00
Left colon 47 (2·6) 1·62 (1·05, 2·51) 0·030 159 (7·4) 0·89 (0·70, 1·12) 0·308
Rectum 76 (4·8) 3·76 (2·43, 5·84) <0·001 211 (7·3) 1·33 (1·05, 1·67) 0·018
Unknown 0 (0) 9 (19) 1·25 (0·52, 2·99) 0·614
Tumour stage 0·606 <0·001
Dukes’ A 28 (3·5) 1·00 46 (4·3) 1·00
Dukes’ B 60 (3·0) 0·92 (0·58, 1·48) 0·742 142 (6·1) 1·05 (0·73, 1·50) 0·791
Dukes’ C 48 (2·9) 0·80 (0·49, 1·30) 0·368 108 (5·4) 0·98 (0·68, 1·42) 0·923
Dukes’ ‘D’ 21 (2·1) 0·55 (0·30, 1·01) 0·054 158 (11·2) 1·79 (1·25, 2·57) 0·002
No resection — 120 (26·0) 3·90 (2·63, 5·79) <0·001
Unknown 5 (4·2) 1·29 (0·48, 3·49) 0·612 17 (10·4) 2·32 (1·25, 4·32) 0·008
Surgeon volume 0·075 <0·001
LVS 63 (3·3) 1·00 290 (11·4) 1·00
MVS 57 (3·1) 0·97 (0·66, 1·44) 0·895 154 (6·4) 0·74 (0·58, 0·93) 0·010
HVS 42 (2·3) 0·59 (0·37, 0·93) 0·024 147 (6·0) 0·66 (0·51, 0·85) 0·002
Hospital volume 0·630 0·896
LVH 50 (2·6) 1·00 208 (8·2) 1·00
MVH 59 (3·3) 1·38 (0·93, 2·04) 0·113 170 (7·4) 0·86 (0·68, 1·09) 0·215
HVH 53 (2·8) 1·40 (0·88, 2·22) 0·161 213 (8·3) 1·11 (0·87, 1·44) 0·403
Values in parentheses are *percentages and †95 per cent confidence intervals. Constants in adjusted analyses for anastomotic leak and operative death
were −3·576 and −2·162 respectively. ASA, American Society of Anesthesiologists; LVS, low-volume surgeons; MVS, medium-volume surgeons; HVS,
high-volume surgeons; LVH, low-volume hospitals; MVH, medium-volume hospitals; HVH, high-volume hospitals. ‡χ2test and χ2test for trend, as
appropriate; §logistic regression.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
least one segment of colon or rectum was carried out in
4616 patients (56·2 per cent). Of these, 736 (15·9 per cent)
had no primary anastomosis and no restoration of bowel
continuity. There was no difference in the rate of
restorative surgery between patients based on surgeon
caseload (LVS: 1093 of 1299, 84·1 per cent; MVS: 1390 of
1635, 85·0 per cent; HVS: 1397 of 1682, 83·1 per cent;
P=0·373) or hospital volume (LVH: 1356 of 1624,
83·5 per cent; MVH: 1222 of 1396, 87·5 per cent; HVH,
1302 of 1596, 81·6 per cent; P=0·138).
Anastomotic leak
Surgery included the formation of a primary anastomosis
in 5581 (75·3 per cent) of 7411 patients. Of these, a
clinically relevant anastomotic leak was recorded in 162
patients (2·9 per cent). This was more common following
rectal resection and primary anastomosis than with colonic
anastomoses: 76 of 1573 (4·8 per cent) versus 86 of 4008
(2·1 per cent) respectively (P<0·001). Leaks were more
common among men, and patients who had emergency
surgery or increased co-morbidity, but were not associated
Tab le 4 Five-year overall survival calculated by Kaplan– Meier method and Cox proportional hazard model for death within 5 years
among all patients who had surgery and those who had potentially curative R0 resection (excluding operative deaths)
All patients (n=7411) Curative R0 resection (n=4818)
Overall survival Cox regression Overall survival Cox regression
Rate (%) P* Hazard ratio P†Rate (%) P* Hazard ratio P†
Sex 0·003 0·273
F49·11·00 65·21·00
M45·11·13 (1·06, 1·21) <0·001 63·01·10 (0·99, 1·22) 0·055
Age (years) <0·001 <0·001
<65 54·81·00 74·41·00
65– 74 50·31·19 (1·09, 1·29) <0·001 67·11·29 (1·13, 1·49) <0·001
75– 84 39·21·56 (1·43, 1·70) <0·001 54·11·94 (1·68, 2·23) <0·001
≥85 25·41·94 (1·70, 2·21) <0·001 40·12·70 (2·21, 3·30) <0·001
ASA grade <0·001 <0·001
I 63·11·00 78·21·00
II 53·81·12 (0·99, 1·26) 0·077 68·01·36 (1·13, 1·63) 0·001
III 35·91·65 (1·45, 1·87) <0·001 53·41·94 (1·60, 2·35) <0·001
IV/V 16·82·50 (2·14, 2·92) <0·001 33·73·06 (2·38, 3·93) <0·001
Unknown 46·81·31 (1·11, 1·53) 0·001 65·91·44 (1·13, 1·84) 0·004
Urgency <0·001 <0·001
Elective 51·31·00 66·01·00
Emergency 28·31·49 (1·37, 1·61) <0·001 51·61·47 (1·28, 1·67) <0·001
Tumour site 0·001 0·006
Right colon 44·61·00 60·71·00
Left colon 50·30·84 (0·77, 0·91) <0·001 68·00·84 (0·74, 0·95) 0·006
Rectum 45·91·03 (0·95, 1·12) 0·461 63·41·15 (1·02, 1·30) 0·021
Unknown 47 0·36 (0·24, 0·54) <0·001 75 0·82 (0·30, 2·19) 0·688
Tumour stage <0·001 <0·001
Dukes’ A 72·91·00 75·91·00
Dukes’ B 63·01·34 (1·17, 1·53) <0·001 67·61·32 (1·13, 1·54) <0·001
Dukes’ C 46·92·29 (2·01, 2·62) <0·001 52·12·43 (2·09, 2·82) <0·001
Dukes’ ‘D’ 12·76·45 (5·65, 7·38) <0·001 — —
No resection 4·713·60 (11·60, 15·90) <0·001 — —
Unknown 60·31·62 (1·23, 2·14) 0·001 67·61·46 (1·28, 1·67) 0·021
Surgeon volume <0·001 <0·001
LVS 41·11·00 60·11·00
MVS 49·20·88 (0·81, 0·96) 0·003 66·30·90 (0·80, 1·02) 0·099
HVS 50·30·93 (0·85, 1·01) 0·090 65·30·94 (0·82, 1·07) 0·355
Hospital volume 0·546 0·709
LVH 46·31·00 63·51·00
MVH 46·60·89 (0·82, 0·97) 0·005 64·50·91 (0·81, 1·03) 0·151
HVH 47·40·92 (0·84, 1·01) 0·077 63·90·95 (0·83, 1·09) 0·451
Values in parentheses 95 per cent confidence intervals. ASA, American Society of Anesthesiologists; LVS, low-volume surgeons; MVS, medium-volume
surgeons; HVS, high-volume surgeons; LVH, low-volume hospitals; MVH, medium-volume hospitals; HVH, high-volume hospitals. *Log rank test;
†proportional hazard model.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
with age or tumour stage (Table 3). In univariable analysis,
surgeon volume was not a significant predictor, but high
surgeon caseload was associated with a significantly reduced
risk of anastomotic leakage in case mix-adjusted analysis.
