national bowel cancer audit annual report 2013 · annual report 2013 supportive ... 11 per cent...

National Bowel Cancer Audit Annual Report 2013

Supportive Document

Copyright © 2013, Health and Social Care Information Centre, National Bowel Cancer Audit 2013. All rights reserved. 2

Contents

1. Methods 3

2. Trust participation, case ascertainment and data completeness 5

3. Description of patients 7

4. Management of patients 8


1.1 Data processingMultiple recordsIt was assumed that multiple tumour and multiple treatment records involved the same tumour episode if their dates fell within a period of two years. If that was the case an algorithm developed by the Project Team was applied to reconcile potentially conflicting information between the multiple records.

Multiple tumour records If multiple tumour records were available, a second tumour diagnosed within two years was considered a duplicate record, irrespective of the tumour site. Second tumours diagnosed more than two years after a first tumour were considered to be separate cancers.

If a second tumour record was present that was diagnosed within two years, the earliest date of diagnosis and the most advanced or most severe results was taken from the available records. In cases where there was conflicting information about tumour site, this was resolved by choosing the site that was compatible with available treatment information; if no treatment record was available, the most distal site was chosen.

Multiple treatment records In cases where there was conflicting information on treatment data, the most recent date and the value that reflected the most advanced or severe results were taken. Procedures and treatments were assumed to have been carried out if they were recorded in at least one of the multiple treatment records. In cases where there was conflicting information about the surgical procedure, the procedure selected was the one that was most compatible with the site recorded in the tumour record.

Once the new dataset is developed there will be one record per tumour and per treatment, avoiding the issue of multiple records.

1. Methods

1.2 Case ascertainmentThe Hospital Episode Statistics (HES) administrative database, containing records of all admissions to English NHS Trusts, was used to estimate the denominator of this proportion. The corresponding Patient Episode Database for Wales (PEDW) was used to estimate the denominator for Wales. A patient was considered admitted for bowel cancer if a bowel cancer diagnosis was coded (C18, C19 or C20 according to the International Classification of Diseases 10th Revision) in the first diagnosis field. It was assumed that it was a first admission if no other records could be identified since 1 April 2006 with a bowel cancer diagnosis in any of the diagnostic fields.

Case ascertainment is also reported at trust and Cancer Network level for England, and at country and MDT level for Wales. However, if hospitals within a trust are part of different Cancer Networks, case ascertainment is reported at hospital level.


1.3 Handling missing dataThe Audit data set does not allow the distinction between patients who have not undergone a surgical procedure and those for whom the data item is missing. This problem was addressed by searching for any information that indicated that a patient had undergone a surgical procedure (eg number of excised nodes, circumferential margins, and post-operative complications). Patients with missing data on type of surgery, but information indicating that they had undergone surgery, were entered into the category “other procedure”. If such information could not be found, we assumed they had not had surgical treatment.

Similar issues arose for diagnostic and staging procedures. For example, it is reported that a CT or MRI scan was carried out if there was information about the patient’s results from the scan or a date of scan. Otherwise it was assumed that no scan had been carried out.

For the four adjusted outcomes reported at the trust level, multiple imputation was used to fill in any missing information on the risk factors. In addition to the variables in the risk-adjustment model, and the outcomes, the following variables were included in the imputation model: surgical urgency, mode of admission according to the Audit, surgical procedure, number of lymph nodes extracted, number of positive lymph nodes extracted, Index of Multiple Deprivation, length of hospital stay, and days from diagnosis to surgery. Amongst patients undergoing major surgery, 11 per cent were missing ASA grade, 11 per cent were missing TNM T-stage, 9 per cent were missing TNM N-stage and 4 per cent were missing distant metastases. Mode of admission and Charlson comorbidity score came from HES and were both missing in all 4 per cent of patients who were not linked to HES/PEDW. Virtually all patients had complete data on sex, age, and site of cancer.

