KOSSMAN, SUSAN P. PhD, RN; BONNEY, LEIGH ANN MSN, RN; KIM, MYOUNG JIN PhD
Electronic health records (EHRs) offer the potential for safer, higher quality, and more cost-effective care by efficiently supporting healthcare providers’ work. However, numerous studies demonstrate that EHR use can create unintended consequences.1,2 Electronic health records change the way practitioners organize and use information.3 The design, organization, and display of information on EHR screens or tools can change workflow in ways that improve efficiency and also in ways that introduce error and create unexpected problems.2,4 Our previous work found that nurses strongly prefer EHRs to paper records but want them to better support work processes, especially thinking and communication.5 These issues require further exploration.
Nurses work in complex, hectic environments with frequent interruptions. In this setting, they need to process clinical data and information quickly to form clinical judgments to communicate with the healthcare team and guide optimal patient care delivery. Nurses obtain needed information from many sources—the EHR, hand-off tools, assessment findings, their knowledge base, expert practitioner advice, and online or printed resources—and use tools that organize information and prompt their memory—data displays, worksheets, and checklists—to guide practice.
The aforementioned source organization and memory tools are cognitive artifacts—devices designed to display information that serve as “mental tools that aid the mind.”6 These tools help nurses stay focused and support clinical judgment and team communication needed to achieve quality outcomes. Electronic health records include cognitive artifacts—screens or tools designed to provide effective data capture, storage, retrieval, and display—designed to support nursing work, with the aim of safer, higher quality, and more cost-effective care. Common EHR cognitive artifacts or tools include problem lists, focused assessment forms, clinical practice guidelines, care plan, medication administration record (MAR), and interdisciplinary summary notes.
Many authors have shown that EHRs can create unintended consequences, negatively affecting workflow and patient care outcomes.1,2,4 In an exploratory study, Kossman and Scheidenhelm5 found that nurses in two community hospitals preferred EHRs over paper records but raised concerns that the EHR-generated cognitive artifact tools had limitations. Specifically, they said that assessment forms limited their critical thinking (not “thinking outside the checkbox”) and that it was difficult to find other health team member notes, thus limiting communication. Others have noted nurses’ challenges with efficiently documenting and finding information in EHR tools7 and the potential that this has to disrupt the flow of patient care information among the healthcare team.8 Nurses want “information tools that work”9(p161) and create paper tools if those generated by the EHR do not address their needs.7–10 Improvised (self-made) paper tools may help nurses’ organization, memory, and thinking but can also disrupt efficient information flow and communication if data are not transferred to the EHR.8,11 These findings suggest that nurses use a variety of paper and EHR-generated cognitive artifacts as they perform nursing work and that some of these tools interfere with communication and clinical judgment instead of fostering these.
Clinical judgment involves identifying, prioritizing, and continuously evaluating complicated, rapidly changing patient problems and solutions.12–14 It requires perception and response to cues from multiple sources: patient/family, data in the EHR, and consultation/communication with nurses and team members. Tanner’s13 Clinical Judgment Model identifies four components of clinical judgment and notes variation along nurses’ experience and knowledge continuums. “Noticing” entails developing a grasp of the situation and is a function of the nurse’s expectations based on knowledge and prior experience. “Interpreting” involves clinical reasoning patterns (which vary depending on the amount and quality of the data) to interpret the meaning of the data and determine the appropriate course of action. “Responding” includes choosing and implementing an appropriate course of action. “Reflecting” includes immediate attending to patient outcomes and later reviewing the clinical judgment process for appropriateness. Clinical judgment and decision making improve with experience, and expert nurses rely both on formal nursing knowledge and tacit knowledge, or intuition.13,15
Each component of clinical judgment could be enhanced with well-designed cognitive artifacts. These “external knowledge representations are purpose-driven data displays created to communicate, augment, and influence human reasoning, understanding, and decision making.”16(p352) Healthcare settings abound with cognitive artifacts, both improvised and computer generated: checklists, task lists, medication records, fall precautions, and clinical practice guidelines. Clinical reasoning and judgment depend on interactions between internal knowledge representation (the person’s formal and experiential knowledge) and external knowledge representation (including cognitive artifacts among other factors).17 McLane et al16(p353) assert that the “information content of these cognitive artifacts directly influences critical thinking, clinical decisions, and the ongoing evaluation and individualization of the plan of care.” In addition, cognitive artifacts can help communication among healthcare team members by providing an accessible vehicle for sharing information.