Hospital volume was not associated with anastomotic
dehiscence.
Operative mortality
Some 591 patients (8·0 per cent) died following surgery.
Death was more often the result of emergency surgery than
elective surgery, and was more common among the elderly
and the unfit. Patients with right-sided colonic cancers
and advanced disease also had higher operative mortality
rates. Mortality was significantly associated with surgeon
caseload, and was lowest among HVS. Hospital volume was
not associated with operative death in univariable analysis,
or after adjustment for all other predictive factors including
surgeon volume (Table 3).
In separate multivariable models including either
surgeon caseload or hospital volume, surgeon caseload
was significantly associated with a reduced risk of death
for MVS (odds ratio (OR) 0·69, 95 per cent confidence
interval (c.i.) 0·55 to 0·87; P=0·002) and HVS (OR 0·70,
95 per cent c.i. 0·56 to 0·88; P=0·002), whereas hospital
volume was not a significant predictor (MVH: OR 0·82,
95 per cent c.i. 0·65 to 1·04, P=0·096; HVH: OR 1·01,
95 per cent c.i. 0·81 to 1·27, P=0·911).
The beneficial effect of higher surgeon volume on
mortality was particularly clear among patients undergoing
elective surgery (MVS: OR 0·56, 95 per cent c.i. 0·42
to 0·76, P<0·001; HVS: OR 0·61, 95 per cent c.i.
0·44 to 0·85, P=0·003), whereas hospital volume had
no significant effect (MVH: OR 1·01, 95 per cent c.i.
0·75 to 1·36, P=0·945; HVH: OR 1·04, 95 per cent
c.i. 0·74 to 1·48, P=0·811). Among patients who had
emergency surgery, neither surgeon volume (MVS: OR
1·23, 95 per cent c.i. 0·83 to 1·83, P=0·299; HVS: OR
0·70, 95 per cent c.i. 0·46 to 1·07, P=0·095) nor increasing
hospital volume (MVH: OR 0·65, 95 per cent c.i. 0·44 to
0·96, P=0·030; HVH: OR 1·29, 95 per cent c.i. 0·88 to
1·89, P=0·186) were significant predictors in the case
mix-adjusted models including both factors.
Tab le 5 Univariable and multivariable analysis of the effects of surgeon caseload and hospital volume on outcomes after surgery for
colonic cancer: case mix-adjusted analysis including both surgeon and hospital volume
Surgeon caseload Hospital volume
Unadjusted Adjusted Unadjusted Adjusted
n*P‡Odds ratio†P§n*P‡Odds ratio†P§
R0 resection <0·001 0·131
Low volume 959 (68·8) 1075 (72·1)
Medium volume 1135 (75·1) 977 (72·3)
High volume 1046 (75·0) 1088 (74·6)
Restorative surgery (elective) 0·706 0·007
Low volume 597 (97·7) 896 (98·0)
Medium volume 999 (97·6) 793 (97·9)
High volume 917 (97·4) 824 (96·5)
Anastomotic leak 0·060 0·455
Low volume 34 (2·7) 1·00 26 (1·8) 1·00
Medium volume 31 (2·2) 1·03 (0·60, 1·75) 0·917 30 (2·4) 1·52 (0·88, 2·63) 0·134
High volume 21 (1·6) 0·66 (0·36, 1·20) 0·175 30 (2·2) 1·44 (0·80, 2·60) 0·220
Operative death <0·001 0·655
Low volume 177 (12·1) 1·00 126 (8·1) 1·00
Medium volume 119 (7·6) 0·89 (0·67, 1·18) 0·405 115 (8·2) 1·11 (0·82, 1·49) 0·508
High volume 75 (5·2) 0·60 (0·43, 0·83) 0·002 130 (8·6) 1·29 (0·95, 1·76) 0·105
Elective length of stay (days) 0·885 0·974
Low volume 10·810·9
Medium volume 10·910·7
High volume 10·810·9
5-year overall survival (%) <0·001 0·936
Low volume 39·31·00 47·81·00
Medium volume 50·10·93 (0·84, 1·03) 0·176 47·41·00 (0·90, 1·11) 0·969
High volume 52·50·84 (0·76, 0·94) 0·003 46·81·09 (0·98, 1·22) 0·120
*Values in parentheses are percentages. †Values are odds ratios with 95 per cent confidence intervals in parentheses, except that hazard ratios are shown
for survival. ‡χ2test, χ2test for trend, ANOVA and log rank test, as appropriate; §logistic regression and Cox regression, as appropriate.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
6054484236302418126
100
90
80
70
60
50
40
30
20
10
0
Overall survival (%)
Time after colonic surgery (months)
a Surgeon caseload
Low volume
Medium volume
High volume
Low volume
Medium volume
High volume
No. at risk
1574
1442
1465
1216
1166
1001
1041
996
821
933
888
704
815
785
620
569
587
457
6054484236302418126
100
90
80
70
60
50
40
30
20
10
0
Overall survival (%)
Time after colonic surgery (months)
b Hospital volume
Low volume
Medium volume
High volume
No. at risk
1409
1519
1553
1073
1155
1155
902
968
989
796
857
872
703
743
774
502
532
579
Fig. 1 Kaplan –Meier overall survival curves for patients with colonic cancer undergoing surgery according to asurgeon caseload and b
hospital volume. aP<0·001, bP=0·936 (log rank test)
Postoperative length of stay
Excluding perioperative deaths, 5629 patients underwent
elective surgery and were discharged from hospital. The
geometric mean length of postoperative stay was 11·6days,
with shorter hospital admissions for HVS than for both
LVSandMVS(LVS11·8days, MVS 11·9days, HVS
11·3days;P=0·002). There was no difference in elective
length of stay between hospital volume groups (LVH
11·6days,MVH11·6days,HVH11·7days;P=0·690).
