Contemporary Nursing UX Issues & Proposed Solutions A View from the Experts

Transcription

1 Contemporary Nursing UX Issues & Proposed Solutions A View from the Experts Executive Summary Background Nurses are not vocal at the nation level about their HIT user experience (UX) problems. The reasons nurses are so quiet is not clear as they comprise the largest volume of EHR users at over 3.4 million in the United States alone. Individual studies are available about nurses interactions with HIT, but no systematic or national view on the topic is yet available. Therefore, the purpose of this project was to identify UX "pain points" for nurses. Project Methods The project team consisted of five members of the HIMSS User Experience Committee and a national long-term care expert. The main phase of the project consisted of interviewing experts in the combined areas of nursing, HIT and UX. We used semi-structured questions with clarifiers to interview 27 experts: 9 nursing and nursing informatics leaders, 4 UX professionals/leaders, 7 nursing researchers, and 7 site leaders. Participants were asked to share their opinion or a summary of research you ve done about nurses critical issues or pain points with HIT. Follow-up questions asked participants to describe the significance of the issues, why they thought UX problems occurred, whether they thought issues were common in other sites and what solutions could rectify the situation. The interviews were conducted and recorded using WebEx. Two listeners independently documented the interviews via extensive notes. Analysis employed conventional content analysis to identify themes and categories.

2 Results. The table below shows the four themes from the analysis with related categories and examples. Themes Categories Sample Quotes UX Pain Points HIT Design We have reached a point where this [UX] is taking away from patient care. EHRs should be useful. They are such a hindrance now. Fit to workflow Nursing is the hub of information. Designs aren't flexible right now, e.g., if nurses wanted to flag information to radiology, that capability may not be available. HIT doesn't match the way nurses think. Importance of the Issues The Responsibility Gap Handoffs Interoperability Lack of information to support the process of care Significance Impacts Threats to patient safety No win for nurses Contractual issues and nondisclosure clauses Training & Education Resources My charting is not my nursing. We tend to get lost in many of the secondary needs, collecting quality measures for example. Handoffs are a pain point, from a nursing standpoint we need to be able to see the data and do something meaningful and actionable with it. Staff has to run around and go into multiple systems. Even when things are "easy" you need to constantly recheck. EHRs support nurses' data entry not retrieval, especially information of interest to nurses. IT staff rounded with nurses and didn't realize how their programming was impacting them. Some came back with tears in their eyes. For nurses the more UX issues they have, the more difficult it is for them to incorporate a system into their practice. Documentation is the saga of a million clicks. Patient safety is #1. Nurses are in a no-win. They have to use the system. Nondisclosure clauses about errors that occur across EHR systems or different facilities need to be removed. Various kinds of training are in use and training standards are highly variable. How do we take advantage of the skills and great things about nursing? Digging Out Solutions Think of EHRs not as a project but as care transformation. UX methods We need to pay more attention to cognitive factors of nursing work and use computers to help. Voice of nursing Every nurse has a role in informatics to move this forward, not just informatics nurses. We need a new EHRs need to be redesigned to focus on the whole vision ecosystem.

3 Primary UX pain points for nurses include the lack of fit between HIT and nurses workflow. Current HIT design does not fit the way nurses think and work. Specific pain points are extensive documentation requirements, the lack of cognitive support for handoffs and HIT workarounds. The work of nursing is non-linear; tasks, such as medication administration, are often clustered for multiple patients. In contrast, HIT is often designed to allow a view of one patient s information at a time. Nurses are knowledge workers acting as information hubs for the healthcare team. Yet, systems continue to be fragmented and not interoperable, meaning nurses forage for information and have difficulty creating the big picture of the patient. Identified UX problems occur across sites and vendors, indicating the importance of UX problems. Current HIT designs result in risks to patient safety. HIT-related inefficiencies can result in delays in care and decision making. Direct results to nurses are increased cognitive loads, time away from patients, disruptions and reductions in overall productivity. One of our participants summarizes the issue: It should not be easier to order a sweater from Lands End than to chart on my patients, but it is. --Teresa Brown (2015), The Shift, p. 117 The voice of nursing is missing in all aspects of HIT from design to purchasing, tailoring and implementation. Organizations may include one or two nurses on strategic efforts but small numbers are not representative of the population of nurses, especially for nurses at the point of care. One participant indicated that HIT systems and tools are seen as a no-win area for nurses who must use whatever simply appears often with inadequate training. Solutions or digging out require that we look beyond the current installed base and think of HIT as process of continual transformation. Partnering is required among vendors, organizations, UX professionals and nurses. We need to create a new vision of HIT as a patient-centered communication system that is part of a triad with the patient, the nurse and HIT. The new vision must consider the whole ecosystem where teams interact, what requirements are similar and those that are unique to the different professions. Solutions include more design standardization, for example for common documentation elements such as an admission assessment. Training and education of the future should reduce variability in quality and include workflow. Continual training is needed versus only prior to go-live. We need to take advantage of existing UX tools such as the HIMSS Usability Maturity Model, usability documents from NIST (the National Institute of Standards and Technology) and SAFER guidelines from the Office of the National Coordinator for Health IT. Conclusions & Next Steps. Every nurse has a role in informatics to move this forward." -Lindsay Steege Clearly, a focus is needed on improving UX for nurses HIT interactions because HIT is integral to nursing practice in any setting. A national home or repository is needed for UX problems and solutions. The timing is right for solutions because nurses comprise the largest group of HIT users nationally and globally. Better HIT designs can result in decreased risks to patient safety, decreased cognitive loads and improved workflow efficiencies. Partnering across vendors, UX professionals, organizations, care teams and nurses can make the difference in nurses work far into the future.

