Educational Innovations

Teaching Dual-Process Diagnostic Reasoning to Doctor of Nursing Practice Students: Problem-Based Learning and the Illness Script Catherine O. Durham, DNP, FNP-C; Terri Fowler, DNP, FNP-C; and Sally Kennedy, PhD, FNP-C, CNE

ABSTRACT Accelerating the development of diagnostic reasoning skills for nurse practitioner students is high on the wish list of many faculty. The purpose of this article is to describe how the teaching strategy of problem-based learning (PBL) that drills the hypothetico-deductive or analytic reasoning process when combined with an assignment that fosters pattern recognition (a nonanalytic process) teaches and reinforces the dual process of diagnostic reasoning. In an online Doctor of Nursing Practice program, four PBL cases that start with the same symptom unfold over 2 weeks. These four cases follow different paths as they unfold leading to different diagnoses. Culminating each PBL case, a unique assignment called an illness script was developed to foster the development of pattern recognition. When combined with hypothetico-deductive reasoning drilled during the PBL case, students experience the dual process approach to diagnostic reasoning used by clinicians. [J Nurs Educ. 2014;53(11):646-650.]

M

any students admitted to a Doctor of Nursing Practice (DNP) program have only recently completed a Bachelor of Science in Nursing (BSN) degree and may have little clinical experience as a bedside nurse. Educating these novice nurses at the doctoral level to become nurse practitioners (NPs) working in primary care is a challenge. Many have not gained the clinical experience necessary to transfer what they have learned from a textbook. Even for the experienced RN, the transition from bedside nursing, where the diagnosis is known, to that of a primary care provider whose job is to diagnose requires reorganization of learned cognitive patterns. Because of the organ-system focus of teaching pathophysiology, a known diagnosis becomes the anchor for building schemas, or mental models, for the bedside nurse. The diagnosis also acts as the trigger for cognitive retrieval when a patient with the same diagnosis is encountered. However, in primary care, the cognitive triggering event changes. Patients in primary care present with signs and symptoms that should bring potential pathological causes to the NP’s mind to correctly diagnose and plan care. For this to occur, the mental models of most BSN students must be restructured or created anew, especially given that multiple diagnoses are often associated with the same symptom.

Variety of Terms Used to Explain the Diagnostic Reasoning Process

Received: March 5, 2014 Accepted: July 15, 2014 Posted Online: October 23, 2014 Dr. Durham is Instructor, Dr. Fowler is Instructor, and Dr. Kennedy is Assistant Professor, College of Nursing, Medical University of South Carolina, Charleston, South Carolina. The authors have disclosed no potential conflicts of interest, financial or otherwise. Address correspondence to Catherine O. Durham, DNP, FNP-C, Instructor, Medical University of South Carolina, College of Nursing, Room 515, 99 Jonathan Lucas St., Charleston, SC 29425-1600; e-mail: durhamc@ musc.edu. doi:10.3928/01484834-20141023-05

646

Research into diagnostic reasoning in medicine dates back to the 1970s, and the descriptive terms have evolved over the years. Initially, two frameworks emerged called hypothetico-deductive reasoning and pattern recognition. Hypothetico-deductive reasoning, now considered the analytic approach (Croskerry, 2009), has also been referred to as forward reasoning (Norman, Brooks, Colle, & Hatala, 1999) and, more recently, System 2 reasoning (Croskerry, 2009). This approach to diagnostic reasoning assumes that reasoning is hypothesis driven and requires careful consideration of all variables before reaching a conclusion. The various terms used over the years have made it difficult to compare and contrast research. To simplify the terminology, the more descriptive term of analytic reasoning will be used throughout this article. A second form of diagnostic reasoning, initially termed pattern recognition, is now called nonanalytic reasoning, a System 1 analysis (Crosskerry, 2009), or backward reasoning (Norman et al., 1999). Pattern recognition involves the often unconCopyright © SLACK Incorporated

EDUCATIONAL INNOVATIONS

scious recognition that patients’ signs and symptoms form patterns that suggest specific disease entities or diagnoses. The recognition of these patterns was once thought to develop over time and is achieved only with repeated clinical experiences, culminating in the intuitive abilities of the expert clinician (de Bruin, Schmidt, & Rikers, 2005). However, pattern recognition is now regarded as a process that both novices and experts use (Ark, Brooks, & Eva, 2006). Because the term nonanalytic reasoning is more descriptive, it will be used throughout this article. The most recent theory of the diagnostic reasoning process, dual-process theory (Croskerry, 2009; Pelaccia, Tardif, Triby, & Charlin, 2011), combines analytical, System 2 theory (the hypothetico-deductive model) with intuitive pattern recognition, a System 1 process. In dual process theory, the complexity of the patient’s presenting complaints, combined with the clinician’s level of knowledge and experience, determines which system would be used.