Five-year survival – all patients
The 5-year overall survival rate for the 7411 patients
who had surgery was 46·8 per cent. Survival was negatively
associated with male sex, increasing age, greater co-
morbidity, emergency surgery and advanced tumour stage,
and was worse for patients with right colonic and rectal
cancers than those with left colonic cancers (Table 4).
PatientsoperatedonbyMVSandHVShadsignificantly
better survival than those operated on by LVS, although
this effect was reduced after case mix adjustment. Among
hospitals, there was no difference in 5-year survival based
on case volume criteria but, following case mix adjustment,
MVH and HVH were associated with a lower risk of death
compared with low-volume units (Table 4).
In separate case mix-adjusted analysis, including either
surgeon or hospital volume, surgeon caseload (MVS:
hazard ratio (HR) 0·87, 95 per cent c.i. 0·81 to 0·95,
P=0·001; HVS: HR 0·89, 95 per cent c.i. 0·83 to
0·97, P=0·006) and hospital volume (MVH: HR 0·88,
95 per cent c.i. 0·81 to 0·95, P=0·001; HVH: HR
0·92, 95 per cent c.i. 0·85 to 0·99, P=0·037) were both
significant.
In stratified multivariable analysis, including surgeon
and hospital caseload, surgeon volume was a significant
predictor of survival in elective surgery (MVS: HR 0·84,
95 per cent c.i. 0·76 to 0·92, P<0·001; HVS: HR 0·88,
95 per cent c.i. 0·79 to 0·98, P=0·012), but hospital
volume was not (MVH: HR 0·93, 95 per cent c.i. 0·84
to 1·02, P=0·104; HVH: HR 0·91, 95 per cent c.i. 0·82 to
1·02, P=0·097). Following emergency surgery, neither
the surgeon (MVS: HR 1·01, 95 per cent c.i. 0·85 to
1·20, P=0·926; HVS: HR 0·99, 95 per cent c.i. 0·83 to
1·17, P=0·861) nor increasing hospital volume (MVH:
HR 0·79, 95 per cent c.i. 0·67 to 0·93, P=0·005; HVH:
HR 0·97, 95 per cent c.i. 0·82 to 1·14, P=0·715) was
significantly associated with survival, although MVH had
better survival after emergency surgery than both low- and
high-volume units.
Among 4818 patients who underwent an R0 resection
and did not die as a result of surgery (operative deaths
excluded), the 5-year survival rate was 63·9 per cent.
Survival was significantly better for both MVS and HVS,
but was not associated with hospital volume (Table 4).
Colonic cancer
Of 4481 patients who had surgery for colonic cancer, 4300
(96·0 per cent) underwent tumour resection. Complete,
potentially curative resection was achieved in 3140
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
Tab le 6 Univariable and multivariable analysis of the effects of surgeon caseload and hospital volume on outcomes after surgery for
rectal cancer: case mix-adjusted analysis including both surgeon and hospital volume
Surgeon caseload Hospital volume
Unadjusted Adjusted Unadjusted Adjusted
n*P‡Odds ratio†P§n*P‡Odds ratio†P§
R0 resection 0·027 0·345
Low volume 645 (71·7) 640 (73·6)
Medium volume 642 (72·1) 576 (70·7)
High volume 644 (76·4) 715 (75·4)
Restorative surgery (elective) 0·026 0·650
Low volume 411 (62·4) 446 (65·1)
Medium volume 489 (69·8) 437 (72·0)
High volume 457 (68·2) 474 (64·3)
Anastomotic leak 0·087 0·380
Low volume 29 (5·8) 1·00 29 (5·5) 1·00
Medium volume 29 (5·2) 0·94 (0·53, 1·67) 0·831 24 (4·6) 0·95 (0·53, 1·77) 0·831
High volume 18 (3·5) 0·61 (0·29, 1·29) 0·194 23 (4·4) 0·97 (0·48, 1·96) 0·936
Operative death 0·045 0·893
Low volume 92 (9·3) 1·00 71 (7·5) 1·00
Medium volume 55 (5·6) 0·61 (0·41, 0·89) 0·012 61 (6·8) 0·94 (0·63, 1·41) 0·766
High volume 64 (7·0) 0·79 (0·51, 1·25) 0·325 79 (7·6) 1·05 (0·67, 1·65) 0·835
Elective length of stay (days) 0·008 0·646
Low volume 13·212·6
Medium volume 12·712·7
High volume 12·212·7
5-year overall survival (%) 0·016 0·204
Low volume 42·81·00 45·71·00
Medium volume 46·60·90 (0·79, 1·02) 0·109 43·60·96 (0·83, 1·09) 0·505
High volume 48·60·99 (0·86, 1·16) 0·975 48·20·85 (0·73, 0·99) 0·036
*Values in parentheses are percentages. †Values are odds ratios with 95 per cent confidence intervals in parentheses, except that hazard ratios are shown
for survival. ‡χ2test, χ2test for trend, ANOVA and log rank test, as appropriate; §logistic regression and Cox regression, as appropriate.
(73·0 per cent) of 4300, with restoration of bowel
continuity in 2513 (97·6 per cent) of 2576 patients who
had an elective R0 or R1 resection. Anastomotic leakage
occurredin86(2·1 per cent) of 4010, and 371 (8·3 per cent)
of 4481 patients died after surgery. The 5-year overall
survival rate was 47·3 per cent.
Complete R0 resections increased with surgeon caseload
(Table 5). There was no association between restoration
of bowel continuity in elective operations and surgeon
volume; an apparent trend towards a lower anastomotic
leak rate with higher surgeon caseload was not significant.
Both operative mortality and long-term survival rates
were significantly better for HVS in both univariable and
multivariable analysis, adjusting also for hospital volume.
