Data quality at the Cancer Registry of Norway: an overview of comparability, completeness, validity and timeliness

IK Larsen, M Småstuen, TB Johannesen… - European journal of …, 2009 - Elsevier
IK Larsen, M Småstuen, TB Johannesen, F Langmark, DM Parkin, F Bray, B Møller
European journal of cancer, 2009Elsevier
AIM: To provide a comprehensive evaluation of the quality of the data collected on both solid
and non-solid tumours at the Cancer Registry of Norway (CRN). METHODS: Established
quantitative and semi-quantitative methods were used to assess comparability,
completeness, accuracy and timeliness of data for the period 1953–2005, with special
attention to the registration period 2001–2005. RESULTS: The CRN coding and
classification system by and large follows international standards, with some further …
AIM
To provide a comprehensive evaluation of the quality of the data collected on both solid and non-solid tumours at the Cancer Registry of Norway (CRN).
METHODS
Established quantitative and semi-quantitative methods were used to assess comparability, completeness, accuracy and timeliness of data for the period 1953–2005, with special attention to the registration period 2001–2005.
RESULTS
The CRN coding and classification system by and large follows international standards, with some further subdivisions of morphology groupings performed in-house. The overall completeness was estimated at 98.8% for the registration period 2001–2005. There remains a variable degree of under-reporting particularly for haematological malignancies (C90–95) and tumours of the central nervous system (C70–72). For the same period, 93.8% of the cases were morphologically verified (site-specific range: 60.0–99.8%). The under-reporting in 2005 due to timely publication is estimated at 2.2% overall, based on the number of cases received at the registry during the following year.
CONCLUSION
This review suggests the routines in place at the CRN yields comparable data that can be considered reasonably accurate, close-to-complete and timely, thereby justifying our policy of the reporting of annual incidence 1 year after the close of registration.
Elsevier