132 Chapter 5 Data collection Three data sources were utilized. First, a questionnaire was administered to nurses to evaluate the acceptability and feasibility of DRAAI. A short questionnaire, developed by the project team members, included 30 questions based on the NASSS framework and the Unified Theory of Acceptance and Use of Technology (UTAUT) [32, 40]. It was made available online via LimeSurvey (LimeSurvey GmbH), and is provided in supplemental file S2 (translated from Dutch). To ensure participant anonymity, no demographic data were collected. The questionnaire was distributed from two up to eight weeks after the rollout. Project team members encouraged nurses to complete it during informal ward visits and educational sessions. Nurses were eligible to participate if they worked on one of the participating wards, with no exclusion criteria applied. Field notes were recorded by project team members, who were actively involved during the entire implementation period. These notes were stored in Siilo and a secure Microsoft Teams environment. All team members had smart phones, allowing them to document observations immediately following their visits. During the implementation, the frequency of joint daily start meetings, educational sessions, informal visits, evaluation meetings, adaptations of the implementation strategies were collected. The number of nurses attending these sessions was also documented. Additionally, the team recorded the types of questions posed by nurses, how nurses responded to questions from the project team’, and other noteworthy observations. The field notes provided insights into the mechanisms of impact. The project team also tracked the time invested in implementing the strategies, excluding preparation time. During the study period, follow-up of DRAAI risk predictions was assessed using routinely collected monitoring data. This was done by reviewing the nursing care plans of all admitted patients. The joint daily stand-up meetings, which began at 07:45 AM, was chosen as the time when DRAAI predictions were presented to users. This timestamp was used in calculating the follow-up time as 'within X hours'. The followup of risk predictions was automatically recorded for quality and control purposes. The follow-up actions were categorized into four groups, 1) At-risk according to nurse judgement prior to or without a prediction; 2) follow-up within 24 hours; 3) follow-up after 24 hours; and 4) no follow-up. No demographic information or specific risk factors of the admitted patients were collected. Data Analysis To analyze the follow-up of the predictions, the following approach was adopted. Since a single patient may have multiple admissions (e.g., a 3-day admission July and a 4-day admission in September), we focused on admissions rather than individual patients.
RkJQdWJsaXNoZXIy MTk4NDMw