Examination of hospital volume revealed significantly
lower rates of restorative surgery in high-volume units,
but there was no association between hospital caseload and
complete resection rate, anastomotic leakage, operative
mortality, length of hospital stay or 5-year overall survival
(Table 5). Kaplan–Meier overall survival curves are shown
for surgeon caseload in Fig. 1a and hospital volume
in Fig. 1b.
Rectal cancer
Some 2883 patients with rectal cancer had surgery. The
primary tumour was resected in 2633 (91·3 per cent) of
these, with a complete R0 resection in 1931 (73·3 per cent).
Among 2030 patients who underwent elective rectal
resection with complete macroscopic (R0 or R1) resection,
bowel continuity at the site of the resection was restored in
1357 (66·8 per cent). Clinically relevant anastomotic leaks
were recorded in 76 (4·8 per cent) of 1573 patients in whom
an anastomosis was formed, and 211 (7·3 per cent) of 2883
patients died following surgery. The overall 5-year survival
rate was 45·9 per cent.
Higher surgeon caseload was associated with improved
rates of R0 resection, restoration of rectal continuity and
length of hospital stay (Table 6). Five-year overall survival
was significantly better for patients of HVS in univariable
analysis, but this association was not significant with case
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
6054484236302418126
100
90
80
70
60
50
40
30
20
10
0
Overall survival (%)
Time after rectal surgery (months)
a Surgeon caseload
Low volume
Medium volume
High volume
No. at risk
982
915
986
783
739
750
656
621
634
560
559
541
472
476
438
344
316
307
6054484236302418126
100
90
80
70
60
50
40
30
20
10
0
Overall survival (%)
Time after rectal surgery (months)
b Hospital volume
Low volume
Medium volume
High volume
No. at risk
897
1037
949
710
826
736
587
694
630
506
613
541
423
519
444
290
353
324
Low volume
Medium volume
High volume
Fig. 2 Kaplan –Meier overall survival curves for patients with rectal cancer undergoing surgery according to asurgeon caseload and b
hospital volume. aP=0·016, bP=0·204 (log rank test)
Tab le 7 Published studies on the effects of hospital volume on outcomes of colonic and rectal cancer surgery
Hospital volume
category Volume– outcome
Reference Country Year Cancer Lowest Highest Outcome association
Birkmeyer et al.9USA 1994 –1999 CC ≤32 ≥125 Mortality Yes
Engel et al.52 The Netherlands 1994– 1999 CRC ≤90 ≥128 Mortality Inverse
Rabeneck et al.11 USA 1991 –2000 CRC <25 ≥25 OS Yes
Rogers et al.22 USA 1996 –1999 CRC <21 >55 Mortality, OS Yes
Schrag et al.32 USA 1991 –1996 CC <10 ≥28 OS Yes
Hannan et al.19 USA 1994 –1997 CC ≤21 ≥64 Mortality Yes
Dimick et al.53 USA 1997 CRC <55 ≥150 Mortality Yes
Urbach and Baxter54 Canada 1994– 1999 CRC ≤53 ≥54 Mortality No
Harmon et al.50 USA 1992 –1995 CRC ≤40 ≥70 Mortality, LoS Yes
Hodgson et al.35 USA 1994 –1997 RC <7>20 Rectal restoration, mortality, OS Yes
Harling et al.55 Denmark 1994 –1999 RC <15 >30 Stoma rate Yes
R0, mortality, OS No
AL Inverse
Wibe et al.10 Norway 1993–1999 RC <10 ≥30 Rectal restoration, OS, LR Yes
Mortality No
Kee et al.18 UK 1990– 1994 CRC ≤23 ≥55 OS No
Meyerhardt et al.37 USA 1988, 1992 CC <47 ≥85 OS Yes
DFS No
Renzulli et al.20 Switzerland 1981–1993 CRC <26 ≥26 OS,DFS,LR Yes
Marusch et al.56 Germany 1999 CC >30 >60 Morbidity Yes
AL, mortality No
Simons et al.34 USA 1988 –1992 RC ≤5>5OS Yes
Marusch et al.57 Germany 1999 RC <20 >40 Morbidity, LoS Yes
Rectal restoration, mortality, AL No
Kapiteijn et al.58 The Netherlands 1988– 1992 RC ≤11 >11 LR No
Simunovic et al.59 USA 1990 RC ≤11 ≥18 Mortality, OS No
Engel et al.60 Germany 1996– 1998 RC <10 ≥30 OS,DFS,LR No
Parry et al.17 UK 1993 CRC ≤30 >110 Mortality, OS No
CC, colonic cancer; CRC, colorectal cancer; OS, overall survival; LoS, length of stay; RC, rectal cancer; R0, curative resection; AL, anastomotic leak; LR,
local recurrence; DFS, disease-free survival.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
mix adjustment. Both the incidence of anastomotic leak and
operative death were lower for surgeons with a medium
and high annual caseload, but this was not significant.
Hospital volume was not associated with R0 resection
and rectal restoration rates, anastomotic leakage, length of
stay and operative mortality, but HVH had a significantly
reduced risk of death within 5 years of surgery in case mix-
adjusted analysis (Table 6). Kaplan –Meier survival analysis
for patients who had rectal cancer surgery is shown in Fig. 2.
Discussion
This study was based on a regional population of
patients who underwent surgery for colorectal cancer; it is
thought to capture the vast majority of these patients,
with the exception of a few treated in the setting of
private healthcare institutions. It carries the recognized
shortcomings associated with an observational study, but
is likely to be a good representation of the UK as the
study population is comparable to others at the time of
data capture14,15,17,44,45, and also resembles that of the
National Colorectal Cancer Audit46 to which NORCCAG
contributes.
In the UK there has been considerable effort in
recent years to improve the management of patients
with colorectal cancer, manifest by the NHS Plan47,48
and facilitated by the publication of national guidance
documents28,29,49. Two of the major recommendations
of this guidance include the discussion of all patients
with colorectal cancer in a site-specific MDT, and the
concentration of colorectal cancer cases in the hands
of specialists who achieve a minimum of 20 curative
colorectal cancer resections per annum28. However,
this minimum number was introduced on a rather
arbitrary basis, and supportive evidence from the literature
remains difficult to establish. Numbers required for
surgeons to qualify as ‘high volume’ has varied widely
between individual studies, ranging from five to 50
colorectal operations14,16–18,20,36,45,50 and three to 45
rectal cancer operations12,13,33,51 per annum. Nonetheless,
many of these studies demonstrated improved outcomes
for surgeons with higher caseloads. As it is impossible to
determine a distinct cut-off point for individual surgeons
above which outcomes improve, the suggested number of
around 20 resections per year at surgeon level appears to
be reasonable and practical for the UK.
