You are here

  • Home
  • Archive
  • Volume 25, Issue 2
  • Understanding optimisation processes of electronic health records (EHRs) in select leading hospitals: a qualitative study

Article Text

Article menu

Download PDFPDF

XML
Research article
Understanding optimisation processes of electronic health records (EHRs) in select leading hospitals: a qualitative study
  1. Mark Chun Moon,
  2. Rebecca Hills and
  3. George Demiris
  1. Clinical Information Systems, Rush University Medical Center, Chicago, IL
  2. Biobehavioral Nursing and Health Informatics, School of Nursing, University of Washington, Seattle, WA
  3. Department of Biobehavioral and Health Sciences, School of Nursing, University of Pennsylvania, Philadelphia, PA and Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA
  1. Author address for correspondence: Mark Chun Moon Clinical Information Systems Rush University Medical Centre Chicago, IL 60612 mrmoon{at}uw.edu

Abstract

Background Little is known about optimisation of electronic health records (EHRs) systems in the hospital setting while adoption of EHR systems continues in the United States.

Objective To understand optimisation processes of EHR systems undertaken in leading healthcare organisations in the United States.

Methods Informed by a grounded theory approach, a qualitative study was undertaken that involved 11 in-depth interviews and a focus group with the EHR experts from the high performing healthcare organisations across the United States.

Results The study describes EHR optimisation processes characterised by prioritising exponentially increasing requests with predominant focus on improving efficiency of EHR, building optimisation teams or advisory groups and standardisation. The study discusses 16 types of optimisation that interdependently produced 16 results along with identifying 11 barriers and 20 facilitators to optimisation.

Conclusions The study describes overall experiences of optimising EHRs in select high performing healthcare organisations in the US. The findings highlight the importance of optimising the EHR after, and even before, go-live and dedicating resources exclusively for optimisation.

  • electronic health records and systems
  • clinical information systems
  • inpatient (inpatient CPOE)
  • system improvement
  • facilitators
  • barriers

Commons license http://creativecommons.org/licenses/by/4.0/

https://doi.org/10.14236/jhi.v25i2.1011

Statistics from Altmetric.com

    Request Permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

    BACKGROUND

    There is little documentation about optimisation of electronic health records (EHR) systems, namely, the process that takes place after implementation to maximise the benefits and utility of the system. Such optimisation following the go-live is critical to successful implementation in ambulatory settings.1–3 Despite the importance of EHR optimisation, little attention is given to this critical process in hospital settings. Too often, implementation of the EHR is considered complete once the system goes live. Optimisation of the system is usually an afterthought. Many studies about EHR, including those on unintended consequences associated with EHR systems, have predominantly focused on implementation and do not take optimisation into consideration.4–6 In reality, however, experts in health information technology say: ‘It’s one thing to go live and a completely different thing to see it through’.7 In addition, early adopters of EHR systems and documented studies emphasise the importance of the optimisation process.7–12

    This study is one of the first attempts to systematically understand and describe EHR optimisation efforts undertaken in hospital settings. As US hospitals continue to adopt EHR systems, documenting the experience of EHR optimisation efforts taking place at leading healthcare organisations will provide valuable insights on how to leverage EHR systems in the post go-live era.13–15 This insight may also benefit healthcare organisations that plan to adopt an EHR system by providing valuable information that helps decision-makers to establish a plan for both the implementation and optimisation phases. Additionally, this study may help to prevent the waste of resources caused by failed EHR implementations.

    Our objective is to understand the optimisation processes undertaken in high-performing hospitals following EHR implementation. The main research questions were: 1) What strategies do hospitals with implemented EHR systems employ to realise the benefits of the deployed systems or to meaningfully use the systems? 2) What advancements are hospitals making, post go-live, to leverage the EHR system? 3) Are there any pattern(s) of optimisation processes in hospitals and, if so, what are they specifically? 4) What are barriers and facilitators to optimisation? 5) How do organisations define and conceptualise EHR optimisation?

    METHODS

    Study design

    Informed by a grounded theory approach,16 a qualitative study was conducted that included in-depth interviews and a focus group with subject matter experts from healthcare organisations across the United States.

    Sampling and recruitment

    In order to include a diverse sample of participants, we pursued a purposive sampling approach. The sampling process is described as shown in Figure 1. We created a list of potential participants from five sources that identified high-performing healthcare organisations, which consisted of the Healthcare Information and Management Systems Society (HIMSS) Davies Award winners (33 healthcare organisations); Baldrige Award recipients (9); the 2014–2015 Best Hospitals Honor Roll by US News & World Report (6); Truven 100 Top Hospitals Winners appearing more than six times (45) and 2013 and 2014 top hospitals by the Leapfrog Group (21). The total number of the organisations was 114, and eight duplicate organisations were excluded, which resulted in a total of 106 candidates. Originally, the sources gave a much larger number of candidates, but exclusion criteria were applied. Any organisation with less than 100 beds was excluded. Also, any organisation located too far (requiring more than 10-hour driving) was removed for practical reasons. However, HIMSS Davies winners were not excluded based on location due to their significance.

    Figure 1
    Figure 1 Sampling process

    From the list, 970 individual participant candidates were identified, using a professional membership directory, literature, official organisation websites and an online professional network. Candidates included clinical information system directors/managers, organisational executives, medical directors, physicians, nurses, clinical staff and information technology (IT) professionals. Age, gender, race and ethnicity were not a factor in selecting participants. An email invitation to participate along with a link to the research website was sent to 946 candidates on 10 August 2015. Nineteen candidates responded. An invitation letter was sent by mail to the remaining 24 candidates out of 970, and one candidate responded. Of the 20 respondents, a total of 15 individuals (sample size = 15) representing 13 healthcare organisations across the United States participated in the study. One participant had dual representation for two separate organisations due to a recent employment change. We conducted 11 interviews and one focus group consists of four individuals. The focus group was offered by one participant site. Considering its relatively small size compared to other participant sites, we accepted it as it could present a richer and more diverse description about optimisation. The participants and organisations were summarised in Table 1. The study was reviewed and approved for exempt status by the Human Subjects Division (HSD) at the University of Washington.