4 Contemporary Nursing UX Issues & Proposed Solutions Project Detail The user experience (UX) of healthcare information technology, especially for electronic health records (EHRs), is a national and global concern for healthcare providers. 1 For example, in 2015 in the U.S., The Joint Commission released an alert after a multi-year analysis found sentinel events related to health information technology (HIT) for user interface, workflow and communication UX issues. 2 Of the health professions in the U.S., physicians have been the most vocal regarding poor EHR usability that negatively impacted their productivity and reimbursements. 1 But the 3.4 million nurses, including advanced practice nurses, constitute the largest professional group interacting with HIT 3 (see Fig 1). Nurses experience different and critical user experience issues with HIT than other healthcare professions. Despite this, nurses are not yet voicing their opinions nationally about the UX issues they encounter daily and the impact on caring for patients. Figure 1. Proportion of RNs to Other Professions in the U.S. * ** Copyright Springer Reprinted with permission Scattered research studies are available on the topic of nurses and UX issues, but a national view does not exist. No systematic methods are yet available to collate and analyze UX issues for any one health professional group. Therefore, the purpose of this project was to describe nurses UX issues with HIT and to develop recommendations for improving the HIT user experience for nurses.

5 Methods This project was completed by members of the HIMSS HIT User Experience (UX) committee and a national long-term care expert with the endorsement of the HIMSS Nursing Informatics Community and the American Nurses Association (ANA). The first phase consisted of requests for case studies or stories about nurses UX issues with HIT, which were sent to informatics and ANA list serves as well as through social media. The second phase used a snowball sampling technique to select UX experts, nursing informatics leaders and nurse informaticists with expertise in UX across both acute and long term care. We obtained experts consent to participate and to audio-record interviews. We used semi-structured interview questions with probes asking them first to describe critical UX issues for nurses and HIT. The probes asked about the perceived significance of the issues, sources of issues and the prevalence of observed issues across sites. Last, we asked experts to identify solutions for the identified UX issues. The interviews were digitally recorded and careful notes created. After two team members verified the detailed notes, we used content analysis to identify categories and themes in the data. 4,5 Results For phase one, the case study requests yielded only 12 uneven submissions, despite the fact that the requests went to thousands of nurses. Therefore, this method was not effective and the resulting data were not representative. The second phase consisted of interviews with 27 experts: UX professionals (4), Nursing informatics/nursing UX leaders (9), UX nurse researchers (7), and Site leaders such as chief nursing/medical informatics officers (7). The experts represented a wide variety of settings across federal and non-federal sites, HIT vendors, academic and non-academic organizations including both acute and long term care. Twenty of the participants were registered nurses, and the modal educational level was a master s degree. The interviews averaged about minutes with a range of minutes.

6 Table 1 presents the results of the analysis, showing themes and representative comments. Four themes emerged from the data: UX Pain Points, The Importance of the Issues, The Responsibility Gap, and Digging Out. Table 1. Themes and Categories Derived from Experts Interviews Themes Categories Sample Quotes UX Pain Points HIT Design We have reached a point where this [UX] is taking away from patient care. EHRs should be useful. They are such a hindrance now. Fit to workflow Nursing is the hub of information. Designs aren't flexible right now, e.g., if nurses wanted to flag information to radiology, that capability may not be available. HIT doesn't match the way nurses think. Importance of the Issues The Responsibility Gap Handoffs Interoperability Lack of information to support the process of care Significance Impacts Threats to patient safety No win for nurses Contractual issues and nondisclosure clauses Training & Education Resources My charting is not my nursing. We tend to get lost in many of the secondary needs, collecting quality measures for example. Handoffs are a pain point, from a nursing standpoint we need to be able to see the data and do something meaningful and actionable with it. Staff has to run around and go into multiple systems. Even when things are "easy" you need to constantly recheck. EHRs support nurses' data entry not retrieval, especially information of interest to nurses. IT staff rounded with nurses and didn't realize how their programming was impacting them. Some came back with tears in their eyes. For nurses the more UX issues they have, the more difficult it is for them to incorporate a system into their practice. Documentation is the saga of a million clicks. Patient safety is #1. Nurses are in a no-win. They have to use the system. Nondisclosure clauses about errors that occur across EHR systems or different facilities need to be removed. Various kinds of training are in use and training standards are highly variable. How do we take advantage of the skills and great things about nursing? Digging Out Solutions Think of EHRs not as a project but as care transformation. UX methods We need to pay more attention to cognitive factors of nursing work and use computers to help. Voice of nursing Every nurse has a role in informatics to move this forward, not just informatics nurses. We need a new EHRs need to be redesigned to focus on the whole vision ecosystem.

