figshare
Browse
ionc_a_1054948_sm8011.pdf (95.31 kB)

Quality assessment of prostate cancer reports to the Danish Cancer Registry

Download (0 kB)
journal contribution
posted on 2015-12-23, 10:06 authored by Inga Jóna Ingimarsdóttir, Ea Rusch, Gerda Engholm, Hans H. Storm, Klaus Brasso

Background. The Danish Cancer Registry (DCR) is the oldest nationwide population-based cancer registry in the Nordic countries. At the time of the study the DCR recorded date of diagnosis, tumor stage and initial treatment. The validity of the clinical information reported to the DCR has never been analyzed.

Material and methods. Patients diagnosed with prostate cancer from 1 May to 31 December 1997 and living in eight Danish counties were identified through the DCR. Clinical data was retrieved from hospital records where date of diagnosis, stage at diagnosis and treatment received were registered.

Results. The mean age at diagnosis was 74.6 years (s.d. 8.6 years). Diagnosis was verified histologically for 87% of cases. Overall 95% of the patients had a difference less than three months between the reported date of diagnosis and the date found in hospital records. Correction of dates of diagnosis had no impact on survival. Hospital records identified 86 patients with T1-2 disease without distant metastases (M0), but only 56 of these patients (65%) were reported to the DCR as having localized disease. According to hospital records a total of 242 patients were confirmed having distant metastases (M1) at diagnosis but only 139 of these cases (57%) were reported to the DCR as such. Considerable “over reporting” of curative treatment was observed.

Conclusion. The DCR has been shown to be reliable in terms of new cases being reported. For the majority of cases there were insignificant differences concerning the date of diagnosis. However, the DCR information on stage and treatment was found to be inaccurate. Since 2004 the DCR registration process, including staging according to the TNM classification, has been carried out electronically from several registers. Future comparison between cohorts of different time intervals or international comparison should be interpreted with caution when clinical information is included.

History