Cervical cancer screening and registration--are they working?

STUDY
OBJECTIVE: The aim of the study was to investigate the accuracy of one year's cervical cancer registrations and to review these women's medical records.
DESIGN: The study was a survey of medical records of women registered as having a malignancy of cervix or uterus in 1985.
SETTING: Cases were drawn solely from the county of Leicestershire, having been registered by the Trent Cancer Registration Bureau. PATIENTS: The study group comprised all 82 women registered as having a malignancy of cervix or uterus (part unspecified) during the study period. In one case the medical records could not be obtained, so 81 were reviewed. MEASUREMENTS AND
RESULTS: Medical records were reviewed; demographic, clinical and tumour characteristics, and screening experience were noted. Errors of registration of invasive cancer resulted in a 22% overestimate: only 54 of 66 cases registered actually had this cancer. Cases over 40 years of age (n = 37) were compared to those under 40 (n = 17). Thirty two women (59%), predominantly in the older age group, had never had a cervical smear. Forty four women (81%) presented with symptoms rather than as a result of screening. There was no difference in the stage of the disease at presentation between the two age groups but eight of nine women under 35 years had poorly differentiated tumours. At 2 year follow up, 12 (71%) of the younger women were alive and asymptomatic, one had a recurrence and four were dead. The corresponding figures for the older women were 18 (50%), 5 (14%) and 13 (36%). Eight screened women had only had normal smears reported in the 5 years preceding the cancer diagnosis. Seven of these were under 45 years old and four had late stage disease.
CONCLUSIONS: It will be impossible to assess the impact of newly implemented computerised call and recall systems for cervical cytology screening unless a more reliable data base can be created.