Early Research in Diagnostic Reasoning Within the framework of analytic reasoning, early research focused on the difference between novices and experts. The early work of Lesgold, Feltovich, Glaser, and Wang (1981) showed that experts in their study not only had an expected larger pool of domain knowledge, they also had a more finely tuned perceptive process that searched their memories more rapidly than novices for schemas that aligned with the task at hand (as cited in Charlin, Tardiff, & Boshuizen, 2000). Two independent groups of researchers (Ark, et al. 2006; Norman, et al., 1999) conducted studies in which psychology students were taught to read electrocardiograms using both analytic and nonanalytic processes. The study by Norman et al. (1999) demonstrated that analytic reasoning resulted in 50% more accurate diagnoses. Ark et al. (2006) built on these findings by testing the dual processing theory; the researchers found that providing the option of using analytic and nonanalytic processes improved diagnostic reasoning versus using one process alone. Both studies delineated that the cognitive approaches of analytic and nonanalytic reasoning could be taught. Most of the research on NPs’ diagnostic reasoning was published in the nursing literature in the 1990s and early 2000s (Victor-Chmil, 2013). Comparisons among those studies are difficult due to the varied descriptions of the processes and labels used, underscoring the evolving understanding of the diagnostic reasoning processes. Burman, Stepans, Jansa, and Steiner (2002) conducted a qualitative study that involved 36 NPs who were asked to reveal their diagnostic reasoning processes as they worked through two clinical scenarios. Although explained as “iterative spiral process” (pp. 60, 62) and searching for “red flags” (pp. 62-63), the dual process of diagnostic reasoning was described. Offredy (2002) compared the diagnostic reasoning processes among 11 general medical practitioners and 11 NPs. During the think-aloud, scenario-based interview process, both groups used the nonanalytic process of pattern recognition as their main decision-making method. In a similarly designed study by Ritter (2003), 10 experienced NPs used think-aloud protocols while working through a case scenario. The results determined that the NPs use of a nonanalytic Journal of Nursing Education • Vol. 53, No. 11, 2014

process only slightly exceeded their use of the analytic process, again confirming the use of dual process reasoning.

A Simplified Perspective The process of diagnostic reasoning can be simplified by thinking of the three variables that interact. They are the: ● Approach to problem solving: analytic versus nonanalytic. ● Clinical experience: along the continuum of novice to expert. ● Perceived level of certainty: uncertainty versus certainty. In the presence of a clinical dilemma, such as making a diagnosis, the nonanalytic process of diagnostic reasoning becomes the method of least cognitive effort and is activated first for both novices and experts (Ark et al., 2006). In the presence of uncertainty after a diagnosis comes to mind, or when a diagnostic pattern is not readily available for recall (increased uncertainty), the clinician will revert to the more laborious process of analytic reasoning—i.e., of evaluating all variables (signs and symptoms and other patient cues) and searching for potential hypotheses. In this sense, the analytic process has a monitoring function over the nonanalytic process that is activated by uncertainty. An additional monitoring process is the gut feeling—a unique aspect of intuition that some authors contend has an affective component (Stolper et al., 2010). For example, an NP student with a clinical background in cardiology, on the basis of his or her experience gained as an RN, may use the nonanalytic process exclusively when a patient presents with symptoms that point to the cardiovascular system. Gut feelings may activate if the patient appears clinically unstable, but the NP student may not be able to pinpoint the origin from the available objective data. Most likely, the gut feelings will cause the student to take action of some kind. However, when the NP student encounters a patient whose chief complaint involves the musculoskeletal system, an area of less certainty due to limited clinical experience as an RN, the NP student’s nonanalytic processes may not be helpful because no schemas have been created for that type of pattern. Consequently, the nonanalytic processes are not viable. The NP student’s cognitive processes turn to the analytic approach— i.e., systematically gathering data and appraising it. Regardless of the approach taken to diagnose a patient’s complaint, possessing domain knowledge is essential. Most domain knowledge is acquired early in the NP program in the core courses; advanced pathophysiology, pharmacology, and physical assessment. However, it is in the clinical courses that students begin to synthesize what they have learned by retrieving this prior domain knowledge and applying it during patient encounters in the clinical setting. This is where the body system approach to learning is not helpful because it is not aligned with how patients present in primary care. In other words, it was not taught according to how the clinician will need to retrieve it. Consequently, those schemas must be restructured. Problembased learning (PBL) is one effective strategy to help students fine tune their analytic reasoning processes through encounters with various disease presentations (signs and symptoms) and thus reorganize cognitive structures for their new role— diagnosing diseases. 647

