Lipford, K., Jones, S. & Johnson, K. (Feb, 2017). Needs Assessment of an Electronic Health Record at an Inpatient Psychiatric Hospital. Online Journal of Nursing Informatics (OJNI), 21(1), Available at http://www.himss.org/ojni
Studies have examined different healthcare organizations’ quest to adopt a meaningful use electronic health record (EHR), but there is a significant lack of studies conducted for inpatient psychiatric hospital settings. The purpose of this mixed design descriptive study was to explore one particular inpatient psychiatric hospital’s EHR and identify facilitators and barriers to the current EHR's use. A convenience sample of physicians, nurse practitioners, and nurses was utilized. A total of 27 completed surveys were returned out of 210 for a response rate of 12.85%. While there may be the potential for clinically significant information, there were no statistically significant findings from the data set. The current EHR’s non-user friendly formats and lack of electronic forms were identified as barriers. Systems that require fewer steps to navigate were viewed as facilitators. The results support the need to solicit information from current users of an EHR system in order to determine the positive and negative aspects of a current system. Doing so will help guide the selection of a new system or revision to a current system. The ultimate goal is the development of a more integrated EHR that demonstrates meaningful use to improve the process of documentation at a facility to potentially reduce patient errors.
A major concern in healthcare is the improvement of patient safety through the use of electronic health records (EHRs), which employ meaningful use (The Joint Commission, 2013). Healthcare organizations must have an EHR in place that demonstrates meaningful use to avoid Medicare payment adjustments (Centers for Medicare & Medicaid Services [CMS], 2016). Loss of funding because of an inadequate EHR would have a devastating effect on a healthcare organization. The original impetus for a meaningful use EHR is the Health Information Technology for Economic and Clinical Health Act (HITECH). The intention is the protection of patients, as well as, the prevention of medical errors (Cohen, Grote, Pietraszek, & Laflamme, 2012).
Significance of Problem to Healthcare
A lack of integration within an EHR leads to numerous problems. The following is an example of a problem with the nonintegrated EHR at the study site: a nurse who administers a pro re nata (prn) narcotic medication must document in the electronic progress note, electronic shift report, tracking pharmacy form, paper medication administration record (MAR), and narcotic flow sheet. Because of the need to document one procedure in five different areas, there is a chance of a documentation error. An integrated EHR would allow the nurse to document the procedure one time and the information would populate in additional needed areas. A primary institutional goal would be to have a system that streamlines patient information. In the quest for an effective EHR at a facility, a needs assessment of the system to ascertain barriers and facilitators of the current EHR has the potential to yield valuable information.
A lack of integration in an EHR leads to omitted and duplicated entries (Samuels & Kritter, 2011). Bar code scanning of medications is another essential piece of a meaningful use EHR. The rationale behind the bar code scanning of medications is the reduction of medication errors in some hospitals from 2.89 errors to 1.48 errors per 10,000 doses of medication after implementation (Richardson, Bromirski, & Hayden, 2012). In the National Healthcare Quality Report, there were 48.7 cases per 1,000 hospital discharges of adverse drug events (United States Department of Health and Human Services [HHS], 2013). An effective EHR is able to reduce errors in the healthcare setting. While the focus of this study was not specific to medication administration, this is another area that would be essential to investigate for possible changes.
The purpose of this mixed design descriptive study was to explore one particular inpatient psychiatric hospital’s EHR and identify facilitators and barriers to the use of the current EHR. In order to have a better understanding of the magnitude of the documentation problems in the current EHR, a needs assessment was recommended. Understanding the positive and negative aspects of a system through the solicitation of information from current system users can aid in the selection of a new system or guide revisions to the current system. Another important point is the necessity of frequent involvement of different members of the healthcare team in the development of a meaningful use EHR (Unger et al., 2014). Therefore, the following research question guided this study: For healthcare providers who document in the EHR, what are the positive and negative aspects of utilizing the current EHR at the hospital?
