A tactical framework for EMR adoption

Introduction

Many hospitals and other health settings (eg, primary care clinics) in Canada are in the midst of upgrades or new Electronic Medical Record (EMR) implementations. These will combine technology and clinical resources, within a change management framework, to successfully deliver new care delivery models via electronic systems. ¹ There is a large volume of research on the benefits associated with EMR implementation and adoption. Buntin et al. ² illustrated that 62% of Health Information Technology (HIT) research showed positive results, with another 30% of research studies showing mixed-positive conclusions overall. Many benefits are discussed in a case study from Ontario Shores Hospital ³ including improvements in evidence-based workflow documentation and enhanced business intelligence, among other tangible benefits.

Complex change initiatives, especially large-scale technology implementations, are expected to have certain obstacles during the introduction of systems. In a complex health setting with numerous and diverse stakeholders, technology can create large degrees of disruption. Kruse et al. ⁴ noted 68 different barriers related to EMR adoptions that were classified into 39 distinct groups from a systematic review of 27 articles. Unfortunately, depending on the research reviewed, between 50% ⁵ and 80% ⁶ of technology implementations fail. This apparent contradiction between benefits and challenges will be explored to provide a background context for a review of different models related to the implementation of HIT. Specifically, existing models of EMR implementation provide theoretical frameworks for implementation but do not provide strategic tactics to help move projects forward. Due to the complex and dynamic health environment, many of the theoretical models are not fully transferrable to hospital or other healthcare environments. Case studies discuss technology disruption throughout the continuum of the change management process ⁷ and emphasize the need for implementation studies to be conducted over time.

Many of the existing studies on HIT implementation are based on specific types of systems such as decision support, computerized physician order entry, or picture archiving and communication systems. While the findings from these studies may be applicable to the broader EMR implementation landscape, the health ecosystem is filled with human and social diversity, and organizational culture that cannot be separated from HIT implementation strategies. We also need to think of HIT implementation as a Learning Health System (LHS) that encompasses both human and digital aspects that evolve over time. ⁸

We cannot separate HIT from the broader context where it is implemented. While implementation and adoption are often used interchangeably, Sheikh et al. ⁹ make an important distinction between implementation and adoption, with the latter referring to the implementation of technology (eg, software, vendor training, interfaces) and the manner in which technology is integrated into diverse organizational settings. Numerous frameworks exist for HIT implementation or adoption; however, a shortcoming with them is they focus on adoption or implementation at a point in time and not the temporal continuum within which implementation occurs. Braithwaite et al. state that implementation strategies in health need to consider the dynamic complexity of health organizations when implementing technology. ¹⁰ This article addresses that issue and reviews existing literature on EMR implementation and adoption to develop a tactical framework for EMR adoption. Our framework extends existing research by focusing on the transition from theoretical EMR implementation and adoption models to the complex nature of modern health delivery.

Benefits and barriers of EMR adoption

Understanding the benefits and barriers to EMR adoption are key factors for developing an effective implementation strategy. Electronic medical records can provide a number of benefits including reducing medical errors and facilitating appropriate treatment or screening strategies. ¹¹ In addition, EMRs can provide organizational and societal benefits such as increasing research ability, averting costs, and improving compliance at both legal and regulatory levels. ³ Canada Health Infoway ¹² found benefits such as decreased adverse drug events, enhanced preventive and chronic care management, and fewer duplicated tests.

Although the above benefits give a strong push for leaders to initiate EMR adoption, there are also many adoption barriers noted in the literature. These barriers are key contributors to low EMR adoption rates in Canada. ¹³ Some of the main barriers are high upfront investment costs, interruptions in workflow, physician buy in/autonomy, security concerns, and other unintended adverse outcomes. ³ In a systematic literature review, Kruse et al. ⁴ grouped over 60 different barriers into 39 distinct categories. These different barriers appeared 125 times in the 27 articles that they reviewed for the study. Another EMR implementation barrier is the need to look at all EMR use across the entire care delivery continuum. In many cases, workarounds are created in the local setting due to the technology itself, linkages with other systems, usage patterns, or conflicts with provider workflow. ¹⁴ One common workaround is continued use of paper to duplicate EMR charting or in some instances a full ghost chart, where a full paper chart is maintained outside the EMR. ¹⁵ Other studies have described various types of unintended consequences to the use of electronic records including workflow, process, and communication issues. ¹⁶ To help reconcile some of these challenges, a number of different models on technology adoption are found in the literature.

