CONFERENCE REPORT: 2009 SUMMER INSTITUTE IN NURSING INFORMATICS
The 19th Annual Summer Institute in Nursing Informatics was held in July 22-25, 2009, in Baltimore, MD. With the conference entitled "Informatics at the Point of Care: A Barrier or a Bridge," the speakers made it clear that it is possible to be the bridge, but fulfilling this goal will require a rethinking of our systems.
Providing ideas for creating this bridge, the first keynote speaker, Dr James Cimino, chief, Laboratory for Informatics Development at the National Institute for Health, in his talk, "Informatics as the Bridge to Knowledge at the Bedside," encouraged us to look at three things: the bedside environment, knowledge needed to support care, and the technology needed to bridge these pieces. At the bedside, we have an interaction among patients, who have the most knowledge about themselves, clinicians who bring their observational skills and knowledge, and technology such as bar coding, the PDA, Internet, and the electronic record. Informatics should be the bridge among these three by integrating all these entities and providing needed information in a timely manner.
Print sources rarely met information needs, and informatics is not yet supplying them. Covell et al1 studied the information needs of practitioners and found that only 30% of physician's information needs were met during the patient visit, and those that were met usually resulted from asking a colleague. Although this study is dated, subsequent studies by Feied et al,2 Corcoran-Perry and Graves,3 and Spath and Buttlar4 have found the same results. There are more unresolved information needs than clinicians realize. Even when computer-based resources are available, they are too often not used because users do not see the need, do not know the appropriate place to find the answer, do not know how to use the resource, and perceive that they do not have time. The result, of course, is that clinicians "wing it."
To fulfill these information needs, clinicians must know and use a standardized terminology to search for needed information. It also requires intelligent data/information retrieval, a good selection of resources, and integration into an easily accessible format. Using a standardized terminology, predesigned queries that direct an accurate, quick retrieval of needed information in real time can be created. System designers, instead of trying to keep information on their system updated, can instead link to current articles in PubMed and allow clinicians to make their decisions based on the literature as well as their knowledge. With the current mandate that all NIH-funded research accepted for publication after April 7, 2008, be openly available 12 months after publication (see http://grants.nih.gov/grants/guide/notice-files/NOT-OD-08-033.html), the number of freely and quickly available full-text articles will increase as each month goes by.
One way to integrate the needed information into work flow is with the use of what Dr Cimino calls Infobuttons. (Great information about this method of putting information at the user's fingertips, including a Power Point presentation by Dr Cimino, can be found by searching the Web for infobuttons.) In conclusion, Dr Cimino pointed out that simply putting the patient, nurse, and technology in the same room is an insufficient use of informatics. To provide needed information, it is necessary to understand the information needs, provide the resources when needed, and integrate them into the work flow.
"Pathways to Translational Informatics for Nursing" was presented by Dr Marceline Harris, senior associate consultant in the Departments of Nursing and Health Sciences Research at Mayo Clinic. Translational informatics, which is an evolving concept, uses informatics to translate research into clinical practice. Translational paths create iterations between basic bioscience, research, and clinical care.
In understanding this concept, it is necessary to differentiate between informatics and information technology. Informatics is a distinct focus of study. It has a unique world view, defined principles and practices, research and theory development, education, and professionalism. Informaticists are "vertical people." That is, they focus on modeling data, information, and knowledge across layers progressing from hardware, through networks, algorithms and software, data structures, tools, human factors, and individuals with specific knowledge to groups with collective knowledge. The information technology person is the person who creates applications in the real world that use these models.
Translational informatics applied to systems can assist nursing in the areas of usability, data retrieval and analysis, and terminology. That this is not yet happening is evident in several studies that report that nurses are not happy with current systems. Common themes are that the systems do not support their work, introduce complexity, and tend to fragment, not coordinate, care. Additionally, nursing is still not integrated into the thinking about electronic records. For example, a model of a fully integrated medical electronic record has no mention of nursing documentation.