    View this table:
    Table 1 Participants and sites

    Data collection and analysis

    In-depth interviews and the focus group session were audio recorded at participants’ office locations (3) or by telephone (9) between 20 August 2015 and 21 October 2015, facilitated by the lead investigator by using the Interview Guide included as Appendix A. The investigator was experienced in facilitating interview and practiced an interview with the guide several times before conducting the interviews and focus group. The Interview Guide was developed during the research design and validated by HSD to elicit answers to the research questions. To control comparability of data, the focus group was performed similarly to an interview directed by the Interview Guide, seeking responses from the members of the group who were best able to answer a particular question. A total of 635-minutes of interview and focus group audio recordings were collected, transcribed and analysed. In addition, the documents that were available to the public such as HIMSS Davies Award applications and internal documents provided by the participants were collected and reviewed. The documents that presented reliable and objective information were selected, coded and analysed.

    We performed qualitative content analysis informed by a grounded theory approach.16 First, the study investigator transcribed the interview and focus group recordings. The investigator repeatedly reviewed the audio recordings and transcripts, having in-depth discussions with the research team throughout the data collection process. Second, we drafted a preliminary coding guide, identifying common themes emerging from the interviews and the transcripts and relating them to the research questions. Third, we developed a comprehensive coding guide by identifying sub-themes and relationships between themes. We further refined the complete coding guide during the initial data analysis with five participants’ data, merging overlapping codes and simplifying codes. Then, we applied the final coding guide to all data. Lastly, we interpreted analysed data by using Data Analysis Framework with Codes (Appendix B) to answer the research questions and understand the whole picture of optimisation. To strengthen the analysis and increase its accuracy, we had regular discussions throughout the iterative analytic process and utilised qualitative data analysis software ATLAS.ti (v.7.5.10).

    RESULTS

    The analysis of data revealed the following characteristics of optimisation efforts undertaken at the selected institutions, including optimisation processes, types (goals) and results of optimisation and its barriers and facilitators.

    Definition of optimisation

    Participants defined optimisation as an ongoing process making the implemented EHR more efficient and usable for end user clinicians that results in improved efficiency in clinicians’ practice and satisfaction.

    Exponentially increasing requests

    After go-live, both requests for fixes and changes increased ‘exponentially’. A developer participant reported requests for EHR enhancements ‘literally flooded in’ for different end users (P1). Another participant said, ‘We have seen … the number of requests go up exponentially. As people feel more familiar with the system, people like to change it, we get more and more requests’ (P4). The exponential surge of requests augmented the participants’ workload with a more flexible timeline even after a smooth go-live. Demands for both fixes and optimisation overflowed continuously.

    And it’s very strange. My workload now is ... it’s actually a lot more than it used to be ... I have a lot more on my plate now than I did during implementation ... it is exponentially more work to do optimisation than it is in implementation (P2).

    Importance of prioritisation emerged

    Participants recognised the importance of prioritising requests. Due to limited resources and staff, information services (IS) had to prioritise the requests to address the most important ones. In this prioritisation, a committee or advisory group that was multidisciplinary with representatives from various departments (e.g. medical staff and nursing) played a key role. A participant said, ‘Optimisation really is about prioritising top best usual resources to improve your system. That’s the hardest part’ (P2).

    The committee or advisory group not only determined prioritisation but also oversaw all system-wide changes that affected everybody within the EHR system. After a comprehensive review, the committee approved a solution, suggested a revision of a solution or recommended further investigation of an issue or request. They prioritised requests according to common fundamental principles: safety, efficiency, return on investment (ROI), quality, regulatory requirements and process improvements, listed in the order of the most cited.

    In addition to the central governance committee, forming a workgroup, team or sub-committee started the optimisation processes. These groups were a taskforce that carried out an optimisation project in support of the central committee. For example, one participant established a team to improve inefficient nursing charting (P5). Two sites had an official optimisation team, while others did not. The optimisation teams consisted of directors or managers, clinical informaticists, EHR application analysts and process improvement personnel (P10 and P12).

    Predominant focus on improving efficiency of EHR

    Improving efficiency of EHR was the predominant focus of optimisation processes. Out of 24 identified optimisation processes, the collective effort in increasing efficiency of EHR systems appeared most frequently, in a rate of 25.0% as shown in Table 2.

    View this table:
    Table 2 Optimisation processes

    Standardisation

    Standardisation was a common theme. Standardising workflows, processes and policies across the organisation was one typical part of the optimisation process. There were two types of standardisation. A top-down approach was a system-wide implementation of optimisation that was approved by the central committee. A bottom-up approach was the opposite. It started at a local facility and was proven with positive results. Then, this optimisation was escalated to the central committee for approval and subsequently, disseminated throughout the integrated health system, even to a national level.

    we want to make certain that the guiding principles say any build work that we do and the optimisations that we do, ensure that we have the gold standard across the organisation (P6).

    Standardisation required thoughtful change management. ‘You get this [change] because you have to have it and you have to prove to us why you have to have it’, said a participant (P2). In addition, there was a local level of standardisation described in a surgery supply optimisation project (P12). It was not an enterprise undertaking, but it standardised how the physicians chose surgery supplies that was previously driven by their preferences. The team unified multiple suppliers into one for cost containment.

    Optimisation timing

    In general, optimisation did not happen immediately after the go-live date. One participant said, ‘Once you get something that works, then you actually start to talk about optimisation’ (P7). However, one site began optimisation even before the go-live date. Because of the considerable investment required for the EHR system, the site made an intentional effort to help realise its full potential and the optimisation yielded great results (P12).

    Types of optimisation

    There were a total of 16 types (or goals) of optimisation, as summarised in Table 3. An optimisation effort often belonged to multiple categories. For example, simplifying a complex nursing flow sheet was categorised as increasing efficiency, improving quality of documentation and improving user satisfaction. The most dominant type of optimisation was increasing efficiency. Increasing efficiency had four sub-categories: making workflow more efficient, minimising time spent with EHR, presenting the right information at the right time during care and general efficiency that did not fit the other three subgroups.