1.4 Statistical AnalysisFunnel plots Funnel plots are used to make comparisons between networks or between trusts/hospitals. The rate for each Cancer Network or for each trust or hospital is plotted against the total number of patients used to estimate the rate. The “target” is specified as the average rate across all Cancer Networks/trusts/hospitals. In this report, those Cancer Networks, trusts or hospitals with results outside the outer (99.8 per cent) funnel limit are considered as potential outliers.

The funnel limits depend on the target rate and the number of patients included in the estimate; rate estimates have greater uncertainty when estimated from fewer patients. Results fall outside the inner limits if they are statistically significantly different from the target at a 0.05 level, and outside the outer limits if they are statistically significantly different from the target at a 0.002 level. The inner funnel limit is the threshold for an “alert” and the outer funnel level is the threshold for an “alarm”. This implies that 95 per cent of the trusts or hospitals are expected to be within the inner funnel limits and 99.8 per cent within the outer funnel limits, if they are all performing according to the target.


Figure 2.1 Case ascertainment by English Cancer Network/Wales

Network/Wales North of England

Yorkshire

Humber and Yorkshire Coast

Lancashire and South Cumbria

Merseyside and Cheshire

Greater Manchester and Cheshire

North Trent

Wales

Greater Midlands

Pan Birmingham

Arden

East Midlands

Anglia

Three Counties

Avon, Somerset and Wiltshire

Thames Valley

Mount Vernon

Essex

North West London

North London

North East London

South West London

South East London

Peninsula

Dorset

Central South Coast

Surrey, West Sussex and Hampshire

Sussex

Kent and Medway

0 10 20 30 40 50 60 70 80 90 100 110

Per cent (%)

2. Trust participation, case ascertainment and data completeness

Table 2.1 Case ascertainment by year for England and Wales

2008-09 2009-10 2010-11* 2011-12

Patients identified in HES/PEDW 32,620 32,935 33,749 34,220

Patients identified in audit 22,972 27,344 29,094 29,445

% case ascertainment 70 83 86 86

* Wales data for 2010-11 is for 1 August 2010 to 31 March 2011, as this was the audit period in use then, and there is no audit data for Wales between 1 April 2010 and 1 August 2010.


Figure 2.2 Percentage of patients undergoing major surgery with complete data on the 7 items from the Audit used in risk adjustment, by English Cancer Network/Wales


Yorkshire





North Trent

Wales

Greater Midlands

Pan Birmingham

Arden

East Midlands

Anglia

Three Counties


Thames Valley

Mount Vernon

Essex

North West London

North London

North East London

South West London

South East London

Peninsula

Dorset

Central South Coast


Sussex

Kent and Medway

0 10 20 30 40 50 60 70 80 90 100

Per cent (%)

Table 2.2 Percentage of patients undergoing major surgery with complete data on the 7 items from the Audit used in risk adjustment, by Audit year

2008-09 2009-10 2010-11 2011-12

N % N % % %

Total patients undergoing major resection 14,522 16,634 17,186 17,250

Complete data on 7 key items 10,174 70.1 12,180 73.2 13,223 76.9 13,612 78.9

Incomplete data on 7 key items 4,348 29.9 4,454 26.8 3,963 23.1 3,638 21.1


3. Description of patients

Table 3.1 Characteristics of all patients with bowel cancer included in the current audit report

Number %

Total number of reported cases 29,445

Total number of surgically treated cases 21,190 72.0

Total number of major resections 17,250 58.6

Sex Male 16,741 56.9

Female 12,703 43.1

Unknown (% of total) 1 (0.0)

Age-group <65 yrs 8,314 28.2

65-74 yrs 8,840 30.0

75-84 yrs 8,819 30.0

85+ yrs 3,472 11.8

Cancer Site Colon 18,500 63.1

Rectosigmoid 1,460 5.0

Rectum 9,340 31.9

Unknown (% of total) 145 (0.5)

Table 3.2 Characteristics of 29,300 patients with a known cancer site

Colon Rectosigmoid Rectal

Number % Number % Number %

Total patients per cancer site 18,500 1,460 9,340

Sex Male 9,804 53.0 876 60.0 5,978 64.0

Female 8,695 47.0 584 40.0 3,362 36.0

Missing (%) 1 (0) 0 (0) 0 (0)