EDUCATIONAL INNOVATIONS

Problem-Based Learning Online PBL was chosen as the primary teaching strategy to increase the clinical exposure for the novice nurse entering the DNP program. PBL combines problem-solving skills with didactic content within a clinical context that is organized to mimic the analytic reasoning process (Gruppen, 1997). According to Bizzocchi and Schell (2009), working through rich PBL cases augments medical education because the process helps students to remember key information and enriches its application to real life. The process also distills the hypothesis-driven, analytic approach to diagnostic reasoning (Schmidt et al., 1996). However, faculty had the additional goal of expanding contact with various pathologies through these rich PBL cases to promote the development of the nonanalytic approach to diagnostic reasoning. To accomplish this, it was important that students experience how one symptom could lead to multiple diagnoses, depending on patient variables. Therefore, the PBL cases were designed so that four scenarios, anchored with the same symptom as the chief complaint, followed different diagnostic paths as the cases unfolded. Because the course was co-taught by four faculty members(C.D., T.F., S.K.), who each worked with 20 to 30 students, each faculty member had only one case to manage, albeit with multiple groups of five to six students. Within an online learning management system (Moodle™), student groups were organized in sets of four groups for each biweekly discussion. Each set included discussions of all four cases. With this organization of case sets, faculty were able to lock students out of the other case sets. Thus, students could not review what was being discussed in other groups that were working on the same case. The process of online PBL followed a step-wise, facultyguided analytic reasoning process that began with students generating multiple hypotheses from prior knowledge, based on a brief chief complaint. In the next step, students associated each hypothesis with expected findings on the patient’s history and physical examination. Their next task was to extract salient data from a provided comprehensive history and examination that also contained extraneous data. After they agreed upon a final diagnosis, they developed an evidence-based treatment plan. Because the goal was for students to understand how a symptom could lead to multiple diagnoses, it was important for them to review the other three cases in the set. To encourage this mindful review and to provide an overall approach to the chief complaint that would promote the nonanalytic process leading to students’ schema development and/or cognitive reorganization, faculty developed an individual assignment called an illness script.

Illness Scripts Striving for a consistent manner to teach diagnostic reasoning, the faculty looked to the seminal work of Barrows and Tamblyn (1980) on illness scripts, a concept similar to schemata. Their work, combined with Bowen’s (2006) conceptualization of educational strategies to teach diagnostic reasoning, resulted in a unique assignment that would summarize the biweekly PBL cases and promote the process of nonanalytic reasoning. Illness scripts are mental models that organize knowledge in a predictable pattern, typically based on, and cued by, a 648

patient’s presenting signs and symptoms. Illness scripts are considered a type of schema. According to the literature, scripts have three components: enabling conditions (patient factors that affect the likelihood of illness, such as the patient’s age, gender, and environmental exposures); faults (pathophysiology); and the clinical consequences (signs and symptoms or the result of a specific fault; Charlin, Tardif, & Boshuizen, 2000). To help NP students recognize these patterns, which would underscore the nonanalytic process, faculty created a grid that included the components of an illness script with signs or symptoms (clinical consequences) as anchors. After reviewing the set of four PBL cases that unfolded from the same symptom, students completed the grid. To simplify the terminology for teaching purposes, the components included in a typical illness script grid are (a) chief complaint, (b) working hypotheses, (c) predisposing conditions, (d) discriminating features, (e) pathophysiological insult, and (f) problem representation (Table). Working through the example in the Table, the chief complaint is written in broad terms, such as a “54 year-old male comes into the office with a chief complaint of right knee pain.” From the limited information that initiated all four PBL cases, students in each group created a list of potential explanations for the symptom that was called the “working hypotheses” to avoid confusing this step with the differential diagnosis. After each group reviewed their unique completed history and physical examination, the list of working hypotheses was pared down, leading to one diagnosis for each group, which was included in the illness script. Although the Table does not include an exhaustive list of potential causes of monoarticular joint pain, it describes the causes commonly seen in primary care. Predisposing conditions that may make an individual more susceptible to the designated pathology include social history, family history, and comorbidities, as well as situational factors such as trauma to the joint. This type of data could be found in the history and physical examination of each patient. The pathophysiological insult is a brief description of the pathophysiology unique to each working hypothesis in enough detail to compare and contrast the working hypotheses. Discriminating features are attributes of the symptoms— i.e., salient features that will distinguish each hypothesis. This is an important step, because it changes the constellation of data that will form the pattern stored in cognitive structures. In this step, key terms, such as acute, chronic, insidious onset, stabbing, or pressure-like, help to highlight the unique aspects of each working hypothesis. The students then develop a brief phrase or problem representation that describes the specifics of the case that is stored in memory. This description will help to trigger the memory of a specific illness script and promote retrieval. For example, a description such as “acute onset of inflammatory monoarticular arthritis” is typically considered as septic arthritis until proven otherwise. The description contrasts with that for osteoarthritis, which may be “insidious onset of monoarticular joint pain without signs of inflammation.” The final step is naming the script that summarizes the common features of all potential working hypotheses, which in this case would be “acute monoarticular joint pain.” The desired outcome is a fluid approach to the chief complaint. The creation of the illness script from the four PBL cases Copyright © SLACK Incorporated