    Efficiency is our biggest priority, you know … making things easier for the providers, for the nurses, to be able to do it themselves in an efficient and effective manner so that they can still see the patients the volume that they’ve always seen and still give good patient care ... (P8.2).

    View this table:
    Table 3 Types of optimisation

    The efforts realised measurable benefits in the participants’ sites. The participants reported achieving outcomes such as reduced physician’s time spent in the EHR by approximately 1 minute per patient (P4); decreased turnaround time by 37% for the emergency department (ED) door to physician and total ED waiting time by 44% for patients being admitted (P12) and cost savings of over half-a-million dollars per year due to reduction in imaging study denials (P9). Such benefits were realised due to other factors as well as process improvement.

    Smarter decision support was the second most common type of optimisation. Participant sites improved the way they utilised decision support. The EHR system did not merely trigger alerts. The EHR helped clinicians by integrating refined decision support into the clinician’s workflow, not as a ‘noise’ or interruption but as an effective reminder (P4).

    So, we want the right information be provided at the point of care … we can build an efficient reminder that has all the information but only the information they need to make effective decision at the time of satisfying the reminder (P5).

    Achieving clinical outcomes was one of the goals for EHR decision support. One participant’s site was optimising a diabetic foot exam workflow (P5). A careful investigation of one facility that had a substantial decrease in the number of foot amputations related to diabetic pressure injuries revealed that the facility did not have an alert triggered by clinical decision support. Instead, the facility had a specific practice – taking socks off of the known diabetic patients – that signalled providers to examine the patient’s foot. Their experience provided a perspective to think about triggering alerts in the EHR. In this case, firing an alert to remind clinicians to perform the exam had already proven ineffective, whereas the physical signal (removing the socks) integrated into the clinical workflow and achieved the desired outcome (doing foot exam).

    You want to get closer and closer to the not developed foot ulcer that means examine the foot and it was practice changing in the clinic that actually fits the outcome … it was the actual practice that fixed the outcome (P5).

    The third most common type of optimisation was improving quality of care. As noted above, this was related to other efforts such as increasing safety or efficiency. Some examples were decreasing sepsis mortality by building a system of the Modified Early Warning Score (MEWS) within the EHR (P6) and better identifying of patients with malnutrition by improving dietary information exchange between dieticians and clinicians (P7).

    Fourth, there were efforts to realise ROI, value and cost-savings in some organisations. They were strong drivers and decision factors to prioritise optimisation projects (P6). These initiatives were typically an organisation-wide committed effort that utilised the EHR as a ‘catalyst’ (P12).

    Our actual benefit was just shy of 35 million [dollars], and that was a direct result of the optimisation projects ... we exceeded our benefits by almost 20 million dollars in the first year which was a huge, huge accomplishment. Still very, you know, exciting to think about (P12).

    Fifth, there was an effort to optimise practice, process or workflow. The goal was optimising workflows, patient experiences or business processes by leveraging an EHR. ‘So it is one thing to say electronic medical record (EMR)/EHR optimisation, but it’s really much bigger than that. We’re talking about optimising the performance of an organisation, using an EHR as a tool, a catalyst’, said a participant (P12). For example, one site optimised the process of obtaining approval for imaging studies from payers that resulted in significant reduction in denials and consequently, increased revenue and satisfaction among all stakeholders (P9). The remaining optimisation types are summarised in Table 3.

    Results of optimisation efforts

    Optimisation initiatives generated positive and productive proceeds such as improved productivity and exceeding financial return. Table 4 summarises 16 results of optimisation efforts. The participants perceived realising ROI, cost-savings and value as the most tangible outcome of optimisation, followed by improved quality of care and enhanced efficiency. Improved efficiency was not the top perceived result although it was the predominant type of optimisation. Each result was not necessarily a direct output of only one effort. An optimisation effort yielded multiple benefits, and they were interdependent by nature.

    View this table:
    Table 4 Results of optimisation

    Barriers to optimisation

    Table 5 summarises 11 barriers to optimisation noted in this study. The participants indicated that resistance to change was the biggest barrier. ‘It’s always people [Laugh] – people who don’t want to change, people who believe they already do things the best way they can…It wasn’t the systems. It wasn’t the leadership as much as it was the staff nurse or the anaesthesiologist who just refused to change’, confessed a participant (P12). Resistance to change was mitigated by getting operation’s support or engaging key stakeholders from the beginning of optimisation projects. These were actually identified as two facilitators to optimisation. Engaging leadership and operations staff, clinicians and users was a key to overcoming the resistance to change barrier. The second barrier was limited resources, in particular, because IS was not a ‘revenue-generating’ department (P6). It was difficult for the IT department to justify increasing resources after a massive EHR implementation that itself required significant resources. Furthermore, a typical request-based allocation of resources added more strain because optimisation did not necessarily require an immediate fix that resulted in low priority on resource allocation. The third barrier was a bureaucratic process that often included multiple layers of approval. Bureaucratic processes were long and arduous and typically reduced agility and efficiency in optimising the EHR (P9). The remaining barriers are summarised in Table 5.

    View this table:
    Table 5 Barriers and facilitators to optimisation

    Facilitators to optimisation

    Table 5 summarises 20 facilitators to optimisation that were identified in this study. Overwhelmingly, the participants recognised dedicated resources for optimisation as the biggest facilitator. A dedicated resource did not necessarily mean an additional resource. It was rather a ‘commitment’ to the optimisation as evidenced by allocating staff/resources/time or forming a team to work on optimisation projects (P10). Dedicating resources was the single most effective facilitator for optimisation.

    What I love about what we had done in the past is… we had a small enough team, I should say a dedicated team to our group [organisation]… And that really worked well, and I worked closely with them and we trusted each other and we got a lot of stuff done (P9).

    Advisory groups, teams or committees that carried out or oversaw optimisation were the second biggest facilitator. These teams facilitated optimisation by performing an optimisation initiative, helping prioritisation and allocation of resources, bridging IT and operations or making system-wide decisions for change. They were ‘the foundation’ of a successful optimisation (P3). The groups met on a regular basis, and having regular meetings was another facilitator of optimisation.

