Background: The absence of meaningful end user engagement has repeatedly been highlighted as a key factor contributing to 'failed' implementations of electronic health records (EHRs), but achieving this is particularly challenging in the context of national scale initiatives. In 2002, the National Health Service (NHS) embarked on a so-called 'top-down' national implementation strategy aimed at introducing commercial, centrally procured, EHRs into hospitals throughout England.
Objective: We aimed to examine approaches to, and experiences of, user engagement in the context of a large-scale EHR implementation across purposefully selected hospital care providers implementing early versions of nationally procured software.
Methods: We conducted a qualitative, case-study based, socio-technically informed, longitudinal investigation, purposefully sampling and collecting data from four hospitals. Our data comprised a total of 123 semi-structured interviews with users and managers, 15 interviews with additional stakeholders, 43 hours of non-participant observations of meetings and system use, and relevant organisation-specific documents from each case study site. Analysis was thematic, building on an existing model of user engagement that was originally developed in the context of studying the implementation of relatively simple technologies in commercial settings. NVivo8 software was used to facilitate coding.
Results: Despite an enduring commitment to the vision of shared EHRs and an appreciation of their potential benefits, meaningful end user engagement was never achieved. Hospital staff were not consulted in systems choice, leading to frustration; they were then further alienated by the implementation of systems that they perceived as inadequately customised. Various efforts to achieve local engagement were attempted, but these were in effect risk mitigation strategies. We found the role of clinical champions to be important in these engagement efforts, but progress was hampered by the hierarchical structures within healthcare teams. As a result, engagement efforts focused mainly on clinical staff with inadequate consideration of management and administrative staff.
Conclusions: This work has allowed us to further develop an existing model of user engagement from the commercial sector and adapt it to inform user engagement in the context of large-scale eHealth implementations. By identifying key points of possible engagement, disengagement and re-engagement, this model will we hope both help those planning similar large-scale EHR implementation efforts and act as a much needed catalyst to further research in this neglected field of enquiry.