Harmon, C. S., Fogle, M. & Roussel, L. (2015). Then and now: Nurses' perceptions of the electronic health record. Online Journal of Nursing Informatics (OJNI), 19 (1).
A quantitative methodology utilizing the Assessment of Nurses' Attitudes Toward Computerization (NATC) questionnaire (Stronge and Brodt, 1985) was conducted to assess nurses' perceptions of an electronic health record (EHR) five years post-implementation. The purpose of this study was to generate evidence that can be applied to support best practices for future EHR implementations and innovations. The following three recommendations were obtained from this study's data analysis: a) improve communication among care givers, b) EHR optimization via clinical decision support, and c) encourage additional time-motion research. An additional secondary analysis was performed to compare these five year, post-implementation results to previous studies conducted in the same facility at pre-implementation and six months post-implementation. Overall, the nurses' attitudes at five years post-implementation were favorable towards the use of EHR in practice.
During the past five decades, information technology has become an important part of the healthcare industry andhas altered the roles, workflow, and functions of nursing practice (Chan, 2007;Saba & McCormick, 2006). Laramee, Bosek, Shaner-McRae, and Powers-Phaneuf's (2012) study reported a gap between the EHR pre-implementation expectations and the post-implementation reality of EHR experiences. This gap determined that nurses' EHR perceptions became less positive at six-months and 18 months post-implementation when compared to pre-implementation. Furthermore, nurses studied by Sassen (2009) suggested the following strategies to improve EHR acceptance: involve nursing in the long-term vision of the EHR, understand nursing perceptions and experiences, and develop a change-management process when workflow challenges arise. EHR adoption rates for nurses improve when communication, training, and goals align and are focused on improving care (Ash & Bates, 2005).
The intentions of this study were to: (a) assess and analyze nurses' perceptions of the EHR at five years post implementation (primary study); (b) conduct a secondary analysis by comparing this study's results to the results of previous studies conducted in the same facility at pre-implementation and six months post-implementation; and (c) formulate recommendations from these results to generate evidence that can be applied to best practices for future EHR implementations and innovations.
Roger's Diffusion of Innovations theory (1995) was used as the theoretical framework to support this study's purpose of generating evidence that can be applied to best practice for future EHR implementations and innovations. Diffusion of Innovations theory seeks to explain the spread of new ideas and technology throughout a culture (Rogers, 1995). In addition, diffusion is the process of communicating an innovation over time to members of a social system. The main elements in the diffusion of new ideas include: innovation, communication, time, and members of a social system. An innovation may be a new idea, practice, or technology. In this study, the EHR can be defined as the innovation. The members of a social system include this study's nursing population. Time is described as five years post-implementation to determine adoption and explore recommendations. Communication allows the transfer of information from one person or organization to another.
This study was performed at a 165-bed hospital located in the urban southeast. Although the facility is considered a community hospital, it is a subset of a large nonprofit healthcare network. Numerous facilities within this healthcare network utilize one standardized EHR, including the study facility.
The study facility had an accessible population of approximately 230 bedside (staff) nurses. Thus, this project's population of interest was bedside nurses who use the EHR. The sample inclusion criteria consisted of registered nurses and licensed practical nurses employed full-time or part-time as staff nurses. The exclusion criteria included individuals who lacked a license as a registered nurse or licensed practical nurse, were not employed at the study's facility, were less than 20 years of age due to facility institutional review board (IRB) recommendations, or were employed at the facility but not in a bedside-nurse position.
There was a 10% response rate from the eligible population. Thirty-nine percent of respondents had 10 to 20 years of nursing experience, 52.2% reported they held a Bachelor's degree in nursing, and 34.8% were between the ages of 30 and 39. Demographic characteristics were important to the investigator, especially during the secondary comparison analysis of the five-year post-implementation study with the pre-implementation and six months post-implementation studies. Furthermore, 82.6% reported using the EHR for approximately two to four years and 60.8% stated that they use the EHR 25 or more hours per week. See Appendix A for the characteristics of study participants.
