Technology Innovations
ABOUT THE AUTHORS
Dr. Rantz is Professor, Dr. Alexander is Assistant Professor, Dr. Aud is As-sociate Professor, Dr. Wakefield is Research AsAs-sociate Professor, and Mr. Miller is Research Associate, Sinclair School of Nursing; Dr. Rantz is also Professor and University Hospital Professor of Nursing, and Dr. Koopman is Assistant Professor, Curtis W. and Ann H. Long Department of Family and Community Medicine, Dr. Popescu is Assistant Professor, Health Management and Informatics Depart-ment, School of Medicine; and Dr. Skubic is Professor, Electrical and Computer Engineering, College of Engineering, University of Missouri, Columbia, Missouri.
The authors disclose that they have no significant financial interests in any product or class of products discussed directly or indirectly in this activity. This re-search was supported by the U.S. Administration on Aging grant 90AM3013 (PI: Rantz) and the National Science Foundation Information Technology Research grant IIS-0428420 (PI: Skubic).
Address correspondence to Marilyn J. Rantz, PhD, RN, FAAN, Professor, S406, Sinclair School of Nursing, University of Missouri, Columbia, MO 65279; e-mail: [email protected].
Posted: December 22, 2009 doi:10.3928/00989134-20091204-02
Developing a Comprehensive Electronic Health
Record to Enhance Nursing Care Coordination, Use
of Technology, and Research
Marilyn J. Rantz, PhD, RN, FAAN; Marjorie Skubic, PhD; Greg Alexander, PhD, RN; Mihail Popescu, PhD; Myra A. Aud, PhD, RN; Bonnie J. Wakefield, PhD, RN; Richelle J. Koopman, MD, MS; and Steven J. Miller, MA
AbsTrAcT
As in acute care, use of health information technology in long-term care holds promise for increased efficiency, better accuracy, reduced costs, and improved outcomes. A comprehensive electronic health record (EHR), which encompasses all health care measures that clinicians want to use— both standard health care assessments and those acquired through emerg-ing technology—is the key to improved, efficient clinical decision makemerg-ing. New technologies using sensors to passively monitor older adults at home are being developed and are commercially available. However, integrat-ing the clinical information systems with passive monitorintegrat-ing data so that clinical decision making is enhanced and patient records are complete is challenging. Researchers at the University of Missouri (MU) are developing a comprehensive EHR to: (a) enhance nursing care coordination at TigerPlace, independent senior housing that helps residents age in place; (b) integrate clinical data and data from new technology; and (c) advance technology and clinical research.
B
y 2030, there will be approxi-mately 72.1 million older adults in the United States, almost double the number in 2007 (U.S. Admin-istration on Aging, 2008). This increase, coupled with the pro-jected shortage of nurses, primary care physicians, and other geriatric health care workers, demands that health care settings find new ways to meet increasing demands. Use of health information technology holds promise for increased efficiency, better accuracy, reduced costs, and improved outcomes. However, withincreased use of technology, the amount of clinical data may become overwhelming.
A comprehensive electronic health record (EHR), which encompasses all health care measures that clini-cians want to use—both standard health care assessments and those acquired through emerging technol-ogy—is the key to improved, ef-ficient clinical decision making. New technologies using sensors to pas-sively monitor older adults at home are being developed and refined (Glascock & Kutzik, 2000; Rantz et al., 2005; Skubic, Alexander, Popes-cu, Rantz, & Keller, 2009) and are increasingly commercially available (Olson, 2008). The challenge remains to integrate the clinical information systems with passive monitoring data, especially in long-term care and home health settings, so clini-cal decision making is enhanced and patient records are complete. This article describes work by researchers at the University of Missouri (MU) who are developing a comprehensive EHR to: (a) enhance nurse care coor-dination at TigerPlace, independent
senior housing that helps residents age in place; (b) integrate clinical data and data from new technology; and (c) advance technology and clinical research.
bAckground
Episodes of acute illness or ex-acerbation of chronic illness often herald functional decline. Delay in recognition of these events and notification of care providers leads to delayed treatment, delayed re-covery, and higher risk of morbid-ity and mortalmorbid-ity (Ridley, 2005).
The key to continued function and independence is to identify prob-lems (e.g., acute illness, exacerba-tion of chronic illness) earlier and offer timely interventions designed to improve functional decline. This kind of monitoring could be accomplished with diligent and consistent human observation; however, for the majority of older adults in the United States, this is neither feasible nor cost effective and could be viewed as undesir-ably intrusive (Demiris et al., 2004).