The Rosswurm and Larrabee model, originally developed in 1999, with a focus on linking problems to interventions is the theoretical model for this study. A major concept of the model is that an intuition driven methodology is to be changed to evidence driven methods of treatment. Actively determining barriers and facilitators of a current EHR, formulating interventions towards solving barriers, and actively engaging in those interventions are all components of an evidence driven solution. The ultimate goal of this study is to implement changes in the EHR that integrates patient information and maximizes meaningful use.
Barriers and facilitators of the use of an EHR can be present in different healthcare settings. Common barriers cited in the literature are (a) fragmentation of the record or interface issues with different areas, (b) problems trying to correct errors, (c) alerting system errors, (d) problems signing off a system, and (e) a preference for paper versus electronic charting (Gates & Roeder, 2011; Xiao et al., 2014). Facilitators identified through research are (a) accurate order entry, (b) less time involved with documentation, (c) quick and easy access to information, (d) format of the system, and (e) fewer transcription errors (Gates & Roeder, 2011; Xiao et al., 2014). A lack of computer expertise did not rate high on the list of problematic areas (Gates & Roeder, 2011).
Regardless of profession, facilitators for user attributes revealed in the literature are (a) motivated learners, (b) strategies users developed to aid in knowledge of the system, (c) experience with computers, (d) effective time management skills, (e) flexibility, and (f) personal initiative (Holden, 2011). User attributes that are barriers include (a) lack of typing ability and experience in technology, (b) lack of system knowledge, (c) inflexibility, (d) negative viewpoint, (e) poor time management, and (f) lack of willingness to ask for assistance (Holden, 2011). System attributes that are considered facilitators included (a) remote and fast access, (b) templates, and (c) access to many computers. System attribute barriers included (a) multiple ways to do the same thing, and (b) not an intuitive system (Holden, 2011).
Study Design and Sample
This study used a descriptive, mixed design with both qualitative and quantitative components. A convenience sample of current users of the EHR at the inpatient psychiatric hospital was used. The participants were physicians, nurses, and nurse practitioners. Inclusion criteria were physicians, nurses, both licensed practical nurses (LPNs) and registered nurses (RNs), nurse practitioners, and physician assistants. Inclusion criteria for participants were current use of the EHR, employee of the hospital, over 19 years of age, with possession of a professional email account. A total of 27 completed surveys were returned out of 210 for a response rate of 12.85%. Only completed surveys were accepted and submissions were excluded for missing data.
An inpatient psychiatric facility in the southeast was the study site for the online survey. The hospital is comprised of approximately 1,000 patients and about 1,500 employees. The hospital is a state run facility and not a private entity. There are six large psychiatric units, both civil and forensic, which are spread across the 100-acre campus. The location of the hospital is in a rural setting.
The qualitative and quantitative questions analyzed in this study were modified from the Baylor EHR user experience survey (Xiao et al., 2014). Written permission was obtained from the authors to use the survey and make modifications. Two faculty experts modified and developed the survey for face validity. The first page of the survey included a copy of the informed consent written at an eighth grade reading level. Demographic information was also collected first after consent was given. Demographic information collected included gender, age, type of healthcare professional, number of years in the profession, and comfort level and number of hours of use of a home computer.
In the original study, the internal consistency of the survey tool was excellent with a Cronbach’s α = .892 (Xiao et al., 2014). There are several different areas of focus on the survey. Thirty-six questions on the survey assess the participant’s experience with the EHR. Three areas are Likert-type items and involved the users’ experience with the current EHR, the frequency the respondents utilize the current EHR, and the impact the current EHR has on the respondents’ work. In addition, narrative areas are present where respondents can include additional information regarding concerns about making errors, difficulty in correcting errors, screen display problems, and areas that require paper documentation first before transferring to the computer. The last portion of the survey has a narrative response portion for other electronic documentation systems the user may have operated, positive experiences with any other EHRs, and negative experiences with any other EHRs.