The NATC questionnaire was used with permission from the original authors (Stronge & Brodt, 1985). Five demographic questions regarding the nurses' age, years of experience in nursing, nursing education, weekly hours of EHR use, and months or years using the facility's EHR were added to the original questionnaire. The original questionnaire is a 20-item survey with a five-point Likert-scale. Scores range from 20 to 100 with higher scores indicating a more positive attitude. The NATC questionnaire has a high reliability (r=0.90) as well as construct and content validity (Brodt & Stronge, 1986; McBride & Nagle, 1996; Stronge & Brodt, 1985).
The five-year post-implementation study used a different survey tool than the pre-implementation and six months post-implementation study. Therefore, six questions from this original survey tool were selected for relationship comparison to the NATC questionnaire. These six questions used interchangeable variables or indicators to ask the nurse about their attitude toward the EHR as it related to their workload. These variables included: positive, difficult, threatened, convenient, efficient, and favorable.
Design and Protocol
A quantitative, non-experimental study design was used. During the study design, each item on the NATC questionnaire was classified into one or more of the following variable categories for relationship correlation analysis: costs, nurses' workload, time, communication, benefits, privacy, efficiency, nursing data, improve patient care, quality nursing care, ease, financial stability, and legal. For the secondary analysis, the construct variables identified for correlation analysis from the pre-implementation and six months post-implementation studies included: positive, difficult, threatened, convenient, efficient, and favorable. Thus, a Pearson two-tailed correlation with a p value of 0.05 was established prior to data analysis to show correlation significance among variables.
The study received approval via the primary investigator's university IRB as well as the study facility's nursing scientific advisory council (NSAC) and IRB. The NATC questionnaire was built on the facility's SharePoint site which afforded accessibility via computers throughout all nursing units, nursing stations, nursing offices, break rooms, and computers on wheels (COWS). The questionnaire was available online for three weeks. Nurses were solicited via email at the beginning of the survey period and again at the end of the second week. Participation was voluntary and participant consent was implied via a completed and submitted survey. Survey results were downloaded onto a Microsoft Excel spreadsheet from SharePoint for data analysis.
Data analysis consisted of three phases. First, the scores for positive or negative attitudes toward computers were calculated. Utilizing the original authors' NATC scale of 20-100, an overall average score was calculated with a score less than 50 being negative and a score greater than 50 being positive (Laramee et al., 2012; Stronge & Brodt, 1985). The NATC questionnaire consists of six positively worded and 14 negatively worded items that were calculated on a five-point scale with the negatively worded items reverse scored to make all responses consistent for analysis. Then, item responses were summarized.
Second, the data was analyzed using Statistical Package for the Social Sciences (SPSS) software. Pearson two-tailed correlations were performed on the NATC item results with a p value less than 0.05 to show significance. If a Pearson rxy value is close to a positive one (+1) a positive relationship is indicated; a value close to a negative one (-1) indicates a negative relationship, and a value close to a zero means there is little or no relationship (Kranzler, 2007). Finally, a secondary comparative analysis was completed to compare the NATC item results from the five-year post-implementation study with the item results of the pre-implementation and six months post-implementation studies' survey tool. This analysis was also completed using SPSS and Pearson two-tailed correlations.
On the NATC scale, the overall average total score for the primary five-year post-implementation study was 68.39. This result indicated an overall positive attitude. The Pearson rxy 0.707 (p=0.000) indicated a significant, positive relationship between 'time spent using a computer is out of proportion to the benefits' and 'a computer increases costs by increasing the nurses' workload'. During the study design, these two NATC items had been classified into the subsequent variables: time, costs, benefits, and nurses 'workload. Another significant, positive relationship was determined between 'computers save steps and allow the nursing staff to become more efficient' and 'computers will allow the nurse more time for professional tasks' with an rxy 0.826 (p= 000). These two NATC items were categorized into the variables: efficient and time. 'Computerization of nursing data offer nurses a remarkable opportunity to improve patient care' had a positive, significant correlation with 'computers make nurses' jobs easier' with an rxy 0.768 (p=000). The variables classified from these two NATC questions consist of nursing data, ease, and improve patient care.
