Dataset in full English is the S2 File title.

Background
The administration of intravenous (IV) medications is a technically complicated and error-prone process. Especially, in the hematopoietic stem cell transplantation (HSCT) setting where toxic drugs are frequently used and patients are in critical immunocompromised conditions, medication errors (ME) can have catastrophic reactions and devastating outcomes such as death. Studies on ME are challenging due to poor methodological approaches and complicated interpretations. Here, we tried to resolve this problem using reliable methods and by defining new denominators, as a crucial part of an epidemiological approach.
Methods
This was an observational, cross-sectional study. A total of 525 episodes of IV medication administration were reviewed by a pharmacist using the disguised direct observation method to evaluate the preparation and administration processes of 32 IV medications in three HSCT wards. We reported errors in 3 ratios; 1) Total Opportunities for Error (TOE; the number of errors/sum of all administered doses observed plus omitted medications), 2) Proportional Error Ratio (the number of errors for each drug or situation/total number of detected errors) and, 3) Corrected Total Opportunities for Errors (CTOE; the number of errors/ Sum of Potential Errors (SPE)).
Results
A total of 1,568 errors were observed out of 5,347 total potential errors. TOE was calculated as 2.98 or 298% and CTOE as 29.3%. Most of the errors occurred at the administration step. The most common potential errors were the use of an incorrect volume of the reconstitution solvent during medication preparation and lack of monitoring in the administration stage.
Conclusion
Medication errors frequently occur during the preparation and administration of IV medications in the HSCT setting. Using precise detection methods, denominators, and checklists, we identified the most error-prone steps during this process, for which there is an urgent need to implement effective preventive measures. Our findings can help plan targeted preventive measures and investigate their effectiveness, specifically in HSCT settings.

背景：静脉给药是一项技术复杂度高且易出现差错的操作流程。尤其在造血干细胞移植（hematopoietic stem cell transplantation, HSCT）场景中，该场景常需使用毒性药物，且患者处于严重免疫功能低下状态，用药错误（medication errors, ME）可能引发灾难性反应乃至死亡等毁灭性后果。由于研究方法学存在缺陷、结果解读复杂，针对用药错误的相关研究往往颇具挑战性。本研究作为流行病学研究方法的核心环节之一，尝试通过可靠的研究方法并定义新的分母指标来解决上述问题。

方法：本研究为观察性横断面研究（cross-sectional study）。本研究由药剂师采用隐蔽式直接观察法，对3个造血干细胞移植病房内32种静脉给药药物的配制与给药流程进行回顾分析，共纳入525次静脉给药操作。本研究通过3个比率指标报告用药差错情况：1）差错总发生机会（Total Opportunities for Error, TOE；差错数/所有观察到的给药剂量数+遗漏给药数）；2）比例差错率（每类药物或场景下的差错数/检测到的总差错数）；3）校正后差错总发生机会（Corrected Total Opportunities for Errors, CTOE；差错数/潜在差错总数（Sum of Potential Errors, SPE））。

结果：在总计5347例潜在差错中，共观察到1568例用药差错。差错总发生机会（TOE）计算值为2.98（即298%），校正后差错总发生机会（CTOE）为29.3%。绝大多数差错发生于给药环节。最常见的潜在差错为药物配制过程中使用了错误体积的复溶溶剂，以及给药阶段未实施监测。

结论：在造血干细胞移植场景中，静脉给药药物的配制与给药过程中常出现用药差错。本研究通过精准的检测方法、分母指标及核查清单，明确了该流程中最易发生差错的环节，亟需针对这些环节实施有效的预防措施。本研究结果可为制定针对性预防方案并评估其效果提供参考，尤其适用于造血干细胞移植场景。