Unfortunately, there is limited evidence to guide the use of technology by nurses. There is a need to understand relationships among nurses' use of technologies (including electronic health records [EHRs]), the moderating and mediating factors that affect use such as social contexts, disciplinary fears, and the potential nurse, patient, and organizational outcomes. New systems are presented with great fanfare, but they too often follow Gartner's hype cycle,5 in which there is a peak of inflated expectations, followed by a trough of disillusion, then a gradual slope of enlightenment as people learn how to use it, and eventually, a "plateau" of productivity, which may or may not be the point where care is improved. For example, expectations in terms of data retrieval and analysis issues have not been met despite billions of dollars spent in data capture and storage. We still have to ask, "Where is the information and new knowledge from these data?"
Dr Paul Tang, internist and vice president, chief medical information officer of the Palo Alto Medical Foundation, and a consulting associate professor at the Stanford School of Medicine, introduced us to "Personalizing Health Care: Partnering With Patients." He opened by challenging us to question the premise of a patient-centered "medical" home by asking where the focus is in this name. He then pointed out that the criteria for certification of a patient-centered medical home assigns only six points out of 100 to Patient Self-Management Support, two of which are for assessing language preference and other communication barriers, while only four points are given to actively supporting patient self-management. The other 94 points all deal with medicine.
His approach is a patient-centered "health" home (note the absence of the word care) that, instead of focusing on medicine, which he calls a "last century" approach, uses informatics to facilitate, or work with, not prescribe, patient health. He referred to the 2001 Institute of Medicine report Crossing the Quality Chasm, which stated that the status quo in healthcare is unacceptable and that we cannot incrementally change what is a failed system. Drawing a parallel with the artificial horizon on an airplane dashboard, which informs a pilot and copilot whether they are flying level or in need of correcting their course, he showed a slide with the copilot (patient) side blacked out. He pointed out that we present information of concern to us but do not provide the patient with the knowledge of what to do with this information, creating a blind copilot.
Today, patients should have unfettered access to their own information. He pointed out that patients seek information about any diagnosis and treatment plan from friends, relatives, librarians, the Internet, and pharmacists. He reported that 62% trust their physicians, but only 11% go to their physician first for information because it is a "pain" to reach this individual. In an effort to improve this situation, he and his colleagues randomly selected patients to participate in focus groups at the Palo Alto Foundation. In the themes that emerged from these groups, they found that patients want custom-tailored information based on their own data and information about what they can do to improve the situation.6 They also found that questions should be answered when they arise, which, too often, is not in the examination room but at home, with friends and family.
To meet these needs, they revamped their system at the Palo Alto Foundation Clinic. Patients have e-mail access to care providers and can ask questions about their conditions and treatments. They can also request a visit, which can usually be accommodated that same day. At home, after a visit, they can review and check their records on the Web and print all the details of the visit. Any treatment plan is linked to information to assist the patient in implementing the plan, such as detailed instructions, a video, or both. All test results carry information about what they mean. Charts are used to illustrate data such as glucose levels, which permits patients with diabetes to view their blood glucose levels over time. This enables them to equate their activities with the readings, for example, a lower reading after they exercise or higher if they indulge. Using approaches like these to improve health can create a change in behaviors because patients perceive that they are accountable for their health. Dr Tang summarized by stating that we need to put patients on an active basis on the health team and personalize healthcare. He closed by challenging us to move not only from a medical-focused record but also from a healthcare model to a health model.
The closing day featured not only two talks but also a panel. The first talk, "Optimizing the Usability of Clinical Systems: Past Work and Future Directions," was presented by Dr Nancy Staggers, professor and director for nursing informatics at the University of Utah College of Nursing. Using the ISO 2006 definition7 as a base, she defined usability as "[horizontal ellipsis]the extent to which a product can be used by specific users in a specific context to achieve specific goals with effectiveness, efficiency and satisfaction." In informatics, this concept is related to human factors, ergonomics, and human-computer interaction. The goals are to increase effectiveness in terms of decision making, decrease errors, fit with work flow, and improve efficiency related to speed and productivity.
