M-H Wu1, C-Y Chiang, R Jou, S-Y Chang, K-T Luh. 1. Reference Laboratory of Mycobacteriology, Research and Diagnostic Center, Centers for Disease Control, Department of Health, Taiwan.
Abstract
OBJECTIVE: To evaluate the quality of sputum smear microscopy in nine Taiwan Centers for Disease Control contract laboratories, an external quality assessment (EQA) programme has been implemented since 2005. DESIGN: A sampling strategy based on the lot quality assurance system was applied to select slides for rechecking. Supervisory visits and technical training were conducted to determine the causes of errors and to take corrective action. RESULTS: Of the 1017 slides sampled in 2005, 637 (63%) had proper smear size, 492 (48%) proper thickness and 884 (87%) proper staining; the corresponding figures were 972 (100%), 748 (77%) and 809 (99.6%) for the 972 slides rechecked in 2006. After training, the quality of size and staining of smear preparation had significantly improved (P < 0.001) in 2006. Rechecking of 981 readable slides in 2005 identified 3 (0.3%) high false-negatives, 3 (16.7%) low false-positives and 26 (2.8%) low false-negatives; the corresponding errors were 3 (0.3%), 8 (28.6%) and 12 (1.3%) for the 972 slides rechecked in 2006. Of the eight laboratories, two (25%) and four (50%) reached 80% sensitivity in 2005 and 2006, respectively. CONCLUSION: Technical training and EQA improved the quality of sputum smear microscopy services.
OBJECTIVE: To evaluate the quality of sputum smear microscopy in nine Taiwan Centers for Disease Control contract laboratories, an external quality assessment (EQA) programme has been implemented since 2005. DESIGN: A sampling strategy based on the lot quality assurance system was applied to select slides for rechecking. Supervisory visits and technical training were conducted to determine the causes of errors and to take corrective action. RESULTS: Of the 1017 slides sampled in 2005, 637 (63%) had proper smear size, 492 (48%) proper thickness and 884 (87%) proper staining; the corresponding figures were 972 (100%), 748 (77%) and 809 (99.6%) for the 972 slides rechecked in 2006. After training, the quality of size and staining of smear preparation had significantly improved (P < 0.001) in 2006. Rechecking of 981 readable slides in 2005 identified 3 (0.3%) high false-negatives, 3 (16.7%) low false-positives and 26 (2.8%) low false-negatives; the corresponding errors were 3 (0.3%), 8 (28.6%) and 12 (1.3%) for the 972 slides rechecked in 2006. Of the eight laboratories, two (25%) and four (50%) reached 80% sensitivity in 2005 and 2006, respectively. CONCLUSION: Technical training and EQA improved the quality of sputum smear microscopy services.