The accuracy of diffusing capacity of the lung for carbon monoxide (DLCO) instruments directly influences clinical decision-making in pulmonary diseases. However, long-term use of these instruments may lead to performance drift and measurement errors. At present, daily calibration using 3-L syringes and biological controls is commonly applied in clinical practice, but their sensitivity is insufficient to detect potential errors. Therefore, there is an urgent need to explore more precise and objective methods for quality evaluation of these instruments.

To evaluate the accuracy of DLCO instruments using a standard diffusion simulator that mimics human single-breath maneuvers, to identify and correct the sources of measurement errors, to observe the stability of calibration over time, and to assess the application value of this method in combination with biological control tests.

From March to July 2021, four DLCO instruments were tested in the pulmonary function laboratory of the National Clinical Research Center for Respiratory Diseases. A standard diffusion simulator (Hans Rudolph series 5560, USA) was employed with three carbon monoxide concentrations (0.08%, 0.10%, 0.13%) and three inhalation volumes (1.5 L, 3 L, 4.5 L) in different combinations to simulate single-breath DLCO tests. An absolute error of <2 mL·min^-1·mmHg^-1 between the measured and target DLCO values was defined as the acceptable range. The accuracy of four DLCO instruments (two brands, two models each) was evaluated as baseline, and the sources of measurement errors were analyzed. Each instrument was then corrected according to its error sources, and changes in accuracy before and after calibration as well as within three months were compared. Biological control testing was also performed to observe inter-instrument differences.

At baseline, 50% (2/4) of the instruments had a mean absolute DLCO error greater than 2 mL·min^-1·mmHg^-1. The error sources varied, including damaged components, incorrect calibration methods, and syringe leakage. After targeted corrections, all instruments achieved a mean absolute DLCO error within 2 mL·min^-1·mmHg^-1 (P<0.001), and this accuracy was maintained for three months. Post-calibration, the differences among instruments in biological control testing were significantly reduced, particularly in the range and coefficient of variation (CV) of DLCO measurements, which showed statistically significant improvements (P<0.05) .

DLCO instruments that have been in long-term clinical use show considerable variability in accuracy, with large measurement differences across instruments. Quality evaluation using a standard diffusion simulator is effective and feasible, enabling objective assessment of instrument accuracy and compensating for the limitations of routine calibration. Regular use of simulators is recommended for instrument quality evaluation and quality control.

肺弥散功能仪器的准确性直接影响临床肺疾病的诊疗决策，但仪器长期使用后存在性能漂移和测量误差。目前，临床多采用日常定标筒校准和生物人验证，由于其敏感性不足，难以发现仪器的潜在误差，因此亟需探索更为精准、客观的仪器质量评估方法。

采用标准弥散功能模拟器来模拟受试者人体进行一口气呼吸法肺弥散功能检查，检测肺弥散功能仪器的准确性，并根据测试结果对仪器的误差原因进行校正，观察校正后的结果和维持的时间，然后结合生物人验证探讨这种方法的应用价值。

2021年3—7月，在国家呼吸系统疾病临床医学研究中心肺功能室对4台肺弥散功能仪进行测试。采用标准弥散功能模拟器（5560型，Hans Rudolph，Inc，美国），使用3种一氧化碳浓度（0.08%、0.10%、0.13%）的混合气和3种吸入容积（1.5、3、4.5 L）分别组合，模拟受试者进行一口气呼吸法肺弥散功能测试，以肺一氧化碳弥散量（DLCO）测试值与目标值的绝对误差<2 mL·min^-1·mmHg^-1为可接受范围，检测4台肺弥散功能仪（2种品牌各2种型号）的准确性并作为基线，同时分析仪器误差来源。然后根据误差进行相应的校正，比较校正前后和3个月内仪器准确性的变化，同时结合生物人验证，观察不同仪器之间测量的差异。

50%（2/4）的仪器在基线状态下的DLCO平均绝对误差>2 mL·min^-1·mmHg^-1。各仪器的误差来源不一，可能是仪器配件损坏、操作方法错误、定标筒漏气等原因；在对误差进行相应的校正后，所有仪器的DLCO平均绝对误差均<2 mL·min^-1·mmHg^-1（P<0.001），且能维持3个月的时间。校正后，不同仪器间生物人测量的差异减小，尤其是DLCO测量的极差和变异系数（CV）比较差异有统计学意义（P<0.05）。

临床长时间使用的肺弥散功能检查仪器的准确性存在较大差异，不同仪器间测量的差异较大。采用模拟器对肺弥散功能检查仪器进行质量评估的方法有效、可行，能够评估仪器的准确性且弥补日常校准的不足，建议定期使用模拟器进行质量评估和质量控制。