No tools exist to objectively, validly, and reliably measure clinical judgment in practicing nurses. Researchers have used self-report, interviews, narrative reports, and direct observation of simulated or real clinical situations to describe nurses’ clinical judgment.13,14 Lasater14 devised a rubric based on Tanner’s13 theoretical model to guide assessment of student nurses’ performance in the four phases of clinical judgment: noticing (important data, patterns, seeking information), interpreting (prioritizing and making sense of the data), responding (communication, intervening, flexibility, skill), and reflecting (self-evaluation, commitment to improvement). Although not developed for practicing nurses, the rubric identifies areas to assess in the evaluation of clinical judgment.
This study sought to answer three research questions: (1) What tools or cognitive artifacts do nurses use to support their clinical judgment and team communication? (2) What are nurses’ perceptions of the utility of selected cognitive artifacts of a full-featured inpatient EHR for support of clinical judgment and team communication? (3) How often do nurses and healthcare team members use these selected cognitive artifacts in patient care?
This descriptive study used a convergent mixed-method design. Following human subjects review and approval, we recruited nurses working on ICU or medical/surgical floors in a tertiary care medical center in the Midwest who had at least 6 months of experience with an EHR to participate in an online survey and/or focus group interviews. The online anonymous survey developed by the research team asked nurses to rate frequency of use and utility of seven cognitive artifacts of clinical judgment and team communication: self-made work lists and EHR problem list, focused assessment forms, clinical practice guidelines, care plan, MAR, and summary note. Survey questions, based on Tanner’s Clinical Judgment Model and Lasater’s operationalization of its four dimensions, asked respondents to rate the utility of these cognitive artifacts for overall communication and clinical judgment as well as attributes reflecting the specific dimensions of clinical judgment: noticing, interpreting, responding, and reflecting. Focus group interviews used open-ended questions to probe nurses about improvised and EHR generated tools they use to support clinical judgment and communication and ways these tools could be redesigned to better support work. In addition, we analyzed EHR usage statistics to identify usage trends of these artifacts by healthcare team members. Quantitative data were cleaned and analyzed using descriptive and inferential statistics. Qualitative data were analyzed using content analysis for identification of themes.
A convenience sample of 50 nurses participated. Of these, 46 responded to the online survey, with 33 (72%) completing it and being included in the analysis. Four nurses participated in the focus group, for a total N = 37. Demographic data for these 37 nurses show that 87% are women, with a mean age of 37.57 years (range, 23–58 years), median of 8 years of experience (range, 1–35 years), median of 6 years on their current unit (range, 2–10 years), and median of 1 year using the current EHR (range, 1–4 year). Most (67%) have a BSN and 62.2% work in an ICU.
Cognitive Artifact Tool Use and Support of Clinical Judgment and Communication
There were no significant differences in cognitive artifact use by education level or work area. Most nurses used all of the tools at least daily (MAR, 97%; self-made work lists, care plan, and summary notes, 87%; focused assessment forms, 76%; clinical practice guidelines, 73%; problem list, 64%). However, most nurses rated only the self-made work lists and MAR as “extremely useful” in supporting overall communication and clinical judgment (Figures 1 and 2).
Nurses rated each tool for usefulness in supporting specific attributes of clinical judgment on a five-point scale ranging from “extremely helpful” to “don’t use.” Table 1 presents the frequency of nurses rating the tools as “extremely helpful” in supporting these attributes, grouped into their respective dimensions of clinical judgment. In the Noticing dimension, self-made work lists helped most nurses in all three attributes and the focused assessment screens helped most recognize patterns. In the Interpreting dimension, the work list helped most with prioritization, although no tool helped most with anticipating potential complications, identifying complications early, or making sense of data. In the Responding dimension, the work list helped most nurses in all the attributes, while the MAR helped most nurses feel they were in control and document completely, and the Summary note helped most monitor patients’ progress. In the Reflecting dimension, none of the tools helped most of the nurses. The work list and MAR assisted most nurses in documentation, an important aspect of team communication.
We computed composite scores for communication, overall clinical judgment, and the four clinical judgment dimensions and conducted inferential tests for significance of association of the tools with these scores, using Friedman tests (nonparametric repeated measures test of association) with Bonferroni correction. If the association between the group of tools and composite score was significant, we conducted post hoc analysis with Wilcoxon test to identify the specific cognitive artifacts that contributed to significance. There was a significant association among the group of cognitive artifacts and composite scores for communication (X2 = 33.01, P < .05), overall clinical judgment (X2 = 44.83, P < .05), and three of the clinical judgment dimensions: Noticing (X2 = 33.1, P < .001), Interpreting (X2 = 29.48, P < .001), and Responding (X2 = 60.82, P < .001). Post hoc analysis identified specific cognitive artifacts that accounted for the significant difference through pairwise comparisons; Bonferroni correction for type I error was applied so that all effects are reported at a .002 level of significance. Self-made work lists and MARs are significantly associated with all composite scores when compared with other tools. The focused assessment and summary note have a more limited effect. They have a significant association with overall Clinical Judgment when compared with the problem list and with Interpreting when compared with the care plan. The focused assessment tool was also significantly associated with Noticing compared with clinical practice guidelines.