The primary data was obtained during two weeks in August of 2015 using an online survey created in Survey Monkey®. Potential participants who met the inclusion criteria were identified and an email group was developed in Microsoft Outlook®. An initial email with a description of the survey was sent out to potential participants. Approximately two days later, a second email was sent with the online Survey Monkey® link, along with the description of the study and directions on how to complete the survey. The survey was available for completion for two weeks. The participants demonstrated consent through completion of the survey. Participation was voluntary and anonymous. Data collected were not associated with users’ email addresses or identified personal numbers (IPN). The Survey Monkey® feature of advanced options was chosen in which anonymous responses from the web link and email invitation area were blocked to prevent collection of personal information. All data was compiled and reported in aggregate format, so no one could be identified as an individual.
International Business Machines Corporation (IBM) Statistical Package for Social Sciences (SPSS) Statistics 23 version was used to analyze the data. Users’ experience with the current EHR, the frequency the respondents utilize the current EHR, and the impact the current EHR had on the respondents’ work was examined using frequency descriptive statistics. Qualitative portions of the survey in which respondents could include additional information regarding concerns about making errors, difficulty in correcting errors, screen display problems, and areas that required paper documentation first and then transferred to the computer were noted and compiled into similar themes, using content analysis techniques. The last areas of narrative responses for other electronic documentation systems the user may have operated, positive experiences with any other EHRs, and negative experiences with any other EHRs were recorded and assembled into similar themes, as well (Xiao et al., 2014).
The mean number of years of work experience was 23.15 years (SD = 11.21) with a range of 3 - 42 years. Characteristics of the study sample are found in Table 1. The majority of the participants were female (81.5%, n = 22), 60-65 years of age (25.9%, n = 7), and RNs (77.8%, n = 21). The mean number of hours per week spent on home computer use was 13.26 (SD = 10.52) with a range of 2 - 50 hours.
Table 1. Characteristics of the study sample (N = 27)a
User Experience with EHR
Frequencies of variables regarding the EHR and user experience that demonstrated the most variety in responses are addressed in Table 2. Full data analysis is available upon request. The item ‘Training received related to the EHR was effective’ demonstrated wide response selection with 48.1% (n = 13) indicating strong agreement/agreement and 29.6% (n = 8) indicating strong disagreement/disagreement. The item ‘EHR supports efficient interdisciplinary communication’ also demonstrated a wide range of responses with 56% (n = 15) indicating strong agreement/agreement and 44% (n = 10) indicating strong disagreement/disagreement. The item ‘Overall satisfaction with the EHR demonstrated variety in participant responses, as well, with 51.5% (n = 14) indicating strong agreement/agreement and 25.9% (n =7) indicating strong disagreement/disagreement. As provided in Table 2, none of the survey respondents selected not applicable (N/A) in the response choice.
Table 2: Frequencies of user experience with the EHR displaying wide variety in responses (N = 27)
Some variables regarding the EHR and user experience demonstrated agreement in responses with less variation. The item ‘The ability to find where to document patient care’ demonstrated the least selection response with 88.8% (n = 24) indicating strong agreement/agreement ‘Confidence in using the EHR also demonstrated less variety with 85.2% (n = 23) indicating strong agreement/agreement. The item measuring opinion regarding ‘Useful alerts/reminders’ was the one variable that displayed the most disagreement/strong disagreement at 40.7% (n=11). Useful alerts/reminders also demonstrated the most neutral responses as well with 25.9% (n = 7).
Respondents Utilization of EHR and Impact on Work
Frequencies regarding the user experience with the EHR and experiences with documentation, assistance, and patient care are found in Table 3. Documentation on time demonstrated the least inconsistency of responses with 85.2% (n = 23) of users being able to most of the time or always achieve timely documentation. Documentation directly onto computer chart without use of paper demonstrated least inconsistency of responses with 74.1% (n = 20) of users being able to always or most of the time chart directly into the EHR. Awareness of changes in the system demonstrated the least inconsistency of responses with 51.9% (n = 14) of users selected most of the time, rarely, to never being aware of system changes.
