C W Yip1, M Y Chan, W F Cheung, K W Yu, H S Tang, K M Kam. 1. Tuberculosis Reference Laboratory, Public Health Laboratory Centre, Centre for Health Protection, Department of Health, Hong Kong SAR, China.
Abstract
BACKGROUND: The Hong Kong TB Reference Laboratory is a high volume laboratory examining around 400 sputum acid-fast bacilli smears daily using fluorescence microscopy (FM). OBJECTIVE: To assess the effectiveness of blinded rechecking applied to FM in a high-throughput laboratory. METHOD: From 2003, 2.5% (5% in 2003 and 2004) of all smears were randomly selected, relabelled and assigned to each technician (rechecker) in turn. These smears were restained and re-examined. Discordance between initial screener and rechecker was resolved by a controller. RESULTS: From 2003 to 2010, low false-negative (LFN) errors (0.10-0.27%) were within the critical values, at 85% (1 year) and 90% (7 years) sensitivity. However, LFN error (0.28-0.62%) among recheckers was prominent. There were also low false-positive (LFP) cases (0.13-0.75%), but subsequent cultures showed these to be mycobacteria culture-positive. This relatively poor performance among the recheckers might be due to background fluorescence increase after restaining and/or inefficiency of the rechecking procedure. CONCLUSION: In a high-throughput laboratory, blind rechecking is a good means of quality assurance. To minimise false LFP, problems due to restaining should be resolved before blinded rechecking can be generally applied in the field for FM where mycobacterial cultures are not routinely performed.
BACKGROUND: The Hong Kong TB Reference Laboratory is a high volume laboratory examining around 400 sputum acid-fast bacilli smears daily using fluorescence microscopy (FM). OBJECTIVE: To assess the effectiveness of blinded rechecking applied to FM in a high-throughput laboratory. METHOD: From 2003, 2.5% (5% in 2003 and 2004) of all smears were randomly selected, relabelled and assigned to each technician (rechecker) in turn. These smears were restained and re-examined. Discordance between initial screener and rechecker was resolved by a controller. RESULTS: From 2003 to 2010, low false-negative (LFN) errors (0.10-0.27%) were within the critical values, at 85% (1 year) and 90% (7 years) sensitivity. However, LFN error (0.28-0.62%) among recheckers was prominent. There were also low false-positive (LFP) cases (0.13-0.75%), but subsequent cultures showed these to be mycobacteria culture-positive. This relatively poor performance among the recheckers might be due to background fluorescence increase after restaining and/or inefficiency of the rechecking procedure. CONCLUSION: In a high-throughput laboratory, blind rechecking is a good means of quality assurance. To minimise false LFP, problems due to restaining should be resolved before blinded rechecking can be generally applied in the field for FM where mycobacterial cultures are not routinely performed.