7 UX Pain Points HIT Design Experts reported on the lack of designs to support the crucial work of the 3.4 million nurses in the U.S. 3 Issues range from large, conceptual ones like overall cognitive support to the more discrete like menu choices. The following material is phrased similarly to the way experts expressed their thoughts. Lack of Support for Nurses Work EHRs are not designed to support nurses' professional practice or the way they think and do work. Understanding nurses work means understanding that it is far more involved than merely accomplishing tasks and documenting them. Nurses are knowledge workers. Yet, little thought has been given to the cognition of nurses. Vendors do not yet understand how to represent nurses work. Nurses don't have a strong voice in system selection and design; systems "just come from somewhere." Nurses do not feel they have a voice in the selection, updates or design of a system. They are just expected to deal with whatever system and designs arrive in their settings. EHRs are built on the untenable assumption of my device, my data, my documentation, my work, my task. The assumption is: one user, one device, one patient, one task. Instead, nurses work often includes caring for groups of patients and typically involves teambased care. EHRs do not support these activities well. Nurses have to do information foraging, navigating across multiple modules in a single system and across multiple systems to find pertinent information to construct complex activities such as handoffs or to prepare for shift report. Shift report, in acute and long term care settings, frames nurses work for the whole shift. In these settings the report process is critical and unique. Current products do not support complex requirements like these for intra- or inter-disciplinary activities. Critical nursing activities are not always related to orderables. Current designs assume a central role for orders management, that orders drive all of nurses work. This is accurate only for inter-dependent nursing activities. In contrast, Electronic Medication Administration Records (emars) and Barcode Medication Administration (BCMA) do not support larger questions nurses think about such as, Is this the right medication for this patient given their current status? The patient s story is difficult to construct. Nurses need to understand what is happening with a patient at any particular point in time. It's difficult to see the big picture of the patient across the disparate forms of documentation and results. What is the pertinent history related to problems heard in shift report? What are the latest relevant problems? What do I need to watch for? What labs are critical? What medications is this patient on and how is the patient tolerating them? Currently, nurses must click across modules to find this information and then construct the patient s story in their heads or try to summarize it on paper. Complex patients are not represented well. For instance, how do we integrate patients with complex needs into increasing their own self-care? For patients who are on multiple medications, how do we help nurses educate patients about their medications, side effects and activities around safe medication delivery? How do nurses manage the information they need to care for patients with complex chronic illnesses? EHRs are generally designed for, and the focus is currently on, physicians. The workflow and responsibilities of nursing differ from those of medicine. The current emphasis has resulted in systems that are not a good fit for nursing and offer little cognitive support for them.

8 Excessive Documentation Documentation is a major pain point for nurses. Clinical documentation can take 25-33% of a nurse s time during a shift. One informaticist counted the number of clicks for an acute care nursing admission assessment and found it took 539 clicks and up to an hour to complete (Patty Sengstack). Admission assessments in other settings such as long term care and hospice can consume hours per assessment. Home hospice has even greater documentation demands. Experts indicate that documentation requirements are onerous. Each agency has its own requirements such as those from The Joint Commission, quality management departments in each hospital, etc. Nurses are documenting requirements for other departments versus concentrating on what they need to provide good nursing care. Thus, documentation has ballooned with questions about patients that are not directly related to nursing care. Typically, no work is done to consolidate on common points for nursing practice or to think about what should be translated into an EHR for nursing practice. Lack of interoperability exacerbated this issue. Sometimes within the same hospital, different modules of an EHR are not interoperable, meaning nurses must toggle between modules and systems to find complete information. Documentation modules are never pruned to remove unneeded documentation. Computers give us the ability to collect all the information we think we might want. So systems balloon with multiple notes about patients and have dropdown menus with many choices but seemingly never the right choice. More important, picklists do not convey how nurses think, provide care and chart. Documentation can include forced choices that don't match reality. Sometimes nurses document something that is not quite true just to get through the required selections even when these may not be appropriate for their patients. Lack of Clean, Intuitive HIT Designs Theresa Brown summarizes this issue: Click, scroll, type, enter. Here's the menu with twenty choices, none of them the one I need. Here's the point where I need information from two different screens, but there's no way to toggle between them. Here's the screen with thirty discrete options to check, but the window it opens up only shows me five at a time. New lab results, X-rays, CT scans, MRIs: none of those generate an alert and the screen is full of minute icons, some of which represent functions I don't use or even understand. 6 Systems in hospitals may not be updated for years, up to 10 years, many lifetimes in terms of technology. One system listed only 4 standard problems in a care plan, meaning care plans could not be individualized or tailored. Acuity of inpatients is increasing, making this sort of limitation inadequate and very frustrating to nurses. It's artificial to have drop down menus be the expression for a patient's condition. "It's so far wrong." For instance, how do dropdown menus capture how someone feels about dying? Implementations often include all the add-ons from a vendor s customer base. No work is done to clean this up ahead of implementation. For example, ordering an oral medication gives too many choices in the menu. Complex choices in menus are a UX pain point. Decisions to customize menus are problematic. One symptom is allowing 20 items on a pick list. One menu for the color of mucous had light green, medium green, dark green, 20 different choices.

