Staff expectations for the implementation of an electronic patient record system: A qualitative study

Global evidence suggests benefits of introducing electronic patient record (EPR) systems to improve patient care. However, implementing EPR into healthcare organisations is complex and in the United Kingdom (UK), uptake has been slow. More research is needed to explore factors influencing successful implementation. This study explored staff expectations for change and outcome following procurement of an EPR system in a large academic acute NHS trust in the UK. Qualitative interviews were conducted with 14 members of hospital staff who represented a variety of user groups across different specialities within an NHS hospital in England. The four components of Normalisation Process Theory (Coherence, Cognitive participation, Collective action and Reflexive monitoring) provided a theoretical framework to interpret and report study findings.

impact of implementation both before and during initial implementation. In the UK, the National Health Service (NHS), is centrally funded and has national strategies that are adopted at local level. Given the major differences in the social, political and economic foundations of the UK health and social care system, it is important to explore whether the benefits derived from EPR are relevant to other contexts.
There is a limited, and emerging evidence base on aspects of context that play out as barriers and facilitators to interventions (36,37). The Medical Research Council's framework for developing and evaluating complex interventions states that taking account of context is crucial to implementation (38). Most effort in preparing for EPR implementation are directed towards organisational readiness, including staff readiness (39). A recent review reported that for 95% of included studies, users were the primary barrier to implementation (40), and an additional review highlighted the importance of human factors in the success of, and barriers to implementation (41). Greenhalgh and colleagues summarised the evidence as to why this might be case and concluded that 'it is not individual factors that make or break a technology implementation effort but the dynamic interaction between them' (42).
Many different approaches have been used in order to produce rich theorisations of implementation of technology-supported health and social care initiatives, but these academic outputs have not been directly accessible to front-line hospital staff. In an effort to plug this gap, several studies have drawn on this literature to produce frameworks to inform implementation (43,44). To ensure successful implementation and avoid further financial waste, we need to understand better how users of EPR can be facilitated with implementation, pay more attention to factors associated with disengagement and translate this understanding across 6 whole healthcare service organisations.

Aim
The aim of this study was to explore health professionals' expectations of change and outcome following procurement of an EPR. In particular, we were interested in testing the relevance of Normalisation Process Theory (NPT) as a framework to understand perceived implementation factors prior to single healthcare organisation-wide EPR implementation.

Study design
A theoretically informed, qualitative exploratory study was conducted to explore health professionals' expectations of EPR implementation at the study site.
Interviews were semi structured and were conducted during November 2016.

Theoretical approach
One recent review of implementation of e-health systems concluded that most of the literature is focused on organisational issues, neglecting the wider social framework that must be considered when introducing new technologies (45). Rather than focusing on predictors of behaviour, Normalisation Process Theory (NPT) focuses on the work that individuals and groups do to integrate interventions into routine practice (36,(46)(47)(48). It can help in understanding why some processes seem to lead to a practice becoming normalised while others do not. NPT proposes that implementing technology can be achieved through 'energising' four mechanisms: coherence (understanding of reasons for implementation and potential value of the technology), cognitive participation (preparedness to engage and commit to use the technology), collective action (ability to do the work to use the 7 technology) and reflexive monitoring (how staff appraise the technology) (46,48,49). It is widely accepted that NPT provides a consistent framework that can be used to describe, assess and enhance implementation potential (49)(50)(51)(52)(53)(54)(55)(56)(57). The mechanisms have high stability across settings and, notwithstanding challenges in applying NPT in terms of managing overlaps between constructs, there is evidence that it is a beneficial heuristic device to explain and guide implementation processes (58).
Recently, NPT has been used to explore implementation of digital health interventions (59,60), including introducing EPRs in specific care settings (49,61).
As far as we are aware, no studies have used NPT to explore users' expectations of change and outcome following procurement of an EPR that will be implemented across an entire healthcare organisation. Using NPT to explore health professionals expectations could generate a better understanding of how they can best be facilitated through the adoption process. This understanding is vital for those managing the change process.

