Impact of early in-hospital medication review by clinical pharmacists on health services utilization.
Impact of early in-hospital medication review by clinical pharmacists on health services utilization.
复制标题
DOI：
10.1371/journal.pone.0170495
复制
发表时间：
2017
期刊：
影响因子：
3.7
通讯作者：
Sobolev B
中科院分区：
文献类型：
--
作者：
Hohl CM;Partovi N;Ghement I;Wickham ME;McGrail K;Reddekopp LN;Sobolev B
药物不良事件是导致急诊和非计划入院的主要原因，并延长住院时间。药物审查干预旨在识别药物不良事件并优化药物使用。以前对住院药物审查的评价集中在出院时的干预措施上，对健康结果的影响尚不清楚。我们评估了早期住院药师主导的药物审查对高危患者健康结局的影响。我们采用了准随机设计，以评估在加拿大不列颠哥伦比亚省的三家医院的质量改进项目。我们将临床决策规则纳入急诊科分诊路径，使护士能够识别药物不良事件的高风险患者。随机选择第一位符合参与条件的患者后，临床药师系统地将后续的高风险患者分配到药物审查或常规护理。药物审查包括获得最佳可能的用药史，并审查患者的药物适当性和药物不良事件。主要结局是30天内住院天数，并使用管理数据确定。我们使用中位数和反向倾向评分加权逻辑回归模型来确定药师主导的药物审查对下游卫生服务使用的影响。在10，807名高危患者中，6，416名接受了早期药剂师领导的药物审查和4，391名常规护理。他们的基线特征是平衡的。与常规治疗相比，药物审查组的中位住院天数减少了0.48天（95%置信区间[CI] = 0.00至0.96; p = 0.058），表示中位住院时间减少了8%。在80岁以下的患者中，中位住院天数减少了0.60天（95% CI = 0.06 - 1.17; p = 0.03），表示中位住院时间缩短了11%。对急诊科复诊、入院、再入院或死亡率没有显著影响。由于我们无法进行随机对照试验，我们受到了限制，但使用了准随机患者分配方法和倾向评分模型来确保治疗组之间的平衡，并使用管理数据来确保盲态结局的确定。我们无法解释护理日的交替水平，因此，可能低估了在等待长期护理期间可能留在医院的虚弱老年患者的治疗效果。在80岁以下的高危患者中，与常规护理相比，早期药师主导的药物审查与病床利用率降低相关，但在老年患者中则不然。我们的评估结果表明，在急诊科药师的药物审查可能会影响住院时间的选择患者人群。
Adverse drug events are a leading cause of emergency department visits and unplanned admissions, and prolong hospital stays. Medication review interventions aim to identify adverse drug events and optimize medication use. Previous evaluations of in-hospital medication reviews have focused on interventions at discharge, with an unclear effect on health outcomes. We assessed the effect of early in-hospital pharmacist-led medication review on the health outcomes of high-risk patients. We used a quasi-randomized design to evaluate a quality improvement project in three hospitals in British Columbia, Canada. We incorporated a clinical decision rule into emergency department triage pathways, allowing nurses to identify patients at high-risk for adverse drug events. After randomly selecting the first eligible patient for participation, clinical pharmacists systematically allocated subsequent high-risk patients to medication review or usual care. Medication review included obtaining a best possible medication history and reviewing the patient’s medications for appropriateness and adverse drug events. The primary outcome was the number of days spent in-hospital over 30 days, and was ascertained using administrative data. We used median and inverse propensity score weighted logistic regression modeling to determine the effect of pharmacist-led medication review on downstream health services use. Of 10,807 high-risk patients, 6,416 received early pharmacist-led medication review and 4,391 usual care. Their baseline characteristics were balanced. The median number of hospital days was reduced by 0.48 days (95% confidence intervals [CI] = 0.00 to 0.96; p = 0.058) in the medication review group compared to usual care, representing an 8% reduction in the median length of stay. Among patients under 80 years of age, the median number of hospital days was reduced by 0.60 days (95% CI = 0.06 to 1.17; p = 0.03), representing 11% reduction in the median length of stay. There was no significant effect on emergency department revisits, admissions, readmissions, or mortality. We were limited by our inability to conduct a randomized controlled trial, but used quasi-random patient allocation methods and propensity score modeling to ensure balance between treatment groups, and administrative data to ensure blinded outcomes ascertainment. We were unable to account for alternate level of care days, and therefore, may have underestimated the treatment effect in frail elderly patients who are likely to remain in hospital while awaiting long-term care. Early pharmacist-led medication review was associated with reduced hospital-bed utilization compared to usual care among high-risk patients under 80 years of age, but not among those who were older. The results of our evaluation suggest that medication review by pharmacists in the emergency department may impact the length of hospital stay in select patient populations.