The US Health Information Technology for Economic and Clinical Health (HITECH) Act of 2009 aimed to reduce healthcare costs, improve quality, and increase patient safety by rewarding providers who adopt and meaningfully use certified electronic health records (EHRs) [
Resistance to EHR adoption and use is not unique to primary care [
In contrast, physicians have demonstrated a reluctance to embrace EHR technology because of personal factors (unwillingness to change practice), professional factors (perceived threat to professional autonomy), equity factors (efficiency benefits accrue to the hospital rather than physicians), and technical factors (limitations of the technologies) [
While there are a number of successful pioneer EHR adopters, the chasm between the early adopters and the early majority is difficult to span [
This study was designed to explore physicians’ and organizations’ perspectives about how adoption and implementation of an electronic health record (EHR) system can be facilitated in ambulatory practice settings. We conducted in-depth qualitative interviews and focus groups to better understand and describe the process through which successful sites implemented their EHR. Our study was guided by the five states of the adoption process outlined by Rogers [
We interviewed 45 physician and organizational representatives across six organizations that were considered as exemplars for successful ambulatory EHR implementation. The sites were selected using an iterative process. First, we generated a list of potential sites based on information from a variety of sources, including (1) winners of the Healthcare Information Management Systems Society (HIMSS) “Davies Award” which recognizes excellence in EHR implementation; (2)
In addition, we conducted six focus groups comprised of 37 physician providers. These providers included physician residents in training as well as attending physicians and physicians in private practice. Additional information about our study participants is provided in Table
Key informants and focus group participants, by role.
Key informants | Number |
| |
Executives | 7 |
Managers | 11 |
IT Professionals | 13 |
Physicians | 10 |
Nurses | 4 |
Total |
|
| |
Focus group participants | Number |
| |
Practicing physicians | 20 |
Physicians in training | 17 |
Total |
|
We designed in-depth interview guides comprised of open-ended questions for both the focus groups and key informant interviews. Our techniques conformed to established standards for rigorous qualitative research outlined by others’ research [
Interviews lasted 30–60 minutes and were recorded and transcribed for later analysis. Focus groups lasted 60–90 minutes and were led by a facilitator with a comoderator available to assist. Focus groups were also recorded and transcribed prior to analysis.
Our analytic process used a grounded theory approach [
Study participants were asked direct questions about the EHR implementation process, including about the challenges of implementation. We categorized responses by type, reflecting personal or system-related barriers, at either the provider or organization level. Personal barriers at the provider level included difficulty in changing work patterns, lack of computer skills, and the nearness of retirement; system-related barriers at the provider level involved providers wishing to customize the system, loss of productivity, and loss of ability to document in detail. At the organization level, personal barriers included general resistance to change, lack of perceived value from the EHR system, and perceptions of insufficient support for EHR use. System-related barriers at this organization level involved challenges such as the system going down, the system’s limitations, and managing system updates. In Table
Personal and system-related barriers challenging EHR implementation.
