This report has been produced with the assistance of
the staff of the Cancer Registry, Hospital Registries, Clinical Epidemiology
Unit, Health Statistics Unit, Information Management Services and other
staff of the South Australian Department of Health.
Staff from the Cancer Council of South Australia have also assisted.
SA DEPARTMENT OF HEALTH
Dr Wayne Clapton
Ms Joanne Bell
Mrs Maria Cirillo
Ms Heather Hall
Ms Mary Merdo
Mrs Teresa Molik
Ms Elaine Morton
Ms Maxene Rosenberg
Ms Christine Scott
Dr Colin Luke
SA HOSPITAL-BASED CANCER REGISTRIES COORDINATION
Ms Lesley Milliken
Mr Graeme Tucker
Mr Kevin Priest
Mr Adrian Heard
Ms Anh-Minh Nguyen
INFORMATION MANAGEMENT SERVICES
Ms Bridget Milanowski
Mrs Sandra Sowerby
STRATEGIC RESEARCH AND ANALYSIS
Mr Chris Gascoigne
CANCER COUNCIL SA
Dr David Roder
This report includes 22 maps. There are
incidence and death maps for the 11 following cancers – breast, prostate,
lung, colon, rectum, melanoma, leukemia, lymphoma, stomach, pancreas and
cervix. All maps show cancer incidence and death for the period
1991-2000. The rates presented in the maps are annualised rates per
The maps all use age-sex standardised data. The
process of direct standardisation ensures that different age sex profiles
for each geographical area are re-calibrated to become directly comparable.
The unit for each dot shown on the map is a statistical local area (SLA),
using the 1998 designated boundaries. For each map 113 SLAs are shown
together with 4 unincorporated areas. There are seven unincorporated
areas which are not shown on the maps as either the population of the area
is zero or residents of the area cannot easily be distinguished from a
neighbouring area (eg. Unincorporated West Coast).
There are a number of problems with mapping cancer
data. Firstly, for some cancers such as lung cancer there is a long
latency period between the exposure which caused the disease and the
diagnosis of cancer. This means that people with lung cancer may move
location, say from a rural area to a regional centre between exposure and
diagnosis, thus inflating the cancer incidence rate of the regional centre.
Secondly, addresses recorded in the South Australian Cancer Registry are not
always detailed enough to accurately assign a SLA. This leads to the
loss of about 2% of data and potentially inaccurate assignment of a further
5% of data.
Caution needs to be taken in putting too much weight
on individual dots on the maps. The population used to assign an
incidence or death rate varies from over 33,000 in the case of
Onkaparinga/Woodcroft in Adelaide’s south to less than 1,000 in
Orroroo/Carrieton in the mid north. With a number of rural SLAs and
unincorporated areas have very small populations, the recording of a small
number of a particular type of cancer can lead to a high incidence rate.
It is where geographical groupings of dots show a consistent pattern that
more accurate interpretations of the data can be made.
There are several clear patterns which emerge on some of the maps.
Firstly, cancers such as breast, melanoma and prostate cancer, which require
screening or a medical check for detection, almost always have higher
incidence rates in high socio-economic status areas such as eastern and
inner southern Adelaide. Secondly, some cancers such as lung and
stomach cancer have a well documented link with low socio-economic status
and incidence and death rates are higher for northern and western Adelaide,
and for rural areas like the Iron Triangle and the Riverland. Thirdly,
for the majority of cancers there a no overall differences in incidence and
mortality between city and country areas. The only exception to this
is prostate cancer where there is higher mortality in country areas than