In the present study, the patient population was divided
into three near-equal groups based on surgeon caseload in
order to make the patient population more homogeneous.
This approach differed from the surgeon-centred analysis
presented in an earlier publication16, which resulted in
large differences in numbers of patients in the individual
groups. In addition, the present approach allowed a more
detailed analysis of the effect of increasing surgeon caseload
among the previously reported group of HVS with a
caseload of more than 18·5 annual cancer operations.
Nonetheless, the group of LVS in the present study,
performing up to 26 annual cancer operations including
palliative and emergency surgery, had a lower R0 resection
rate, a higher stoma rate in elective rectal cancer surgery,
significantly higher operative mortality and lower 5-year
survival than surgeons who operated on a higher number of
patients. For most outcomes, there was very little difference
between the MVS and HVS groups, indicating that the
volume–outcome relationship is not linear.
Traditionally, many general surgeons and non-
colorectal specialists in the UK provided emergency
surgical services and also performed occasional elective
colorectal surgery. With increasing concentration of such
cases within the MDT and between specialist surgeons,
the contribution of non-colorectal surgeons is diminish-
ing. Although specialization of surgeons and high-volume
care are closely linked, the authors have previously demon-
strated the benefit of a volume-based comparison between
surgeons over the use of a surrogate definition of a col-
orectal specialist such as membership of the Association of
Coloproctology16. Thus, adjustment for caseload of indi-
vidual surgeons was preferred to enable the simultaneous
examination of both hospital and surgeon volume in this
study.
The effect of hospital caseload on colorectal cancer
outcomes has been a contentious and occasionally
confusing issue, with much of the evidence for a significant
volume–outcome relationship originating from North
America (Table 7). Current recommendations in the UK
suggest that the annual caseload for each colorectal cancer
MDT unit should not be fewer than 60 new cases per year,
and that the average expected number of new cancer cases
in a district general hospital is around 120 per annum28.
Taking into account that around 10–20 per cent of new
cancers do not come to surgery61, this indicates that most
UK MDT units have an annual operative workload that
is at least equal to, if not substantially higher than, that
reported elsewhere (Table 7), and the annual number of
colorectal cancer operations in UK LVH and HVH also
compares well.
For most of the outcomes presented here, hospi-
tal caseload was not a significant predictor of out-
come. Of interest, HVH had a proportionally greater
emergency caseload. The rates of colorectal restoration
were broadly comparable to previously reported figures
(63–70 per cent), and the absence of an association with
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
hospital volume is supported by others17,55,59. However,
reports from Norway10 and North America38 described
a significant relationship between increasing hospital
caseload and higher rates of sphincter preservation.
The incidence of anastomotic dehiscence in this study
was not associated with hospital volume, as found in a
recent meta-analysis30. A population-based study from
Denmark found a significantly higher anastomotic leak
rate in hospitals with more than 30 rectal cancer resections
per annum55. However, univariable analysis showed no
differences in the rates of curative resection or stoma, and
the leak rate may reflect other factors specific to that study.
The operative mortality rate in this study is comparable
with that of other population-based studies from the
UK14,17,45 and other countries9,52,55. The present results
did not show an association with hospital volume. In
a recent meta-analysis30 evidence for a hospital–volume
relationship was found for colonic but not for rectal cancer
surgery. It is possible, as suggested by the authors of the
largest European population-based registry study52,that
a more complex case mix in high-volume units may even
increase the risk of operative death.
Few studies have examined the length of hospital stay
following colorectal cancer surgery and its association
with hospital volume. One American study found that
the average length of stay decreased from 12·7to
10·8 days with increasing hospital caseload50. European
studies on rectal cancer57 and laparoscopic colorectal
resections62 reported significantly shorter hospital stays
when patients were treated in high-volume units. Length
of hospital stay is an overall poor marker of quality
of care as there are intercountry, intercontinental and
socioeconomic reasons for heterogeneity. However, in this
study, conducted within a relatively homogeneous patient
population, surgeon volume, not hospital caseload, was a
significant determinant of the length of hospital stay.
It may be that the benefits of cancer care in high-volume
units are more important for long-term outcome. There is
varying evidence for an association between higher hospital
caseload and long-term survival31,63. In the present study,
overall survival was marginally better than that reported
from an earlier UK audit14. Once adjustment had been
made for differences in case mix, survival of patients
operated on in hospitals with the lowest caseload was
significantly poorer than that in medium- and high-volume
units. This effect was largely due to the subgroup of patients
with rectal cancer.
Since the period covered by this study, several regional
low-volume units have merged with other units for
all colorectal cancer surgery; these changes in regional
service structure have already resulted in fewer units
providing colorectal cancer care, and may well have
addressed the reasons for this survival difference. Further
centralization of colorectal cancer care beyond that already
implemented in the hospital restructuring process cannot
be recommended. However, this study provides further
evidence that case volume is an important surgeon-related
factor, and supports the notion that specialists should
perform a minimum number of around 25 colorectal
cancer operations per annum. The data presented here
are based solely on patients with malignant disease,
and a similar relationship between volume and outcome
may exist for benign disease. The recent suggestion by
the Association of Coloproctology64 to include benign
cases in the minimum caseload requirement appears very
reasonable.
Collaborators
Steering group of NORCCAG at the time of this study:
I. Bain, D. M. Bradburn, J. Corson, W. J. Cunliffe,
S. El-Rabaa, A. Gunn, K. Gunning, P. J. Hainsworth,
I. Hawthorn, S. B. Kelly, R. Kirby, J. Palmer, C. Pritchett,
I. L. Rosenberg, J. S. Varma, D. Ward, M. Weaver,
M. Whittaker, R. G. Wilson, S. Plusa (surgery); P.
Atherton, N. Bailey, W. Dobrowsky, U. K. Mallick, J.
J. Nicoll, I. D. Pedley, T. Podd, H. van der Voet, N. Wadd
(clinical oncology); P. Barrett, M. Bennett, C. Bloxham,
S. Chawla, P. Geddy, R. Jones, J. McElroy, O. Mohamdee,
J. Shrimankar (histopathology); J. Burn (human genetics);
M. Welfare (medicine).
The following surgeons contributed patients to the
study:H.Hashimi,A.M.F.Hassn,T.Layzell,J.G.
