Rapid review (partial rescreening) of cervical cytology. Four years experience and quality assurance implications.

AIMS: To determine the sensitivity of the partial rescreening method of rapid review for internal quality control of cervical cytology; to determine which staff members are most suited to undertake it; and to investigate the cell patterns of false negative smears previously detected by the method.
METHODS: As a prospective study 9517 cervical smears were partially screened by four cytotechnologists using the 'step' method prior to conventional screening and the results compared with the final report. As a retrospective study 62 false negative smears that had been identified by the method over four years were reviewed.
RESULTS: A detection rate for dyskaryosis of 86% (range 82-91%) was achieved. Sixteen abnormal smears were missed on conventional screening that had been detected by prescreening. Review of the 62 false negatives revealed three patterns: (1) scanty abnormal cells; (2) abundant dyskaryotic cells presenting as "microbiopsies"; and (3) abundant, readily recognisable abnormal cells.
CONCLUSIONS: Partial rescreening enables the detection of errors due to both fatigue and misinterpretation. In this laboratory the method has, together with targeted full rescreening, reduced the false negative report rate from 5.0% to 0.4%. For laboratories using a rapid review method to reduce false negative reports, a prescreening trial is recommended in order (1) to select the most effective review method and the staff most suited to undertake it; and (2) to determine the laboratory's sensitivity with the method, as this is required for accurate estimation of the false negative rate.