テキストマイニングを用いた人工呼吸器の内的要因と外的要因に関連した不具合分析

抄録

In this study, malfunctions of mechanical ventilators were categorized into internal and external factors, with the objective of analyzing each type of malfunction using text mining. The data used in this study consisted of malfunction reports of mechanical ventilators published between 2010 and 2023 by the Ministry of Health, Labour and Welfare and the Japan Council for Quality Health Care. The number of reported malfunctions included 2,864 cases of medical device malfunction reports, 1,984 near-miss incidents, and 255 cases of medical accident information. The analysis revealed that internal factors included hardware and software failures as well as technical malfunctions. Malfunctions related to external factors were commonly reported in relation to routine operations, procedural errors, human errors, and patient condition management. Specific malfunctions reported in near-miss incidents involved equipment setting errors and attachment issues, while malfunctions related to alarms and emergency responses were reported in medical accident information.