A qualitative study of physician perspectives on adaptation to electronic health records

To improve our understanding of this next phase of use, adaptation to EHR technology, we conducted a qualitative study of physicians’ perspectives related to adaptation to use of an EHR to identify factors that facilitate adaptation during initial implementation as well as adaptation required when the EHR system is upgraded. Our goal was to obtain a range of perspectives across specialties to identify common adaptation factors rather than to compare factors relevant to one specialty to another.

We conducted semi-structured interviews with hospital-based physicians at a large Academic Medical Center (AMC) in the Midwest. The AMC implemented the Epic Systems EHR (Verona, WI) system-wide in 2011. The EHR is in use in all inpatient and outpatient practices. The AMC operates on a regular upgrade schedule in which minor upgrades may take place quarterly with larger upgrades bundled into an annual upgrade and provides notification to all care team members through multiple venues including email, login screen notification and discussion at departmental meetings.

We focus on physicians from any specialty practicing in the inpatient setting because use of the EHR may be different in the inpatient and outpatient setting, with the inpatient setting requiring potentially different interaction with the EHR than may be necessary in an outpatient setting. In addition, the roll out process used for inpatient and outpatient settings was different which may have influenced adaptation. Because workflow and EHR use patterns may vary by specialty, we sought to include a range of physician specialties in our sample to obtain as broad of a view as possible. Interviewees were identified by a snowball sampling method in which each interviewee was asked to identify colleagues who might be willing to discuss their perspectives on incorporating the EHR into their practice. The research team met regularly throughout the data collection process to determine when saturation of concepts was reached.

The semi-structured interview guide included questions focused on 5 main topics: physician background information including area of specialty, years of use of EHR and self-described comfort with EHR; frequency, format and perceptions related to training in using the EHR; impact of EHRs on communication with departmental colleagues, colleagues across the AMC and outside of the AMC; impact on work productivity; and perspectives about technology-related policies. Interviews were conducted by telephone and lasted approximately 30 min. All interviews were audiorecorded and transcribed verbatim.

We utilized both inductive and deductive methods using the constant comparative analytic approach [23]. Two members of the research team developed a preliminary coding dictionary based on the semi-structured interview guide and applied this dictionary to double code five interviews, reconciled any coding differences, then coded the remaining interviews independently. Coders met to discuss each coded transcript and compared themes, expanding the coding dictionary as new themes were identified. The entire study team met regularly to discuss coding progress, identify and resolve discrepancies and reach consensus. We used Atlas.ti (version 6.0) qualitative data analysis software to support our analysis.

Table 1 provides a demographic description of interview participants. Interviewees represented a range of physician specialties, had been practicing medicine for between 6 and 28 years, and were slightly over 50% male. All except one participant reported being moderately or very comfortable with the EHR. Our analysis did not identify differences in the themes described below by specialty, years of practice or comfort with the EHR.

We found that physicians noted specific benefits to using an EHR. For example, one physician noted the ease of using the EHR while conducting rounds, “As it relates to patient care, it makes it much easier. I was actually on the floor rounding when you called initially. It makes it easier. Certainly we can pull up any data any time that we need it. We have those little computers on the walls and stuff. It’s not like we have to go back to the main nursing desk to get the information.” Another noted that communication with patients and outside the health system can happen significantly faster, “It probably gets there faster. In the old days, I used to dictate a letter. In the 1990s I had a Dictaphone. I would see a patient, I’d dictate a letter, and then it would be transcribed. Then I had to sign it, and it was sent out. That might take a week.” Finally, another noted the increase in efficiency of communication among colleagues, “I think it gets a—the information a little bit more at your fingertips and embeds it in a place you’re already looking for the information, anyhow. I think it’s truncated the notes, so we’re not getting as verbose in the notes. I think that’s been one benefit where some of the communication between partners to be able to flag those kinds of things.”

In addition, we identified four major themes related to managing change from EHR implementation: impact of changes to the EHR, dealing with changes, factors that facilitate adaptation, and impact on patient care. Below we describe subthemes identified in each of these major themes with example quotations.

The impact of attitude on an individual’s response to change can be significant. A positive, accepting attitude toward change has been shown to result in engagement with changes required and more positive work behaviors [27, 28]. In our study, physicians identified that accepting the necessity of change in using EHRs positively influenced their ability to adapt to their use. Many noted that while they experienced frustrations with EHRs, they also recognized the benefits and thus felt more positive toward use. Other studies have identified strategies such as protecting time in the workday that does not involve interaction with the EHR and taking care of oneself outside of work as ways to maintain a perspective about adapting to EHR use [29, 30].