Technology, in the form of sensor networks, offers an alterna-tive approach that may be more feasible, more cost effective, and less intrusive. Researchers at MU evaluated an aging in place model of care to help community resi-dents maintain their independence and found that community-based care with RN care coordination improved participant outcomes (i.e., activities of daily living, cog-nition, depression, incontinence) when compared with individuals of similar acuity in
institutional-based long-term care (Marek et al., 2005). On the basis of these findings, TigerPlace, an indepen-dent senior housing complex, was created to help older adults age in place (Rantz et al., 2005). TigerPlace was built by Americare Systems, Inc., a long-term care corporation, in partnership with the MU Sinclair School of Nurs-ing. Built to nursing home stan-dards but operated as independent living and specially designed to promote independence, TigerPlace is the only licensed aging in place facility in the country. Residents live in independent apartments with services such as two meals per day, housekeeping, a variety of so-cial activities, and transportation; residents remain in their home environment, receiving increasing care services as needed through the end of life.
Sinclair Home Care, a home care agency, provides the care at TigerPlace, including a com-prehensive health assessment at admission and every 6 months, private home care visits, access to a wellness center, health promotion activities, and social work assis-tance to help with life transitions. An RN is on call 24 hours per day, 7 days per week. The Sinclair Home Care RN care coordinator coordinates all of the residents’ medical care with their physicians, other health care providers, and family members to ensure their health care needs are addressed.
dEscrIPTIon of ThE InTEgrATEd sysTEm
When TigerPlace opened in 2004, a commercially available EHR was adopted, but not all standardized assessments were able to be entered. Databases were needed to enhance the EHR, which link the EHR data with databases containing standardized assessment information and data from the sensor system under de-velopment at TigerPlace (Skubic et
The key to continued function and independence
is to identify problems...earlier and offer timely
interventions designed to improve functional decline.
al., 2009). To address these issues, the research team developed a new EHR that provides standardized assessment forms and information from the sensor technology.
To enhance the care provided by Sinclair Home Care, an MU interdisciplinary research team has installed passive sensor networks in the apartments of volunteer res-idents at TigerPlace and is linking the data retrieved from the sen-sors with the resident’s health data stored in the EHR (Skubic et al., 2009). The project was approved by the MU Institutional Review Board, and residents provide informed consent on admission to TigerPlace to have their health record used for research. Residents who also participate in the sen-sor research provide consent to have their sensor data used in the research in addition to their health record so that sensor data can in-form the health record and health data can inform sensor displays.
The integrated system under development at MU is shown in
Figure 1. The system includes:
l A physiological sensor
net-work including motion sensors, a stove sensor, and a bed sensor developed by collaborators at the University of Virginia (Alwan et al., 2003).
l A video sensor network,
which focuses on preserving the resident’s privacy by extracting anonymous silhouettes or other graphical representations of the residents used to capture falls and gait information.
l A behavior reasoning
com-ponent that combines the sensor and video data, analyzes routine patterns, and generates alerts of potential problems.
l An activity database for
stor-ing sensor data, as well as custom-ized alert conditions.
l The EHR database including
all of the standardized assessments completed on the residents of TigerPlace.
l Secure web-based interfaces
with customizable access to dis-play the data for the researchers, clinicians, family members, and residents.
The system architecture in-cludes a strong link between the activity database, which contains the sensor data, and the EHR da-tabase. This connection allows the sensor data to be annotated with significant health events, and alerts generated from the sensor data can be logged in the EHR.
The physiological sensor net-work, including the bed sensor, motion sensors, and stove sensor, has been installed in the
apart-ments of 26 residents at Tiger-Place. The bed sensor is a padded pneumatic strip that lies on top of the mattress and under the lin-ens; it detects presence in bed and measures restlessness, as well as qualitative breathing and pulse as the person sleeps (Mack, Alwan, Turner, Suratt, & Felder, 2006). Motion sensors are installed in every room in the apartment to detect presence and infer activities. For example, sensors installed in kitchen cabinets and the refrigera-tor, together with the stove sensor, infer cooking activity.
A password-protected web-based interface was developed
Figure 1. The integrated system of sensor and electronic health record databases.
Figure 2. Screen shots of web-based interface showing histograms of bed restlessness and pulse aggregated to a daily level.
to graphically display the sensor data for clinicians and research-ers from a variety of disciplines. All stakeholders within Tiger-Place participated in the design to ensure it was user friendly, clinically relevant, and easy to interpret (Alexander et al., 2008). To use the interface, users select a participant and a time frame. The user may then select options to see data from smaller time frames or from individual sensors. Sensor data are grouped by category: mo-tion, restlessness, breathing, pulse, and stovetop sensor. Data may be aggregated into increments rang-ing from 15 minutes to daily and displayed in a variety of graphical formats, including histograms, pie charts, and line graphs.