EDUCATIONAL INNOVATIONS

TABLE Illness Script Grid Chief Complaint: 54-Year-Old Male Patient With Right Knee Pain

Working Hypothesis

Predisposing Condition

Pathophysiological Insult (Describes Key Pathology)

Discriminating Features (Expected Findings on History, Physical Examination, or Diagnostic Tests)

Defining Features (Qualifiers)

Problem representation for infectious arthritis: acute onset of monoarticular inflammatory arthritis associated with fever and prostration. Infectious arthritis

Seeding of joint from puncture wound may be evident

Acute infectious process

Acute onset, warm, red joint, effusion present, febrile and ill, single episode

Acute, monoarticular, painful, single episode

Problem representation for acute gouty arthritis: acute onset of a recurrent, exquisitely painful, monoarticular inflammatory process Gout

Acute ethanol (alcohol) ingestion in susceptible individuals, diet, trauma, surgery

Acute, recurrent inflammatory process caused by monosodium urate crystal deposition in a joint

Acute onset, episodic, warm violaceous joint, effusion present, exquisitely painful, more common in males, monosodium urate crystals found on aspirate

Acute, monoarticular, recurrent

Acute onset, warm, red joint, effusion present, calcium pyrophosphate deposition crystals found on aspirate

Acute, may begin as monoarticular but will involve multiple joints eventually

Problem representation for pseudogout: acute onset of monoarticular joint pain Calcium pyrophosphate deposition disease, pseudogout

Trauma, surgery

Acute, recurrent inflammatory process caused by calcium pyrophosphate crystal deposition in a joint

Problem representation for osteoarthritis: acute monoarticular onset without signs of inflammation Osteoarthritis

Trauma (acute or chronic)

Noninflammatory process of cartilage destruction caused by wear and tear

enables students to synthesize pertinent data from each case, identify the features that distinguish each case, and compare and contrast the cases in one succinct document. First, this approach promotes the construction or reorganization of prior knowledge in a manner that will serve students in clinical practice. Second, the PBL cases broaden exposure to various pathologies that NP students may not encounter during their precepted clinical experiences, yet they will be expected to manage after having graduated. Third, the process of PBL reinforces the step-wise diagnostic reasoning process of analytic reasoning. Finally, culminating the PBL process in an illness script assignment introduces students to the nonanalytic reasoning process as a meaningful way to organize the discriminating features of a symptom. This approach to teaching diagnostic reasoning promotes the dual process of diagnostic reasoning that will serve the NP in clinical practice.

Lessons Learned PBL has been used as the main teaching method in the DNP program at the authors’ institution since 2010. The initial pilot of the illness script assignment was completed with a mixed group Journal of Nursing Education • Vol. 53, No. 11, 2014

Chronic process, although acute pain can occur, bony proliferation without heat, redness, small effusions can occur

Chronic, decline in functioning, often polyarticular

of 63 adult and family NP students in the spring 2013 semester. Four sets of PBL cases, with four scenarios in each set, were assigned that semester. This meant that during the semester, students participated in four PBL cases that spanned 2 weeks each, and each student was expected to review the other three cases in her or his set. The first and third PBL sets culminated in an illness script assignment. A one-page instruction sheet with the grid was provided to students, along with an exemplar of right knee pain. The largest milestone was training faculty so that expectations of students were clear and grading was equitable. A simple grading criteria was used to grade each script. Overall, students did well on this assignment. On the end-of-course evaluations, one student commented, “The scripts provided a way to organize and simplify the research and differentiate between diagnoses.” Several students commented that they wanted more opportunities to develop illness scripts following the PBL cases. One student felt the illness scripts were “quite time-consuming” but felt they were “a great learning tool.” With a faculty-to-student ratio of up to 1 to 30, and given that faculty provide significant individual feedback to promote learning, these assignments are time consuming to grade. How649