Personal barriers | System-related barriers | |
---|---|---|
Provider-level barriers |
|
|
| ||
Organization-level barriers |
|
|
Using the conceptual framework we established for this exploratory study, we sought to characterize distinct elements of EHR implementation process that could be associated with the Implementation Phase of the Rogers’ Innovation-Decision process [
Making the Ambulatory EHR Adoption Decision
“A formal strategic initiative was developed and submitted and required approval for funding since obviously it was huge, the funding dollars that were involved.” ~Physician “It was across the entire system, it was not something that was in any way shape or form limited to simply just the IT department.” ~Executive Vendor Selection and Support
“There was more your standard vendor analysis process that went on to figure out what the best product was in relation to what [the health system’s] needs were, and so forth.” ~IT Professional “We leveraged a very strong partner, an outsource vendor Implementation Planning
“So what we do then is we evaluate each of the people and we assess them as far as, ok now we’ve identified each of the people that we think are going to need maybe some extra training versus others. And then it just gives you a feel as far as the doctors. “Do I feel comfortable as far as using an electronic medical record? Do I think an electronic medical record is going to benefit my practice? What are the detriments of a medical record? Do you think it’s going to help the communication?” Based on some of those answers we can tell the attitudes and we can tell where we’re going to find some challenges.” ~Manager
“There was a really aggressive roll-out plan to do it. There were different scenarios: two years, three years, four years, five years.” ~Manager “[On the project plan is] everything you need to do starting from the walkthrough of the hardware, doing the tutorial, getting all the hardware in place. And we have a check off—is this done? Is this done? Turning on interfaces, end user training, any clinical content, end user training, go live. And any follow-up and any process improvements that we do along the way. So yeah, everything.” ~Manager Implementation Go-Live
“We were in meetings for every week for hours upon end doing these screen designs and talking through the flow.”~Physician “We really were learning that and she understood the flow and we understood our flow. XX understood the electronic flow. So she would find out how she needed to make the system work for us and knew the system’s limitations and so then she’d say, ‘Well you’re not going to be able to do that in the system. We need to work within these limitations.’ And so we’d have to bring that back to the group and say, ‘Well we can’t do that, so how are we going to
“If those charts aren’t preloaded, it not only affects the provider, it also affects the nurse, it affects the staff-lab, because lab now is backed up.” ~IT Professionals “As far as best practice goes,
“Part of what we anticipated during the implementation would be a reduction in physician productivity. We gave them a six-week period of time to do that—to go back and build their skill set. Six week period—they were given a guarantee in terms of their capitation during that period of time.” ~Manager
“And then just a really good support staff that’s going to be able to help them through this process. And I need quality people to be able to send out there. I can’t send somebody who’s marginal out to a practice and expect to have successful go-live.” ~Manager
The first phase of implementation described across all sites involved the decision to adopt an ambulatory EHR system. Part of this decision was the articulation of a clear rationale, vision, and goals for EHR adoption, and these were clearly communicated to key stakeholders early and consistently. Across sites, informants emphasized the importance of early physician buy-in, with several sites noting that, ideally, the physicians themselves should drive the EHR adoption decision as opposed to the health system. Early and strong physician buy-in was deemed critical in order to ensure successful implementation and minimize resistance. During this adoption decision phase, informants also noted that organizations must recognize the significant investment required and that the investment involved a sustained commitment of both financial and nonfinancial resources (e.g., time, commitment to change).
The second phase we characterized in the implementation process involved selecting an EHR system vendor for system implementation and ongoing support. Several best practice sites viewed their EHR vendor as an active “partner” in the EHR implementation process, but this was not articulated as a firm requirement. Informants did note, however, that physician users should be an integral part of the vendor selection process to ensure that the product selected could support their clinical needs/preferences.
Prior to implementation, successful organizations took a strategic approach to minimize disruption to the organization. Several organizations incorporated a “readiness to change” assessment as part of this planning phase, using structured and formal processes to assess “readiness for implementation” at both the clinic (e.g., current use of IT, clear understanding of change needed) and individual clinician levels (e.g., typing skills). Information from the change readiness assessment informed decision-making and focused support efforts as the implementation progressed (e.g., making a decision to support typing skills training).
Additionally, best practice sites discussed how they used a project management approach to implementation that included several key components: (1) having a designated project leader; (2) establishing clear timeframes and accountabilities; (3) incorporating a systematic approach to feedback and improvement; and (4) ensuring clear and consistent communication. While the strategy was consistent, the personnel charged with leading components varied across sites. In some sites, the health system assigned IT-based project managers to partner with individual practices and clinics. In other sites, individual leaders within the practice were identified to lead the project. Regardless of the model used, informants from all sites highlighted the need to dedicate resources to the EHR implementation.
