Related Experiment Video
Updated: May 31, 2026

Safety Precautions and Operating Procedures in an (A)BSL-4 Laboratory: 3. Aerobiology
Published on: October 3, 2016
Performance evaluation of biological safety cabinets: a real-world analysis
Tao Song1, Shaohua Yin1, Zhenlin Liu1
1Department of Medical Engineering, Peking University Third Hospital, Beijing, China.
Introduction:
Biological safety cabinets (BSCs) are essential for protecting personnel, samples, and the environment in clinical and research laboratories. Evidence on the real-world performance of Class II BSCs in China remains limited, particularly regarding differences between domestic and imported equipment.
Methods:
We conducted a cross-sectional evaluation of 360 Class II BSCs in Beijing using 1,803 performance testing records from 2018 to 2023. Performance indicators, including noise, illumination, cleanliness, inflow velocity, downflow velocity, HEPA filter integrity, and airflow smoke pattern were measured according to JG 170-2005, YY 0569-2011 and SN/T 3901-2014 standards. BSCs were assessed and classified as fully qualified performance (FQP) if the indicator met the standard without equaling the threshold, marginally qualified performance (MQP) if the indicator equaled the threshold, or unqualified if the standard was not met. Qualified rates were compared by manufacturing region, hospital level, using sector, and service life using χ 2 tests and non-parametric analyses.
Results:
Among 360 Class II BSCs, the median service life was 7 years (IQR 5-10). The overall qualified rate was 70.4%, with no difference between domestic and imported BSCs (71.3% vs. 69.2%; χ 2 =0.909, P=0.340). Domestic BSCs showed higher FQP rate (55.4% vs. 40.5%; χ 2 =11.625, P=0.0007) and lower MQP rate (15.8% vs. 28.7%; χ 2 =15.087, P=0.0001), with similar pattern for noise, inflow velocity, and downflow velocity. The inflow velocity, downflow velocity, and illumination showed relatively lower qualified rates, with only inflow velocity having a higher qualified rate in domestic BSCs (85.5% vs. 80.9%; χ 2 =6.845, P=0.009). Imported BSCs performed slightly better in several key indicators, including inflow velocity, downflow velocity, and HEPA filter integrity, although differences were not statistically significant.
Discussion:
These findings suggest that BSC reliability reflects an interplay between manufacturing-related characteristics and in-use performance, and that the MQP helps identify BSCs operating near minimum safety margins for proactive monitoring and timely intervention.

