A decision tree approach for investigating the background of research activity of community and hospital pharmacists in Mie Prefecture: a retrospective questionnaire-based survey

Abstract

BackgroundThe support system for research activities has not been sufficiently established in clinical settings. A survey should be conducted to identify the causes of low research activity among pharmacists and the characteristics of pharmacists who could serve as mentors to build a support system at the regional level.MethodsA retrospective cross-sectional survey was conducted with 156 pharmacists, including hospital and community pharmacists, who attended a webinar on research ethics held once a year in Mie Prefecture. Decision tree (DT) analysis was performed to extract the low research activities and pharmacists who could serve as mentors in research activities using independent factors identified by multivariate logistic regression analysis.ResultsThe questionnaire response rate was 72.4% (113/156), and most respondents were community pharmacists (81.4%). In the DT model, pharmacists who did not belong to academic societies (78%, 46/59) or those who belonged to one or two academic societies but had no certifications (100%, 5/5) had low research activities. Pharmacists who read papers more than once a month and had a nearby mentor (73%, 11/15) were more likely to become mentors in research activities.ConclusionsThe combination of the number of academic societies and the presence of certifications determines the efforts in research activities. In addition to reading at least one paper monthly, the presence of a mentor for writing research papers may also be a crucial factor in becoming a mentor. The proposed DT model may be helpful in building a support system for research activities at the regional level.