Age-group <65 yrs 4,662 25.2 422 28.9 3,180 34.0

65-74 yrs 5,506 29.8 435 29.8 2,852 30.5

75-84 yrs 5,919 32.0 445 30.5 2,427 26.0

85+ yrs 2,413 13.0 158 10.8 881 9.4

TNM T-stage T1 826 6.9 76 9.3 643 13.9

T2 1,302 10.9 123 15.0 1,291 27.9

T3 6,041 50.7 434 53.0 2,290 49.5

T4 3,755 31.5 186 22.7 403 8.7

Missing (% of total) 6576 (35.5) 641 (43.9) 4713 (50.5)

TNM N-stage N0 6,517 55.3 464 57.3 3,008 64.5

N1 3,087 26.2 213 26.3 1,076 23.1

N2 2,191 18.6 133 16.4 581 12.5

Missing (% of total) 6705 (36.2) 650 (44.5) 4675 (50.1)

Distant metastases No 11,528 75.1 810 70.4 5,094 76.2

Yes 3,822 24.9 340 29.6 1,589 23.8

Missing (% of total) 3150 (17) 310 (21.2) 2657 (28.4)

Liver metastasis Liver metastasis 2,471 18.9 245 24.3 1,054 15.2

Normal Liver 9,810 75.0 703 69.7 5,462 78.8

Liver uncertain 798 6.1 61 6.0 412 5.9

Missing (% of total) 5421 (29.3) 451 (30.9) 2412 (25.8)


4. Management of patients

Table 4.1 Description of management of the 29,300 patients with a known cancer site



Total patients per cancer site 18,500 1,460 9,340

Patients undergoing surgery 13,957 1,050 6,038

Discussed at multidisciplinary team meeting Yes 17,761 97.6 1,407 98.5 8,946 98.0

No 430 2.4 21 1.5 187 2.0

Missing (% of total) 309 (1.7) 32 (2.2) 207 (2.2)

Seen by clinical nurse specialist Yes 12,984 87.0 1,020 88.1 6,818 89.0

No 1,933 13.0 138 11.9 845 11.0

Missing (% of total) 3583 (19.4) 302 (20.7) 1677 (18.0)

Had CT scan* Yes 16,532 89.4 1,289 88.3 8,316 89.0

No 1,968 10.6 171 11.7 1,024 11.0

Surgery type Major resection 11,670 83.6 835 79.5 4,615 76.4

Local excision 493 3.5 45 4.3 480 7.9

Non resectional procedure 536 3.8 96 9.1 419 6.9

Other procedure 1,258 9.0 74 7.0 524 8.7

No surgery (% of total) 4543 (24.6) 410 (28.1) 3302 (35.4)

Urgency of operation Elective 7,962 58.9 622 61.6 4,205 71.8

Scheduled 2,075 15.4 194 19.2 1,130 19.3

Urgent 1,640 12.1 96 9.5 350 6.0

Emergency 1,834 13.6 97 9.6 169 2.9

Missing (% of total) 446 (2.4) 41 (2.8) 184 (2.0)

No surgery (% of total)

Laparoscopy Open 6,008 53.1 383 46.9 2,334 50.3

Laparoscopic then open 287 2.5 30 3.7 166 3.6

Laparoscopic converted to open

497 4.4 60 7.3 244 5.3

Laparoscopic completed 4,520 40.0 344 42.1 1,892 40.8

Missing (% of total) 2645 (14.3) 233 (16.0) 1402 (15.0)

No surgery (% of total) 4543 (24.6) 410 (28.1) 3302 (35.4)

*Yes if patient has a result of CT scan or date of CT scan


Table 4.2 Description of the 17,120 patients who underwent major surgery by cancer site



Total patients undergoing major resection 11,670 835 4,615

Sex Male 6,164 52.8 520 62.3 3,038 65.8

Female 5,505 47.2 315 37.7 1,577 34.2

Missing (% of total) 1 (0.0) 0 (0.0) 0 (0.0)

