Verdoorn S, Kwint H F, Hoogland P, Gussekloo J, Bouvy M L
Division of Pharmacoepidemiology & Clinical Pharmacology, Utrecht Institute for Pharmaceutical Sciences (UIPS), Utrecht, The Netherlands.
SIR Institute for Pharmacy Practice and Policy, Leiden, The Netherlands.
J Clin Pharm Ther. 2018 Apr;43(2):224-231. doi: 10.1111/jcpt.12637. Epub 2017 Oct 2.
To facilitate the identification of drug-related problems (DRPs) during medication review, several tools have been developed. Explicit criteria, like Beers criteria or STOPP (Screening Tool of Older Peoples' Prescriptions) and START (Screening Tool to Alert doctors to Right Treatment) criteria, can easily be integrated into a clinical decision support system (CDSS). The aim of this study was to investigate the effect of adding a CDSS to medication review software on identifying and solving DRPs in daily pharmacy practice.
Pre- to post-analysis of clinical medication reviews (CMRs) performed by 121 pharmacies in 2012 and 2013, before and after the introduction of CDSS into medication review software. Mean number of DRPs per patient, type of DRPs and their resolution rates were compared in the pharmacies pre- and post-CDSS using paired t tests.
In total, 9151 DRPs were identified in 3100 patients pre-CDSS and 15 268 DRPs were identified in 4303 patients post-CDSS. The mean number of identified DRPs per patient (aggregated per pharmacy) was higher after the introduction of CDSS (3.2 vs 3.6 P < .01). The resolution rate was lower post-CDSS (50% vs 44%; P < .01), which overall resulted in 1.6 resolved DRPs per patient in both groups (P = .93). After the introduction of CDSS, 41% of DRPs were detected by the CDSS. The resolution rate of DRPs generated by CDSS was lower than of DRPs identified without the help of CDSS (29% vs 55%; P < .01). The two most prevalent DRP types were "Overtreatment" and "Suboptimal therapy" in both groups. The prevalence of "Overtreatment" was equal in both groups (mean DRPs per patient: 0.84 vs 0.77; P = .22), and "Suboptimal therapy" was more frequently identified post-CDSS (mean DRPs per patient: 0.54 vs 1.1; P < .01).
The introduction of CDSS to medication review software generated additional DRPs with a lower resolution rate. Structural assessment including a patient interview elicited the most relevant DRPs. Further development of CDSS with more specific alerts is needed to be clinical relevant.
为便于在药物治疗评估过程中识别药物相关问题（DRP），已开发了多种工具。明确的标准，如Beers标准或STOPP（老年人处方筛查工具）及START（提醒医生正确治疗的筛查工具）标准，可轻松整合到临床决策支持系统（CDSS）中。本研究的目的是调查在药物治疗评估软件中添加CDSS对日常药房实践中识别和解决DRP的影响。
对2012年和2013年121家药房在将CDSS引入药物治疗评估软件之前和之后所进行的临床药物治疗评估（CMR）进行前后分析。使用配对t检验比较各药房在引入CDSS前后每位患者的DRP平均数量、DRP类型及其解决率。
在引入CDSS之前，共在3100名患者中识别出9151个DRP，在引入CDSS之后，共在4303名患者中识别出15268个DRP。引入CDSS后，每位患者识别出的DRP平均数量（按药房汇总）更高（3.2对3.6；P <.01）。引入CDSS后的解决率更低（50%对44%；P <.01），总体而言两组中每位患者解决的DRP数量均为1.6个（P =.93）。引入CDSS后，41%的DRP由CDSS检测到。CDSS生成的DRP的解决率低于未借助CDSS识别出的DRP的解决率（29%对55%；P <.01）。两组中最常见的两种DRP类型均为“过度治疗”和“治疗不充分”。两组中“过度治疗”的发生率相当（每位患者患者的平均DRP数：0.84对0.77；P =.22），且引入CDSS后“治疗不充分”被更频繁地识别出来（每位患者的平均DRP数：0.54对1.1；P <.01）。
将CDSS引入药物治疗评估软件会产生额外的DRP，且解决率更低。包括患者访谈在内的结构化评估能找出最相关的DRP。需要进一步开发具有更具体警示功能的CDSS，使其具有临床相关性。