She pointed out that although the press has been mainly positive about the impact of computerized provider order entry (CPOE), too often, it is not integrated into the full system or work flow. Currently, a Health Information Management Systems Society 2008 survey found that only 0.1% of agencies have a full EHR, 1.9% have CPOE with a clinical decision support system, and 32.9% have electronic clinical documentation, leaving much work to be done to realize the full benefits of CPOE. Current EHRs lack cognitive support for clinicians, do not integrate well into clinical work flow, and do not take advantage of human-computer interaction principles. This leads to poor designs and potential errors. The end result is that health information technology (HIT) could actually worsen healthcare. To avoid these pitfalls, it is necessary that informatics specialists understand and apply usability concepts.
In an extensive review of past research results from 1980 to 2009 of the usability of user interfaces for clinical technology,8 a potential for minor and catastrophic errors with medical device design was found. Some of the causes of these are confusing labels, poor information visibility, faulty data synchronization, and poor navigation systems. Additionally, poor designs take twice or thrice as long to use. When systems are inflexible, the navigation is poor, or there is limited visibility of system status, satisfaction is low. Users want interfaces with visible formats (MM/DD/YYYY), consolidated information (high-level information first), fewer levels, ability to customize to their work, and integrated applications. They also found that experienced users want dense screens and that graphical designs improve efficiency. It is imperative that potential buyers evaluate devices for navigation tasks and, to avoid potentially dangerous situations, evaluate what triggers alarms.
Future studies should expand the types of devices studied and EHR interfaces, as well as move testing to actual clinical settings instead of laboratories. Studies must also look at how work is done in interdisciplinary teams. Systems today are too often rich in data but poor in information.
Change of shift report is another area in need of more exploration. In a qualitative study9 that explored the context and content of change of shift report, four themes were found: dance of report (matching the report to the recipient, eg, changing the depth of information based on the familiarity of the oncoming nurse with the patient, or level of experience), passing of facts, information related to professional nursing practice, and lightening the load or team building.
The panel for the topic "Creating Usable Systems for Nurses: Recommendations From the TIGER (Technology Informatics Guiding Education Reform) Usability Collaborative" (see http://tigerusability.pbworks.com/Final-Summary-Report for an outline of this report) featured Gregory Alexander, assistant professor at the Sinclair School of Nursing; Cheryl Parker, senior clinical informatics specialist for Motion Computing; Nancy Staggers; and Denise Tyler, clinical specialist at Kaiweah Delta Health Care District. The first presenter, Dr Alexander, reported that this TIGER work group performed a comprehensive literature review about usability, collected case studies, and developed recommendations for HIT vendors and practitioners. It had the highest number of volunteers, indicating its importance to practicing nurses.
Ms Tyler presented a case study of both a successful and a painful system implementation. Two key factors that determine which outcome a system has are the involvement of end users and integration with existing systems. Successful systems have a connected and engaged leadership, a multidisciplinary team that evaluates systems, all end users involved in all processes, a partnership with the vendor, and, most important, from system selection, to design, through testing and education. Other factors in the case study with successful implementation that she reported were customized education sessions that were not system driven and could thus qualify for continuing education credits, a rapid response to users, and an ongoing investment in the system. In the case study with a not so successful implementation, there were issues discovered during go-live that resulted from a lack of work flow analysis, such as missing key pieces of documentation. Additionally ordered equipment had not been installed until during go-live, neonatal ICU vent interfaces did not work despite a checklist saying they did, and there was no support for go-live. There was no end-user involvement, and the only clear vision was to implement on time. Despite all these problems, the chief information officer and vendor reported a successful project.
Dr Parker looked at recommendations for providers and vendors. Key recommendations for providers include an interdisciplinary team that designs and discusses all new functionality, engages users early and often in selection and design, and tests using real-life scenarios. Additionally, there needs to be enough money budgeted for training, as well as a plan for go-live support. Reports need to provide the ability to locate, manipulate, and aggregate data quickly and downtime procedures for all possible scenarios, from the mundane to the truly catastrophic. This report also advises using free text very judiciously because of the difficulty and time involvement in coding and analyzing free text.