This finding also has implications for how EHRs and the accompanying upgrades are presented to physicians. Communication that helps physicians to see the benefits of EHRs and changes to the EHR could facilitate physician adaptation. For example, working with physician champions within a department who can model a positive attitude may help address specialty-specific concerns may facilitate adaptation among colleauges, a strategy that has been successful in facilitating interventions across various domains of health care [31, 32]. However, while physican champions may be effective for encouraging a positive attitude toward changes, they alone may not be sufficient. Organizations also need strong institutional leadership and a culture that supports clinicians in adapting to the changes associated with EHRs and associated upgrades [33, 34].

Incorporating stakeholders in the design process has gained acceptance across a range of disciplines [35]. Specific to EHRs, research shows that when clinicians are able to make significant contributions to the development of clinical content for EHR systems, they report higher levels of satisfaction when reviewing clinical content within their domain of knowledge [36]. Farley et al. suggest that EHR systems should be designed with clinicians involved at the front-line to ensure alignment with clinician perception and decision making [37], and no mechanism exists that allows or encourages end users to provide feedback about ongoing issues or concerns in the EHR. Furthermore, a growing body of literature exists that identifies poor EHR system design, rather than user error, as the more frequent cause of medical errors [37, 38]. In response to the lack of support in the traditional EHR development process, a non-profit organization created the open EHR approach [39]. Through this approach, clinicians are able to engage in the development and reviewing of standardized clinical content models, thus contributing to improved communication and quality patient care [36, 39]. Another study yielded similar findings using the open EHR approach, suggesting that successful EHR design is dependent on both technical and clinical competence, and clinicians and developers working together [40]. A user centered design approach could help realize the benefits of clinician involvement in EHR development and improve clinician satisfaction and quality of care [41, 42].

In addition to physician involvement in EHR design, including physicians in EHR training so that it can be tailored to their specialty can also help with adaptation. In the Technology Acceptance Model (TAM), training is an important concept because it provides an opportunity to influence the users’ perceived ease of use of the technology, thus increasing their acceptance of the technology [22]. In the context of EHRs, training that focuses on the specific applications to each specialty and workflow can improve perceived ease of use. In our study, physicians expressed a desire for training that focused on the particular ways physicians use the EHR, as compared to general training provided to all users, and reported seeking out representation on EHR-related committees in order to have their perspective included. In this way, training designed in collaboration with physicians to address physician-specific needs can increase perceived ease of use and help physicians adapt to using the EHR.

In our study, physicians noted concerns about communication about upgrades. While physicians expressed mixed views about how upgrades effect their productivity, with some expressing frustration and some accepting that upgrades are necessary and their productivity will recover, most wanted additional means of learning about upgrades and incorporating them into new workflows. Email and alerts at sign in were the most commonly mentioned means of communicating about upcoming changes to the EHR, but physicians frequently remarked that they did not find these means to be particularly effective. Because of the potential disruption caused by having to adapt to frequent and sometimes significant upgrades, healthcare systems need to find additional ways to support physicians in the process. The user centered design approach suggested above for development of the EHR can also be applied to communication related to EHR upgrades.

Finally, the impact of EHR use on patient care is still a concern. Physicians in our study noted that patient care was at the center of their desire to practice medicine and simply the presence of the EHR caused them to adapt how they manage these interactions. They noted descreased patient interaction and the need to adapt how they interacted with the patient. These concerns have also been noted elsewhere in the literature. For example, several studies have found adverse effects of EHR use on flow of conversation between the patient and physician. Margalit and colleagues found that physicians’ use of orientation statements (i.e., instructions and directions to the patient) decreased as their screen gazing and keyboard activity increased [43]. Similarly, additional studies discovered that entering data into the computer during the patient consultation was associated with lower levels of patient trust and patient satisfaction [44‐46]. Other research has found that the style of physician-patient interaction shifted from a conversational style to a more “blocked” style due to data entry during the consultation, thus limiting the ability to build or maintain good rapport [47, 48]. These concerns are present not only during the initial EHR implementation but also resurface when upgrades change the way physicians interact with the EHR. The result is that physicians must not only adapt their workflow to incorporate the requirements of the EHR; they must also adapt their communication style in the visit to accommodate the presence of computer on which they use the EHR. When upgrades to the EHR change the user interface, the physician must become familiar with this while still engaging with the patient.