Study setting
The study site was an NHS teaching hospital trust in the North of England. In November 2016, when the study took place, the hospital served a population of over half a million people. Annually there were around 135,000 attendances at Accident & Emergency, 121,000 in-patients, and around half a million out-patients. The population had increased by over 10% in the last 10 years, and was predicted to have increased by almost a third over the next 15 years.
There were approximately 6,000 people employed by the trust of which around 5,000 would be expected to use EPR during the course of their daily practice. It was estimated that approximately two-thirds of the workforce had little or no experience 8 of using EPR prior to their engagement with the local implementation strategy. A great deal of the patient record work of the trust was completed on paper. At least 44 clinical systems and clinical data repositories were being used in distinct areas within the trust and previous attempts at introducing a version of an EPR or systems that had some of functionality of an EPR had not been joined up. To remedy this, the trust developed a strategy for implementation of a single, trust wide EPR system.

Description of preparing for EPR use
A local, comprehensive procurement assurance process commenced in March 2015.
The EPR system came 80% preconfigured -the remaining 20% of the EPR could be adjusted so that it met local requirements. The components of the strategy that were designed to prepare mass users for use are the focus of this paper.
Essentially, these included clinically-led development of the strategy, strategy road shows across areas which included raising awareness of the EPR and its potential benefits to staff, as well as allowing staff to identify areas of concern, formal on site demonstrations and training, additional interactive demonstrations and e-learning materials including an online 'play-domain' which was essentially a dummy version of the system that staff could use to practice what they had learned and additional training to staff designated as 'EPR friends' or 'super-users' to support the 'go-live' weekend.
In the original iteration of the plan, the hospital trust were planning to go live in November 2016. The implementation phase lasted until the immediate preparation for the go-live weekend, which happened in September 2017. This study was conducted prior to an official announcement of a decision to delay going live, and as such this overview of the implementation phase is concerned with strategies that 9 were used prior to the original go live date.

Sampling and recruitment
Data was collected from health professionals and members of the EPR implementation team. A purposive sampling frame was used to recruit hospital staff who represented a range of staff groups and grades. Participants were recruited from four case wards, which were considered too broadly represent the key services provided by the hospital: Accident and Emergency (A&E), General Surgery; Rheumatology-outpatients and Elderly care (in-patient). Permission was sought from ward managers and/or senior staff at each ward to recruit staff to interview.
Recruitment emails were then sent either by the senior member of ward staff and/or qualitative researcher and asked staff to contact the research team directly should they wish to take part. Some snowball sampling was also used to recruit participants, either by asking health professionals to recommend colleagues at the end of each interview, or by the researcher being approached directly whilst visiting the wards for data collection.

Participants
Fourteen members of staff who represented a variety of staff groups (7 doctors and 5 nurses of different grades, a hospital manager and a ward clerk) and departments (Accident and Emergency, Assessment, Outpatients, General Medical, Anaesthesia, Neurology and Stroke) within the study site took part in semi-structured interviews.
Three interviewees were involved in EPR implementation, as members of the support team or super-users.

Interview design and content
Qualitative, face-to-face, semi-structured interviews were conducted and lasted between 11 and 40 minutes. The majority of participants were interviewed individually with three participants interviewed together. A topic guide (Additional file 1) provided the framework for the interviews, and was informed by the four mechanisms in NPT (48,62,63). Participants were asked about their perceptions of the benefits and barriers to implementation of EPR, as well as the perceived impact of the system upon their practice and patient safety.

Analysis
With consent from participants, all interviews were audio-recorded and transcribed verbatim. Analysis was facilitated by use of the qualitative data management programme NVIVO 12. CM conducted the analysis, with regular discussions held with JB and AS during coding and theme development. Analysis was conducted in two stages. The first involved initially analysing all data thematically, following guidance as outlined by Braun and Clarke (64). Theme and sub-theme development was largely deductive, using a-priori codes dictated by interview questions (e.g.

Results
The four core mechanisms of NPT provided a structure for interpreting the findings.
It is important to consider that activities in all four domains may occur concurrently, 11 and relations between these core concepts are not linear. Even so, they focus attention down on "how the work gets done" (47). An overview of the coding framework is shown in Additional File 2.