Focus Group Themes
My Organization Supports My Thinking
Improvised work sheets are created by the nurse (with information, which they acknowledge is contained in the EHR), organized and displayed in ways that support their work style. “I have a paper I use… I’ve been using it for 8 years…helps me stay organized… [the EHR tools] help me fill in the gaps.” They can find information easier on their lists than on the EHR tools: “if the patient had a CT, you’d have to go into orders or something…but it’s right here [on her list] instead of scrolling. That’s just easier for me.” They report difficulty finding some items in the EHR that can be highlighted on their lists: “the potassium/magnesium protocol is a big one that gets missed frequently because you have to go into your orders and find it.”
Can’t Live Without Electronic Health Record, But It Could Be Simpler to Use
Nurses commented that it takes time to learn to use EHRs well, “the more you use it the easier it becomes,” and assessment screens especially are time-consuming to complete. They stated that if the EHR system went down, they would “cry,” “pray,” and “have a panic attack.”
Conform Instead of Create
When asked how they might redesign EHR tools to better support their work, nurses were at a loss. “I wouldn’t even know how to answer that.” One nurses said, “we’re like robots. We conform. Whatever they give us to conform to…you just learn to use what they tell you to.”
A limited descriptive analysis of 1 day of usage statistics of EHR tools by healthcare professionals in roles of nurse, nurse practitioner, physician, pharmacist, dietitian, social worker, physical therapist, occupational therapist, speech pathologist, and pastoral care included a rank ordering of frequency of use of the selected cognitive artifacts. Professionals in all roles viewed summary notes and flow sheets extensively. Summary notes were ranked as the first or second most frequently used tool by all except nurses, for whom it was third, following flow sheets and the MAR. The MAR ranked as first or second most frequently used tools by nurses, physicians, and pharmacists. The care plan was viewed by all roles except for physicians and pharmacists, but at a lower frequency (third or fourth). The problem list was viewed at a low frequency by nurses, physicians, and pharmacists, while clinical guidelines were viewed only by nurses.
Full-featured EHRs offer users many design elements and tools that should support clinical judgment and communication, including the ability to customize screens, to view information from multiple screens, to link to clinical practice guidelines and focused assessment screens tied to a patient’s clinical diagnosis, and to document care in an interdisciplinary location accessible to all providers such as a summary note. These tools and features should enhance clinical judgment and team communication, but their promise has not been fully realized. Our findings indicate that most nurses used all the cognitive artifacts at least once daily and found them to be helpful in supporting overall clinical judgment (except for the problem list) and team communication (except for the care plan) but rated their self-made work lists as more helpful than any EHR tool except for the MAR. Nurses rated their self-made work lists and the MAR as the most helpful cognitive artifacts; statistical analysis verified the significant association between these tools and communication, clinical judgment, and three of its dimensions (noticing, interpreting and responding). Looking more deeply, improvised work lists and EHR tools varied in their utility to support specific clinical judgment attributes. It is concerning that most nurses rated none of the cognitive artifacts as “extremely helpful” to key pieces of nursing work such as anticipating potential complications and making sense of patient data (aspects of “interpreting”) and evaluating care through determining if the patient met outcomes and adjusting treatment to achieve goals (aspects of “reflecting”). The findings suggest that the organization or display of information on these tools is not supporting important aspects of nursing work.
Findings indicate that health professionals use EHR tools differently. Frequency of use of EHR tools by team members suggests the value of EHR for team communication. Usage statistics show that health professionals in all roles view the summary notes and (except physicians and pharmacists) the care plan more frequently than nurses do, indicating that these may be more useful tools for team communication than nurses perceive.
Nurses rate their improvised cognitive artifacts as more useful for overall clinical judgment and communication than any of the EHR tools except the MAR. They use a variety of names for these—work list, report sheet, hand-off sheet, and “my list.” Comments indicate that the organization and display of information on improvised tools aid workflow by prompting memory and highlighting needed information in an easily accessible place. This finding needs more exploration. What information do nurses include on these tools and how are they displayed? How do the information elements nurses capture on self-made work lists map to EHR-generated cognitive artifacts? How can we redesign these EHR tools to better reflect nursing information needs and work? To answer these questions, we need nurses to identify what best supports nursing work. The tendency to quietly accept tools that do not support nursing work (conform to it, be “robots”) or develop workarounds can create obstacles to developing better tools.