Further, communication was repeatedly emphasized as critical to success during the Implementation Planning phase. A clear understanding, across all parties about their role(s), the timeframe for implementation, and the expected impact of the new system, was identified. Messaging was also important, for instance, the need to consistently communicate the message that this transition was not one that could be completed “on the side.” By managing expectations and creating a system for constructive dialogue, negative word-of-mouth exchanges that reduce the likelihood of successful implementation could be addressed before they influenced attitudes towards EHR adoption, acceptance, and use.
The final phase of implementation representing the switchover to the new system had six subelements: (1) phasing; (2) workflow analysis; (3) training; (4) preloading; (5) mitigation; and (6) support. First, the process of implementation phasing in these sites consisted of determining the pace and scope of the implementation “go-live,” ranging from a “big bang” approach in which the EHR was implemented “overnight” within a site (or across sites), to a more conservative approach in which various functions of the EHR were phased in sequentially. Each approach appeared to have both benefits and drawbacks and we found no approach that was the most successful for phasing the implementation.
Second, as informants across sites are noted, a detailed workflow analysis and redesign were critical in order to ensure that the EHR could be utilized effectively; without this detailed process, organizations reportedly risked transferring bad practices from the paper environment into the electronic world and missed opportunities which the EHR provides to improve practice. Most sites viewed the workflow analysis process as a way to review and improve clinical practice and operations.
Third, while all sites required physicians to participate in a formal training prior to implementation go-live, the most successful trainings were conducted within a few weeks of go-live. Many informants noted that giving physicians the opportunity to “play around” with the system prior to going live was very valuable, but informants also reported they typically found that many physicians were unwilling to dedicate the time to learning about the EHR until it was absolutely necessary. One health system with a very sophisticated training model had evolved their training approach into a “competency-based” model. In this organization, physicians had mandatory e-learning requirements and then had to demonstrate competency with basic EHR functions before they could use the new EHR system for patient interactions (e.g. place an order, document patient history and physical, etc.).
Fourth, although time-consuming, many sites reported that having physicians abstract their own charts and “
Fifth, all sites planned for reduced productivity levels during the go-live period, and reduced productivity was usually expected for 1-2 months. Productivity levels were typically reduced to 25% or so of original levels during an initial go-live phase, and then were bumped up incrementally (e.g., to 50%, then to 75%, etc.) as users gained proficiency with the system. Although most sites reported a return to full productivity, others had reportedly struggled to return to original levels. It was suggested that the link between physicians’ compensation and productivity expectations could make a difference. Several sites noted that compensating physicians for training time helped increase participation rates, but not all sites had the financial resources or flexibility to provide this support.
Sixth, support was a critical component of the implementation go-live period. We found that most sites offered on-site IT support (vendor and/or internal IT) during the go-live period (e.g., “red coats” who were available on-site during go-live to answer questions and help troubleshoot on the spot). It was reportedly critical that the on-site support persons have strong technical knowledge of the EHR system as well as clinical knowledge. Informants across sites reported that a capable support team was an essential facilitator of successful implementation and noted that in-person support was particularly wellreceived by physicians anxious about the transition to the EHR.
The final key component of successful EHR implementation involved an emphasis on optimization of EHR system use. Study participants noted two time frames during which optimization could be particularly valuable. First, a focus on optimization immediately after the system went live reportedly minimized frustration at a time when users were on a steep learning curve. Then, once physicians had some experience using the EHR system, a second opportunity arose when users were more engaged in the system and interested in “optimizing” their use of the EHR. Ideally, EHR implementation itself was described as a process of ongoing learning and improvement, and recognizing opportunities for optimization was noted to be an important part of this process.
We categorized four subthemes related to optimization in EHR implementation: (1) acknowledging optimization as a specific component of implementation; (2) supporting optimization of EHR use; (3) EHR user recognition of the value of optimization; and (4) the need to view optimization as an ongoing process. We present representative quotations associated with these four subthemes involving optimization in the list below.