    Interacting with users and business owners played an important role in optimisation and was the third most common facilitator. This connection between end users or operations and the IT team was essential for successful optimisation efforts. It helped the IT department to understand users’ issues and their actual workflows. The optimisation people conducted regular user-developer conferences, meetings, conversations or observed users and business owners. Based on common ground, they collaborated to develop a solution that received a wide and effortless adoption among end users.

    I think the big key is we left IT and corporate and we lived in the hospital. And that was I think the biggest and it’s biggest for me. I think it’s what made us the most successful (P12).

    Informatics professionals who connected IT to clinicians and operations were the fourth facilitator. One informatics nurse said, ‘Specifically, I’m kind of the communication conduit… I’m kind of the conduit for requests and changes and needs’ (P8.1). The remaining facilitators are listed in Table 5.

    DISCUSSION

    The main finding of this study was recognising the importance of optimisation following, and even before, implementation of EHR systems. We found there were overflowing requests largely related to increasing efficiency of EHR after implementation. This need to make EHR more efficient and usable is real, as evidenced by failing design and usability of implemented EHR systems.17–20 Improvement does not necessarily follow implementation, contrary to an assumption that a smooth go-live will automatically make clinicians’ jobs easier and subsequently improve clinical outcomes.10,21–24 To some extent, there are immediate benefits after implementation, but in practice, it takes an effort of cultivation to ensure such promised results are actually realised.9–12,25,26 Optimisation is a hallmark of successful implementation as McAlearney et al.2 discovered. From the evaluation standpoint, in accordance with Hadji et al.,27 this refinement of the system by optimisation becomes a major determinant of user satisfaction.

    The diffusion of innovation theory by Rogers28 is commonly used as a framework to study adoption of innovations related to information technology. According to this theory, there are four main determinants of success of an innovation: communication channels, the attributes of the innovation, the characteristics of the adopters and the social system. In our findings, we see that some respondents would be considered ‘early adopters’ of the optimisation approach (the innovation in this context); an organisation had started the optimisation process even before their EHR implementation phase. The EHR system was the largest financial investment this organisation had made, and ROI was a significant driver in adopting the optimisation approach as early as possible. The ‘relative advantage’ of the optimisation, namely the degree to which an organisation perceives benefits by adopting the innovation, is one of the user-perceived qualities that define an innovation and can affect adoption.28 In this case, the expectation of and commitment to anticipated benefits led to early adoption of optimisation. As Moore and Benbasat29 point out, the more an innovation can address the needs and expectations of potential adopters, the greater the likelihood for adoption. Our findings demonstrate that organisational commitment to maximising benefits from the implementation of an EMR may indeed affect the timing and extent of optimisation efforts.

    Second, the present study highlighted the importance of dedicating resources solely for optimisation, validating Cooley et al.30 request not to underestimate the resources necessary to support computerised physician order entry after implementation. Dedicated resources were the biggest facilitator and the second biggest barrier to optimisation. The participant organisations that devoted resources exclusively to optimisation had seen great returns. Cost of dedicating resources was relatively small in comparison to a typical huge investment in implementation. It was largely cost of human resources – a team of about five experts. After go-live, organisations tend to be occupied with maintenance requests, ‘putting out fires all the time’, thereby consuming most of their limited resources (P9). These types of requests are usually urgent because something is broken, but they are not necessarily the most important from an organisational standpoint. Even if there is attention to optimisation, it may not be optimal because in the end, the same people work both in the maintenance and optimisation. The staff may experience divided attention and burnout. In order to move optimisation opportunities to the next level, dedicating resources and staff are required. Optimisation should ideally be separated from maintenance and support of EHR.

    Third, this study found there are barriers and facilitators specific to optimisation. Most of the identified barriers and facilitators, for example, resistance to change,31,32 engaging leadership and end users,33,34 importance of informatics professionals33,35 and interdisciplinary committees30 are well documented in many studies.30,32–34,36–41 Despite this overlap, we uncovered additional barriers unique to optimisation such as a bureaucratic process requiring multiple layers of approval for change, poor communication and lack of standardisation. We also appreciated and highlighted those documented success factors in the optimisation perspective, particularly regarding dedicating resources.

    Fourth, this study revealed an outcomes-driven approach to clinical decision support. Drawing from the insight of a participant’s experience, clinical decision support should drive actual clinical outcomes (P5). Conventionally, clinical decision support consists of triggering many alerts and sometimes hard-stops, which interrupt the workflow of clinicians and cause alert fatigue. Mandated hard-stops dictate what clinicians should do, leaving them feeling less autonomous and dissatisfied. The ultimate goal is to ensure that clinical care is delivered at the right time, not just reminding clinicians of doing it. Actual clinical intervention, not a reminder, is the closest proxy to a desired clinical outcome. Busy and hardworking clinicians are drowning in a flood of useless alerts. Alerts should be refined and wisely integrated into the clinician’s workflow to drive actual intervention.

    Fifth, this study provided insights on defining EHR optimisation. Participants recognised optimisation as an ongoing process improving EHR systems. We noted there were two different approaches to optimisation. One was user-driven. Starting from issues or requests, it continually brought refinement and enhancement to the system, thus, making it more useful and efficient. The second was organisation-driven; it was an organisational project, not an IT project. It was truly about optimising processes, practices, workflows and performance of the organisation by leveraging the implemented EHR system. It was an intentional commitment to realise actual benefits of the EHR implementation. This finding is comparable to a conceptual understanding of EHR optimisation noted in the literature in the absence of any direct definition.3,42 Blavin et al.42 recognise EHR optimisation as ‘continually modifying technology for optimal use as better able to use technology to meet an organisation’s performance goals’ in a comprehensive literature review. Blavin et al.42 acknowledge optimisation as one of four distinct stages of the EHR implementation process, which are ‘planning and vendor selection’ (acquisition), ‘workflow and software design’ (implementation), ‘training and user support’ (implementation) and ‘optimisation and modification’. Notably, Blavin et al.42 discovered a need to examine EHR optimisation with emphasis on outcomes. Drawing from the findings from the study and literature, optimisation of EHR should be defined as an ongoing commitment with dedicated resources to improve the EHR system and realise its benefits to the fullest by achieving measurable outcomes both after and before implementation.

