診療報酬情報を用いた中心静脈カテーテル挿入に伴う重症気胸事例検出法

• Published in: Journal of the Japan Society for Healthcare Administration Vol. 53; no. 4; pp. 217 - 225
• Main Authors: 鳥羽 三佳代, 森脇 睦子, 佐瀬 裕子, 尾林 聡, 伏見 清秀
• Format: Journal Article
• Language: Japanese
• Published: Japan Society for Healthcare Administration 2016; 一般社団法人 日本医療・病院管理学会
• Subjects: Adverse event; DPC Data; DPCデータ; Incident report; Medical safety; Pneumothorax associated with central venous catheterization; インシデントレポート; 中心静脈カテーテル挿入に伴う医原性気胸; 医療安全; 有害事象
• ISSN: 1882-594X; 2185-422X
• DOI: 10.11303/jsha.53.217

［Introduction］　This study represents an attempt to elaborate a method for detecting cases of severe pneumothorax associated with placement of a central venous catheter using information on fee-for-medical-service records, to realize conformation of internal audits with actual medical service situations.　［Methods］　Cases with calculated medical fees for central venous catheterization and continuous thoracic cavity drainage were extracted from the EF file information.　In this study population, the detection sensitivity of the following methods for the above-mentioned adverse event cases were compared:Method 1, detection by reference to the name of the disorder occurring after hospital admission;Method 2, detection by reference to the diagnosis stated on the medical bill;Method 3, detection based on a survey of the medical records;Method 4, early detection by reference to the diagnosis stated on the medical bill without prior extraction from the EF files.　［Results］　The detection sensitivities of the four methods were found to be 15.4%, 42.3%, 69.2% and 53.8%, respectively, the incidence of central venous catheterization-associated severe pneumothorax was 0.8%, and the incident report submission rate was 61.5%.　［Conclusion］　Detection of adverse events based on Fee-for-medical-service information involves problems that need to be resolved, such as the detection sensitivity being dependent on the data input accuracy and the subject events being limited.　Concurrent use of survey of the medical records, however, enabled even detection of events that could not be identified from incident reports. 【緒言】診療実態に即した内部監査実現のため，診療報酬情報を用いた中心静脈カテーテル挿入に伴う重症気胸事例の検出法開発を試みた。【方法】EFファイル情報から中心静脈カテーテル挿入と持続的胸腔ドレナージが算定されている症例を抽出後，入院後発生疾患名で検出する第1法，レセプト病名で検出する第2法，診療録調査から検出する第3法，EFファイルによる抽出は行わずレセプト病名のみから検出する第4法の感度を比較した。【結果】それぞれの検出法の検出数と感度は15.4%，42.3%，69.2%，53.8%，中心静脈カテーテル挿入に伴う重症気胸事例発生率0.8%，インシデントレポート提出率は61.5%であった。【結論】診療報酬情報を用いた有害事象検出は，入力精度により感度が左右される，対象事象に制限がある，などの課題はあるが，診療録調査を併用することでインシデントレポートでは収集できなかった事象が検出できた。