Detection of false negative pap smears by rapid reviewing - A metaanalysis

OBJECTIVE: To explore the diagnostic validity of rapid reviewing (RR) as a quality control method in cytologic laboratories. STUDY DESIGN: Fourteen studies dealing with the detection of false negative Pap smears by RR were included in a metaanalysis. RESULTS: The overall additional yield of positive slides, expressed as the percentage of all reviewed slides, is: 0.18% (95% confidence interval [CI]: .14-.21) for all cytologic abnormalities; 0.07% (CI:.05-.09) for squamous i ntraepithelial lesions (SIL) and 0.02% (CI:.01-.03) for high grade SIL. The false negative rate of primary screening, evaluated by RR, was 2.0% (CI: 1 .5-2.6) for all cytologic abnormalities and 1.4% (CI:.8-2.1) for high grade SIL. The specificity of rapid rescreening was estimated as 97.2% (CI: 96.4 -98.1). The positive predictive value of suspicion at RR is about 8.8%. Sev en references contained historical data on full rescreening of a random sam ple of slides reported originally as negative. The results were also pooled and compared with RR. Complete rescreening is more sensitive, but if appli ed on only 10% of the negative workload, it would yield, on average, 4.7 ti mes fewer extra positives, 5.6 times fewer SIL and 7.9 times fewer high gra de SIL in comparison with RR of all sides. CONCLUSION: RR of all smears initially reported as nonpositive is a more ef fective and a fortiori a more cost effective quality control method in comp arison with full rescreening of a 10% random sample.