    Lastly, this study validated findings of prior studies. There was dissatisfaction among clinicians following EHR implementation. The decreased productivity, burden of documentation, sensory and cognitive overload and increased time required to get things done in EHR systems contributed collectively to clinicians’ widespread dissatisfaction and frustration over EHR systems.5,43–45 The reduced productivity was consistent with previous findings.46,47

    The study findings present numerous implications. Recognising a strong demand to improve efficiency of EHR systems as evidenced by surging requests, the healthcare organisations planning on, in the midst of implementation or post go-live should allocate adequate resources to optimisation. Despite limited resources, even small healthcare organisations could still make an intentional commitment to optimisation within their capacity. Ideally, a dedicated optimisation team should be established. The organisations should also cultivate an optimisation-friendly environment by eliminating or mitigating the barriers to optimisation while promoting the facilitators. Additionally, during the design and implementation phase, they should ensure more robust engagement of clinicians (e.g. usability test, seeking clinicians’ iterated feedback on EHR system design) to incorporate the clinicians’ optimal workflows and insights into development of EHR system which will result in more refined system. Finally, an evaluation system measuring actual outcomes of optimisation and implementation should be established and employed. Making the EHR system ‘go-live’ is just the beginning. Real success of EHR implementation or optimisation must be measured by its realised benefits.

    This study has limitations to consider. First, the collected data may not fully represent the participant organisations’ perspectives, because most organisations were represented by a single participant. Additionally, a 1-hour interview may not be enough time to capture the full experience of a large organisation. However, strong efforts were made to mitigate this risk by referring to publicly available information (e.g. HIMSS Davies Award applications) for validation of findings. Second, a single researcher completed the study interviews, transcriptions and large part of coding and analyses due to inadequate resources, thus, possibly introducing a personal bias and violating the best practice in qualitative research. In order to overcome this limitation, a rigorous study methodology was designed and followed and there was regular frequent discussion among the research team throughout the study. During the coding and data analysis, other team members reviewed the coding, refined development of the complete coding guide and data analysis framework and validated data analysis and interpretation. The team’s detailed involvement is documented in Authors’ Contributions. Third, the sample size is small (N = 15) with a low response rate (20 respondents out of 997 contacted individuals). Although we could recruit more participants with follow-up invitations or phone calls, we intentionally did not pursue it due to practical considerations and also, considering the qualitative nature of the study and the fact that data saturation was reached. Expanding sample size was not feasible, for this study was carried out under very limited resources. Lastly, the research team did not include patients and caregivers who would be an ultimate recipient of EHR optimisation benefits.

    Future research should focus on measuring the impact of optimisation on clinical outcomes. This study was not designed to identify a concrete connection between optimisation efforts and actual clinical outcomes for the patient. This is one area that needs improvement and further study.12

    CONCLUSION

    This study describes experiences of optimising EHRs in select high performing healthcare organisations in the US. Optimisation included prioritising exponentially increasing requests, focus on improving efficiency of EHR, standardisation and forming teams and advisory groups to guide optimisation processes. Sixteen types of interdependent optimisation efforts were identified and discussed as well as barriers and facilitators to optimisation. These results highlighted the importance of optimising EHR after, and in some instances before, go-live and dedicating resources exclusively for optimisation. We found that optimisation means more than improving an EHR system; it also includes the optimisation of workflows, processes, practices and performance of the organisation by leveraging EHR as a ‘catalyst’.

    Acknowledgements

    The researchers are very grateful to the participants who generously volunteered their time and shared their views and experiences. The researchers appreciate the colleagues who willingly provided valuable feedback and review of manuscripts.

    APPENDIX A Understanding optimisation processes of electronic health records in select hospitals

    Interview Guide

    Date ___________

    Participant Name ______________________ Organisation __________________________

    INTRODUCTION

    Thank you for participating in this study; I do appreciate your time. I need to give you a little background before asking if you agree to being interviewed. This interview is focused on your experience handling the optimisation of the EHR in your organisation after the system went live. During this interview, I’ll ask you questions about your background and position, the EHR implementation, and largely optimisation and your specific experience on post go-live. If you agree, I would like to record this interview; the audio file will be used to help with note taking. All information from this interview will be confidential. Your name will be known only to me. If names come up during the interview they will be removed from my notes to maintain your anonymity and the confidentiality of all information you share. This interview will take approximately 30–60 minutes. You do not need to answer every question. If at any time you wish to stop the interview, please let me know and we will end the session. I appreciate your candid responses to my questions.

    Do you agree to being interviewed? If participant agrees, recorder is turned on.

    Thank you for agreeing to be interviewed.

    BACKGROUND

    What is your title?

    Please describe your role within your organisation.

    How long have you been in this position?

    What is (was) your role in relation to implementing or supporting and maintaining the EHR system in your organisation?

    EHR IMPLEMENTATION

    Implementation process

    Please describe the overall implementation process of the EHR system in your organisation.

    EHR OPTIMISATION

    After go-live of the EHR system

    What has become your priority regarding the EHR after go-live?

    What workload change have you seen?

    What are any users or organisational expectations on post go-live?

    Definition of EHR optimisation

    How would you define EHR optimisation?

    EHR optimisation processes

    What has your organisation been doing to leverage the implemented EHR system?

    What processes or initiatives has your organisation undertaken to improve the quality and efficiency of care or lowering the cost of care with relation to the implemented EHR system?

    Analysis of EHR optimisation processes

    What results/outcomes has your organisation been seeing through these efforts? Why these results/outcomes are important to your organisation?

    What specific roles did the EHR optimisation play to achieve such outcomes/results?

    What are the barriers and facilitators you have seen during the EHR optimisation? How has your organisation overcome the barriers? How has your organisation promoted the facilitators?

    Who have played key roles for the successful optimisation? Please describe their roles in the EHR optimisation.

    Advice for embarking on EHR optimisation

    Do you think it is important to plan and execute the EHR optimisation, with dedicated resources, as an integral part of the EHR implementation? If so, why do you think so?

    What would you do differently to ensure successful EHR optimisation if you have to do it again?

    For those who are embarking on a journey of the EHR optimisation after go-live, what advice would you give to them?