“When you think about going forward, I hear there [are] human factor things that are going to come up as the implementation is wrapping up. There is the optimization. What are other things that are on the drawing boards for you all going forward?”~Manager “We’ve actually been doing optimization for, after all the clinics went live, for a little period of time after they went live, maybe 6 to 10 weeks we did a little touch points but now, a lot of them it’s a year past go-live and we’re still out there doing the optimization.”~IT Professional
“The optimization on the clinics side is coming out of the clinic education department. These are half a dozen people—they do the training for some of the doctors.”~Manager “So they’re really out there kind of on the cutting edge of looking for more things to use the system for.”~IT Professional
“The people that are here now have really embraced it and they’re more looking… they love the optimization. They love it “When you go out there and show them something different it’s a big “aha!” moment and they really embrace it immediately because they can see that it’s a win.” ~IT Professional
“They are bringing back those optimization ideas and we are trying to spread them throughout the clinics as better work processes.” ~Manager
Ford and colleagues recently noted that qualitative studies were needed to better understand HIT implementation issues: “during this transitional phase, systematic qualitative studies would be the most valuable to both policy makers and leaders wishing to explore the management of information systems” [
Our results support the maxim that accomplishment is detail dependent and success is interpersonal. Specifically, many of the successful implementation strategies we identified were the result of both meticulous planning and clear mechanisms to deal with unforeseen or unintended consequences. Further, the interpersonal issues are crucially important—EHR adoption is a cultural revolution [
Research conducted by Lapointe and Rivard [
From a management theory standpoint, this research highlights the complexity of the Implementation Stage of Rogers’ Innovation-Decision process and proposes a distinct “optimization” phase of implementation that can enable organizations to progress to the Continuation Stage of the Innovation-Decision. While this characterization shares elements of others’ discussions of innovation modification during the adoption stage of the innovation-decision process (e.g., [
Finally, productivity hits are to be expected, but they can be planned. There is a learning curve associated with EHR implementations, and the time for this learning must be accommodated. In addition, prior research has shown that incremental implementation strategies are associated with better productivity outcomes [
Several limitations to the current research need to be addressed. The first is the small number of hospitals that were contacted for the study. It should be noted, however, that given the small number of ambulatory practices that have been recognized for having fully implemented an EHR system that meets the new federal regulations [
As meaningful use regulations have incentivized more practices to adopt EHR systems, we expect to see more discussion of the successes and failures of these efforts. It remains to be seen whether these first adopter sites will adopt strategies that are entirely different from strategies used by later sites, for instance, by selecting the alternative that is least costly in the short term in an effort to take advantage of the meaningful use program’s phased reward program only to see long-term costs increase as a result of that short sightedness. Alternatively, practice leaders may simply choose to take on the project in one fell swoop and use the single vendor approach. Future qualitative research will be crucial to the exploration of these adoption dynamics.
As the healthcare system moves toward greater and more meaningful use of EHR systems, individual organizations benefit from understanding not only barriers to implementation but also critical elements of the implementation process. This study examines organizations that have come to the latter two stages of rogers’ diffusion theory, and, in many ways, these two stages represent the light at the end of the tunnel for those embarking on or traversing through the initial stages of their adoption journey [
Ethical approval was obtained from the Behavioral and Social Sciences Institutional Review Board, The Ohio State University.
The authors are extremely grateful to the organizations and informants who participated in this study and to the health system members of our Project Advisory Team. They also thank their research team members, Drs. Paula Song, Deena Chisolm, research associates Dr. Annemarie Hirsch, Maria Jorina, Trevor Young, and Emily Kathryn Orcutt Knecht, and clinical consultants, Drs. John Mahan, David Rich, and Scott Holiday, all of whom were affiliated with The Ohio State University during the study. This research was funded by the Center for Health Management Research. There is no conflict of interests associated with this paper.