Table 3: Frequencies of user experience, documentation, assistance, & patient care (N = 27)
Frequencies of the user experience with the EHR and those that had the most impact on work are addressed in Table 4. N/A was selected for six of the seven variables and included slowness of computers 11.1% (n = 3), having to reboot computers 18.5% (n = 5), finding a computer to use 48.1% (n = 13), duplicating entries 44.4% (n = 12), login or passwords 18.5% (n = 5), and some forms are paper and not electronic 7.4% (n = 2). Some forms are paper and not electronic was identified by 55.6% (n = 15) users who noted this had an impact more than once a day to daily. Multiple steps with accessing electronic information was identified by 92.6% % (n = 25) of users as having an impact more than once a day to daily.
Table 4: Frequencies of user experience with the EHR and impact on work (N = 27)
Qualitative Descriptive Data
String variables were a part of the survey to allow participants the opportunity to provide further narrative format information about the EHR. Not all participants shared a narrative response. Frequencies of the responses to the different qualitative portions of the survey are addressed within Table 5.
Table 5: String variable responses on survey
The most common areas of concern with making errors were progress notes, discharge summaries, and one participant mentioned a lack of connectivity between different areas of the EHR which leads to errors. In addition, the patient progress notes were the most problematic with correcting an error. Screen display issues centered on labs, medical consents, too much information on one screen, and maneuverability issues. Paper documentation before computer notation was required for vital sign documentation, admission notes, monthly summaries, and clinical nursing report. Lack of availability was also noted to be an issue that had an impact on paper versus computer documentation. The majority of participants had not used another EHR. Those who had used another system identified the systems Computer Programs and Systems Inc. (CPSI), VISTA, Meditech, McKesson, Paragon, and Clientware.
If participants had used another electronic documentation system, the positive and negative experiences with the other systems were also solicited. Of the participants who responded,55.5% (n = 15) listed ‘none’ because other systems had not been used. One participant who had not used another system listed positive and negative comments about the EHR at the study site. Eleven participants (40.7%) listed actual positive aspects of the other systems. Some participants listed more than one positive or negative comment. In regards to negative experiences, 11 participants who had used other systems provided comments. The most common positive and negative experiences with other systems are summarized in Table 6.
Table 6. Positive and negative aspects of other EHRs
After a review of the frequencies of responses by a faculty member experienced in statistics, an Analysis of Variance (ANOVA) was conducted in order to analyze the difference in the number of years of work experience and the training received related to the EHR was effective. These two areas were specifically targeted because there appeared to be less variability in responses from a review of the frequencies. While the ANOVA of these two areas was not significant (F (4, 22) = 1.165, p > .05), clinically significant trends were noted. As the number of years of work experience increased, the less agreement was displayed with the amount of training received on the EHR. An ANOVA was also conducted in order to analyze the difference in the number of years of work experience and the awareness of changes and improvements to the EHR. While the ANOVA was not significant (F (4, 22) = 1.537, p > .05), clinically significant trends were noted. As the number of years of work experience increased, the less agreement was displayed with awareness of any changes made to the EHR.
Another recommendation by the faculty expert was to take two variables, more useful alerts/reminders and overall satisfaction with the EHR and recode the data in order to collapse strong agreement/agreement, and strong disagreement/disagreement categories to determine if statistically significant data was present. ANOVAs were conducted to examine differences in these variables and the number of years of work experience. There was no significant difference in number of years of work experience and overall satisfaction with the EHR (F (2, 26) = .905, p > .05). Again, while none of the differences were statistically significant, clinically significant trends were noted. In both variables, the higher the number of years of work experience, the greater the disagreement.