    CONCLUSION

    Do you have any other concerns, thoughts or recommendations regarding EHR optimisation in your organisation or in general? Could you recommend any other persons in your organisation, who can best address the questions we just went over?

    That was my last question. Thank you for your time. What questions do you have for us?

    APPENDIX B Data analysis framework with codes

    View this table:

    References

    1. 1.↵
      1. McAlearney AS,
      2. Sieck C,
      3. Hefner J,
      4. Robbins J,
      5. Huerta TR
      . Facilitating ambulatory electronic health record system implementation: evidence from a qualitative study. BioMed Research International 2013;2013629574
      OpenUrl
    2. 2.↵
      1. McAlearney AS,
      2. Song PH,
      3. Robbins J,
      4. Hirsch A,
      5. Jorina M,
      6. Kowalczyk N, et al
      . Moving from good to great in ambulatory electronic health record implementation. Journal for Healthcare Quality 2010;32(5):41-50. doi:10.1111/j.1945-1474.2010.00107.x
      OpenUrlCrossRefPubMed
    3. 3.↵
      1. Pandhi N,
      2. Yang W-L,
      3. Karp Z,
      4. Young A,
      5. Beasley JW,
      6. Kraft S, et al
      . Approaches and challenges to optimising primary care teams’ electronic health record usage. Informatics in Primary Care 2014;21(3):142-51. doi:10.14236/jhi.v21i3.57
      OpenUrlCrossRef
    4. 4.↵
      1. Ash JS,
      2. Berg M,
      3. Coiera E
      . Some unintended consequences of information technology in health care: the nature of patient care information system-related errors. Journal of the American Medical Informatics Association 2004;11(2):104-12. doi:10.1197/jamia.M1471pmid:14633936; pmcid:PMC353015
      OpenUrlCrossRefPubMed
    5. 5.↵
      1. Campbell EM,
      2. Sittig DF,
      3. Ash JS,
      4. Guappone KP,
      5. Dykstra RH
      . Types of unintended consequences related to computerized provider order entry. Journal of the American Medical Informatics Association 2006;13(5):547-56Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC1561794/doi:10.1197/jamia.M2042pmid:16799128; pmcid:PMC1561794
      OpenUrlCrossRefPubMed
    6. 6.↵
      1. Ash JS,
      2. Sittig DF,
      3. Poon EG,
      4. Guappone K,
      5. Campbell E,
      6. Dykstra RH
      . The extent and importance of unintended consequences related to computerized provider order entry. Journal of the American Medical Informatics Association 2007;14(4):415. doi:10.1197/jamia.M2373pmid:17460127; pmcid:PMC2244906
      OpenUrlCrossRefPubMed
    7. 7.↵
      1. Leventhal R
      . Trend: EHR optimization. Post-implementation advancements. Leaders from various healthcare organizations explain how they have been moving forward with their EHRs following implementation. Healthcare Informatics 2014;31(2):30Available from: https://www.ncbi.nlm.nih.gov/pubmed/24812989pmid:24812989
      OpenUrlPubMed
    8. 8.
      1. Terry K
      . Rev up your EHR: how to optimize performance: learn ways to increase revenue, improve practice efficiency and quality. Medical Economics 2011;88(22):S4
      OpenUrlPubMed
    9. 9.↵
      1. Baron RJ
      . Quality improvement with an electronic health record: achievable, but not automatic. Annals of Internal Medicine 2007;147549-52Available from: www.annals.orgdoi:10.7326/0003-4819-147-8-200710160-00007pmid:17938393
      OpenUrlCrossRefPubMedWeb of Science
    10. 10.↵
      1. Desroches CM,
      2. Campbell EG,
      3. Vogeli C,
      4. Zheng J,
      5. Rao SR,
      6. Shields AE, et al
      . Electronic health records’ limited successes suggest more targeted uses. Health Affairs 2010;29(4):639. doi:10.1377/hlthaff.2009.1086pmid:20368593
      OpenUrlAbstract/FREE Full Text
    11. 11.
      1. Ryan A,
      2. Bishop T,
      3. Shih S,
      4. Casalino L
      . Small physician practices in New York needed sustained help to realize gains in quality from use of electronic health records. Health Affairs 2013;32(1):53-62. doi:10.1377/hlthaff.2012.0742pmid:23297271; pmcid:PMC3864096
      OpenUrlAbstract/FREE Full Text
    12. 12.↵
      1. Spetz J,
      2. Burgess JF,
      3. Phibbs CS
      . The effect of health information technology implementation in Veterans Health Administration hospitals on patient outcomes. Healthcare 2014Accessed 12 December 20152(1):40-7Available from: http://www.sciencedirect.com/science/article/pii/S2213076414000062
      OpenUrl
    13. 13.↵
      1. Adler-Milstein J,
      2. DesRoches CM,
      3. Kralovec P,
      4. Foster G,
      5. Worzala C,
      6. Charles D, et al
      . Electronic health record adoption in US hospitals: progress continues, but challenges persist. Health Affairs 2015;34(12):2174-80Available from: http://content.healthaffairs.org/content/34/12/2174.abstractdoi:10.1377/hlthaff.2015.0992pmid:26561387
      OpenUrlAbstract/FREE Full Text
    14. 14.
      1. Charles D,
      2. Gabriel M,
      3. Furukawa MF
      . Adoption of electronic health record systems among U.S . Non-federal acute care hospitals : 2008–2013 [Internet]2014. Available from: http://www.healthit.gov/sites/default/files/oncdatabrief16.pdf. Accessed 20 November 2015.
    15. 15.↵
      The Office of the National Coordinator for Health Information Technology (ONC)Office of the Secretary USD of H and HS. Report to Congress: Update on Adoption of Health Information Technology and Related Efforts to Facilitate the Electronic Use and Exchange of Health Information [Internet] 2014Accessed 20 November 20153001Available from: http://www.healthit.gov/sites/default/files/rtc_adoption_and_exchange9302014.pdf
    16. 16.↵
      1. Strauss AL,
      2. Corbin JM
