Abstract
BACKGROUND: /Aims De-implementation, including the removal or reduction of unnecessary or inappropriate prescribing, is crucial to ensure patients receive appropriate evidence-based health care. The utilization of de-implementation efforts is contingent on the quality of strategy reporting. To further understand effective ways to de-implement medical practices, specification of behavioural targets and components of de-implementation strategies are required. This paper aims to critically analyse how well the behavioural targets and strategy components, in studies that focused on de-implementing unnecessary or inappropriate prescribing in secondary healthcare settings, were reported. METHODS: A supplementary analysis of studies included in a recently published review of de-implementation studies was conducted. Article text was coded verbatim to two established specification frameworks. Behavioural components were coded deductively to the five elements of the Action, Actor, Context, Target, Time (AACTT) framework. Strategy components were mapped to the nine elements of the Proctor's 'measuring implementation strategies' framework. RESULTS: The behavioural components of low-value prescribing, as coded to the AACTT framework, were generally specified well. However, the Actor and Time components were often vague or not well reported. Specification of strategy components, as coded to the Proctor framework, were less well reported. Proctor's Actor, Action target: specifying targets, Dose and Justification elements were not well reported or varied in the amount of detail offered. We also offer suggestions of additional specifications to make, such as the 'interactions' participants have with a strategy. CONCLUSION: Specification of behavioural targets and components of de-implementation strategies for prescribing practices can be accommodated by the AACTT and Proctor frameworks when used in conjunction. These essential details are required to understand, replicate and successfully de-implement unnecessary or inappropriate prescribing. In general, standardisation in the reporting quality of these components is required to replicate any de-implementation efforts. TRIAL REGISTRATION: Not registered.