9 Lack of Clean, Intuitive HIT Designs, cont d Non-intuitive icons. One example is a stat order. A physician verbally advised a nurse about a new, stat order. The nurse went to the computer and continually pressed the "refresh" button until the order appeared. She said that it was months until she realized a small icon indicated a new order. "It's so easy to miss these." New care models are not well represented. An example is diabetes care. The task of knowing when a meal is coming is important for a diabetic patient who gets insulin based on caloric intake. In the EHR, meal timing is not linked to either serum glucose or insulin dosages. The nurse s task is to be able to assess when the greatest risk is for hypoglycemia which includes the link between the timing of meals, serum glucose and insulin. The EHR should be able to help nurses with this but how? Fit to Workflow Experts thought workflow is one of the most critical UX pain points for nursing and HIT. Nurses are the ones interacting with EHRs on a constant basis; more emphasis is needed on what nurses need and the way nurses think. Information was not integrated during HIT design about the complexities of nurses mental models. Nurses work is not just a linear task flow, although most EHR systems are linear in how information is documented and used. Nurses working on inpatient nursing units need to cluster activities to be efficient and effective. Nurses work is often grouped around multiple patients rather than linear around a single patient, e.g., nurses can be administering medications to a number of patients all while assessing the patients and providing education. This kind of workflow and decisions surrounding it is not currently supported. Vendors must understand the different workflows and information required in different settings and different specialty areas and levels of nursing. HIT tools are limited in support of assessments, problems, and the human response to health issues. While screen design is important, the critical point is that systems overall don t support the cognitive workflow of nurses. Nurses have to jump between screens, and they can't see the right data in context to be able to identify patterns. Being able to document in real time is a big issue, e.g., organizing and administering medications requires considerable walking and can be a highly inefficient use of time, making documentation in real time difficult. HIT implementations without workflow assessments result in HIT on top of messy workflows and documentation, resulting in difficult-to-use systems. Workflows are hardwired into the system and it becomes difficult to do things differently when needed. Hybrid systems (paper and computer combined) mean there are multiple places to look for information and disparate, non-integrated applications. Institutions often use screens from the vendor as is to speed implementations instead of taking the time to understand workflow and tailoring screens to match workflow. The result is that users learn to ignore screens that are not pertinent. This is inefficient and contributes to the increase in nurses cognitive loads. Increased cognitive burden. Increased cognitive burden results when nurses cannot determine where to document in the EHR so others will take notice of pertinent information. No clearly identifiable location is available. Nurses spend time trying to figure out where to document so others can find it, e.g., responses to medications.

10 Handoffs A special case for HIT design relates to support handoffs and transitions in care. Handoffs are typically still a manual process and electronic designs are not meeting the mark as only some nurses use an electronic template; the common method for handoffs is for nurses to write down information on paper, e.g., tasks due, pertinent vital signs or labs. Electronic handoffs templates and methods do not match nurses' mental models. Handoffs and EHRs do not align. It makes it difficult to coordinate care. Nurses talk about their brains, a piece of paper that is a summary, created by them, of what care they need to be providing on that shift (and events that happen during a shift). This is a covert work system with a cross-patient view representing critical information, care planning and care priorities. One expert completed research on a nurse giving a handoff. A patient experienced three events, but the oncoming nurse missed two of the three events. The available HIT did not convey the information to support the nurse in thinking about these events. The important pain point is that nurses use critical thinking skills to synthesize information to develop the patient s story. There is also a lack of consistency in how to do things, often multiple ways in the system. That means information ends up in multiple places in a system, making it hard to construct the patient s story and communicate it to other members of the healthcare team. Interoperability Lack of interoperability is a severe usability issue for nurses. Many systems are still stand-alone. That means staff has to run around and go into multiple systems. Even when things are easy, you need to constantly recheck to be sure you don t miss something. EHRs promised efficiency and communication benefits, but systems don't talk to each other. In an example from long term care, one participant said, it s amazing that we're still faxing in healthcare (across institutions). Another major pain point is network connectivity. When technology is taken outside of institutional settings like hospitals, connectivity is not reliable. Then, information cannot be shared or integrated. Not all physicians use EHRs in rural sites, resulting in dual documentation and fragmented systems. Moreover, not all the pertinent patient information is contained in EHRs. Integration is needed with non-ehr HIT, such as smart pumps, point-of-care devices, and BCMA or other barcoded devices that might avoid the nurse having to be the data copier. Currently information is siloed in separate systems. Lack of Information to Support the Process of Care The EHR supports data entry not information retrieval. Moreover, the data nurses are able to retrieve often does not support the processes of care. Some nurses ignore the screens because the answers and questions of interest to them are not in the system. Coordinating patient care is a critical issue. The picture of a patient can change rapidly over an episode of care. HIT is fairly static and supports cross-sectional data better than longitudinal views. HIT supports a linear process, whereas patient conditions are dynamic, with varying rates of change.

