TY - JOUR T1 - Assessing the ability of the Drug-Associated Risk Tool (DART) questionnaire to stratify hospitalised older patients according to their risk of drug-related problems: a cross-sectional validation study JF - BMJ Open JO - BMJ Open DO - 10.1136/bmjopen-2017-021284 VL - 8 IS - 6 SP - e021284 AU - Dominik Stämpfli AU - Fabienne Boeni AU - Andy Gerber AU - Victor A D Bättig AU - Rebekka Weidmann AU - Kurt E Hersberger AU - Markus L Lampert Y1 - 2018/06/01 UR - http://bmjopen.bmj.com/content/8/6/e021284.abstract N2 - Objectives The Drug-Associated Risk Tool (DART) has been developed as a self-administered questionnaire for patients with the aim of stratifying patients according to their risk of drug-related problems (DRPs). We aimed to validate the ability of the questionnaire to distinguish between hospitalised patients showing lower and higher numbers of DRPs.Design Cross-sectional study assessing the questionnaire’s concurrent criterion validity.Setting Five geriatric and the associated physical and neurological rehabilitation wards of a Swiss regional secondary care hospital with 617 beds.Participants We recruited 110 patients from a total of 437 admissions. Exclusion criteria were insufficient knowledge in spoken or written German, medical conditions preventing meaningful conversations and already receiving pharmacy services.Interventions Comprehensive pharmacist-led clinical medication reviews were performed, including patient interviews, to identify potential and manifest DRPs. A cluster analysis was conducted to assess the discriminatory potential of the DART to group patients according to number (low and high) of identified DRPs. A subsequent discriminatory function analysis was performed to reduce the number of items. We determined which DART items may be used to trigger what type of medication review.Results Recruited patients had a median age of 79 years and were prescribed a median of 11 drugs. Patients with a median DART score of 10 and a median of 3 DRPs represented one cluster, whereas patients with a median DART score of 15 and a median of 8 DRPs represented another cluster. Discriminatory function analysis reduced the questionnaire to five items with a moderate to strong correlation with the number of DRPs per patient (Spearman’s rank correlation ρ=0.44). Additional items were associated with patients benefiting from interviews.Conclusions As a self-administered questionnaire for patients, the DART may be used to stratify hospitalised non-acute older patients in groups of having low and high likelihood of DRPs. The analyses showed that a short form of the DART can be used instead of the full tool to identify older inpatients at risk for DRPs. Additional eight items from the DART may be used to initiate additional clinical pharmacy services. The linkage between certain DART questions and type of medication review enables pharmacist resource allocation. ER -