文献速递药物不良反应与药物警戒

药物治疗质量与药物相关入院评估的评分者间信度研究

Inter-rater reliability of assessments regarding the quality of drug treatment, and drug-related hospital admissions.
PubMed 文献 · Br J Clin Pharmacol · 2021-03-19

瑞典Sahlgrenska大学医院3名专科医生和2名住院医生独立及共识评估30例随机患者(50%女性,中位年龄72岁)的药物治疗质量及其与入院的关系,用Gwet AC1评价信度。入院时210种药物中200种(95%)、出院时245种中238种(97%)被所有评估者评为合理。入院药物与入院关联评估在16/30例患者中一致(AC1全组0.67)。结论认为信度从弱到几乎完美不等,评估可靠性是方法学问题,应纳入专科医生。

要点

  • 药物治疗质量评估的评分者间信度从弱到几乎完美不等,需关注方法学可靠性。
  • 入院药物关联评估全组AC1为0.67,住院医生仅0.54,专科医生0.74。
  • 疑似处方错误致入院的3例中评估者均未达成一致。

核心结论

  • 药物治疗质量及不良后果评估的评分者间信度从弱到几乎完美不等,可靠性是方法学关注点。
  • 为获得可接受的可靠结果,应纳入专科医生参与评估。

研究设计与样本

  • 3名专科医生和2名住院医生独立及共识评估30例随机患者(50%女性,中位年龄72岁)的药物治疗质量及其与入院的关系。
  • 患者于2018年4月入住瑞典Sahlgrenska大学医院,信度用Gwet AC1评价。

主要发现

  • 入院时210种药物中200种(95%)、出院时245种中238种(97%)被所有评估者评为合理药物治疗。
  • 入院时无药物、出院时2种药物被所有评估者评为不合理药物治疗。
  • 合理药物治疗AC1:全组入院0.88、出院0.94;专科医生0.86、0.95;住院医生0.92、0.92。
  • 药物与入院关联评估在16/30例患者中一致,AC1全组0.67、专科医生0.74、住院医生0.54。
  • 疑似处方错误致入院的3例中评估者均未达成一致。

临床怎么用

  • 评估药物治疗质量及药物相关入院时,应纳入专科医生以提高信度。
  • 对疑似处方错误致入院等不良后果的评估需谨慎,因信度可能较低。

局限性

  • 样本量仅30例随机患者,评估药物数量有限。
  • 仅纳入两家医院(Sahlgrenska大学医院)的医生,外推性受限。

以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。

原文

AIMS: To investigate inter-rater agreement on the quality of drug treatment, and the relationship between the drug treatment and hospital admission. METHODS: Three specialist physicians and two resident physicians determined, independently and in consensus, the quality of drug treatment from an overall medical perspective, and its association with admission, in 30 randomly selected patients (50% female, median age 72 years) admitted to Sahlgrenska University Hospital, Sweden, in April 2018. The inter-rater agreement was evaluated with Gwet's agreement coefficient (AC1 ). RESULTS: In all, 200 (95%) out of 210 drugs at admission and 238 (97%) out of 245 drugs at discharge were assessed as reasonable drug treatment by all assessors. Conversely, none of the drugs at admission, and two at discharge, were assessed as unreasonable drug treatment by all assessors (AC1 : 0.88 and 0.94 [all], 0.86 and 0.95 [specialists], 0.92 and 0.92 [residents], respectively). The assessments regarding the association between the drug treatment and the hospital admission (not related or main/contributory reason) were consistent between the assessors for 16 out of 30 patients (AC1 : 0.67 [all], 0.74 [specialists], 0.54 [residents]). In none of the three cases where the hospital admission was considered possibly attributable to a prescribing error did the assessors make consistent assessments. CONCLUSIONS: As the inter-rater agreement ranged between weak and almost perfect, the reliability of assessments of drug treatment quality, as well as adverse consequences, appears to be a methodological concern. To yield acceptably reliable results regarding both drug treatment aspects at issue, specialist physicians should be involved.

来源:Br J Clin Pharmacol,查看原文。版权归原作者所有。

引用

  1. Lönnbro J, Holmqvist L, Persson E, et al. Inter-rater reliability of assessments regarding the quality of drug treatment, and drug-related hospital admissions. Br J Clin Pharmacol. 2021. PMID: 33609324. doi:10.1111/bcp.14790

供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。