Discussion of Findings/Outcomes
Information from the needs assessment of facilitators and barriers shared by users was then shared with the executive leadership team at the hospital. Some of the barriers were problematic areas of navigation of the system, multiple screens, lack of useful alerts and reminders, and too much information in one area of the chart. Particular mention was made of issues with documentation of vital signs, admission assessments, and progress notes. Careful search for a system that allows more integration to prevent documentation of the same treatment in multiple areas, while paying attention to current problematic areas is important to current users. As previously noted a PRN order must be documented in five separate areas. An integrated EHR would allow information to be populated into multiple areas at the same time. If the search committee for a new system or changed system is partially composed of current users, a more useful EHR can be purchased to meet the needs of the hospital (Unger et al., 2014). This information was shared with the executive leadership team who verbally agreed these areas were important to consider as the hospital moves forward in the selection process.
Another important consideration is to review how training regarding the EHR and awareness of changes made to the EHR is implemented across the healthcare organization. Participants in the current project varied greatly regarding both of these areas. Possible causes of these differences may be due to how information is disseminated within the individual units. As previously mentioned, the study site is a large inpatient psychiatric hospital which spreads across 100 acres and has multiple, individual units or buildings. Once information is obtained and reviewed, a consideration should be made to streamline the training process of how notifications of changes are made across the hospital into one cohesive method. When this information was presented to the executive leadership team there was a verbal acknowledgment that dissemination of information and training was a process that needed to change. The leadership team agreed there were gaps in how training is completed and how information is disseminated across the hospital.
The lack of integration in a hospital EHR can have a negative impact on nursing and the organization. One national patient safety goal encompasses improvement of patient safety through the use of EHRs (The Joint Commission, 2013). Prevention of medical errors is a very important part of healthcare. Improvements in an EHR can have a positive impact on the prevention of medical errors. If an improved EHR can be found to prevent documenting the same skill in more than one area, there is less chance of making a documentation error.
As well, the implementation of cohesive training and dissemination of information at a hospital can also improve the healthcare system. Adequate training of staff to recognize all of the features of an EHR will allow users to maximize usage of the system. Ignorance of changes could lead to mistakes being made by users of the EHR. Key changes with the initiation of a new integrated EHR have the potential to improve the healthcare system at the hospital. Because the hospital receives patients from all over the state, improvement in services at the hospital has the ability to create a positive process change at a state level.
To ensure a positive practice change at the hospital the committee selecting the new EHR or revised EHR should include current users. As well, any changes which are incorporated into the current EHR needs to be clearly communicated to users of the system. The hospital has an Intranet and employees have Microsoft Outlook® accounts so the infrastructure is in place to improve communication. The last recommendation is to initiate EHR training that is not currently in place at the hospital. The hospital has a central training department and individual unit trainers to develop and disseminate training on the EHR which can facilitate this process.
A small sample size is a limitation for this study. Out of 210 email invitations, 33 participants initiated and 27 participants fully completed the study survey. With a response rate of 12.85% correlations would be difficult to accurately measure. The sample was limited to healthcare providers but there are other providers who use the current EHR. The hospital is an inpatient, state funded, rural psychiatric hospital. Additional limitations would also include the fact that the study was done in one rural hospital located in the southeast region of the United States and not in multiple hospitals in multiple locations across the country. Because of the specific nature of the hospital itself, generalizations to other institutions may be affected.
Additional information could be beneficial in the determination of a change to an existing EHR or in the search for a new EHR. The study could be expanded to include all current users of the EHR at the hospital and not just healthcare providers. Useful information regarding the current system may be obtained from other users of the documentation system. Because the study was conducted in a very specific location and healthcare setting, investigation into other psychiatric settings such as outpatient clinics or urban locations would yield important further information.
Healthcare providers have useful information to share regarding facilitators and barriers of an EHR system. Evidence based practice as described by the Rosswurm and Larrabee (1999) model with a focus on linking problems to interventions can be guided by the useful information shared from the current users. In addition, sharing information about changes to a system and training can have positive impacts on the practice setting. A possible benefit to the hospital is a more integrated system that incorporates the essential component of meaningful use.