A Pearson rxy 0.785 (p=000) demonstrated a positive, significant relationship between 'computers can cause nurses to give less time to quality nursing care' and 'computers cause a decrease in communication between hospital departments'. The variables from these two NATC items include: time, quality nursing care, and communication. 'A computer increases costs by increasing the nurses' workload' correlated to 'computers can cause nurses to give less time to quality nursing care' with a rxy 0.773 (p=000). These NATC questions were classified into the following variables: time, costs, nurses' workload, and quality nursing care.
The five-year post-implementation study used a different survey tool than the pre-implementation and six-month post-implementation study. Therefore, the responses to the NATC items were compared to the responses to the pre-implementation and six-month post-implementation study tool items. The following results were analyzed for the variables from the pre-implementation and six-months post-implementation study tool: positive rxy 0.452 (p=0.030), convenient rxy 0.432 (p=0.040), efficient rxy 0.432 (p=0.040), and favorable rxy 0.481 (p=0.020) which indicate a positive, significant relationship with the NATC item: 'computers make nurses' jobs easier' from the five-year post-implementation study.
With an overall average rating of 68.39 on the NATC scale, the nurses' attitudes at five years post-implementation were favorable towards the use of EHR in practice. When comparing this study's average score of 68.39 to scores obtained from two other NATC studies, it is noted that all scores ranged between 60 and 75. Laramee and colleagues (2012) reported a 74.2 at pre-implementation, 65.9 at six months post-implementation, and 67.7 at 18 months post-implementation. In addition, Smith, Smith, Smith, Krugman & Oman (2005) obtained a score of 70.7 at pre-implementation and 61.4 at one year post-implementation.
The correlations among questions indicated positive, significant relationships within the primary five years post-implementation study analysis and the secondary analysis with the pre-implementation and six-month post-implementation study. Six relationships emerged from this analysis (see Appendix B). These relationships can be formulated into the following variables as pre-defined during the design phase to inform recommendations: time, benefits, costs, nurses' workload, efficient, ease, nurses' data, improve patient care, quality nursing care, positive, favorable, convenient, and communication. These variables inform the subsequent recommendations: a) improve communication among care givers, b) EHR optimization via clinical decision support, and c) encourage additional time-motion research (see Appendix C).
Clinical Decision Support
The recommendation for clinical decision support derived from the correlation of the following questions: 'computers make nurses' jobs easier' and 'computerization of nursing data offer nurses a remarkable opportunity to improve patient care.' The subsequent variables obtained from these two questions were used to inform the recommendation of clinical decision support: nursing data, ease, and improve patient care. Clinical decision support eases healthcare professionals' jobs by quickly compiling data from the EHR to generate advice at the point of care to improve patient care (Heba & Czar, 2013). Clinical decision support is an optimization or advancement of the average EHR. Garg and colleagues (2005) completed a systematic review of 100 studies with 64% of studies reporting clinical decision support improved practitioner performance. Furthermore, Kawamoto, Houlihan, Balas & Lobach (2005) did a systematic review of 70 studies with 68% of studies significantly improving clinical practice as a result of decision support systems.
The recommendation to improve communication stemmed from the correlation of 'computers can cause nurses to give less time to quality nursing care' and 'computers cause a decrease in communication between hospital departments.' Communication, time, and quality nursing care consist of the variables acquired from these two questions to inform the recommendation to improve communication within the organization via an organization-based communication skill training program such as TeamSTEPPS®.
TeamSTEPPS® has proven effective in improving skills and communication among health care teams (Agency for Healthcare Research and Quality [AHRQ], 2013). Teams trained with TeamSTEPPS® provide safer and higher quality patient care by eliminating barriers, resolving conflict, clarifying team roles, and optimizes people, information, and resources (AHRQ, 2013). Briefs, huddles, and debriefings are some of the methods utilized by TeamSTEPPS®. Briefs are used for planning, huddles are used for problem solving, and debriefings assist with process improvement.