Age-group <65 yrs 3,147 27.0 269 32.2 1,782 38.6

65-74 yrs 3,702 31.7 289 34.6 1,569 34.0

75-84 yrs 3,777 32.4 236 28.3 1,093 23.7

85+ yrs 1,044 8.9 41 4.9 171 3.7

ASA 1 1,369 13.2 117 15.4 764 18.5

2 5,557 53.7 428 56.5 2,477 59.2

3 3,033 29.3 191 25.2 863 20.6

4 or 5 383 3.7 22 2.9 80 1.9

Missing (% of total) 1328 (11.4) 77 (9.2) 431 (9.3)

TNM T-stage T1 554 5.2 57 7.7 408 10.3

T2 1,169 11.0 117 15.7 1,139 28.6

T3 5,553 52.5 402 54.1 2,080 52.3

T4 3,309 31.3 167 22.5 349 8.8

Missing (% of total) 1085 (9.3) 92 (11) 639 (13.8)

TNM N-stage N0 5,844 55.1 423 57.1 2,649 63.8

N1 2,781 26.2 200 27.0 981 23.6

N2 1,978 18.7 118 15.9 522 12.6

Missing (% of total) 1067 (9.1) 94 (11.3) 463 (10.0)

Distant metastases No 9,732 86.6 698 87.0 4,024 92.0

Yes 1,510 13.4 104 13.0 348 8.0

Missing (% of total) 428 (3.7) 33 (4.0) 243 (5.3)

Mode of admission (from HES) Elective 8,246 78.3 681 89.3 3,994 95.5

Emergency 2,289 21.7 82 10.7 190 4.5

Missing (% of total) 1135 (9.7) 72 (8.6) 431 (9.3)

Cancer site Caecum/ascending colon 4,752 40.7 0 0.0 3 0.1

Hepatic flexure 719 6.2 0 0.0 0 0.0

Transverse colon 1,078 9.2 0 0.0 1 0.0

Splenic flexure/descending colon 1,039 8.9 0 0.0 4 0.1

Sigmoid colon 4,082 35.0 0 0.0 35 0.8

Rectosigmoid 0 0.0 835 100.0 398 8.6

Rectal 0 0.0 0 0.0 4,174 90.4

Comorbidities 0 6,630 59.6 496 62.2 2,959 66.6

1 3,204 28.8 213 26.7 1,165 26.2

2+ 1,285 11.6 89 11.2 321 7.2

Missing (% of total) 551 (4.7) 37 (4.4) 170 (3.7)


Figure 4.1 Distant metastases around time of initial treatment in all patients by English Cancer Network/Wales


Yorkshire





North Trent

Wales

Greater Midlands

Pan Birmingham

Arden

East Midlands

Anglia

Three Counties


Thames Valley

Mount Vernon

Essex

North West London

North London

North East London

South West London

Peninsula

Dorset

Central South Coast


Sussex

Kent and Medway

0 5 10 15 20 25 30 35

Per cent (%)


Figure 4.2 Major surgery carried out as an urgent or emergency procedure by English Cancer Network/Wales


Yorkshire





North Trent

Wales

Greater Midlands

Pan Birmingham

Arden

East Midlands

Anglia

Three Counties


Thames Valley

Mount Vernon

Essex

North West London

North London

North East London

South West London

South East London

Peninsula

Dorset

Central South Coast


Sussex

Kent and Medway

0 5 10 15 20 25 30

Per cent (%)

Published by the Health and Social Care Information Centre

This publication may be requested in large print or other formats.

For further information:

www.hscic.gov.uk0845 300 6016 [email protected]

Copyright © 2013 Health and Social Care Information Centre. All rights reserved.

This work remains the sole and exclusive property of the Health and Social Care Information Centre and may only be reproduced where there is explicit reference to the ownership of the Health and Social Care Information Centre.

This work may be re-used by NHS and government organisations without permission.

national bowel cancer audit annual report 2013 · annual report 2013 supportive ... 11 per cent...

Documents