60 Chapter 3 Data collection A data extraction form was developed, pilot tested and redefined after assessing the first five studies. Information on study characteristics e.g., design, setting and type of hospital, study duration, country of origin, as well as details about the FPIs and comparators, were collected. The data collection for the FPIs was organized according to the four domains of the Prevention of Falls Network Europe (ProFaNE) taxonomy [16]. Data were extracted and checked by at least two reviewers (DS, TL, FH and EI). To answer question 1: ‘What implementation strategies are used to implement FPIs in hospitals?’ implementation strategies were identified based on the described actions. When classifying the strategies, we adhered to three decision rules: Firstly, FPI components were not classified as implementation strategies; for example, patient education was often a component of the FPI. Second, regardless of level of detail, any kind of change initiating action was considered an implementation strategy. This approach was adopted to ensure that even minimally described strategies would be included in the data analysis. Identified strategies were classified using Expert Recommendations for Implementing Change (ERIC) taxonomy [17, 18]. The third decision rule in order to classify all implementation strategies, was that not all elements of the ERIC definition had to be reported. The detailed definitions of the 73 ERIC implementation strategies were used as a guide, for example, for the strategy ‘remind clinicians’, the definition requires development of reminder systems. All reminders were classified as a reminder, regardless of whether there was a system behind them. If the reviewer was uncertain which an ERIC strategy fitted, it was categorized as ‘Not in ERIC’ until EI and DS reviewed these strategies and selected the ERIC strategy that fitted best. We defined tailoring according to Powell et al. [17] as ‘Tailor the implementation strategies to address barriers and leverage facilitators that were identified through earlier data collection’. In the Prospero protocol, the EPOC taxonomy [19] was initially described to classify the implementation strategies; however, we transitioned to the ERIC Taxonomy. In our view, the ERIC taxonomy offers a more comprehensive description of the strategies’ content compared to the EPOC taxonomy. For instance, the strategy ‘organizational culture’ in EPOC is rather nonspecific, while in ERIC, specific elements aimed at changing the organizational culture are possible, e.g. ‘recruit, designate, and train for leadership’. Additionally, the use of implementation theories, models, and frameworks from the included studies were collected and categorized according to the article by Nilsen et al. [20].
RkJQdWJsaXNoZXIy MTk4NDMw