Theoretical models of EMR adoption

Several theoretical models have been generated to help conceptualize technology adoption both inside and outside healthcare. Examples of such models include Actor Network Theory, ² Holistic eHealth Framework, ¹⁷ Sociotechnical Framework, ¹⁸ adoption models such as Technology Adoption Model, ¹⁹ and other review frameworks for EMR adoption. ⁴ Lau et al. ¹⁷ developed a pan-Canadian EMR adoption strategy across macro, meso, and micro levels. While the approach identifies EMR adoption strategy needs, there are a number of top-down, bottom-up, and lateral influences within the adoption framework that are not accounted for in the model.

Overall, the above models have allowed us to understand factors associated with implementation of technology in healthcare. A shortcoming with theoretical models is they do not always scale down to the dynamic health environments where technology is used. For example, the Delone and McLean model ²⁰ was used by the Canada Health Infoway ¹² to develop a benefits framework for EMRs. The model states that key contextual factors including strategy, culture, and business processes are out of scope; however, it is these contextual factors that differentiate implementation from adoption. Our attempts to account for these dynamic contexts when implementing technologies such as EMRs are often completed with varying degrees of success according to Unertl et al. ²¹ Another shortcoming with existing EMR implementation frameworks is they typically focus on a selected subset of implementation factors ⁵ rather than the diverse spectrum of technical (eg, workflow, usability, interoperability) and non-technical (eg, governance, stakeholder negotiation, business continuity) factors that differentiate EMR implementation and adoption. Moving forward we need to extend theoretical frameworks such as ^5,17 to represent a broader range of EMR implementation factors and then provide guidance for how to transition from EMR implementation to adoption.

Tactical framework for EMR adoption

The main contribution from this article is a tactical framework for EMR adoption (Figure 1). The framework contains six categories of an integrated EMR adoption strategy. The adoption framework was developed by analyzing the literature on EMR implementation and adoption coupled with the authors years of experience in the medical device and health sectors. For example, Kruse et al. ⁴ found that other than cost or technical issues, change management to support workflow and training was a key driver of successful EMR adoption. Technology and change management are two categories in the framework. The workflow category was derived from numerous studies such as Ash et al., ¹⁶ Novak et al., ¹⁸ and Unertl et al. ²¹ The other categories were similarly derived from literature on unintended consequences ¹⁶ and findings from EMR adoption and implementation literature. ^3,7,17

OPEN IN VIEWER

Figure 1.

Tactical framework for EMR adoption. EMR indicates electronic medical record.

The six categories should not be thought of as mutually exclusive as factors may interact across the categories. A list of three seed questions are provided for each category as a starting point to help facilitate the dialogue for developing an overall strategy for EMR adoption.

The seed questions within each category provide a means to start the conversation to transition from theoretical frameworks of EMR implementation to a more tactical approach of EMR adoption within a specific healthcare context. Our tactical framework contains key implementation considerations such as workflow, training, and user-centered design, but the seed questions glean insight to enable transition from implementation to adoption. Our framework also acknowledges the overlap across categories. For example, the change management category will help determine the overall approach for change management strategies and how other categories such as training and environmental design should be done. The workflow category will be dependent on best practices, vendor selection, technology capability, and design factors. The dynamic nature of both healthcare delivery and technology adoption will require negotiation between stakeholders to answer the questions both within and across categories. Addressing these questions will allow for a more comprehensive approach to adoption. Finally, EMR adoption is a prime example of a LHS ⁸ and the answer for many of these questions will evolve over time, creating the need for an ongoing adoption strategy.

Conclusion

Electronic medical record adoption continues to be a complex problem. While research in the area of EMR implementation has identified the major barriers faced during implementation, ^4,13,14 EMR adoption research needs to incorporate change management approaches to help operationalize the support the transition from implementation to adoption of EMRs and other health information technologies. ⁶

A number of theoretical models exist that look at various aspects of EMR implementation, often from a high level. While these models provide meaningful insight, they do not provide a tactical approach to assist with the transition from EMR implementation to adoption. This article addressed that shortcoming and presented a tactical framework for EMR adoption. Our framework contains six categories and a set of seed questions within each category that provide a starting point for discussion to strategize EMR adoption at a tactical level. The more planning and consideration of how context will affect EMR implementation, the better chance we have of having a success implementation. The categories in our framework provide a means for obtaining the considerations of different stakeholders to facilitate successful technology adoption.

The complex nature of healthcare means that the EMR implementation process is a dynamic and challenging endeavour. Further, there is no one size fits all EMR adoption solution but rather they need to be tailored to the individual context of where the EMR will be used. We did not rank the categories such as by importance or other relevance, as the purpose of our framework is to enable conversations about EMR adoption in a particular context. The relevance of the categories for a specific EMR context will emerge through the conversations enabled by our framework. Shortcomings of our framework is that we have not tested it in any practical settings. That task is the next stage of our research.