11 Training & Education Healthcare organizations may not account for nurses time to learn a system. Nurses may not have extra staff support during system implementations, meaning they have the same number of patients to care for while they're learning a new system or adapting to changes during optimization of an existing EHR system. With multiple systems, various kinds of training are in use and training standards are highly variable. It's difficult to evaluate the effectiveness of training and how well people can subsequently use systems. Every interaction is an opportunity for patient education and/or patient assessment. HIT can alter the way we do work in that nurses become more task focused so that activities can be recorded as listed in the computer. This micro-focus is important, but participants indicated nurses should not lose sight of the bigger context of patient engagement. The Importance of the Issues UX difficulties create consequences across a number of areas including: risks to patient safety, increased cognitive burdens and inefficiencies. Most importantly, these issues occur across all sites and vendors according to participants. When HIT is designed well, one expert explained, nurses may be able to work to a higher level of practice. Right now, HIT is a stressful part of work. It takes an emotional toll and takes away from thinking about patients. Significance Experts agreed UX issues are pervasive across vendors and sites. A number of experts work for vendors or as consultants visiting multiple sites and seeing nurses interactions with HIT across the U.S. For nurses the more UX issues they have, the more difficult it is for them to incorporate a system into their practice. Usability comes up all the time at the CNIO forum one participant attends. Nurses really have to be able to use these systems efficiently. Without that they cannot do evidence-based practice or use data to make wise decisions. Products can be too flexible and sites can configure systems in ways that are not best practice. The prevalence is across EHRs and sites, especially in that workflow is not supported. These problems also extend across different patient populations, e.g., pediatrics. For instance, systems may not have comprehensive knowledge bases such as those for pediatrics or they may lack pertinent material, e.g., growth charts, pediatric medication dosing or lab references appropriate for children. Issues are fused between technology, implementation and adoption. There is often a disconnect between the goals of clinicians and administration for HIT. Clinicians foci are on care coordination, patient safety and comprehensive care while administrators may have purchased HIT for its purported return on investment or administrative efficiencies. Impacts include increased cognitive load, increased work-arounds and compromised patient safety. An example is BCMA and associated barcode scanning issues. In one site, nurses did not complain about scanning issues because they had developed a work-around to avoid problematic scans, although scanning issues were still occurring. That meant the BCMA goal of patient safety was lost and the actual goal became to get the meds out on time. EHRs create so many extra steps nurses may get lost and even forget the goal of what they're trying to do. Frequent logging in and out creates workarounds like leaving notes open all day so nurses can continue to update them versus signing them in real time (which they can forget to sign altogether).

12 Impacts One expert required IT staff unfamiliar with the impact their programming might have on nurses to accompany nurses on their clinical rounds. Some programmers returned with tears in their eyes. Nurses are treating HIT as correct even though it may not be. Nurses may overly trust that the HIT is correct even if it is not. Designs and products can be very different due to implementation methods. Implementations can take so long that content can be outdated by the time the system goes live. The HIT experience of nurses is extremely frustrating. The problems are fundamentally changing the way nurses practice. More and more documentation seems to return less and less value. Nurses are evaluated by their documentation now, e.g., whether they administer medications on time or enter a note every four hours. Documentation is "hanging over us." Maybe because it's easier to measure nursing contribution and productivity by measuring documentation. Working with multiple applications and differing interfaces slows the nurses down. Staff is at risk of making mistakes. We cannot begin to understand the health of our communities if we don't have good connectivity (especially in long-term care). Technological solutions are available, but financial resources, human resources, or intellectual capital stand as barriers to implementing technology with good connectivity. EHRs can be used as a way to shame nurses and make them look bad in the work they do. In a recent retreat about nursing burnout, nurses wanted to talk about how EHRs contributed to burnout, but the organizers recommended shifting to different topics, because you can't do anything about EHRs. Documentation is pulling clinicians away from the bedside, preventing them from spending adequate time with patients, a common stressor and component of burnout. If technology works well then stress and fatigue should be reduced. Connecting disparate systems from one site's database and merging it with other databases is a major pain point because there is no common interface. Accessing data from where it is located rather than moving it from point to point is a much preferred method for using data. Merging these types of datasets is important for data mining and population health. There is such a reliance on technology. Younger nurses and physicians don't know what to do when systems go down. In one example, a physician did not know how to write a paper order for pain meds; nurses did not know what to do without the system. Pain meds were delayed by 1.5 hours so a patient was in pain unnecessarily. Increased Cognitive Burdens We need help with better patient care decision making. Tools that were intended to reduce cognitive load have instead increased it. There is always something to help us remember how to use the tool that was supposed to help us remember. When there is little consideration for nurses cognitive load, HIT designs can produce multiple types of unpredictable errors.