      . Basics of Qualitative Research : Techniques and Procedures for Developing Grounded Theory. Thousand Oaks, CA: Sage Publications, 1998.
    17. 17.↵
      1. Chan J,
      2. Shojania KG,
      3. Easty AC,
      4. Etchells EE
      . Does usercentred design affect the efficiency, usability and safety of CPOE order sets?. Journal of the American Medical Informatics Association 2011;18(3):276-81Available from: https://academic.oup.com/jamia/article-lookup/doi/10.1136/amiajnl-2010-000026doi:10.1136/amiajnl-2010-000026pmid:21486886; pmcid:PMC3078654
      OpenUrlCrossRefPubMed
    18. 18.
      1. Lambooij MS,
      2. Koster F
      . How organizational escalation prevention potential affects success of implementation of innovations: electronic medical records in hospitals. Implementation Science 2015;11(1):75Available from: http://implementationscience.biomedcentral.com/articles/10.1186/s13012-016-0435-1doi:10.1186/s13012-016-0435-1pmid:27206920; pmcid:PMC4875635
      OpenUrlCrossRefPubMed
    19. 19.
      1. Khajouei R,
      2. Jaspers MWM
      . The impact of CPOE medication systems’ design aspects on usability, workflow and medication orders a systematic review. Methods of Information in Medicine 2010;493-19pmid:19582333
      OpenUrlCrossRefPubMedWeb of Science
    20. 20.↵
      1. Chan J,
      2. Shojania KG,
      3. Easty AC,
      4. Etchells EE
      . Usability evaluation of order sets in a computerised provider order entry system. BMJ Quality & Safety 2011;20(11):932-40Available from: http://qualitysafety.bmj.com/lookup/doi/10.1136/bmjqs.2010.050021doi:10.1136/bmjqs.2010.050021pmid:21807652
      OpenUrlAbstract/FREE Full Text
    21. 21.↵
      1. Agha L
      . The effects of health information technology on the costs and quality of medical care. Journal of Health Economics 2014Accessed 20 November 20153419-30Available from: http://www.sciencedirect.com/science/article/pii/S0167629613001720
      OpenUrlCrossRefPubMed
    22. 22.
      1. Hahn K,
      2. Ohman-Strickland P,
      3. Cohen D,
      4. Piasecki A,
      5. Crosson J,
      6. Clark E, et al
      . Electronic medical records are not associated with improved documentation in community primary care practices. American Journal of Medical Quality 2011;26(4):272-7. doi:10.1177/1062860610392365pmid:21266596; pmcid:PMC3137715
      OpenUrlCrossRefPubMedWeb of Science
    23. 23.
      1. Huntington MK,
      2. Shafer CW
      . EHR implementation adversely affects performance on process quality measures in a community health center. American Journal of Medical Quality 2010Accessed 26 July 201425(5):404-5Available from: http://www.ncbi.nlm.nih.gov/pubmed/20833986
      OpenUrlCrossRefPubMedWeb of Science
    24. 24.↵
      1. Zhou L,
      2. Soran CS,
      3. Jenter CA,
      4. Volk LA,
      5. Orav EJ,
      6. Bates DW, et al
      . The relationship between electronic health record use and quality of care over time. Journal of the American Medical Informatics Association 2009;16(4):457-64Available from: http://jamia.oxfordjournals.org/content/16/4/457.abstractdoi:10.1197/jamia.M3128pmid:19390094; pmcid:PMC2705247
      OpenUrlCrossRefPubMed
    25. 25.↵
      1. Price M,
      2. Singer A,
      3. Kim J
      . Adopting electronic medical records: are they just electronic paper records?. Canadian Family Physician 2013;59(7):E322pmid:23851560; pmcid:PMC3710063
      OpenUrlPubMed
    26. 26.↵
      1. Liang LL,
      2. Berwick DM,
      3. Permanente K
      . Connected for Health: Using Electronic Health Records to Transform Care Delivery. San Francisco, CA: Jossey-Bass, 2010.
    27. 27.↵
      1. Hadji B,
      2. Martin G,
      3. Dupuis I,
      4. Campoy E,
      5. Degoulet P
      . 14 Years longitudinal evaluation of clinical information systems acceptance: the HEGP case. International Journal of Medical Informatics 2016;8620-9
      OpenUrl
    28. 28.↵
      1. Rogers EM,
      2. Singhal A
      Salwen M, Stacks D. Diffusion of innovations. An Integrated Approach to Communication Theory and Research 20035thMahwah, NJ: LEA409-19
    29. 29.↵
      1. Moore GC,
      2. Benbasat I
      . Development of an instrument to measure the perceptions of adopting an information technology innovation. Information Systems Research 1991;2(3):192-222. doi:10.1287/isre.2.3.192
      OpenUrlCrossRef
    30. 30.↵
      1. Cooley TW,
      2. May D,
      3. Alwan M,
      4. Sue C
      . Implementation of computerized prescriber order entry in four academic medical centers. American Journal of Health-System Pharmacy 2012;692166-73. doi:10.2146/ajhp120108pmid:23230041
      OpenUrlAbstract/FREE Full Text
    31. 31.↵
      1. Boonstra A,
      2. Versluis A,
      3. Vos JFJ
      . Implementing electronic health records in hospitals: a systematic literature review. BMC Health Services Research 2014;14(1):370Available from: http://www.biomedcentral.com/1472-6963/14/370https://doi.org/10.1186/1472-6963-14-370pmid:25190184; pmcid:PMC4162964
      OpenUrlCrossRefPubMed
    32. 32.↵
      1. Poon EG,
      2. Blumenthal D,
      3. Jaggi T,
      4. Honour MM,
      5. Bates DW,
      6. Kaushal R
      . Overcoming barriers to adopting and implementing computerized physician order entry systems in US hospitals. Health Affairs 2004;23184-90. doi:10.1377/hlthaff.23.4.184
      OpenUrlAbstract/FREE Full Text
    33. 33.↵
      1. Gellert GA,
      2. Hill V,
      3. Bruner K,
      4. Maciaz G,
      5. Saucedo L,
      6. Catzoela L, et al
      . Successful implementation of clinical information technology: seven key lessons from CPOE. Applied Clinical Informatics 2015;6(4):698-715. doi:10.4338/ACI-2015-06-SOA-0067pmid:26767065; pmcid:PMC4704039
