Clinical reasoning is essential in practice-based disciplines, in the way critical thinking be applied in clinical situations [17]. Clinical reasoning is the clinician's main task in diagnosis and treatment [25, 60, 61]. Developing clinical reasoning skills is a critical part of a bigger, unified identity that learner will need to bring to clinical experiences in order to participate in caring for patients and work in teams [60, 61]. Although clinical reasoning is required to clinical practice on a daily basis, it is not directly taught during didactical stage, but is given deliberately through clinical practice [2,8,62−64]. Previously, dental education emphasized on how to think critically during clinical decision making; what, when and how to determine dental procedure for oral complaint. Therefore, the portion of learning biomedical knowledge during didactic stage is less than clinical procedural or instrumental knowledge [65–67]. According to Academic Guidelines of Faculty of Dentistry, University of Gadjah Mada, from the total curriculum credits, approximately twenty percent are dedicated to basic knowledge and the other twenty percent to biomedical knowledge along with behavioral science. All the teaching and learning processes take place in the early stage of didactic educational. Later the teaching and learning process emphasizes on procedural knowledge and practicing to train psychomotor skills [68, 69].
The clinical findings in hypothetical clinical cases related to the scope of oral medicine, one of the clinical sciences of dentistry. Previously the course starts from semester 5 to semester 7, but now started on 4th semester in the newly integrated dental curriculum. The initial course topic is Oral Diagnosis, followed by Diagnosis of Oral Disease, Treatment of Oral Disease, and Dental Management of Medically Compromised Patients. All learning materials require good understanding and comprehension of basic science and biomedical knowledge. In addition to routine practicum, to implement 'early clinical exposure', educator apply student-based learning, or specifically refer to case-based learning. Learners in groups are asked to solve clinical problems from hypothetical cases, recap it as a concept map for easy understanding, presented and conduct a discussion. This condition considered as naturalistic experiment of qualitative studies, the commonly research design used for studying clinical reasoning process. Situations where the intervention occurs naturally without planning refers to naturalistic conditions, resembling experimental requirements [70].
There were varied responses from participant in this study after they were skimming the initial information from the hypothetical clinical case. Two of them determine several probable hypothesis or diagnosis, and they regarded on applying a hypothetico-deductive pattern of clinical reasoning, and the rest were presented a forward-reasoning pattern. The clinical reasoning is a problem-solving process commonly used by clinicians. The effectiveness and efficiency of applying it determines how well a clinician's knowledge when providing patient care [2, 71]. Key elements of clinical diagnostic reasoning consist patient’s story, data obtained, accurate presentation of the problem, generation of hypotheses, search and selection of scripts from memory and determination of diagnoses [72]. It is assumed that memory function involves the use of abstract cognitive structures [44] and knowledge is organized in elaborate memory network [46, 73]. According to structural paradigm, medical knowledge is organized and store in memory as mental representation of diseases. At least four distinct models of mental representations have been proposed: prototype, exemplars, semantics networks and illness script. Illness script is particular sets of symptoms or phrases become associated with particular diagnoses. The illness script connected to semantic qualifiers to form problem representation, that activates clinical memory and allows knowledge to became usable for diagnostic reasoning. Semantic qualifiers used to describe clinical cases as descriptors, which is usually paired and used to compare and contrast clinical phenomenon. It associated to certain disease [2, 38, 44, 46, 74, 75].
On acquiring expertise in medicine, learners progress through several transitory stages, characterized by distinctively different knowledge structures. The structure of knowledge in learner’s memory naturally develops from beginner, intermediate to expert. The learner gains a lot of basic and biomedical knowledge in the early stage of didactic year. These concepts are linked together in a knowledge network, then gradually, more concepts are added, refined and more formed sophisticated knowledge network, with increasing length of study [13, 33, 40]. At every stage, the clinical reasoning process can be characterized by a line of reasoning consisting of a chain of small steps based on detailed biomedical knowledge. Initial development of clinical reasoning process marked by reduced knowledge followed by dispersed, elaborated causal, scheme and script [33] (Figure 6).