Coherence -do staff understand the reasons for EPR implementation and
the potential value of incorporating EPR into their routine work?
The extent to which health professionals understood the value in implementing EPR was strong amongst participants, with digitisation seen as a normal part of working in the modern NHS. Despite universal acceptance of the potential value of EPR, staff groups varied in their perceptions of the intended purpose of EPR. For nurses and ward clerks, the purpose of EPR was perceived as data-centric (data storage and sharing): Nurse (Sister): "It's going to become more of a task-orientated job, where you're having to input stuff into EPR, rather than just getting on and carrying on with your clinical work like you normally would…" (10:54-57).
Clinicians, on the other hand, viewed EPR as treatment-centric and an aid to patient flow and decision making: Doctor (Consultant): "…you will be able to get a better overview of the department, so to run the department will probably be easier." (3:289-290), and Manager: "[It will] reduce clinical variation, improve the safety of care for patients and drive decision-making… it tells them [clinicians] what to do so we get consistency in practice." (7:181-185).
The idea of using an EPR in routine practice was strongly supported by the majority of participants, with a number of anticipated benefits proposed. Participants were particularly enthusiastic about the prospect of having all information in one place.
The majority of views coincided with the 'official perspective' of the anticipated rewards of EPR implementation. The presentation of information in a standardised, legible format was particularly well-received: Doctor: "The thing that I am looking forward to most is being able to read the consultant's writing, which I, personally, struggle with at the moment, whereas if it's dictated and typed there is no, sort of, room for error. So, that is the best part of it for me." (2:114-116).
Use of EPR was expected to improve efficiency of transfer of information between different specialties, leading to improved prescribing and test requests: "…the electronic prescribing, the electronic requesting, those things will be better". (Consultant,. One participant, with direct responsibility for hospital governance, expected that over time, use of EPR would improve capacity for audit and research. This would optimise opportunities to produce robust evidence for good quality care, as well as highlighting areas for improvement: Manager: "…once it settles in…the benefits of what the outputs are from the system…I think it will prove that actually we deliver high quality care across the board. And then we'll know the areas where we don't and we can target them." (7:161-173).
Beyond improving access to and legibility of information, the anticipated benefits of EPR varied across and between staff groups and services. EPR implementation was expected to be of most benefit to the working practice of junior doctors. For example, it was expected that the risks of missing important information or steps required within clinical decision making processes would be minimised through prompts to enact specific protocols within EPR: Doctor: "…when you try to do a ward round for a person, or clerk somebody in, you physically can't do anything until you do a VT prophylaxis, until you put their weight to prescribe a drug…if you prescribe a blood thinning medication… somewhere it forces you to do a certain score of their risk of bleeding…things that basically can be missed out quite often if we are doing paper versions." (2:138-144).
It was less clear how nurses would benefit, particularly with regards to the volume of information that they would need to record into the system. Nurses were concerned as to how important information could be safely passed on to their colleagues: Nurse: "…we don't physically know how we are going to give handover…people worry about how that's going to happen safely, for the information to be passed on safely from one shift to the next…because there's a lot going on, tests and results chasing, all that sort of thing…" (1:47-54).
Nurses were also concerned that using EPR could take them away from the business of nursing: Nurse (Sister): "We're all a bit scared of is it going to be task oriented, taking you away from your patient care… taking time away from the patient so we can tick all the boxes on the system…" (10:20-26).
Senior clinicians, who were not members of the EPR support team, expected to benefit least from implementation, primarily because use was perceived to have the potential to slow down their pace of work: A cumulative effect of least benefit existed between senior clinicians and outpatient services. The relatively fast pace of patient flow in clinics, and a 14 perception that the staff working in these services were less computer literate than their acute services colleagues, meant that the introduction of EPR was perceived as a potential threat to service delivery: Doctor (Consultant): "… I think outpatients will be an absolute disaster..." (4:516).
Cognitive participation -are staff prepared to engage and commit to using EPR?
All participants viewed EPR as central to delivering patient care, and were motivated to invest in implementation. Participants with previous experience of using EPR (mainly junior doctors and members of the EPR support team) were relatively confident in the benefits to be derived from change in their usual practice: Doctor: "…the consistency in care with things that we miss out quite often will obviously be a big benefit." (2:152-154). For other participants (mainly senior clinicians and nurses), they were concerned that they were ill-prepared to use EPR.
Their concerns were based around four main issues: lack of consultation/preparation for context-specific needs and wants, equipment and usability, formal training and support for introduction of use.

