Using Cancer Registry Data to Assist Efforts in Cancer Control

1
Using Cancer Registry Data to Assist Efforts in
Cancer Control

• Professor Graham Giles
• Victorian Cancer Registry
• Melbourne, AUSTRALIA

2
LATITUDE 330SydneyTripoliDamascus
330
SYDNEY
DAMASCUS
TRIPOLI
EQUATOR
3
Presentation Content

• Presentation 1
• Reviews the purpose of cancer registries what
  information they collect and how it can be used
  to help cancer control initiatives
• Presentation 2
• Reviews cancer registry publications outputs to
  identify good examples for emulation

4
QUESTION 1
Why do we have cancer registries?
5
Chapter 17 the role of cancer registries
6
Cancer Control
The task of cancer registries is to guide and
evaluate cancer control activity on 1 or more
theme -

• Primary prevention
• Early detection screening
• Treatment Palliative care

And to facilitate aetiological research
7
START HERE
Cancer Registry Roles
8
Cancer Control Questions

• How big is the problem?
• new cancer diagnoses
• deaths from cancer
• prevalent cancers
• survival
• What are the trends in cancer?
• Is cancer concentrated in a particular population
  group?
• Are cancer patients being treated appropriately?

9
QUESTION 2
How are the data from cancer registries used?
10
Uses of registry data

• To plan the provision of services
• To ensure that services are equitably provided
• To assess the impact, if any, of public health
  interventions
• To estimate costs
• To assist research
• identifying subjects
• identifying outcomes

11
QUESTION 3
Who uses the data from cancer registries? It is
important to know who your users are - and what
they will want to know
12
Dont over estimate your readers abilities
13
Users of registry data

• Government Departments
• Politicians
• Researchers

Although each user has different requirements,
all need to be provided with simple, accurate,
and interpretable information
14
Cancer Registry Roles

• FIRST to collect complete and accurate data that
  comply with international standards (CANREG)
• SECOND to analyse and present data in ways that
  are appropriate, easily understood and useful to
  the planning and evaluation of cancer control
  programs

15
Data collection

• FIRST task is to collect complete and accurate
  data that comply with international standards
• NB 1 the range of data will depend on the local
  setting, the aims of the registry and resources
• NB 2 data should be limited to items that can be
  obtained in a complete and accurate manner

16
Only collect complete and accurate data
remember GIGO - garbage in - garbage out!
17
New feature coming soon to CANREG
18
Range of data

• Personal identifiers hospital / doctor
  details for follow up record linkage
• Demographic details age, sex, ethnic group /
  tribe
• Diagnostic details basis of diagnosis, screening
  diagnosis
• Tumour details topography, morphology,
  behaviour, stage, markers
• Treatment details surgery, radiotherapy,
  chemotherapy

19
Remember, each additional data item costs
resources
20
Role of Cancer Registry

• Monitoring trends
• Descriptive data analysis
• Gap analysis
• Follow up
• With respect to each of the Cancer Control topics

21
Monitoring trends

• Incidence
• Mortality
• Prevalence
• Survival
• Variations in above indicators by demographic sub
  groups

22
Descriptive analysis

• counts
• rates
• proportions
• errors and confidence intervals
• adjustment / standardisation
• comparisons

23
Counts

• New diagnoses
• Deaths
• Potential years of life lost
• Prevalence

24
The size of the problem
Victoria 2004 Breast Prostate New
cases 2900 3500 Median age 58 70 Deaths 725 800 P
otential years lost 7,100 1,900 Average years
lost 9.8 2.4
25
Cancer Profile Australia
26
Breast Cancer Incidence, Victoria
27
Female cancer in Victoria 2003
28
1987
2003
PROSTATE
29
Rates

• incidence
• mortality

30
Breast Cancer Trends 1982-2000
31
Cervical cancer In UK Ireland 1991-1999
32
(No Transcript)
33
In situ breast cancer, 1982-2000
34
Prostate and breast cancer Incidence and
mortality Australia
35
Lung cancer trends by age
Male
Female
36
(No Transcript)
37
Still Awake?
38
Standardisation

• standardisation
• comparisons

39
Migrant population pyramids
Source Australian Bureau of Statistics, 1996
Census
40
Proportions

• survival
• variation in subgroups

41
Cancer Mortality Comparisons
YEARS OF POTENTIAL LIFE LOST
PROPORTION OF ALL DEATHS
42
Cancer compared with other causes of death
Proportion of all deaths from cancer, heart
attacks and motor vehicle crashes
Cancer
Heart attack
Motor vehicle crashes
43
Causes of Cancer
Source Doll 1998
44
Causes of Cancer
45
Standard errors confidence intervals

• standard errors
• confidence intervals

46
A
u
s
t
r
a
l
i
a
(
r
a
t
e


1
7
.
4
5
/
1
0
0
,
0
0
0

m
e
n
)
a
l
i
t
y

r
a
t
e

r
a
t
i
o
(
w
i
t
h

9
5


C
.
I
.
)
47
Comparisons

• counts
• rates
• proportions
• errors and confidence intervals
• standardisation
• comparisons

48
Risk of Prostate Cancer
49
Breast Cancer Incidence
50
Breast cancer age-specific incidence
51
Female cancer mortality trends
52
Follow up
Following up patients is important for certain
aims

• Survival analysis
• Second primaries
• Clinical management
• Availability for research

53
5-year relative survival, females
ALL CANCER 60 Breast 82 Colon
55 Rectum 57 Melanoma 94 Lung 12 NHL
55 Hodgkins Disease 84 Uterus 81 Ovary
39 Cervix 74
for women diagnosed 1990-1997
54
ALL CANCER
55
COLON CANCER
56
LUNG CANCER
57
Time for tea?