Fig-ure 2 contains examples from the
web-based interface including bed restlessness and pulse displayed in histogram format.
Other methods for displaying the data such as motion density maps, which show computed motion densities over 1 month, have been developed (Figure 3). Densities are calculated as the total number of motion sensor events during 1 hour divided by the time at home during that hour (Wang & Skubic, 2008). The passive infrared motion sensors generate an event every 7 seconds if there is con-tinuous motion. Thus, the motion
density events relate to a resident’s level of activity (i.e., a low den-sity indicates a sedentary lifestyle, whereas a higher density indicates a more active lifestyle within the apartment).
Colors are used to signify the level of activity (density of sensor
events); the more vibrant the col-ors, the higher the density. Black is used to note time out of the apart-ment. White indicates no sensor firings. The X axis represents time of day, and the Y axis indicates the day of the month. Regular patterns typically emerge by examining the motion densities over time. For example, Figure 3 shows the den-sity maps for two residents for 1 month. The resident in Figure 3a exhibits a regular pattern of get-ting up at approximately the same
time every day; routinely eating breakfast, lunch, and dinner in the Tiger Place dining room (noted by the three black vertical strips), and going to bed at approximately the same time. Figure 3b is from a cognitively impaired resident who would frequently leave her apartment at irregular intervals and wander around the apartment when she was at home. As seen in Figure 3b, the resident did not have a regular routine. In addi-tion to the visualizaaddi-tion, a com-putational method for comparing density maps has been developed so alerts can be generated if a resident’s pattern changes (Wang, Skubic, & Zhu, 2009).
Retrospective review of the sensor data leading up to and after significant health events, such as falls, emergency depart-ment visits, and hospitalizations, reveals that the sensors are detect-ing changes in health status that traditional health care assessment missed (Rantz, Skubic, & Miller, 2009; Rantz et al., 2009). To use these data prospectively in clinical practice requires quick and easy methods to access the data and review them in conjunction with the health record.
dIscussIon
As technology that can help as-sess clinically relevant parameters such as physical or mental func-tion, vital signs, and gait, progress-es, the need for comprehensive EHRs that incorporate more than text-based data grows. Clinicians typically want as much assessment data as they can get to inform their decisions, but with technology, the amount of data can quickly become overwhelming (Alexander & Staggers, 2009). Efficiently dis-playing the data in ways that are meaningful and quickly accessible is critical for user-friendly EHRs. Forcing clinicians to access mul-tiple databases to retrieve the data can be too time consuming for
al-Figure 3. Motion density maps of two individuals exhibiting different behavioral patterns. The image on the left (Figure 3a) depicts a resident with a regular pattern of activity. The image on the right (Figure 3b) shows data from a resident with an irregular pattern of activity.
Efficiently displaying
the data in ways that are
meaningful and quickly
accessible is critical for
user-friendly electronic
ready overworked staff. Expecting busy clinicians to view multiple screens to integrate and interpret multiple data sources is unrealistic and can be prone to error.
At MU, it was necessary to develop a new EHR database to integrate sensor data, standardized health assessments from a variety of fields, and medical records. No one system met all of the needs of the interdisciplinary research and health care teams. The teams were involved from the beginning to ensure the system is easy to use and the sensor data is displayed in a way that is clinically relevant and easy to interpret (Alexander et al., 2008).
conclusIon And ImPlIcATIons
There are several implications for long-term care facilities or home care agencies that are con-sidering purchasing an EHR. First, it is important to understand how data are stored in the EHR for in-tegration into another system and how data from other technologies, such as motion sensors, could be incorporated into the EHR. Link-ing these databases so they can inform each other is essential for clinical practice. Technology tools that require clinicians to log into a separate system that is outside the clinicians’ usual workflow generate dissatisfaction and are unlikely to be used (Huvane, 2008). Clinicians need to easily view and use the technology data to inform clinical decisions and to do so efficiently within the EHR. In addition, the system must be able to incorporate any standardized assessments that clinical staff decide to use. Stan-dardized assessments are essential for valid risk assessments for skin, falls, and other clinical problems.
Results of the assessments need to be easily viewed and interpreted by clinicians so risk-reducing in-terventions can be provided in the course of care delivery.