Concerns raised about using EPR
The perceptions that some participants held about the way in which the implementation programme had been enacted impacted negatively on their engagement with EPR. The trust had put in place strategic planning for uploading data into EPR, yet several participants lacked knowledge of these and were anxious that ultimately front-line staff would be required to complete the majority of this work: Nurse: "…we don't have any ward clerks…we have to wait for admissions to do it… so we're waiting to put a patient actually on to the system before we can do anything really…" (13:351-357).
Despite a positive appraisal for the perceived benefits of the system, some health professionals felt unprepared to operationalise the system within their usual work practice. Senior staff reported a lack of engagement with them as to how the system could best work for them: Those participants who believed that the go-live weekend was imminent were concerned that they lacked access to computer equipment or lacked physical space in which to operate computers. Additional challenges related to the practicalities of agency staff using the EPR system. For example, for wards that depend on agency staff, there was concern that these staff may not know how to use the system, and that this deficit would lead to an increased workload for nurses. Despite online training provision for agency staff, participants were concerned this pre-requisite would put some agency staff off coming into the hospital, thereby reducing further 16 the numbers of staff available:

Training and support
To support staff commitment and engagement with EPR, the trust provided mandatory training events and additional resources including play domains (simulations of the EPR, which allow staff to practice using the system), and superusers (a group of health professionals that received additional training on the system). Participants were divided as to the impact of their engagement with training on their expectations of the EPR system. Junior doctors were relatively confident in their skills and abilities to use EPR, with one junior doctor reporting that they had treated formal training sessions as an opportunity to ask questions that they had generated through using the EPR play domains. However, others felt they had not received enough training, or found it too intense or generic: Nurse: "We are not trained enough to be sure we know what to do… I don't feel confident to back up somebody who doesn't know what they are doing." (11:218-224).
Many participants were not experienced in using computers in their daily work practice and reported a lack of opportunity to move beyond the classroom setting. There was dissonance between staff expectations and training objectives. One senior member of staff suggested that: "…the knowledge of the system is now ready, the skill of how you use it will only happen when we go live…" (Manager,. However, lack of capacity during shift hours and lack of access to play domains impeded some participants' ability to engage with EPR. Where they were able to practice on play domains, some participants found there was inadequate simulation of what they would do in practice: Several participants reported that they were efficient in performing 'little tasks' using EPR, yet were anxious as to how they would integrate EPR into their usual This was compounded by uncertainty over the level of support that would be available to them, particularly during the early implementation phase. Some participants were suspicious that plans for additional resources would not materialise, and they would be pushed to deliberately fail in order to gain access to additional support: Nurse: "…I think we have to fail in a way in order to…get loads of screens in there." (11-12:416-422).

Collective action -do staff feel able to do the 'work' to use EPR?
All participants believed that they had completed the official training programme, and had, to varying degrees, engaged with the additional resources that were available to them. The extent to which they perceived that this had prepared them for EPR implementation was influenced by perceived compatibility of EPR with existing work practices. Similar to findings reported earlier, the perceptions expressed by junior doctors indicated that they were least concerned about the impact of EPR on their working practice. Other participants reported concerns for perceived changes in their working relationships, patient flow and available information which may impact their ability to do the work of using EPR to improve patient care. However, participants were unanimous that they would have to find ways to make EPR 'work' for them in practice: Nurse: "…we have in our practice found out that you don't have to fill them all out, so we're already cutting corners." (1:545-546).

