EARLY RESCREEN RECALL IN THE UK NATIONAL-HEALTH-SERVICE BREAST SCREENING-PROGRAM - EPIDEMIOLOGIC DATA/

Objective-In the United Kingdom in 1994-95 about 16 500 women aged 50 to 64 were asked to come back earlier than the usual three yearly scre ening interval for further mammography (early rescreen (ES)) or for a range of further investigations at an assessment centre (early recall (ER)). This study aimed at providing epidemiological data on ES/ER, in particular, how often and why it is used, and what the outcome is of using it. Setting-National Health Service breast screening programme i n the United Kingdom. Method-All breast screening units were invited t o complete a postal questionnaire. Two reminders were sent. The units were asked retrospectively to provide breast screening data about wome n aged 50 to 64 from their local computer systems. Women placed on ES/ ER were followed up for at least one year. Results-The response rate w as 71% (69/97), of which 14% (10/69) were unable to provide the requir ed data, leaving 59 completed questionnaires (61%). The rate of placin g women on ES/ER was above 1.0% for 26/54 (48%) responding breast scre ening units and above 2.0% for 13/54 (24%) units. Women were placed on ES/ER because of (a) previous breast cancer (21% of cases; cancer det ection rate 2.4%), (b) diagnostic uncertainty (51%; cancer detection r ate 3.0%), or (c) family history (27%; cancer detection rate 0.6%). Br east screening units with a high rate of placing women on ER were sign ificantly more likely to have a high recall rate (r(s)=0.63; n=53; p<0 .0005) or a high benign surgical biopsy rate (r(s)=0.33; n=49; p<0.05) , or both. The cancer detection rate of ES/ER tended to decrease with increasing ES/ER rates (r(s)=-0.37; n=51; p<0.01). Conclusions-A relat ively large number of women were placed on ES or ER for a range of rea sons. If the recommendations given are followed, the number of women p laced on ER may be reduced while maintaining the cancer detection rate at the required level. The option of ES should not be used.