Stephen, S. E. Stock (Bishop Auckland General Hospi-
tal, Bishop Auckland); F. Hinson, J. D. Holdsworth, J. G.
Palmer, S. Raimes, J. E. G. Shand, M. R. Williams (Cum-
berland Infirmary, Carlisle); P. M. Atkinson, R. Brookstein,
N. B. Corner, S. Debrah, K. A. Gunning, M. Mowafi, R.
C. O’Murchu, M. G. Whittaker (Darlington Memorial
Hospital, Darlington); J. Chamberlain, R. M. Charnley, P.
J. Hainsworth, A. F. Horgan, N. A. G. Jones, D. Manas,
P. Wright (Freeman Hospital, Newcastle upon Tyne); R.
Bryan, I. A. Cheema, M. Edwards, M. Hashishi, O. O. M.
Nugud, A. F. J. Shepherd, D. Simons, D. C. Ward (Fri-
arage Hospital, Northallerton); A. K. Agarwal, R. Kirby,
G. R. McLatchie, H. A. Naqesh-Bandi, C. P. L. Wood
(University Hospital of Hartlepool, Hartlepool); K. M.
Abusin, D. Gatehouse, M. Majid, J. D. O’Callaghan, A.
Said, R. M. Weaver (Hexham General Hospital, Hexham);
J. R. Bell, D. Clarke, A. E. Clason, W. M. Cooke, W.
A. Corbett, P. Durning, S. Papagrigoriadis, S. Wakefield,
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
R. G. Wilson (James Cook University Hospital, Middles-
brough); E. L. Gilliland, C. Hennessy, M. J. Higgs, D. J.
Leaper, A. L. G. Peel, I. L. Rosenberg, M. A. Tabaqchali,
M. Yassin (University Hospital of North Tees, Stockton-
on-Tees); F. S. Bassili, J. Black, I. A. Goulbourne, F. J.
Holmes,L.F.Horgan,S.B.Kelly,R.Pollard,B.J.Slater
(North Tyneside General Hospital, North Shields); H. Y.
Ashour, R. Bliss, D. A. Browell, W. J. Cunliffe, R. Farrell,
A. Mudawi (Queen Elizabeth Hospital, Gateshead); F. C.
Campbell, H. Gallagher, S. M. Griffin, C. D. Griffith, A.
B.Griffiths,N.Hayes,D.Karat,T.W.J.Lennard,S.M.
Plusa, J. S. Varma (Royal Victoria Infirmary, Newcastle
upon Tyne); T. Armitage, V. Joypaul, C. J. Pritchett, I.
M. Rogers, K. Wynne (South Tyneside District Hospital,
South Shields); O. Adedeji, M. Bashir, L. H. Boobis, R.
J. Corson, I. L. Crighton, P. Dunlop, M. Gray, W. K.
J. Huizinga, J. M. Lennox, A. J. Rich, D. Sarson, P. K.
Small, P. Surtees, S. Vetrivel (Sunderland Royal Hospital,
Sunderland); I.Bain, K. Clark, A. I. M. Cook, P. Cullen, S.
Green, I. E. Hawthorn, D. Herring, J. R. Mason (Univer-
sity Hospital of North Durham, Durham); S. Bawa, D. M.
Bradburn, M. Carr, J. Guest, D. Matar, S. Mills (Wans-
beck General Hospital, Ashington); N. Abdullah, S. M. A.
El-Rabaa, D. Irving, N. C. Keddie, C. Metcalf-Gibson, D.
G. Richards, A. Sowinski, M. Walker (West Cumberland
Hospital, Whitehaven).
Acknowledgements
The authors are indebted to Sir Len Fenwick, Chief
Executive of the Newcastle Group of Hospitals, and
Professor John Burn, Director of the Institute of Human
Genetics in Newcastle upon Tyne, who provided office
accommodation for audit staff at the Freeman Hospital
and in the Cancer Research UK Office at the Institute of
Human Genetics, Newcastle upon Tyne; and to Alastair
Gunn, who initiated and led NORCCAG during the years
included in the present study. They acknowledge audit
coordinators, Mrs Astrid Adams, Mrs Jane Robinson,
Mrs Dorothy Simms and Mrs Joan Openshaw, and the
help given by the audit departments in the contributing
hospitals.
NORCCAG was initially funded by the Purchasers
Clinical Audit Group of the former Northern Region
of England, the Northern Cancer Network, the Cancer
Care Alliance and all participating Trusts of the former
Northern Region. Currently, NORCCAG is funded
entirely by the strategic health authority for the North
of England. D.W.B. and B.B. received financial support
from Northumbria Healthcare NHS Trust. The authors
declare no conflict of interest.
References
1 Begg CB, Cramer LD, Hoskins WJ, Brennan MF. Impact of
hospital volume on operative mortality for major cancer
surgery. JAMA 1998; 280: 1747–1751.
2 Killeen SD, O’Sullivan MJ, Coffey JC, Kirwan WO,
Redmond HP. Provider volume and outcomes for
oncological procedures. Br J Surg 2005; 92: 389–402.
3 Chowdhury MM, Dagash H, Pierro A. A systematic review
of the impact of volume of surgery and specialization on
patient outcome. Br J Surg 2007; 94: 145–161.
4Andr
´
en-Sandberg A, Neoptoloemos JP. Resection for
pancreatic cancer in the new millennium. Pancreatology 2002;
2: 431–439.
5 Bachmann MO, Alderson D, Edwards D, Wotton S,
Bedford C, Peters TJ et al. Cohort study in South and West
England of the influence of specialization on the
management and outcome of patients with oesophageal and
gastric cancers. Br J Surg 2002; 89: 914–922.
6 Gillison EW, Powell J, McConkey CC, Spychal RT.
Surgical workload and outcome after resection for carcinoma
of the oesophagus and cardia. Br J Surg 2002; 89: 344 –348.
7 McCulloch P, Ward J, Tekkis PP. ASCOT group of
surgeons; British Oesophago-Gastric Cancer Group.
Mortality and morbidity in gastro-oesophageal cancer
surgery: initial results of ASCOT multicentre prospective
cohort study. BMJ 2003; 327: 1192–1196.
8 Thompson AM, Rapson T, Gilbert FJ, Park KGM. Hospital
volume does not influence long-term survival of patients
undergoing surgery for oesophageal or gastric cancer. Br J
Surg 2007; 94: 578–584.