Working relationships
The role of junior doctors was expected to respond to and evolve with EPR implementation. The dynamic of ward rounds was perceived to change from consultants documenting clinical decisions to junior doctors having a more active role in care plans. Junior doctors expected to be doing most of the documentation, most of the time, which led to some concern that they would become clerks for their consultants and result in missed learning opportunities: Doctor (Consultant) "…one of my issues with junior doctors is that they will spend time being clerks on the computer rather than being a junior doctor… they won't be behind the curtain with [the patient]… I think it will have a significant impact on their potential training on the job". (8:512/551-555).
It was anticipated that some members of staff would require more support to use EPR than others. With the introduction of EPR, some participants were concerned that junior doctors would be left to: "sort their own selves out…and get themselves up to a certain level" (Nurse, 1:94-95). There was variation in understanding of the anticipated change to working relationships between different professions, with some staff unclear as to how their role would evolve: Ward Clerk: "… but apparently there are other things that we're going to be doing instead [of filing paper records], which I don't know…" (5:116-117). Unfortunately, some participants anticipated that staff may leave the NHS on the back of implementation as they would find introduction of EPR too cumbersome: Nurse: "…some staff on the ward are older and are frightened of the computer, even in this day and age. Two staff may leave on the back of this, because I think they will find it too much…" (1:27-30).
One participant suggested that where there was strong team cohesion, and were confident that they would 'ride the storm': Nurse: "…we're a good team on here, and I think if they can't manage on here then they're not going to manage anywhere else; and we know that it's doable…" (1:105-107).

Patient flow
With the introduction of EPR, consultations, including ward rounds, were expected to take more time to complete. Usual practice on in-patient wards is for junior doctors to complete lists of tasks for different patients after the daily ward round.
However, with the introduction of EPR, staff will be required to complete tasks such as recording allergies, ordering tests, and prescribing medications during the ward 20 round, which was expected to increase their duration and alter the dynamic:

Doctor: "… typing it all out, and drop down boxes, and searching… which is just a long drawn out version of what we do at the moment. So it will take longer…" (2:200-206).
Participants accepted that compulsory completion of templates may reduce the risk of important information or decisions being missed. However, anecdotal reports from a neighbouring hospital who recently implemented the same electronic system caused concern. Specifically participants discussed the potential for EPR to increase duration of ward rounds, which may delay discharges, affecting A&E waiting times and in turn pose risks to patient safety. They also based their perceptions on experiences in primary care following the introduction of EPR.
Participants were also concerned that sometime after implementation in primary care, wait times had not returned to pre-implementation levels:

Doctor (Registrar): "You go back to GPs…When their electronic records came in
years ago they were on six and two third minute appointments. They changed to ten minute appointments and they've never been able to go back…" (14:120-123).
Similarly, in out-patient services, participants were concerned that EPR use would limit and slow down productivity in services which were 'working flat-out' (Consultant,4:66). Longer wait times as the staff got used to using the EPR system were anticipated-with services considered unprepared to respond. Although there was a planned 25% reduction in clinic referrals for the first 2 weeks of EPR going live, some participants believed that this did not allow enough time for the system to be fully embedded. This was compounded by an observation that the majority of staff working in out-patients were comparatively slow typists and so EPR implementation was, to a point, considered an unjustified additional use of time. As a result, the initial implementation period was predicted to be: Doctor (Consultant): "…horrendous…" (9:326) and "… there's no turning back now, it's going to happen… we wait with baited breath" (8:820/828).

Available information
Implementation of EPR required changes to be made to the nature and type of information that could be recorded. This was perceived to be particularly complex for nurses who record lots of different types of information from different sources.
Participants were concerned that important information that could impact patient care would be lost, due to the sheer volume of information that nurses acquire and are required to record: Nurse: "…you could take a phone call from some relatives who were concerned about their mum, and you could be on the phone for 45

minutes, and you are getting all sorts of information thrown at you…you could have 4 or 5 of these conversations in one day… Most of us are only two-finger or one-finger typists…
We're worried about how long it is going to take up to record accurately their concerns… so that nothing gets missed." (1:419-430).
Similarly, clinicians were concerned that they would not be able to provide a comprehensive picture of their thinking around patient care, which may change the nature in which clinical opinion is communicated. Some participants were worried that although they could find ways to work around this issue, the rationale underpinning their clinical decisions would be lost through use of EPR. The loss of information on clinical opinion was considered to potentially result in a lack of transparency as to how patient care is carried out:

Doctor (Registrar): "You will lose a lot of information… you really need all that information in there… because it is a clear record of what story we were given, what examination findings we were given and what is the clinical opinion. And that is still a really vital part of what we do…there is a danger of losing some of that
information…" (14:166-171).

