Methodological issues in the development of the Canadian Cancer Incidence Atlas.
Stat Med
; 19(17-18): 2437-49, 2000.
Article
in En
| MEDLINE
| ID: mdl-10960864
ABSTRACT
The Canadian Cancer Incidence Atlas is among recent national atlases using incidence rather than mortality data. Methods used to assess the significance and spatial correlation of the age-standardized rates (ASIRs) for the 290 census divisions are described. The expected number of cases by area was used to determine cancer sites with sufficient cases to be mapped. ASIR significance was assessed using a simulation based on a Poisson distribution. The consistency of the observed case distributions with the Poisson distribution was examined. The bootstrap confidence interval (CI) for the ASIR developed by Swift was used in the atlas. Spatial correlation was assessed with Moran's I/I(max) and the significance determined by a simulation in order to allow for area population variation. Data quality indicators typically used for cancer registries were presented, supplemented by a registry questionnaire.
Search on Google
Collection:
01-internacional
Database:
MEDLINE
Main subject:
Atlases as Topic
/
Cluster Analysis
/
Data Collection
/
Maps as Topic
/
Neoplasms
Type of study:
Incidence_studies
/
Prognostic_studies
/
Risk_factors_studies
Limits:
Female
/
Humans
/
Male
Country/Region as subject:
America do norte
Language:
En
Journal:
Stat Med
Year:
2000
Document type:
Article