Additionally, clinicians are trained to use tools such as the SBAR, a mental model that stands for Situation, Background, Assessment and Recommendation (Haig, Sutton & Whittington, 2006), bedside handoff communication, and conflict resolution skills to stop procedures that could cause a patient harm.
A time-motion study is recommended due to the correlation of questions classified into the following variables: time, costs, nurses' workload, benefits, ease, efficient, positive, favorable, convenient, and quality nursing care (see Appendix B and C). Of the four correlations that recommend a time-motion study, the variable time arose three times and efficient appeared twice. A time-motion study will define how nursing spends or focuses their time during work. For example, Dwibedi and colleagues (2011) used a time-motion study to observe and measure nurses' workflow with defined factors such as direct patient care activities, indirect patient care activities, administrative activities, and miscellaneous activities. A time-motion study observes the nurses' workflow at the point of care; thus determining quantitatively what and how much of a nurses' time is spent on specific tasks throughout the shift. Shadowing nurses may not only define how much time is spent on tasks but it may also identify cumbersome processes and workarounds related to the EHR. Furthermore, a time-motion study can be used as the basis to implement Lean methodologies to reduce inefficiencies and decrease wastes.
Risks and Limitations
Although the surveys do not have any identifiers, a slim risk of participant identification exists through demographic data. Study limitations include: convenience sampling, self-reporting methodology, possibility of ballot box stuffing, and the study was limited to one facility and one EHR. Additionally, the survey samples are not matched and represent the nursing facility population at pre-implementation, six months post-implementation, and five years post-implementation of an EHR. Therefore, different nurses may have participated in each sample collection.
Implications to Nursing
The purpose of this study was to generate evidence that can be applied to support best practices for future EHR implementations and innovations. Per study analysis, the following EHR recommendations still exist five years post-implementation: improved communication among care givers, EHR optimization via clinical decision support, and additional time-motion research. These recommendations could improve the effective use of the EHR and contribute to evidence-based practice. As well, this study adds to the limited informatics research related to nurses' perceptions of EHR use and provides preliminary recommendations related to these perceptions.
A snapshot of current nurses' perceptions at five years' post-implementation of an EHR informs recommendations for clinical decision support and improved communication within the organization. In addition, a time-motion study was recommended from both the five years post-implementation snapshot as well as during the secondary analysis of the pre- and six months post-implementation studies. A time-motion study may provide a better understanding and insight regarding nursing workflow in relationship to HER use. Overall, the nurses' attitudes at five years post-implementation were favorable towards the use of EHR in practice.
Agency for Healthcare Research and Quality. (2013). How to use the evidence rating. Retrieved from http://www.innovations.ahrq.gov/evidencerating.aspx
Ash, J. S., & Bates, D. W. (2005). Factors and forces affecting EHR system Adoption: Report of a 2004 ACMI discussion. Journal of the American Medical Informatics Association, 12(1), 8-12. Retrieved from http://search.proquest.com/docview/220785140?accountid=35812
Chan, M. F. (2007). A cluster analysis to investigating nurses' knowledge, attitudes, and skills regarding the clinical management system. CIN: Computers, Informatics, Nursing, 25(1), 45-54.
DiCenso, A., Guyatt, G., & Ciliska, D. (2005). Evidenced-based nursing: A guide to clinical practice. St. Louis, MO: Mosby.
Dillon, T. W., Blankenship, R., & Crews, T. (2005). Nursing attitudes and images of electronic patient record systems. CIN: Computers, Informatics, Nursing, 23(3), 139-145.
Dwibedi, N., Sansgiry, S. S., Frost, C. P., Dasgupta, A., Jacob, S. M., Tipton, J. A., & Shippy, A. A. (2011, June 1). Effect of bar-code-assisted medication administration on nurses' activities in an intensive care unit: A time-motion study. American Journal of Health-System Pharmacy, 68, 1026-1031. http://dx.doi.org/10.2146/ajhp100382
Gale, B. V., & Schaffer, M. A. (2009, February). Organizational readiness for evidence-based practice. The Journal of Nursing Administration, 39(2), 91-97.