      OpenUrlCrossRefPubMed
    34. 34.↵
      1. Ovretveit J,
      2. Scott T,
      3. Rundall TG,
      4. Shortell SM,
      5. Brommels M
      . Implementation of electronic medical records in hospitals: two case studies. Health Policy 2007;84(2–3):181-90Available from: http://www.ncbi.nlm.nih.gov/pubmed/17624470doi:10.1016/j.healthpol.2007.05.013pmid:17624470
      OpenUrlCrossRefPubMedWeb of Science
    35. 35.↵
      1. Ash JS,
      2. Stavri PZ,
      3. Dykstra R,
      4. Fournier L
      . Implementing computerized physician order entry: the importance of special people. International Journal of Medical Informatics 2003;69(2–3):235. doi:10.1016/S1386-5056(02)00107-7
      OpenUrlCrossRefPubMedWeb of Science
    36. 36.↵
      1. Ash JS,
      2. Stavri PZ,
      3. Kuperman GJ
      . A consensus statement on considerations for a successful CPOE implementation. Journal of the American Medical Informatics Association 2003;10229-34. doi:10.1197/jamia.M1204
      OpenUrlCrossRefPubMed
    37. 37.
      1. McGinn CA,
      2. Grenier S,
      3. Duplantie J,
      4. Shaw N,
      5. Sicotte C,
      6. Mathieu L, et al
      . Comparison of use groups perspectives of barriers and facilitator to implementing EHR—a systematic review. BMC Medicine 2011;946. doi:10.1186/1741-7015-9-46
      OpenUrlCrossRef
    38. 38.
      1. Sengstack PP,
      2. Gugerty B
      . CPOE systems: success factors and implementation issues. Journal of Healthcare Information Management 2004;18(1):36-45Available from: http://www.ncbi.nlm.nih.gov/pubmed/14971078pmid:14971078
      OpenUrlPubMed
    39. 39.
      1. Weir C,
      2. Lincoln M,
      3. Roscoe D,
      4. Turner C,
      5. Moreshead G
      . Dimensions associated with successful implementation of a hospital based integrated order entry system. Proceedings the Annual Symposium on Computer Application [sic] in Medical Care 1994653-7pmid:7950009; pmcid:PMC2247902
      OpenUrlPubMed
    40. 40.
      1. Safdari R,
      2. Ghazisaeidi M,
      3. Jebraeily M
      . Electronic health records: critical success factors in implementation. Acta Informatica Medica 2015;23(2):102-4Available from: http://www.ncbi.nlm.nih.gov/pubmed/26005276%5Cn; http://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4430004doi:10.5455/aim.2015.23.102-104pmid:26005276; pmcid:PMC4430004
      OpenUrlCrossRefPubMed
    41. 41.↵
      1. Scott JT,
      2. Rundall TG,
      3. Vogt TM,
      4. Hsu J
      . Kaiser Permanente’s experience of implementing an electronic medical record: a qualitative study. BMJ 2005;3311313-6. doi:10.1136/bmj.38638.497477.68
      OpenUrlAbstract/FREE Full Text
    42. 42.↵
      1. Blavin F,
      2. Ramos C,
      3. Shah A,
      4. Devers K
      . Lessons from the literature on electronic health record implementation: a study funded by the Office of the National Coordinator for Health Information Technology of the U.S. Department of Health and Human Services [Internet]2013. Available from: http://www.healthit.gov/sites/default/files/hit_lessons_learned_lit_review_final_08-01-2013.pdf. Accessed 20 November 2015.
    43. 43.↵
      American Medical AssociationImproving care: priorities to improve electronic health record usability [Internet]2014. Available from: https://download.ama-assn.org/resources/doc/ps2/x-pub/ehr-priorities.pdf. Accessed 6 October 2014.
    44. 44.
      1. Friedberg MW,
      2. Chen PG,
      3. Van Busum KR,
      4. Aunon FM,
      5. Pham C,
      6. Caloyeras JP, et al
      . Factors Affecting Physician Professional Satisfaction and Their Implications for Patient Care, Health Systems, and Health Policy [Internet]. Santa Monica, CA: RAND Corporation, 2013.Available from: http://www.rand.org/pubs/research_reports/RR439.html
    45. 45.↵
      1. Koopman RJ,
      2. Steege LMB,
      3. Moore JL,
      4. Clarke MA,
      5. Canfield SM,
      6. Kim MS, et al
      . Physician information needs and electronic health records (EHRs): time to reengineer the clinic note. The Journal of the American Board of Family Medicine 2015;28(3):316. doi:10.3122/jabfm.2015.03.140244pmid:25957364
      OpenUrlCrossRefPubMed
    46. 46.↵
      1. Bonner LM,
      2. Simons CE,
      3. Parker LE,
      4. Yano EM,
      5. Kirchner JE
      . “To take care of the patients”: qualitative analysis of Veterans Health Administration personnel experiences with a clinical informatics system. Implementation Science 2010;563Available from: http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2931450/doi:10.1186/1748-5908-5-63pmid:20727182; pmcid:PMC2931450
      OpenUrlCrossRefPubMed
    47. 47.↵
      1. Jamoom EW,
      2. Patel V,
      3. Furukawa MF,
      4. King J
      . EHR adopters vs. non-adopters: impacts of, barriers to, and federal initiatives for EHR adoption. Healthcare 2014Accessed 12 December 20152(1):33-9Available from: http://www.sciencedirect.com/science/article/pii/S2213076413000845
      OpenUrlCrossRefPubMed

    Footnotes

    • Competing Interests The authors declare they have no conflicts of interest in the research.

    • Funding The authors declare the study was not funded.

    • Authors’ Contributions Mark Chun Moon conceived and designed the study, recruited study participants, collected data, performed data analysis, drafted and revised manuscript. Rebecca Hills refined the study design, reviewed the Interview Guide, the coding guide and the coding, data analysis and revised a significant portion of manuscript. George Demiris supervised overall conduction of the research, reviewed and approved the study design, sampling, data acquisition, analysis and interpretation and revised a significant portion of manuscript.