Using technology for health care assessment is growing; having a comprehensive EHR will be a necessity for all health care pro-viders. With some forethought and planning, agencies and software companies can be prepared for the inevitable: the need to incorporate new technologically generated data for clinical use with the EHR.
rEfErEncEs
Alexander, G., & Staggers, N. (2009). A sys-tematic review of the designs of clinical technology: Findings and recommenda-tions for future research. Advances in Nursing Science, 32, 252-279.
Alexander, G.L., Rantz, M., Skubic, M., Aud, M.A., Wakefield, B., Florea, E., et al. (2008). Sensor systems for monitoring functional status in assisted living facil-ity residents. Research in Gerontological Nursing, 1, 238-244.
Alwan, M., Kell, S., Dalal, S., Turner, B., Mack, D., & Felder, R. (2003, December).
In-home monitoring system and objective ADL assessment: Validation study. Paper presented at the International Conference on Independence, Aging and Disability, Washington, DC.
Demiris, G., Rantz, M., Aud, M., Marek, K., Tyrer, H., Skubic, M., et al. (2004). Older adults’ attitudes towards and perceptions of “smart home” technologies: A pilot study. Medical Informatics and the Inter-net in Medicine, 29(2), 87-94.
Glascock, A.P., & Kutzik, D.M. (2000). Be-havioral telemedicine. A new approach to the continuous nonintrusive monitoring of activities of daily living. Telemedicine and e-Health, 6, 33-44.
Huvane, K. (2008). Sign of the times: Adopt-ing sAdopt-ingle sign-on can lead to improve-ments in workflow, clinician satisfaction, and security, all with one step. Healthcare Informatics, 25(5), 28, 30-33.
Mack, D.C., Alwan, M., Turner, B., Suratt, P., & Felder, R.A. (2006). A passive and portable system for monitoring heart rate and detecting sleep apnea and arousals: Preliminary validation. In Proceedings of the 1st Distributed Diagnosis and Home Healthcare (D2H2) Conference. Retrieved
from the Medical Automation Research Center, University of Virginia Health System website: http://marc.med.vir- ginia.edu/pdfs/library/NAPS%20D2H2-06.pdf
Marek, K.D., Popejoy, L., Petroski, G., Mehr, D., Rantz, M., & Lin, W.C. (2005). Clini-cal outcomes of aging in place. Nursing Research, 54, 202-211.
Olson, E. (2008, May 25). High-tech de-vices keep elderly safe from afar. New York Times. Retrieved from http://www. nytimes.com/
Rantz, M.J., Marek, K.D., Aud, M., Tyrer, H.W., Skubic, M., Demiris, G., et al. (2005). A technology and nursing collab-oration to help older adults age in place.
Nursing Outlook, 53, 40-45.
Rantz, M.J., Skubic, M., Alexander, G., Aud, M., Wakefield, B., Koopman, R., et al. (2009). Improving nurse care coordina-tion with technology. Manuscript submit-ted for publication.
Rantz, M.J., Skubic, M., & Miller, S.J. (2009). Using sensor technology to augment tra-ditional healthcare. In Proceedings from the 31st Annual International Conference of the IEEE Engineering in Medicine and Biology Society (pp. 6159-6162). Minne-apolis, MN: IEEE.
Ridley, S. (2005). The recognition and early management of critical illness. Annals of The Royal College of Surgeons of Eng-land, 87, 315-322.
Skubic, M., Alexander, G., Popescu, M., Rantz, M., & Keller, J. (2009). A smart home application to eldercare: Current status and lessons learned. Technology and Health Care, 17, 183-201.
U.S. Administration on Aging. (2008). A pro-file of older Americans: 2008. Retrieved from http://www.aoa.gov/AoARoot/ Aging_Statistics/Profile/2008/docs/ 2008profile.pdf
Wang, S., & Skubic, M. (2008). Density map visualization from motion sensors for monitoring activity level. In Proceed-ings of the IET International Confer-ence on Intelligent Environments (pp. 64-71). Retrieved from the Center for Eldercare and Rehabilitation Technology, University of Missouri website: http:// eldertech.missouri.edu/files/Papers/ Wang/Density_Map_Visualization_ from_Motion_Sensors_IEEE.pdf Wang, S., Skubic, M., & Zhu, Y. (2009).
Activity density map dis-similarity comparison for eldercare monitoring. In Proceedings from the 31st Annual International Conference of the IEEE Engineering in Medicine and Biology So-ciety (pp. 7232-7235). Minneapolis, MN: IEEE.