Reflexive monitoring -how staff appraise the EPR
Participants appraised the EPR by identifying a number of advantages and disadvantages to using the system.

Advantages
All participants perceived long-term benefits, which coincided with the official Participants believed that patient safety and quality of care would be improved through use of the system. They expected that EPR use would result in a reduction in risk of errors, particularly around prescribing. They also anticipated transparency in errors and safer practice as all information would be legible and collected in a consistent manner:

Doctor: "…I think with prescriptions and prescribing, often it [EPR] flags up errors.
So I am hoping that if…you try to prescribe…five hundred grams of amoxicillin which I have seen…it will flag that up and say, that is not an appropriate dose for a drug…" (2:158-163).

Disadvantages
Some participants (those not involved in EPR set-up) were concerned that the potential for intelligent problem solving was missing. There was a tension between standardisation and localisation of the system. Users' (clinicians) could not communicate with software developers directly and they believed that the EPR friends, who were mainly administrative staff, could not enhance the system directly. The individual needs of specific specialties, and a perceived complex chain of command in making changes to resolve such issues and the way in which the system could be customised was not transparent: Use of EPR was expected to expose further frustrations in the hospital system and that blame would falsely be apportioned to EPR. Participants were also concerned 24 that patients could be harmed as people did not know how to use the system. To off-set this, EPR implementation was ultimately perceived as moving towards 'paper-light' as opposed to a paper-less system: Manager: "I am anxious that we'll harm patients because people don't know how to use system, haven't got the skill. But the mitigation to that is that the patient takes priority, the system is just there. If you can't get it to work or you don't know how to do it, you write on a piece of paper…" (7:265-269).