Centers for Medicare and Medicaid Services [CMS]. (2016). Medicare and Medicaid EHR incentive program basics. Retrieved from https://www.cms.gov/regulations-and-guidance/legislation/ehrincentiveprograms/basics.html
Cohen, S. B., Grote, K. D., Pietraszek, W. E. & Laflamme, F. (2012). Increasing consumerism in healthcare through intelligent information technology. The American Journal of Managed Care, 16(1), 37-44.
Gates, M. L. & Roeder, P. W. (2011). A case study of user assessment of a corrections electronic health record. Perspectives in Health Information Management, 8(Spring), 1-20. Retrieved from: http://www.perspectives.ahima.org
Holden, R. J. (2011). What stands in the way of technology-mediated patient safety improvements? A study of facilitators and barriers to physicians’ use of electronic health records. Journal of Patient Safety, 7, 193-203. doi: 10.1097/PTS.0b013e3182388cfa
Richardson, B., Bromirski, B. & Hayden (2012). Implementing a safe and reliable process for medication administration. Clinical Nurse Specialist 26(3), 169-176. doi: 10.1097/NUR.0b013e3182503fbe
Rosswurm, M. A. & Larrabee, J. H. (1999). A model for change to evidence-based practice. Journal of Nursing Scholarship, 31(4), 317-322. doi: 10.1111/j.1547-5069.1999.tb00510.x
Samuels, J. G. & Kritter, D. (2011). Pain management documentation. CIN: Computers, Informatics, Nursing 29(9), 512-518. doi: 10.1097/NCN.0b031e31821a1582
The Joint Commission. (2013). 2014 National patient safety goals hospital accreditation programs (pp. 1-17). Washington, DC: The Joint Commission.Retrieved from http://www.jointcommission.org
Unger, M. D., Aldrich, A. M., Hefner, J. L., & Rizer, K. (2014). A journey through meaningful use at a large academic medical center: Lessons of leadership, administration, and technical implementation. Perspectives in Health Information Management, 11, 1-12. Retrieved from http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4272440/
United States Department of Health and Human Services, Agency for Healthcare Research & Quality (2013). National healthcare quality report 2013 (pp. 1-231). Retrieved from http://www.ahrq.gov
Xiao, Y., Montgomery D. C., Philpot, L. M., Barnes, S. A., Compton, J. & Kennerly, D. (2014). Development of a tool to measure user experience following electronic health record implementation. Journal of Nursing Administration, 44(7/8), 423-428. doi: 10.1097/NNA.0000000000000093
Karen Lipford PMNHP-BC, DNP, EdD (primary author)
Originally graduated from Chipola College Associate Degree in Nursing (ADN) in 1986. MSN as a Family Nurse Practitioner and a post-master’s degree as a psychiatric nurse practitioner. I will graduate in less than two weeks on May 6, 2016, with second doctorate degree, a Doctor of Nursing Practice (DNP) from Troy University. First doctorate is a doctorate of education (EdD) from University of West Florida. Currently the Dean of the School of Health Sciences at Chipola College and works part-time as a psychiatric nurse practitioner at Life Management.
Dr. Stacey Jones DNP, FNP-BC
Assistant Professor Troy University
DNP, University of Alabama at Birmingham 2011
MSN, Troy University, 2001
BSN, Troy University, 1995
Committee chair of DNP project by Karen Lipford
A published author since 2012 and presenter since 2008. Honors and awards include recipient of the Nurse Educator Scholarship from the Alabama Board of Nursing for 2009-2010 and 2010-2011. Recipient of the Nurse Faculty Loan Aware through University of Alabama at Birmingham sponsored by HRSA for 2009-2010 and 2010-2011. Clinical focus is family practice.
Dr. Kelly Johnson, DNP, RN Assistant Professor Troy University
Faculty committee member of DNP project by Karen Lipford
DNP, MSN, and BSN from Troy University. A published author since 2008 and grants and projects author. Honors and awards from National League of Nursing and Sigma Theta Tau. Clinical focus is critical care nursing.