EDUCATIONAL INNOVATIONS

ever, faculty agreed that the assignment had merit; thus, they have continued requiring at least one illness script each semester. It is difficult to assess the impact of the illness script assignment on students’ ability to learn the dual process of diagnostic reasoning. During the didactic courses that introduce the illness scripts, students also are precepting in clinical sites. This is truly a time when students synthesize what they have learned in previous courses and apply that learning to managing patients. It is a time of rapid professional growth in the NP role. Faculty would like to think that developing illness scripts has filled the void of learning the dual process of diagnostic reasoning.

References Ark, T.K., Brooks, L.R., & Eva, K.W. (2006). The benefits of flexibility: The pedagogical value of instructions to adopt multifaceted diagnostic reasoning strategies. Academic Medicine, 41, 281-287. Barrows, H.S., & Tamblyn, R.M. (1980). Problem-based learning: An approach to medical education. New York, NY: Springer. Bizzocchi, J., & Schell, R. (2009). Rich-narrative case study for online PBL in medical education. Academic Medicine, 84, 1412-1418. doi:10.1097/ ACM.0b013e3181b6ead0 Bowen, J.L. (2006). Educational strategies to promote clinical diagnostic reasoning. New England Journal of Medicine, 355, 2217-2225. Burman, M.E., Stepans, M.B., Jansa, N., & Steiner, S. (2002). How do NPs make clinical decisions? The Nurse Practitioner, 27(5), 57-64. Charlin, B., Boshuizen, H.P., Custers, E.J., & Feltovich, P.J. (2007). Scripts and clinical reasoning. Medical Education, 41, 1178-1184. Charlin, B., Tardif, J., & Boshuizen, H.P. (2000). Scripts and medical diagnostic knowledge: Theory and applications for clinical reasoning instruction and research. Academic Medicine, 75, 182-190.

650

Croskerry, P. (2009). A universal model of diagnostic reasoning. Academic Medicine, 84, 1022-1028. doi:10.1097/ACM.0b013e3181ace703 de Bruin, A.B., Schmidt, H.G., & Rikers, R.M. (2005). The role of basic science knowledge and clinical knowledge in diagnostic reasoning: A structural equation modeling approach. Academic Medicine, 80, 765773. Gruppen, L.D. (1997). Implications of cognitive research for ambulatory care education. Academic Medicine, 72, 117-120. Norman, G.R., Brooks, L.R., Colle, C.L., & Hatala, R.M. (1999). The benefit of diagnostic hypotheses in clinical reasoning: Experimental study of an instructional intervention for forward and backward reasoning. Cognition and Instruction, 17, 433-448. Offredy, M. (2002). Decision-making in primary care: Outcomes from a study using patient scenarios. Journal of Advanced Nursing, 40, 532541. Pelaccia, T., Tardif, J., Triby, E., & Charlin, B. (2011). An analysis of clinical reasoning through a recent and comprehensive approach: The dualprocess theory. Medical Education Online, 16, 5890. Retrieved from http://med-ed-online.net/index.php/meo/article/view/5890 Ritter, B.J. (2003). An analysis of expert nurse practitioners’ diagnostic reasoning. Journal of the American Academy of Nurse Practitioners, 15, 137-141. Schmidt, H.G., Machiels-Bongaerts, M., Hermans, H., ten Cate, T.J., Venekamp, R., & Boshuizen, H.P. (1996). The development of diagnostic competence: Comparison of a problem-based, an integrated, and a conventional medical curriculum. Academic Medicine, 71, 658-664. Stolper, E., Van de Wiel, M., Royern, P., Bokhoven, M., Van der Weijden, T., & Dinant, G.J. (2010). Gut feelings as a third track in general practitioners’ diagnostic reasoning. Journal of General Internal Medicine, 26, 197-203. doi:10.1007/s11606-010-1524-5 Victor-Chmil, J. (2013). Critical thinking versus clinical reasoning versus clinical judgment: Differential diagnosis. Nurse Educator, 38, 34-36. doi:10.1097/NNE.0b013e318276dfbe

Copyright © SLACK Incorporated

Teaching dual-process diagnostic reasoning to doctor of nursing practice students: problem-based learning and the illness script.

Accelerating the development of diagnostic reasoning skills for nurse practitioner students is high on the wish list of many faculty. The purpose of t...
73KB Sizes 1 Downloads 6 Views