Walker Anne E, Grimshaw Jeremy, Johnston Marie, Pitts Nigel, Steen Nick, Eccles Martin
Health Services Research Unit, University of Aberdeen, Aberdeen, UK.
BMC Health Serv Res. 2003 Dec 19;3(1):22. doi: 10.1186/1472-6963-3-22.
Biomedical research constantly produces new findings but these are not routinely translated into health care practice. One way to address this problem is to develop effective interventions to translate research findings into practice. Currently a range of empirical interventions are available and systematic reviews of these have demonstrated that there is no single best intervention. This evidence base is difficult to use in routine settings because it cannot identify which intervention is most likely to be effective (or cost effective) in a particular situation. We need to establish a scientific rationale for interventions. As clinical practice is a form of human behaviour, theories of human behaviour that have proved useful in other similar settings may provide a basis for developing a scientific rationale for the choice of interventions to translate research findings into clinical practice. The objectives of the study are: to amplify and populate scientifically validated theories of behaviour with evidence from the experience of health professionals; to use this as a basis for developing predictive questionnaires using replicable methods; to identify which elements of the questionnaire (i.e., which theoretical constructs) predict clinical practice and distinguish between evidence compliant and non-compliant practice; and on the basis of these results, to identify variables (based on theoretical constructs) that might be prime targets for behaviour change interventions.
We will develop postal questionnaires measuring two motivational, three action and one stage theory to explore five behaviours with 800 general medical and 600 general dental practitioners. We will collect data on performance for each of the behaviours. The relationships between predictor variables (theoretical constructs) and outcome measures (data on performance) in each survey will be assessed using multiple regression analysis and structural equation modelling. In the final phase of the project, the findings from all surveys will be analysed simultaneously adopting a random effects approach to investigate whether the relationships between predictor variables and outcome measures are modified by behaviour, professional group or geographical location.
生物医学研究不断产生新的发现，但这些发现并未常规地转化为医疗保健实践。解决这一问题的一种方法是开发有效的干预措施，将研究成果转化为实践。目前有一系列实证干预措施可供使用，对这些措施的系统评价表明，没有一种单一的最佳干预措施。这个证据库在常规环境中难以使用，因为它无法确定哪种干预措施在特定情况下最有可能有效（或具有成本效益）。我们需要为干预措施建立科学依据。由于临床实践是一种人类行为形式，在其他类似环境中已被证明有用的人类行为理论可能为开发将研究成果转化为临床实践的干预措施选择的科学依据提供基础。本研究的目的是：用卫生专业人员的经验证据来扩充和充实经过科学验证的行为理论；以此为基础，采用可复制的方法编制预测性问卷；确定问卷的哪些要素（即哪些理论结构）能预测临床实践，并区分符合证据和不符合证据的实践；并根据这些结果，确定可能是行为改变干预主要目标的变量（基于理论结构）。
我们将编制邮寄问卷，测量两种动机理论、三种行动理论和一种阶段理论，以探讨800名普通医生和600名普通牙医的五种行为。我们将收集每种行为的表现数据。每次调查中预测变量（理论结构）与结果指标（表现数据）之间的关系将使用多元回归分析和结构方程模型进行评估。在项目的最后阶段，将同时采用随机效应方法分析所有调查的结果，以研究预测变量与结果指标之间的关系是否因行为、专业群体或地理位置而有所不同。