Discussion
This study explored health professionals' expectations of EPR implementation. NPT provided a framework that has characterised a range of factors that staff perceive to impact their understanding of the purpose of (coherence), engagement with (cognitive participation), anticipated use (collective action) and appraisal of (reflexive monitoring) their preparedness for the EPR.
Our study identified that health professionals perceive potential value in using EPR, and that benefits of use would be reaped in the future. They were willing to engage and commit to EPR use, but for some staff the opportunities for them to do so were limited. There was variation in health professionals perceptions on their ability to do the work required for successful implementation. Despite most staff believing that EPR implementation was imminent, there were still challenges with acceptance.
The most obvious finding to emerge from the analysis was variation in preparedness for change at an individual, professional and system level. Opinions differed as to the anticipated impact of EPR use on roles, relationships and interactions. A variety of perspectives were expressed about sufficiency of training and support in 25 preparing staff to be able to do the work. A recurrent theme in the interviews was that acute services and junior doctors were perceived to be the main benefactors of EPR. This manifested as services which could be responsive to the system. Outpatients services, nurses and those senior doctors who were not involved in the implementation support team, would have to react to the system. This finding is consistent with post-implementation literature that has identified barriers and facilitators that are specific to professional and individual priorities (28,65) and suggests that electronic records are often viewed as a set of clinical systems for primarily clinical users (11).
In their accounts of perceived impact of use of EPR on ability to do their work, some staff anticipated reduced performance with the introduction of EPR. Previous literature has highlighted that early users of EPR systems experience a performance dip as they struggle with an unfamiliar system (66). Common negative impacts reported include changes to workflow and work disruption (1). The disruption to workflow and changes required are significant challenges for users, particularly in systems that have limited modularity and configurability (67). In the past, failure to implement has occurred because health professionals found that the systems did not meet their needs and required work-arounds in order to complete work procedures (68). Electronic transmission of referrals, requests and reports for example were reported as making some workflows faster overall. However, individual stages of these workflows could become more or less time-consuming than the work system that was previously operational, with a range of consequences for the different staff involved (11). There is a growing body of evidence that a new technology is easier to embed if there is a mutually supporting relationship between technical, social and organisational factors in which new, often unanticipated ways 26 of working are allowed to emerge (69). Some workarounds may, in some cases, result in more efficient ways of working (70). Whilst it is beyond the scope of the current study to explore the extent to which certain workarounds were encouraged by the organisation, it can be assumed that concerns about reduced performance that are reported here are potentially preventable through a perceived capacity for more effective design and tailoring of the EPR to meet local requirements. Further work is required to establish why and how preparation to use EPR worked well for some staff and professional groups, and not others.
Another important finding was that nurses were concerned as to how they would record the volumes of information that they usually record. There is an emerging, but limited evidence base that has reported on the negative impact of EPR implementation on nursing practice when documenting crucial patient information (27,(70)(71)(72). Nurses have in the past had minimal influence in the design of systems (70). The perceived resistance to EPR has been explained as defying poorly designed systems that fail to meet the needs of documentation of nursing practice (72). There is at present a lack of studies that explore nurses' experiences of EPR in hospitals, and work is needed to understand the specific needs of nurses with regards to using EPR. Surprisingly, there is also a lack of literature on the accounts of senior clinicians (who were not involved in developing implementation strategies) who are responsible for specific services within hospitals. The findings reported here have also drawn attention to the potential loss of information usually recorded in the clinical decision making process of patient care. There is at present no literature that reports on how this perceived loss impacts patient care. One possible explanation for these findings is "a lack of correspondence between the design of technological properties and the culture of professionals" (73), and adds 27 to the growing literature advocating a cultural approach to the study of technology in organisations (73,74). This suggests that implementation of EPR could be improved through efforts to ensure that professional group and service context are much more visible within the implementation preparation plan. The findings also indicate that the traditional roles of some health professional groups could be altered, significantly, through introducing organisation-wide EPRs. Further research is needed to understand the nature of any change to traditional roles and how these changes can be translated into factors associated with engagement.
Despite concerns, a common view amongst staff was that they were confident that they would find ways to make the system work for them. A recurrent theme within their narratives was that more could have been done to support them so as to make the transition to EPR less onerous. Our study has identified a perceived lack of user involvement in preparing for implementation. In the past, EPR implementation programmes have been criticised for being too centralised, for not engaging with trusts and their healthcare professionals, and for flawed procurement processes (13). Failure is often linked to implementations where health professionals perceived that they were not involved or listened to about specific requirements.
Conversely, a high degree of user involvement is associated with successful implementation (75,76). The use of pilot testing phases has been linked with successful implementation, where user feedback on their requirements informed the implementation strategy (77). In a rare discussion on successful implementation we reported on a bottom-up, user led development of an EPR at a different NHS hospital (78). The EPR was developed by an in-house team and evolved over an extended period (at a fraction of the cost of commercial EPRs) and illustrated the value of user involvement in EPR design. A positive feedback loop in which users were listened and responded to supported development and further growth of the system. Characteristics of success were high levels of user activity by large numbers of diverse users who report that they are getting significant benefits from its use.
Some staff were concerned that use would cause risks to patient safety. The narratives of staff suggested that their level of preparedness mitigates patientsafety concerns. The relationship between staff narratives on their preparation for use and patient safety concerns is depicted in Figure 1: Health professionals' perceptions on factors that influence successful implementation of EPR. In the period preparing for a hospital-wide EPR implementation health professionals' main concern was the impact on their ability to provide safe care for patients. They expected that implementation would have potential benefits and identified events and actions that could mitigate potential risks to providing safe, quality care.
Perceptions varied across different staff groups and specific care contexts with regards to how their training and perceived levels of support impacted patient safety. There is growing interest in the relationship between EPR and patient safety within the literature (14,30,(79)(80)(81)(82)(83) . As UK systems mature it will be possible to further explore impact to patient safety. A recent analysis of patient safety incidents in fully digital hospitals shows that human-computer interaction is associated with most health information technology incidents (84). Nevertheless, there exists a perception that there are associated risks with implementation, most notably through incomplete or unavailable information (9,81), which warrants further investigation. The findings from our study draw attention to the way in which EPR potentially makes incidents that threaten patient safety more visible.
Implementation of EPR systems may also change the nature of risk and introduce new failure modes/incident types. Work is required to understand how EPR use influences detection of and management of patient safety concerns.

