AIMS AND BACKGROUND: The aim of this study was to assess the performance of routine breast cytology in a group of Italian laboratories, as a basis for interlaboratory quality control. METHODS: A multicenter retrospective consecutive series of 23,063 aspirates from breast lesions (5,130 invasive carcinoma, 247 intraductal carcinoma, 2,386 benign lesions [histological], 13,144 benign lesions [follow-up], 2,156 other [no follow-up]) was reviewed. Standard accuracy measurements (inadequacy rate, sensitivity, specificity, predictive value) were determined, as well as the contribution of diagnostic tests such as physical examination and instrumental diagnostic procedures (mammography, sonography). RESULTS: Overall, inadequacy rate was 6.5% for malignant and 23.2% for benign lesions. Sensitivity was 92.2%, specificity was 95.3%, and the positive predictive value of an equivocal or malignant report was 53.4% or 96.9%, respectively. Sensitivity was associated with histological type (intraductal vs invasive: 77.0% vs 92.9%, p < 0.000001; lobular vs other invasive: 85.5% vs 94%, p = 0.0025) as well as with tumor size (pT1 = 91.6%, pT2 = 94.3%, pT3-4 = 96.0%, p < 0.000001). Observed differences in performance indicators between centers may be explained by variability in sampling or reading accuracy, and warrant the need for some adjustment in single centers, especially, as far as the criteria to report an equivocal smear are concerned. Cytology contributed to detect otherwise unsuspected carcinomas, although in some centers this benefit was poor and not cost-effective because a large number of unnecessary biopsies were caused by false equivocal/positive reports. CONCLUSIONS: Although centers scored, on average, well within standards recommended by national authorities, this study suggests the opportunity of interlaboratory quality control to achieve more homogeneous criteria.
AIMS AND BACKGROUND: The aim of this study was to assess the performance of routine breast cytology in a group of Italian laboratories, as a basis for interlaboratory quality control. METHODS: A multicenter retrospective consecutive series of 23,063 aspirates from breast lesions (5,130 invasive carcinoma, 247 intraductal carcinoma, 2,386 benign lesions [histological], 13,144 benign lesions [follow-up], 2,156 other [no follow-up]) was reviewed. Standard accuracy measurements (inadequacy rate, sensitivity, specificity, predictive value) were determined, as well as the contribution of diagnostic tests such as physical examination and instrumental diagnostic procedures (mammography, sonography). RESULTS: Overall, inadequacy rate was 6.5% for malignant and 23.2% for benign lesions. Sensitivity was 92.2%, specificity was 95.3%, and the positive predictive value of an equivocal or malignant report was 53.4% or 96.9%, respectively. Sensitivity was associated with histological type (intraductal vs invasive: 77.0% vs 92.9%, p < 0.000001; lobular vs other invasive: 85.5% vs 94%, p = 0.0025) as well as with tumor size (pT1 = 91.6%, pT2 = 94.3%, pT3-4 = 96.0%, p < 0.000001). Observed differences in performance indicators between centers may be explained by variability in sampling or reading accuracy, and warrant the need for some adjustment in single centers, especially, as far as the criteria to report an equivocal smear are concerned. Cytology contributed to detect otherwise unsuspected carcinomas, although in some centers this benefit was poor and not cost-effective because a large number of unnecessary biopsies were caused by false equivocal/positive reports. CONCLUSIONS: Although centers scored, on average, well within standards recommended by national authorities, this study suggests the opportunity of interlaboratory quality control to achieve more homogeneous criteria.