13 Inefficiencies The data entry burden is ridiculous. It is the saga of a million clicks. This comes in part from excessive documentation to reduce legal and regulatory risk. Excessive documentation hinders nurses ability to streamline and improve. Nursing workflows around documentation and information seeking is inefficient. Required documentation only increases. There is no consolidation. No re-envisioning what documentation should be. The direction is toward more and more complexity. There are always new things to document but nothing drops off. This adds to the burden of documentation with no sense of consolidation. Leaders say, "It's just one [more] thing we are asking you to do " There is also a reduction in productivity; nurses who are busy with documentation can't spend as much time with patients. Time is the main issue. The nurse is at the computer rather than with patient or with other members of the team doing patient care. People are still using paper in a significant way, e.g., in the ED when the functions are not available in the EHR. High risk activities need interventions, e.g., IV insulin calculator embedded in the Electronic Medication Administration Record. The old process was labor intensive, prone to confusion and error. New process is accurate, built in, decreasing nursing time to calculate and increasing dosing accuracy. Nursing pain points include the amount of time nurses spend on the computer because of documentation requirements. For example, a second example of the documentation burden was an inpatient assessment had 500 lines and took over an hour (for the nurse) to complete. Anything that takes the nurse away from facetime with the patient will be a pain point. Nurses need a more streamlined system that is integrated and does the task in fewer steps Reduce dual and triple documentation requirements. This is an issue especially in rural settings. These settings have constrained resources and these issues in combination with Meaningful Use requirements may cause small hospitals to go under. Nurses are gifted at finding workarounds. They will also find the fastest path to where they want to go even if that is not the best path. Given the waste in the system (screens, clicks), nurses will do what they can to find a quicker way to accomplish what is purposeful and important. Free texting is also used instead of filling out other fields. It s quicker but can compromise information needed for data analyses, for quality, etc. This makes it difficult to show what nursing does.

14 Threats to Patient Safety HIT creates patient safety issues, work-arounds, sentinel events. They occur across vendors, and healthcare delivery sites. One expert found patient mortality increased during two large implementations efforts (CPOE and clinical documentation). A possible reason: Big implementations caused significant disruptions for nurses and their processes, both workflow and cognitive workflow/planning. EHRs were designed for adults but healthcare organizations are using them for pediatrics. Weight based dosing is a problem in EHRs designed for adults but being using for children, e.g., rounding errors were a problem as kids need precise dosages. Armbands that require scanning on multiple children (Baby A, Baby B, Baby C, etc.) can be a problem and could introduce error if the system does not support more than two children per mother, for example. We need use cases and scenarios for pediatric patients to illustrate error and design of EHRs. One case of a safety risk was a set of triplets where armbands could not be used because the EHR allowed only a single baby assigned to each mother, making it hard to uniquely identify each baby. This led to huge errors. Workarounds include things like not actually signing off on notes until the end of shift, so nurses can keep editing. Then notes do not get signed. No one uses nursing notes. Interestingly, when nurses want to find out information, they review MD notes, not nurses notes. Are nurses notes even important? The EHR supports data entry, not retrieval. What data nurses do retrieve does not support the process of care. Nurses rely then on verbal handoffs, which are also fraught with errors. Both are threatening patient safety. Interdepartmental handoffs are an issue because an EHR may not necessarily allow a nurse access to the patient's record until the patient is physically on the unit. Thus, nurses cannot prepare for patients before they actually arrive, creating a rush and potential omissions when they do arrive. Critical information is often written on nurses' paper handoff forms (which are thrown away and not part of the patient's record), this results in an increased risk to patient safety through increased risk of human error. Work-arounds are plentiful. For example, using patient ID bands in a list on a clipboard versus from the patient's actual wristband. Long notes in documentation create patient safety issues, shortcuts and other work-arounds such as copy/paste inaccuracies. The Responsibility Gap Nurses are in a no-win situation. They have to use the system that appears, but they cannot make any changes whatsoever. Their voice is not being considered either in system purchase, design or optimization. Nurses just have to absorb the new work of new EHR systems and their lack of integration. New staff are often not hired to help even during systems implementations or upgrades. Pain points are tolerated by the nurses because they see the benefits of these tools, such as finding important information. Unintended consequences have also been identified, those being new demands for communication, multiple devices, etc. More studies are available about physicians and HIT such as merging practice guidelines; however, literature is not available reflecting similar efforts for nursing.