The chain of steps as presented in Figure 6 resembles the level of knowledge structure assumed by the concept map in the SOLO taxonomy (Table 2). Learner with ‘pre-structural level of knowledge’ seems to have scattered bits of information obtained but in unorganized structure and less comprehension of knowledge [50]. The pre structural level is seem equivalent to the first steps of the development of knowledge structure. At this stage, learner with ‘reduced knowledge’ presents little knowledge about disease names and their manifestations and shows to guess when solving problem [33]. Chan [58] describes ‘wild guessing' as one of the explanations for pre-structural level of the SOLO Taxonomy. Based on this, it can be assumed that participant #1 applies guessing when determining the hypothesis. This is proved with extended time, more questions and statements of nescience showed by participants during the interview. The participant #1 has an attempt to recalled some textbook-related clinical fact, looking at available clinical data and trying to match the hypothesis or diagnosis that has been established. The clinical reasoning pattern applied is closer to ‘deductive or backward-oriented reasoning’, because the initial statement is more general in nature and moves towards a specific conclusion. The addition of new information led to a certain diagnosis [76, 77]. A similar pattern is also found in study by Nafea [55], where deductive or backward reasoning is applied when novice learner solves clinical problem. The backward reasoning pattern is regarded as characteristic of novice learner’s reasoning [28, 78].
By the end of the first stage of knowledge acquisition, the initial knowledge network built that allows learner directly connect the lines of reasoning between different concepts within network. In this stage, learner will have ‘dispersed knowledge’, which mean they have less knowledge about disease, such as knowing the name of many diseases, but few manifestations about each disease [13, 33]. It is similar with ‘the uni-structural level of knowledge’, where the learner manifests a correct grasp of one or two relevant pieces of information obtained directly from the problem but lack of appropriate relations to each other (Chan et al., 2002; Bakouli and Jimoyiannis, 2016). The learner at this stage commonly uses hypothetical deductive reasoning when solving problem [33]. Participant #3, who showed have these structure and level of knowledge, applies an ‘elaborated model of hypothetico-deductive reasoning’, which learner reasoned by first generating few hypotheses very early as soon as the first pieces of data became available and then testing a set of hypotheses to account for clinical data. The learner selectively collected data focusing only on the relevant data and applied hypothetico-deductive process in the end [79].
With increasing knowledge during skill development, learners would demonstrate ‘elaborated causal network’ which has rich knowledge about diseases and their manifestations with detailed cause-effect links exist in memory. Furthermore, they can clearly explain the causes and consequences of disease in terms of general underlying pathophysiological process [13, 80]. The structure of knowledge level 3 (multi-structural) SOLO taxonomy can represent end stage of ‘dispersed knowledge’, while the structure of knowledge level 4 (relational) represents early stage of ‘elaborated causal network’. Both of them can shows slight connectivity between one or two aspects, even they were still novice. The multi-structural level of knowledge has an incomplete understanding obtained due to focus on several relevant concepts, and seems recognize the relation between them but without further elaboration [58, 81]. Meanwhile, the relational level of knowledge shows an understanding of various concepts and their integration to form a coherent structure [50]. Participant #2 who have multi-structure level and participant #4 who have relational level of knowledge, both applied ‘inductive or forward-oriented reasoning’ when solved the problem. A forward-oriented reasoning pattern is reasoning strategy from data to hypothesis. The initial information as reflected on chief complaint does not immediately lead to the formation of a hypothesis or diagnosis, but needed more clinical data are required to determine it. Later, one by one the new clinical data obtained leads to the generation of hypotheses, and by adding new clinical data and considering its relevance, the possibility of a diagnosis can be made [28].
Clinical problems solving is not easy task for under-graduate dental students. As health professional, they must have and could demonstrate competence in clinical knowledge and skills. Given the limited previous clinical experience with real patients, dental undergraduate students were naturally concerned with symptoms and signs as it told by patient or written medical record or hypothetical case, later recall and retrieve their knowledge of physical manifestations of diseases from memory [24, 72]. Basically, theoretical conceptual knowledge has been obtained during didactical stage, that take places in the first four-five years. The teaching and learning process is largely based on textbook, hence the resulting perspective on disease is prototypical, with only limited comprehension to the variability disease manifests in reality [28]. Prototype refers to organization of clinical categories in memory around particular exemplars and serve as anchors for other members of the category [46]. On the other hand, clinical procedural knowledge taught is limited to theory and simulation during practicum [33, 82]. This fact can explain that only some characteristics of the knowledge structure from the stages of clinical reasoning development were found in this study, but no ‘scheme’ and ‘script’. Both of this level of knowledge structure develops when learner has increased their expertise, whether they intermediate or expertise. Clinical procedural knowledge along with clinical experience with real patient have a major role in developing expertise [8, 82].