Accuracy of registration of invasive cervical cancer

The quality of the data recorded by the British Columbia Cancer Registry for 521 new cases of invasive cervical cancer was evaluated. The registry's pathological diagnosis in all new registrations of invasive cervical cancer diagnosed in British Columbia between 1977 and 1979 was compared with a best estimate of the true diagnosis, which was determined from the results of the provincial cervical cytology screening program and the clinical charts at the Cancer Control Agency of British Columbia. The registry's data overestimated the true incidence of invasive cervical cancer by approximately 55%, since 184 (35%) of the cases were incorrectly registered. Of the 184, 141 (77%) were cases of preinvasive cervical cancer, 26 (14%) did not meet the criteria for a true case (i.e., they were not newly diagnosed in British Columbia between 1977 and 1979) and 17 (9%) were cases of invasive cancer of another primary site. In addition, 28 cases of invasive cervical cancer diagnosed in the province during the study period had not been reported to the registry. Thus, both over-reporting and under-reporting occurred. There is a need for constant evaluation of registry data if cancer registries are to fulfil their potential contribution to cancer control programs and research.