15 Contractual Issues The nondisclosure clauses, common in vendor EHR contracts need to be removed. This would provide transparency for the public, would likely improve care and reduce errors. By not sharing information about EHR errors and problems between organizations the same incidents keep happening. The opportunities to reduce risk to patients and improve care more broadly get missed. Resources Bedside nurses do not understand usability or user experience. They just know the computer is hard to use. Bedside nurses do not understand or use CMIOs, Nursing Informaticists and what they might offer Lack of knowledge about available resources like nurse informaticists. Digging Out Solutions Think of EHRs not as a project but as care transformation. We need to go past initial adoption/implementation of systems and really work on processes after implementation. Change can be successful using an incremental change model like those used in other areas, small continual changes towards improvement rather than large, episodic change. The way forward is partnering between organizations and vendors. It will take a partnership between vendors, healthcare organizations, plus any third party organizations such as content managers. All stakeholders need to make a joint effort (and share responsibility) to improve UX for HIT. Implement a framework to look at the implementation process and how professional practice is done. Honor how each profession is practiced and how they work in patient care. Assign ownership of UX (usability) to someone (a senior leader) in the organization. Define clear lines of ownership for EHR UX. The system UX should reflect both clinical and IT components. Make them own it. Right now it is confusing about whether UX belongs to IT or clinical? Really needs a clinical champion in addition to IT. A second wave is occurring to move toward more HIT standardization and organizations are learning how this can benefit them. Solutions should represent nursing workflow. Many EHR systems are designed around other workflows; often nursing functions are added later under the assumption that nursing workflows resemble other workflows. Understanding nursing and what information they need and when it is needed as well as their nursing cognitive processes is critical.

16 Solutions, Cont d Standardize! o Employ more standardized designs, picklists. o Embrace the opportunity to design modules, documentation for cross-site use. o Include workflow in the design of HIT training. Clinicians must work to adopt tools into real use in their practice. Pay more attention to the content versus the tool. The web of connections means the team and their HIT along with patients and their devices. Focus on designs that will make a difference. Have everyone get together to focus on key points to make an effective handoff form, for example. Fix interoperability issues. It's unreasonable for legacy systems to be redesigned all at once, but sections or modules could be designed using this new vision. Google uses this approach: define the goal and then design each new piece to be congruent with that goal. We should be reducing documentation. The dyad of CMIO and CNIO is helpful in large organizations (harder to effect in smaller organizations.) Need better trained Nursing Informaticists. They need better analytic skills, including workflow analysis knowledge and skills. UX Methods Need to pay more attention to cognitive factors of nursing work and how to use computers to help. Must understand the workflow better before computerization. We still don't fully understand what it is to be a patient in a setting. The setting needs to support the individual and their technology too. Look at 6 key components of user experience: Infrastructure (downtime, etc.). Reliability has not been there; system response time is not predictable either due to design and hardware issues. Optimization. Very hard to imagine an EHR designed perfectly out of the box; all go through a process of maturing with the organization and nursing workflow. Integration. Integrate with other non-ehr HIT like smart pumps, barcoding. Try to avoid nurses being the data copier between systems. Policies. The EHR becomes the amplifier for a lack of policy or standardization, for example what to document. Examine and clarify policies around the value of the documentation. Training. Often training is not fully considered. Leaders think that if a system is designed well, a whole lot of training isn t needed, while the need is actually huge because of the large amount of skills that have to be learned. Support. Organizations often do not achieve good support due to quality of hardware. Need a conceptual framework for thinking about usability. We need the 3 U's of thinking about a technology: Usability making things easy to use, Usefulness, Understanding the work. Have analysts observe and redesign processes, e.g., use a requirements analyst who focuses on what requirements that are needed to solve the problem, versus focusing on what the technology can do. Then, the requirements analysts look across the enterprise (the military) and elsewhere to see what's available that might as a solution. Understand workflow before purchase, how patients flow through your system. Creating design guidelines for EHR vendors that are driven by clinician user needs. Create designs for areas like the Emergency Department that function differently than the inpatient and outpatient areas. Need to do usability evaluations throughout the lifecycle of systems, give feedback to vendors, institutions to help. Having a consistent look and feel is important. Order and colors, etc. should be consistent.

