Transcript Slide 1

Biogrid: Evaluating and informing evidencebased practice in oncology
Dr Kathryn Field
Colorectal cancer
Treatment options
Clinician choices
Patient choices
Are we following the evidence?
Are we contributing to the
evidence?
Issues with clinical trials
Patients treated in trials are




Younger (on average 10 years)
Fitter (better performance status)
Fewer comorbidities
More likely to have normal organ function
than patients in routine clinical practice
1. Elderly cancer patients
 renal
 liver
function
function
 Polypharmacy
 Comorbidities
2. Overweight/obese patients
• 1.2 million women
•Aged 50-64 during 1996-2001
•Follow up av 5.4y (cancer incidence) and 7.0y (cancer
mortality)
BMJ 6th November 2007
How do we capture this?
 Improve
data collection in routine clinical
practice
 Link
databases to cross check and verify
data and improve completeness and
accuracy
ACCORD database
Biogrid Australia
 Individual
clinicians “own” the data
 Data
can be linked for research & audit
 Data
linkage and analysis support
BioGrid Linkage
Query/Reporting tools
AustinRadiotherapy
RMHChemo
St Vs
Chemo
Peter Mac
Surgery
1.Is Australian colorectal cancer
management up to international
standards?
Lymph node yield for CRC
surgery
Influenced by:
•Surgical factors
•Pathologist factors
•Patient/tumour
factors
Lymph node yield (2)
Large studies already carried out internationally
Year of diagnosis
20
08
20
06
20
04
20
02
20
00
19
98
19
96
19
94
19
92
19
90
19
88
Median LNY
Median LNY
16
14
12
10
8
6
4
2
0
2. How should we treat elderly
patients on the basis of current
evidence?
7x RCT
3351 patients
Stage II/III
X-ACT
Study
N=1987
Median age= 62 (80% <70yo)
Dose reductions in 42%
Stage 3 colon
cancer
5FU
(node positive)
–2000pts
XELODA
Median age
245 patients, stage II/III, Jan 2003 – Feb 2008
70
68
66
64
62
60
58
56
capecitabine
5FU/LV
FOLFOX
Dose reductions
80%
70%
60%
50%
0
40%
1
>1
30%
20%
10%
0%
capecitabine
5FU/LV
FOLFOX
•US survey of surgeons/medical oncologists
•Hypothetical patients:
•Stage III colon cancer
•55yo versus 80yo
•Comorbidities: none/moderate/
severe CCF
Overall
(n=252)
<60 years
(n= 66)
60-70 years
(n= 77)
71-80 years
(n=67)
>80 years
(n=42)
Chemotherapy not given
58 (23.0%)
1 (1.5%)
4 (5.2%)
18 (26.9%)
35 (83.3%)
1. Not recommended
-Age
-Comorbidity
-Age & comorbidity
13 (22%)
10 (17%)
17 (29%)
0 (0%)
0 (0%)
0 (0%)
0 (0%)
1 (25%)
1 (25%)
1 (6%)
4 (22%)
3 (17%)
12 (34%)
5 (14%)
13 (37%)
2. Patient declined treatment
14 (24%)
1 (100%)
1 (25%)
8 (44%)
4 (11%)
3. Other reason
4 (7%)
0 (0%)
1 (25%)
2 (11%)
1 (3%)
•4 Melbourne hospitals
•Jan 2003-Feb 2008
•Stage III
3. Are we looking after
overweight/obese patients
appropriately?
Dosing in obese patients
Cumulative incidence mortality by BMI category
A = following colon cancer events
HR 1.36
B = other deaths
Age and obesity
70
68
66
64
62
60
58
<21
21-24.9
25-29.9
30-34.9
>35
BMI and participation in clinical trials
50.00%
45.00%
40.00%
35.00%
30.00%
Clinical trial
25.00%
stage IV patients overall
20.00%
15.00%
10.00%
5.00%
0.00%
<21
21-24.9
25-29.9
BMI (kg/m2)
30-34.9
>/=35
4. Can we trust the available
evidence?
Benefits of linkage
Br J Cancer. 1994 Dec;70(6):1229-31.
The impact on colorectal cancer survival of cases registered by 'death certificate only': implications for
national survival rates.
Pollock A, Vickers N
Med Care. 2004 Nov;42(11):1111-6.
Using Medicare data to estimate the number of cases missed by a cancer registry: a 3-source capturerecapture model.
McClish D, Penberthy L
N Z Med J. 2002 May 24;115(1154):246-7.
Can cancer centres in New Zealand help the cancer registry generate survival data? A pilot study in
prostate cancer.
Evans TK, Lamb DS, Cornes DA et al.
M staging
Survival
Matching of VCR death dates to
hospital death data
How is Biogrid helping?

Excellent resource for clinical research

Linkages within and between hospitals
enhances opportunities, allows
comparison with international data

Linkage between organizations
enhances accuracy and reliability
Thank you