For vendors, they recommend that consideration be given to all levels of expertise when designing systems and to allow individual users to select the level of support they want. Early in the development of a product, vendors must do usability testing, including user observation of several team members at all levels and disciplines. There needs to be an understanding not only of the work flow but also the reasoning behind it. Products should be designed with the end user in mind, making it easy to do the right thing and hard to do the wrong thing, but also with error traps for occasions when users still manage to do the wrong thing.
To provide the full benefit of EHR systems, it is necessary to increase nursing input into their design and implementation. There must be collaborative efforts to define the crucial aspects for nurses, as well as to define and apply key concepts of usability in nursing-intensive environments. Paying attention to usability will result in increased user acceptance and system adoption with improved patient safety and more timeliness of information collection, reporting, and use of the information. Good usability is no longer a choice but a mandate to support safe, effective decision making.
The concluding talk by Suzanne Bakken, alumni professor of nursing and professor of medical informatics at Columbia University, "Reducing Health Disparities Through Informatics," started with Dr Bakken stating that health disparities abound in many areas including preventative medicine, disease prevalence, and infant mortality. They result from racial, ethnic, sex, education, income, location, and sexual orientation characteristics. She then presented the question of whether informatics is part of the solution or whether it will contribute to increased health disparities. The digital divide, Dr Bakken reported, can be seen not only in terms of access but also in the ability to understand information (computer, functional, numeracy, and health literacy). Although racial and ethnic differences have decreased, those between rural and metropolitan areas, as well as age, are still present.
In 2001, Columbia University formed the Center for Evidence-Based Practice in the Underserved. In 2004, they expanded to include community and public health systems as well as economic analysis. One of the current efforts is the improvement of self-management. One attempt at this is a randomized controlled trial of clinical encounters of advanced practice nurse (APN) students with the hypothesis that having decision support, as compared with having no decision support, will result in greater adherence to guidelines by practitioners. Decision support for screening and guideline-based management of depression, obesity, and smoking cessation was provided using a type of infobutton with PDAs and cellular phones. The decision support was knowledge based, centrally controlled and updated, standards based, and integrated into the APN work flow. Additional analysis of the data is continuing, but to date, they have found that there was increased screening, but in some cases, there was no increase in plans of care.
Another intervention from the Center for Evidence-Based Practice in the Underserved is the Tailored Intervention for Management of Depressive Symptoms program. This was designed to provide tailored, computer-based education on self-care for depression symptoms for patients with HIV/AIDS in Harlem.10 They have found that acceptance of the program is associated with perceived ease of use and usefulness. Another program that the center has started is with Latin American adolescents. They observed that 4 years after arriving in America, many of these adolescents became obese. They are using Facebook to promote physical activity among Latino immigrants. Techniques being investigated include a user profile with goal setting, nonmonetary rewards for reaching goals, peer modeling, and social support. In closing, Dr Bakken stated that careful attention must be paid to usability issues for those with low levels of computer, functional, and health literacy.
The many competing concurrent sessions again made institute participants make difficult choices because all were inviting (abstracts of prizewinning presentations and posters can be found in the "ANI Connections" section of the September/October 2009 issue). Information about some of these, as well as the keynote talks, can be found in Peter Murray's blog (with help of Scott Erdley) at http://ucru.wordpress.com/. The poster session Friday afternoon gave conference participants ample time to talk with the owners and network for help on situations.
The conclusion from the conference is that informatics can indeed be a bridge at the point of care. To achieve this, however, we must be willing to develop and use clinical decision support, assess and meet patient needs where and when they occur, and focus more on usability. Nursing, with our health and person focus, is ideally placed to support and lead in implementing these approaches.
Contributed by Linda Q. Thede, PhD, RN
CONTRIBUTORS TO THIS ISSUE[horizontal ellipsis]
Linda Q. Thede, PhD, RN, is the Editor of CIN Plus.
William Perry, MA, RN, is Adjunct Instructor, Wright State University, Dayton, OH.
REFERENCES