Evaluation of data quality in the National Prostate Cancer Register of Sweden

Abstract

BackgroundData in cancer quality registers are increasingly used for quality assurance, benchmarking, and research. Materials and methodsData in the National Prostate Cancer Register (NPCR) of Sweden were evaluated for completeness, timeliness, comparability and validity. Completeness and timeliness were assessed by cross-linkage to the Swedish Cancer Register, comparability was examined by comparing registration routines in NPCR with national and international guidelines, and validity was assessed by re-abstraction of data from medical charts for 731 men diagnosed with prostate cancer (Pca) in 2009. Furthermore, data on treatment were validated by record linkage to the Swedish Patient Register and The Prescribed Drug Register. ResultsNPCR captured 98% of Pca cases in the Cancer Register and the mean value for completeness of the 48 evaluated variables was 90% (range 64–100%). Timeliness increased substantially from 2008 to 2012 with 95% of cases reported within 12months after diagnosis in 2012. NPCR complied with national and international coding routines. Overall, the agreement between original data and re-abstracted data from 731 charts was high. For example, the correlation between original and re-abstracted data was 1.00 for date of surgery, and 0.97 for serum levels of prostate specific antigen and exact agreement was 97% for Gleason score at biopsy, 83% for clinical local T stage and more than 95% of the androgen deprivation therapies registered in NPCR had a corresponding filled prescription. ConclusionRecord linkages with other data sources and re-abstraction of data showed that data quality in NPCR is high.