Strengths and limitations
A key component of our analysis was understanding how contextual and human factors influence preparedness for EPR use within the case analysed and as such generalisation to other contexts and programmes should be undertaken with caution. However, the findings build on existing evidence of EPR implementation and map onto the constructs of NPT -a recognised theory for understanding change -and may be considered transferable to other hospitals implementing EPRs. The two-staged approach to analysis and the quality checks on the coding frameworks that were developed during the analysis ensured rigour in the research process.
The small study sample may potentially have only included participants who were directly invested in successful implementation and those that had strong concerns about EPR. However, the purposive sampling frame and decision to undertake data collection across four case wards, which broadly represented key areas of hospital activity, ensured that we obtained diversity in clinical setting, professional group and grade.
NPT provided a useful set of conceptual tools to aid understanding of preparing for EPR use as a dynamic process. Using NPT enabled insights to be gathered on the 'work' that is involved in implementation (50).
In their recent review of the use of However, as acknowledged by May and colleagues, NPT places undue emphasis of individual and collective agency without explicitly locating this within, and as shaped by, the organisational and relational context in which implementation occurs (50). Additionally, we do not know the extent to which some mechanisms are more important than others in determining implementation process outcome.

Recommendations
A clearer understanding as to how staff groups and services differentially interact with EPR as they go about their daily work is required. This should include adaptation of the system to reflect this understanding. More opportunities for nurses and senior clinicians to engage in preparation are needed. The methods through which staff find ways to make EPR work for them in practice need to be better understood. Research that measures preparedness for change and factors that mitigate common and unique challenges to implementation should be prioritised. The heterogeneity of implementation programmes poses challenges for synthesising evidence for successful implementation. Detailed case studies are the cornerstone for understanding how technologies get embedded into healthcare (43) and the NPT framework offers the potential to compare implementation elements across different settings.

Conclusions
Hospital staff were motivated to invest in EPR implementation and perceived benefits to use, after an initial embedding period. Perceptions varied across different staff groups and specific care contexts with regards to how training and support impacted their preparedness for EPR use. This variation meant that some 31 staff would be responsive to the system, whereas other would be reactive. These differences were related to the perceptions that staff held around their opportunities to engage in preparation. The four core mechanisms of NPT provided a useful framework to explore individual and group expectations for change and outcome following procurement of an EPR. Given the difficulties often seen in implementation, and the political pressure to move forward with the implementation agenda, more research is needed not only on the effectiveness of EPR, but importantly, on what can be done to facilitate the implementation of EPR.

Key learning points
Health professionals' perceived potential value in using EPR and that benefits to use would be reaped after an initial embedding period.
Health professionals were motivated to invest in implementation.
There was variation across staff groups and services on the perceived impact of EPR use on ability to carry out their role.
Junior doctors and acute services could be responsive to the system. Outpatient services, nurses and senior clinicians would have to react to the system.
There was variation across staff roles and services in perceived opportunities to facilitate their engagement. Nurses and senior clinicians perceived that they were least prepared, and that opportunities for them to engage in preparation were limited.
There was consensus that staff would find ways to make EPR work for them in practice and that this would likely involve a move towards being a paper-light, rather than a paper-less system.
The four core mechanisms of NPT provided a useful framework to explore individual and group expectations for change and outcome following procurement of an EPR.

Consent for publication
All participants gave informed consent for anonymised quotes from their interviews to be included in any publications of the findings from the study.

Availability of data and material
The data-set which we have acquired will not be shared as a supplementary file. Our ethical approval does not permit the sharing of the entire dataset which we have acquired. Additional file 2 provides an overview of the coding framework from the analysis of interview transcripts.

Competing interests
The authors declare no competing interests.

Funding
The project was supported by funding form the National Institute of Health Research

Authors' contributions
CM drafted the manuscript and led on analysis of data. JB contributed to analysis and interpretation of findings and helped to draft the manuscript. OJ oversaw the development of the paper and reviewed multiple drafts. AS conceived of the study, undertook recruitment and data collection, contributed to data analysis and drafting the paper.