The Tanner Clinical Judgment Model provided a useful framework for studying clinical judgment in practicing nurses. With Lasater’s operationalization of attributes of the model’s four dimensions, the concept was deconstructed into concrete examples of nursing work. The survey developed from this framework prompted nurses to differentially rate the utility of the cognitive artifacts to aspects of clinical judgment and provided more depth of insight than questions about clinical judgment in general. The survey needs revision, however. The number of noncompleters raises questions about the survey’s length, and nurses commented that some of the names of the cognitive artifacts were unclear. This could be a mismatch between vendor names for the tools and those used in everyday practice, which may vary among nursing units. Matching terms to those in common use will increase the accuracy of results.
Nurses are the largest group of health professionals using EHRs and, thus, can leverage the potential of EHRs to enhance quality care and patient outcomes. This will happen only if EHR tools support rather than interfere with nursing work. This study investigated nurses’ use of paper and EHR-generated cognitive artifacts and views of their utility in supporting clinical judgment and team communication. While nurses used most of the EHR tools daily, they found only the MAR to be as useful as improvised tools in supporting clinical judgment and communication. Other team members used the summary notes and care plans more frequently than nurses did, indicating that these tools hold more potential for team communication than nurses perceive. To optimize efficiency and information flow, there is a need to further explore nurses’ use of EHR tools and to enhance their design so that they better reflect nurse’s information needs and support clinical judgment, team communication, and improved patient outcomes.
1. Ash J, Sittig D, Dykstra R, Campbell E, Guappone K . The unintended consequences of computerized provider order entry: findings from a mixed methods exploration. Int J Med Inf. 2009; 78:(Suppl 1): S69–S76.
2. Koppel R, Wetterneck T, Telles J, Karsh B . Workarounds to barcode medication administration systems: their occurrences, causes, and threats to patient safety. J Am Med Inform Assoc. 2008; 15:(4): 408–423.
3. Patel VL, Kushniruk AW, Yang S, Yale JF . Impact of a computer-based patient record system on data collection, knowledge organization, and reasoning. J Am Med Inform Assoc. 2000; 7:(6): 569–585.
4. Karsh B, Weinger M, Abbott P, Wears R . Health information technology: fallacies and sober realities. J Am Med Inform Assoc. 2010; 17:(6): 617–623.
5. Kossman SP, Scheidenhelm S . Nurses’ perceptions of the impact of electronic health records on work and patient outcomes. Comput Inform Nurs. 2008; 26:(2): 69–77.
6. Norman D . Things That Make Us Smart: Defending Human Attributes in the Age of the Machine. New York: Basic Books; 1994; .
7. Stevenson J, Nilsson G . Nurses’ perceptions of an electronic patient record from a patient safety perspective: a qualitative study. J Adv Nurs. 2012; 68:(3): 667–676.
8. Keenan G, Yakel E, Dunn Lopez K, Tschannen D, Ford Y . Challenges to nurses’ efforts of retrieving, documenting, and communicating patient care information. J Am Med Inform Assoc. 2013; 20:(2): 245–251.
9. Staggers N, Clark L, Blaz J, Kapsandoy S . Nurses’ information management and use of electronic tools during acute care handoffs. West J Nurs Res. 2012; 34:(2): 153–173.
10. Staggers N, Clark L, Blaz J, Kapsandoy S . Why patient summaries in electronic health records do not provide the cognitive support necessary for nurses’ handoffs on medical and surgical units: insights from interviews and observations. Health Inform J. 2011; 17:(3): 209–223.
11. Hyde E, Murphy B . Computerized clinical pathways (care plans): piloting a strategy to enhance quality patient care. Clin Nurse Spec. 2012; 26:(5): 277–282.
12. Banning M . A review of clinical decision making: models and current research. J Clin Nurs. 2008; 17:(2): 187–195.
13. Tanner C . Thinking like a nurse: a research-based model of clinical judgment in nursing. J Nurs Educ. 2006; 45:(6): 204–211.
14. Lasater K . Clinical judgment development: using simulation to create an assessment rubric. J Nurs Educ. 2007; 46:(11): 496–503.
15. Simmons B, Lanuza D, Fonteyn M, Hicks F, Holm K . Clinical reasoning in experienced nurses. West J Nurs Res. 2003; 25:(6): 701–719.
16. McLane S, Turley J, Esquivel A, et al. Concept analysis of cognitive artifacts. Adv Nurs Sci. 2010; 33:(4): 352–362.
17. Zhang J, Norman D . Representations in distributed cognitive tasks. Cogn Sci. 1994; 18:(1): 87–122.