OBJECTIVETo evaluate the results of the active follow-up among registered cancer patients in 2002 - 2005 in urban areas of Beijing.

METHODSA number of 63 997 cancer patients diagnosed during 2002 - 2005 were selected from the surveillance database of Beijing Office for Cancer Prevention and Control. By matching the identity information of the patients with the death surveillance database built by the vital statistic department in Beijing, 29 223 patients were confirmed to be alive.1149 cases were removed from the study due to lack of exact key variables, such as address and telephone numbers. 28 074 patients were, at last, included in the active follow-up study. The investigators and the inspectors, who accepted standard training program, investigated each patient's status of census register and survival condition by phone calling, household interview and visits at local police station or residential committee. The loss ratio of follow-up and the constituent ratio of the withdrawal reasons were calculated.

RESULTSAmong the 28 074 patients selected in active follow-up, 21 696 patients were followed successfully; 1453 of whom didn't have the census register of Beijing, which accounted for 6.70%. Out of the other 20 243 Beijing residents, 4715 patients (23.29%) were already dead and 84.22% (3971/4715) of them replenished the failure to report by passive follow-up. Among all the 4715 dead cases, 4405 (93.43%) patients were died from cancer. The follow-up study helped to replenish the vital statistics in different districts, the ratio ranged from 4.87% and 8.85%. 6378 patients were withdrawn from the study. The loss ratio was 22.72% (6378/28 074), and the total loss ratio was 12.03% ((6378 + 1149)/(63 997 - 1453)). Of these withdrawal cases, 3041 (47.68%) were lost to follow-up in that the investigators can't find the patients or the relatives of the patients according to the registered phone number or address information. The other reasons included: the patients removed to other areas (1199 cases, 18.80%), the patients and their family members were temporarily not at home (127 cases, 1.99%), the patients and their family members rejected to answer the interview (292 cases, 4.58%), and other reasons (1719 cases, 26.95%).

CONCLUSIONThe method of active follow-up towards registered cancer patients can replenish the missing information which could not be collected from passive follow-up procedure; and therefore effectively improve the quality of data in cancer registration.