Garg, A. X., Adhikari, N. K., McDonald, H., Rosas-Arellano, M. P., Devereaux, P. J., & Beyene, J. (2005). Effects of computerized clinical decision support systems on practitioner performance and patient outcomes. The Journal of the American Medical Association, 293(10), 1223-1238. http://dx.doi.org/10.1001/jama.293.10.1223
Heba, T., & Czar, P. (2013). Handbook of informatics for nurses and healthcare professionals (5th ed.). Boston, MA: Pearson.
Kawamoto, K., Houlihan, C. A., Balas, E. A., & Lobach, D. F. (2005, March 31). Improving clinical practice using clinical decision support systems: A systematic review of trials to identify features critical to success. British Medical Journal , 330(7494): 765.
Kranzler, J. H. (2007). Statistics for the terrified (4th ed.). Upper Saddle River, NJ: Prentice Hall.
Laramee, A. S., Bosek, M., Shaner-McRae, H., & Powers-Phaneuf, T. (2012, October). A comparison of nurse attitudes before implementation and six and 18 months after implementation of an electronic health record. CIN: Computers, Informatics, Nursing, 30(10), 521-530.
Lee, J., Kuo, Y. F., & Goodwin, J. S. (2013). The effect of electronic medical record adoption on outcomes in US hospitals. BMC Health Services Research, 13(39). http://dx.doi.org/10.1186/1472-6963-13-39
National Institutes of Health. (n.d.). Theory at a glance: A guide for health promotion practice. Retrieved from http://www.cancer.gov/cancertopics/cancerlibrary/theory.pdf
Rogers, E. M. (1995). Diffusion of innovations (4th ed.). New York: The Free Press.
Saba, V. K., & McCormick, K. A. (2006). Essentials of nursing informatics (4th ed.). New York: McGraw-Hill.
Sassen, E. J. (2009, September/October). Love, hate, or indifference: How nurses really feel about the electronic health record system. CIN: Computers, Informatics, Nursing, 27(5), 281-287.
Carolyn S. Harmon, DNP, RN-BC serves as nursing faculty at York Technical College and University of South Carolina Lancaster cooperative nursing program. Dr. Harmon also teaches masters and doctoral nursing informatics courses at Capella University. She is board certified as an informatics nurse by the American Nurses Credentialing Center. Dr. Harmon received her Doctor of Nursing Practice at the University of Alabama at Birmingham.
Maureen Fogle Ed.D, RN, NE-BC, is the Director for Education & Professional Practice at Carolinas Medical Center which is a part of the Carolinas Healthcare System. Dr Fogle has been with Carolinas Healthcare for twenty-five years and has held positions in critical care nursing, education and administration. She also is responsible for the current curriculum design used for senior nursing care as part of Nurses Improving the Care of Health System Elders (NICHE) program out of New York University's Hartford School of Nursing. Dr Fogle also holds nursing advisory board positions with the Mercy School of Nursing (MSON) in Charlotte, North Carolinas and Pfeiffer University in Meisenheimer, North Carolina. In addition she a nurse researcher and a member of the North Carolina Organization of Nurse Leaders (NCONL) state wide Nursing Research Committee.
Linda Roussel, PhD, RN, NEA-BC, CNL, is a Professor at the University of Alabama Birmingham (UAB) School of Nursing, and serves as Doctor of Nursing Practice (DNP) Program Director. Dr. Roussel teaches courses in leadership, translational and improvement science, and Scholarly Project Design and Implementation. Her scholarly initiatives focus on clinical nurse leadership, academic-clinical partnership, and frontline engagement. Dr. Roussel has authored and co-authored nursing textbooks including Management and Leadership for Nurse Administrators, Initiating and Sustaining the Clinical Nurse Leader Role,Project Planning and Management, A Guide for CNLs, DNPs, and Nurse Administrators, and Evidence-Based Practice, An Integrative Approach to Research, Administration, and Practice