9 Birkmeyer JD, Siewers AE, Finlayson EV, Stukel TA,
Lucas FL, Batista I et al. Hospital volume and surgical
mortality in the United States. NEnglJMed2002; 346:
1128–1137.
10 Wibe A, Eriksen MT, Syse A, Tretli S, Myrvold HE,
Søreide O et al. Effect of hospital caseload on long-term
outcome after standardization of rectal cancer surgery at a
national level. Br J Surg 2005; 92: 217–224.
11 Rabeneck L, Davila JA, Jessica A, Thompson M,
El-Serag HB. Surgical volume and long-term survival
following surgery for colorectal cancer in the Veterans Affairs
health care system. Am J Gastroenterol 2004; 99: 668–675.
12 Porter GA, Soskolne CL, Yakimets WW, Newman SC.
Surgeon-related factors and outcome in rectal cancer. Ann
Surg 1998; 227: 157–167.
13 Martling AL, Cedermark B, Johanson H, Rutqvist LE,
Holm T. The surgeon as a prognostic factor after the
introduction of total mesorectal excision in the treatment of
rectal cancer. Br J Surg 2002; 89: 1008–1013.
14 Smith JA, King PM, Lane RH, Thompson MR. Evidence of
the effect of ‘specialization’ on the management, surgical
outcome and survival from colorectal cancer in Wessex. Br J
Surg 2003; 90: 583–592.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

D. W. Borowski, D. M. Bradburn, S. J. Mills, B. Bharathan, R. G. Wilson, A. A. Ratcliffe and S. B. Kelly
15 McArdle CS, Hole DJ. Influence of volume and
specialization on survival following surgery for colorectal
cancer. Br J Surg 2004; 91: 610–617.
16 Borowski DW, Kelly SB, Bradburn DM, Wilson RG,
Gunn A, Ratcliffe AA et al. Impact of surgeon volume and
specialization on short-term outcome from colorectal cancer
surgery. Br J Surg 2007; 94: 880–889.
17 Parry JM, Collins S, Mathers J, Scott NA, Woodman CB.
Influence of volume of work on the outcome of treatment for
patients with colorectal cancer. Br J Surg 1999; 86: 475–481.
18 Kee F, Wilson RH, Harper C, Patterson CC, McCallion K,
Houston RF et al. Influence of hospital and clinician
workload on survival from colorectal cancer: cohort study.
BMJ 1999; 318: 1381–1385.
19 Hannan EL, Radzyner M, Rubin D, Dougherty J,
Brennan MF. The influence of hospital and surgeon volume
on in-hospital mortality for colectomy, gastrectomy, and lung
lobectomy in patients with cancer. Surgery 2002; 131: 6–15.
20 Renzulli P, Lowy A, Maibach R, Egeli RA, Metzger U,
Laffer UT. The influence of the surgeon’s and the hospital’s
workload on survival and local recurrence after colorectal
cancer surgery. Surgery 2006; 139: 296–304.
21 Schrag D, Panageas KS, Riedel E, Cramer LD, Guillem JG,
Bach PB et al. Hospital and surgeon procedure volume as
predictors of outcome following rectal cancer resection. Ann
Surg 2002; 236: 583–592.
22 Rogers SO Jr, Wolf RE, Zaslavsky AM, Wright WE,
Ayanian JZ. Relation of surgeon and hospital volume to
processes and outcomes of colorectal cancer surgery. Ann
Surg 2006; 244: 1003–1011.
23 Stocchi L, Nelson H, Sargent D. Impact of surgical and
pathological variables in rectal cancer: a United States
community and cooperative group report. JClinOncol2001;
19: 3895–3902.
24 Soljak M. Volume of procedures and outcome of treatment.
BMJ 2002; 325: 787–788.
25 McArdle CS, Hole DJ. Outcome following surgery for
colorectal cancer: analysis by hospital adjustment for
case-mix and deprivation. Br J Cancer 2002; 86: 331–335.
26 Kingston RD, Walsh S, Jeacock J. Colorectal surgeons in
district general hospitals produce similar survival outcome to
their teaching hospital colleagues: review of 5-year survivals
in Manchester. JRCollSurgEdinb1992; 37: 235 –237.
27 Fielding LP, Stewart-Brown S, Blesovsky L, Kearney G.
Anastomotic integrity after operations for large bowel cancer:
a multicentre study. BMJ 1980; 281: 411–414.
28 National Institute for Health and Clinical Excellence
(NICE). Guidance on Cancer Services – Improving Outcomes in
Colorectal Cancer – Manual Update. NICE: London, 2004.
29 Association of Coloproctology of Great Britain and Ireland.
Guidelines for the Management of Colorectal Cancer (3rd edn).
Association of Coloproctology of Great Britain and Ireland:
London, 2007.
30 Iversen LH, Harling H, Laurberg S, Wille-Jørgensen P.
Influence of caseload and surgical speciality on outcome
following surgery for colorectal cancer: a review of evidence.
Part 1: short term outcome. Colorectal Dis 2006; 9: 28 –37.
31 Iversen LH, Harling H, Laurberg S, Wille-Jørgensen P;
Danish Colorectal Cancer Group. Influence of caseload and
surgical specialty on outcome following surgery for colorectal
cancer: a review of evidence. Part 2: long-term outcome.
Colorectal Dis 2006; 9: 38–46.
32 Schrag D, Cramer LD, Bach PB, Cohen AM, Warren JL,
Begg CB. Influence of hospital procedure volume on
outcomes following surgery for colon cancer. JAMA 2000;
284: 3028–3035.
33 Holm T, Johansson H, Cedermark B, Ekelund G,
Rutqvist LE. Influence of hospital- and surgeon-related
factors on outcome after treatment of rectal cancer with or
without preoperative radiotherapy. Br J Surg 1997; 84:
657–663.
34 Simons AJ, Ker R, Groshen S, Gee C, Anthone GJ,
Ortega AE et al. Variations in treatment of rectal cancer: the
influence of hospital type and caseload. Dis Colon Rectum
1997; 40: 641–646.
35 Hodgson DC, Zhang W, Zaslavsky AM, Fuchs CS,
Wright WE, Ayanian JZ. Relation of hospital volume to
colostomy rates and survival for patients with rectal cancer.
J Natl Cancer Inst 2003; 95: 708–716.
36 Schrag D, Panageas KS, Riedel E, Hsieh L, Bach PB,
Guillem JG et al. Surgeon volume compared to hospital
volume as predictor of outcome following primary colon
cancer resection. J Surg Oncol 2003; 83: 68–78.