17 UX Methods, Cont d Resources are available from NIST National Institute Standards and Technology. Resource: Technical Report 7865 Human Factors guide to enhance EHR usability and Technical report 7804 Technical evaluation and testing of the EHR. Available on the NIST website at: SAFER guides. Government should mandate the use of these guides. Have someone in the organization own this. Companies who will succeed will observe work versus asking users what they do. During one observation, the ICU nurses were writing patient's weight on a sticky note because the wrong weight was in the EHR. Having people follow and observe nurses so they can see how what they are developing and doing is utilized in real life. Have managers shadow nurses. That was a tipping point for one expert. Then, managers see how much nurses are asked to do. This is better than having IT folks shadow nurses. Managers can pressure vendors for change. Meet regularly with end-users about usability, educate end-users about the topic so they have the tools to communicate and be understood with more clarity. The Voice of Nursing Every nurse has a role in informatics to move this forward, not just informatics nurses (Linda Steege). Nurses also need to understand why they are interacting with EHRs. Vendors need to understand what nurses need to use these tools at different levels. Nurses are the majority of the workforce so they must be included for success. The biggest pain point remains: Nurses do not actively participate in the development and customization of HIT or in making decisions in their organizations. Their voice is just not heard enough. Nursing needs a digital strategy (Linda Harrington). Leaders need more knowledge about informatics or we risk losing our voice or our future jobs. How do we empower the CNO and nurse informaticists to play an important role? One example is ICD-10. A leader indicated that ICD-10 doesn't impact nursing; however, 126 codes are available for pressure ulcer documentation. Nurses document on pressure ulcers so EHR documentation will be used to record ICD codes. Nurses need to know they have a voice. They need to be empowered, they need to know that systems can changed and they need to submit system change requests. Nurses must be included at all levels as end-users, middle managers and executives. Nurses are natural leaders of interdisciplinary teams that will ultimately clean up these implementations. There is a fundamental triad in healthcare, the nurse, the person getting care and the technology. Nursing IT needs should be better identified in system design with clinicians being at the table with vendors during design phases. A need is to understand what "brains" are, what they do, why nurses will come in 30 min before their shift to create their own brains? Developers need to understand this. Educate bedside nurses - roles/people/resources that are available to them with respect to HIT/EHRs. Organizations should allow and pay for nurses to attend a professional meeting on informatics so they can see what they don't know and what they need to learn. Nurses would speak up, in their role as patient advocate, if they were educated and if there were a process identified for them to follow. We have not been successful enough in integrating HIT understanding into nursing curricula. We need to educate the educators so that curricula can integrate HIT topics. This will enable nurses to speak at the table when these issues are discussed.

18 We Need a New Vision EHRs need to be redesigned to focus on the whole healthcare ecosystem including how providers interact together and separately. Redesign to focus on the patient versus the department, be more patient centered. Redesign interfaces and EHR organization. EHRs should all be organized around the knowledge needed to work with the patient. We need to re-conceptualize the EHR as a communication tool rather than a documentation tool. Users can then realize they are communicating information and craft better messages for communication. Big lesson for low acuity systems (long-term care) is to create a sense of team between the on-site and remote providers. EHR systems need to be centered on the scope of practice for all disciplines for inter-professional team-based care. Have an innovative lab experiences (e.g., simulations) with nurses to create the future. Could leverage the data in the EHR more fully that way. The solutions are three parts: software, information content and then post-implementation processes. Need a focus on simplicity, efficiency and interoperability. Embed more task automation into workflow to decrease costs and improve outcomes. Link disparate types of data and apply analytics to predict barriers to patient outcomes, i.e., predict/prevent delirium due to over medicating. Change the types and methods of training. Be sure nurses are aware their thought processes will be disrupted and change after implementations. We just teach steps now. Need continual training, not just at the beginning, but also when changes are made. Teach that disruptions during implementation will impact the way nurses think and organize their work. Vendors also need to provide a way for organizations to do training, currently training is highly scripted with little or time for exploration. Nurses learn one method, no matter how inefficient it may be and stick with it. Hardware support should become more automated. An unused workstation, for example, should trigger an alert to systems support. Data visualization is becoming an important area of using data. We need more concise and meaningful ways of displaying desired information. Conclusions & Next Steps The analysis of experts interviews reveals that nurses face significant HIT usability issues that seriously impact nurses ability to care for patients, many of which differ from other healthcare providers. Current HIT design does not support nursing workflow and nurses cognitive processes. User experience problems are seen across vendors and sites. The impacts of these issues include increased risk to patient safety, outcomes and quality care as well as stark inefficiencies. The voice of nursing was missing in all phases of the HIT lifecycle. Currently, nurses at the point-of-care just have to live with whatever technology shows up in their environments. Solutions include those that look beyond the design and configuration of the installed base. Experts indicate a new vision is needed for EHRs, one that is more patient-centered and inter-professional and one that better supports team communication. Going forward, developing partnerships between healthcare organizations, vendors, healthcare professionals and UX professionals is the means for reaching the suggested solutions. The voice of nursing must be included in all phases of the systems lifecycle. Nurses UX needs a home in healthcare organizations as well as at the national level among organizations that represent nursing.