Completeness of case ascertainment at the Irish National Cancer Registry.

BACKGROUND: Cancer registries play a major role in cancer control, including quantifying the burden of cancer in the population, informing health service planning, and evaluating service provision. Registries must achieve a high level of completeness of case ascertainment. AIM: Completeness of case ascertainment for the National Cancer Registry (NCR) in Ireland was estimated.
METHODS: For all invasive cancers, and the four most common sites, we used two quantitative methods which use information from death certificates to obtain estimates of completeness of registration-the Lincoln-Petersen (LP) estimator (which provides a lower bound estimate) and the flow method (which estimates completeness at a fixed time-point, in this case 5 years). Data were extracted at the end of 2010. Independent case ascertainment was used to assess completeness of breast cancer registration in the 50-64 age group and of rectal cancer in surgical patients diagnosed in 2007.
RESULTS: Completeness of case ascertainment was >99 % for the breast screening age-group and for surgical rectal cancer cases. A lower bound on completeness of case ascertainment for all invasive cancers excluding non-melanoma skin cancers is 94.3 % using the LP estimator. The estimate from the flow method is 97 % after 5 years, [95 % confidence interval (96.2 %, 97.7 %)]. Five-year levels of completeness for the four commonest cancers using the flow method were: colorectal, 97.4 %; lung, 98.7 %; prostate, 95.8 % and breast, 98.0 %.
CONCLUSION: The estimated levels of completeness for the NCR are satisfactory. Further work is required to compare the methods used.