Introduction

The deployment and use of clinical information systems (CISs) in healthcare facilities differ between developed and developing countries [1-4]. In the United States, the HITECH Act enacted under Title XIII of the American Recovery and Reinvestment Act (ARRA) of 2009 was followed by a dramatic increase of CIS coverage in hospitals as well as general practices. According to the Health Information Management System Society (HIMSS), by Q4 2017, 73.1% of US hospitals had reached level 5 to 7 of the HIMSS/EMRAM (Electronic Medical Record Adoption Model) maturity level (32.9% level 5, 33.8% level 6, and 6.4% level 7) [2]. As of May 2017, more than 525,000 healthcare providers received payment for participating in the Medicare and Medicaid EHR Incentive Programs [3]. In Canada, Europe, and Asia, deployments have been much slower, with mean EMRAM scores of approximately 3 or 4 in 2017. Counter positive examples exist in Denmark or the Netherlands, which have reached adoption levels close to the USA [4]. CISs intend to automate the execution of clinical processes so that health professionals can benefit from clinical decision support tools and spend more time with their patients. High CIS maturity achievement is expected to improve institution financial efficiency, increase the quality of care, and reduce the incidence of medical errors. It is also expected to foster clinical and translational research through data reuse directly from the EHR databases or from associated data warehouses [5]. However, benefits are not linearly related to the EMRAM maturity level, and there could be a tipping point around level 5-6 that corresponds to the extensive management of protocols and clinical pathways and to the full coverage of drug ordering and delivery loop processes [6]. This applies to the representation of the top performing hospitals in terms of quality metrics excellence of the Joint Commission [1], the representation of hospitals with a “A” Leapfrog safety grade [7], or the Value-Based Purchasing (VBP) clinical score initiative of the Centers for Medicare & Medicaid Services (CMS) initiative [8]. Achieving a high EMRAM level might be a necessary but insufficient condition of quality improvement. In the 2015 Leapfrog Hospital Survey, for example, hospitals’ CPOE systems failed to flag 39% of all potentially harmful drug orders and 13% of potentially fatal orders [7]. If a reduction in medical errors is the major reason for implementing a CPOE, CIS users are now well aware of the unintended and negative consequences of running clinical information systems [9-11]. They include, among others, the excessive time devoted to data entry, particularly when a comprehensive CPOE is used, workflow issues and the risk of asynchronous communication between end-users in an urgency context, as well as alert fatigue in front of overly reactive decision support systems. Physicians who are likely to use the broadest scope of CIS functions are the most prone to resist their deployment [12-15]. In this difficult and slowly evolving context, permanent evaluation of deployed systems is a prerequisite to their continuous improvement. CIS use and acceptance need to be measured for each category of health professional user and at each phase of the CIS lifecycle, i.e., planning, implementation, deployment, and consolidation [14,16-20].