37 Meyerhardt JA, Catalano PJ, Schrag D, Ayanian JZ,
Haller DG, Mayer RJ et al. Association of hospital procedure
volume and outcomes in patients with colon cancer at high
risk of recurrence. Ann Intern Med 2003; 139: 649–657.
38 Meyerhardt JA, Tepper JE, Niedzwiecki D, Hollis DR,
Schrag D, Ayanian JZ et al. Impact of hospital procedure
volume on surgical operation and long-term outcomes in
high-risk curatively resected rectal cancer: findings from the
intergroup 0114 study. JClinOncol2004; 22: 166–174.
39 Dripps RD, Lamont A, Eckenhoff JE. The role of anesthesia
in surgical mortality. JAMA 1961; 178: 261–266.
40 National Confidential Enquiry into Perioperative Deaths
(NCEPOD). The Report of the National Confidential Enquiry
into Perioperative Deaths 1992/93. NCEPOD: London, 1995.
41 Turnbull RB Jr, Kyle K, Watson FR, Spratt J. Cancer of the
colon: the influence of the no-touch isolation technic on
survival rates. Ann Surg 1967; 166: 420–427.
42 Kaplan EL, Meier P. Nonparametric estimation from
incomplete observations. JAmStatSoc1958; 53: 457–481.
43 Cox DR. Regression models and life tables. JRStatSoc1972;
34: 187–220.
44 Kee F, Wilson R, Currie S, Sloan J, Houston R,
Rowlands BJ et al. Socioeconomic circumstances and the risk
of bowel cancer in Northern Ireland. J Epidemiol Community
Health 1996; 50: 640–644.
45 Mella J, Biffin A, Radcliffe AG, Stamatakis JD, Steele RJ.
Population-based audit of colorectal cancer management in
two UK health regions. Colorectal Cancer Working Group,
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd

Volume–outcome analysis of colorectal cancer-related outcomes
Royal College of Surgeons of England Clinical Epidemiology
and Audit Unit. Br J Surg 1997; 84: 1731–1736.
46 Tekkis PP, Poloniecki JD, Thompson MR, Stamatakis JD.
ACPGBI Colorectal Cancer Study 2002 Part A: Unadjusted
Outcomes. Association of Coloproctology of Great Britain and
Ireland: London, 2002.
47 Department of Health. The NHS Cancer Plan: a Plan for
Investment, a Plan for Reform. Department of Health:
London, 2000.
48 Department of Health. NHS Plan: a Plan for Investment, a
Plan for Reform. The Stationery Office: London, 2000.
49 Association of Coloproctology of Great Britain and Ireland.
Guidelines for the Management of Colorectal Cancer. Association
of Coloproctology of Great Britain and Ireland: London,
2001.
50 Harmon JW, Tang DG, Gordon TA, Bowman HM, Choti
MA, Kaufman HS et al. Hospital volume can serve as a sur-
rogate for surgeon volume for achieving excellent outcomes
in colorerectal resection. Ann Surg 1999; 230: 404–411.
51 Hermanek P, Mansmann U, Staimmer D, Riedl S,
Hermanek P. The German experience: the surgeon as a
prognostic factor in colon and rectal cancer surgery. Surg
Oncol Clin N Am 2000; 9: 33–49.
52 Engel AF, Oomen JL, Knol DL, Cuesta MA. Operative
mortality after colorectal resection in the Netherlands. Br J
Surg 2005; 92: 1526–1532.
53 Dimick JB, Cowan JA, Upchurch JR, Colletti LM. Hospital
volume and surgical outcomes for elderly patients with
colorectal cancer. JSurgRes2003; 114: 50–56.
54 Urbach DR, Baxter NN. Does it matter what a hospital is
‘high-volume’ for? Specificity of hospital volume– outcome
associations for surgical procedures: analysis of administrative
data. BMJ 2004; 328: 737–740.
55 Harling H, B¨
ulow S, Møller LN, Jørgensen T; Danish
Colorectal Cancer Group. Hospital volume and outcome of
rectal cancer surgery in Denmark 1994–1999. Colorectal Dis
2005; 7: 90–95.
56 Marusch F, Koch A, Schmidt U, Zippel R, Lehmann M,
Czarnetzki HD et al. Effect of caseload on the short-term
outcome of colon surgery: results of a multicenter study. Int J
Colorectal Dis 2001; 16: 362–369.
57 Marusch F, Koch A, Schmidt U, Pross M, Gastinger I,
Lippert H. Hospital caseload and the results achieved
in patients with rectal cancer. Br J Surg 2001; 88:
1397–1402.
58 Kapiteijn E, Marijnen CA, Colenbrander AC, Klein
Kranenbarg E, Steup WH, van Krieken JH et al. Local
recurrence in patients with rectal cancer diagnosed between
1988 and 1992: a population based study in the west
Netherlands. Eur J Surg Oncol 1998; 24: 528–535.
59 Simunovic M, To T, Baxter N, Balshem A, Ross E,
Cohen Z et al. Hospital procedure volume and teaching
status do not influence treatment and outcome measures of
rectal cancer surgery in a large general population.
J Gastrointest Surg 2000; 4: 324–330.
60 Engel J, Kerr J, Eckel R, G ¨
unther B, Heiss M,
Heitland W et al. Influence of hospital volume on
local recurrence and survival in a population sample of
rectal cancer patients. Eur J Surg Oncol 2005; 31:
512–520.
61 Smith JJ, Cornish J, Tekkis PP, Thompson MR. The
National Bowel Cancer Audit Project 2007: Quality Improvement
&OpenReporting. Association of Coloproctology of Great
Britain and Ireland: London, 2008.
62 Kuhry E, Bonjer HJ, Haglind E, Hop WC, Veldkamp R,
Cuesta MA et al. Impact of hospital case volume on
short-term outcome after laparoscopic operation for colonic
cancer. Surg Endosc 2005; 19: 687–692.
63 Salz T, Sandler RS. The effect of hospital and surgeon
volume on outcomes for rectal cancer surgery. Clin
Gastroenterol Hepatol 2008; 6: 1185–1193.
64 Carr ND. Letter to Members of the ACPGBI regarding 20
Resections Rule. Association of Coloproctology of Great
Britain and Ireland: London, 2008.
Copyright 2010 British Journal of Surgery Society Ltd www.bjs.co.uk British Journal of Surgery
Published by John Wiley & Sons Ltd