Status: Refereed paper
Clinical Decision Making: Issues in Teaching
The evolution of nursing as a discrete profession with a distinctive knowledge base and orientation towards health care, has emphasised and increased the importance of clinical decision making. Implicit within clinical thinking are two types of knowledge: theoretical knowledge and applied knowledge as it relates to the clinical setting.
The theoretical component is articulated in many texts but unfortunately the component derived from practice - the knowing how, has been found mostly in the "folk" milieu of the practice setting. The intrinsic nature of this knowledge has lead to an inadequacy in the record of the clinical experience. Benner's work in 1982, little had been written and the impetus for action remained hidden as tacit knowledge.
Teaching the processes underlying clinical decision making has always proved difficult. While many frameworks have been devised to simplify the steps and make the process more explicit, the problems seem to continue. This leaves us to ponder on the issues surrounding the teaching of clinical decision making with a subsequent challenge of being able help integrate students' clinical decision making ability into the practice setting.
Many students have difficulty articulating and understanding the decision making process. Compounding the problem is the expectation that students will gain an understanding of the theories underlying the concepts and will apply these in the practice setting. Consequently, if the student has not grasped the concepts and is unable to reflect on the action taken in practice, much of the knowledge of decision making processes will remain obscure and confused. An added complication in nursing is the fact that traditionally and to assure patient well being it is necessary for the nurse to be competent enough to engage in the diagnostic reasoning and treatment decisions in at least two domains - nursing and medicine (Carnevali 1984)
The combined domains of decision making used in nursing greatly affect the clinical reasoning process. The nurse must make considered diagnosis in both domains with the added complication that in the biomedical domain, data are often not as well defined. This is because the nurse is either not privy to or is unable to gather all data required to make a definitive diagnosis. However, as necessary as it may be for nurses to make decisions in both domains, it seems folly to teach beginning practitioners the decision making steps of another domain before they are cognisant with their own.
Nursing - Decision Making in Two Domains
Nurses identify and solve client problems in the nursing domain as well as being aware of, identifying and carrying out treatment (under medical supervision) of client problems in the biomedical domain. Carnevali, in 1984 stresses the fact that by tradition and training, nurses have been biased to direct problem solving primarily and explicitly towards the biomedical domain. This may have been, but with the shift to holistic care and a changing paradigm in health care, nurses make appropriate and cost-effective choices about when to direct their decision making and judgement to daily living as related to health. These daily living requirements are central to the domain of nursing practice.
The Domain of Nursing Practice
2.1.1 The Domain of Nursing Practice
The Nursing practice domain consists of many subdomains. (Benner 1984, p.46) lists these as:
These sub-domains portray the nursing responsibilities but both domains of clinical decision making are commonly used in clinical practice settings.
Combining the Domains
In the biomedical domain nurses have delegated responsibilities to make accurate and appropriate clinical judgements on the patients' pathophysiological health status. However, the nursing domain focuses on daily living, including the environment within which daily living takes place and functional health status of the patient. These elements are viewed in their relationship to each other and are the two major components of diagnostic reasoning in nursing (Carnevali 1984).
However, nurses do not only work in institutional settings. Those nurses working as independent practitioners, in industrial clinics, home care or midwifery make decisions on referral for medical diagnosis and treatment. They use clinical judgement when recommending continued self care, independent nursing action or retention of the client under nursing management.
Nurses do not for the most part engage in systematic scanning of body systems, as used in the biomedical domain, but rather take a holistic view of the client. For the nurse in an institutional setting, the environment in which the data are gathered and the pre-encounter data also helps to shape the search field. If the encounter is in an antenatal clinic, for example, the nurse would be privy to the information gathered previously. She/ he would probably have prepared and formulated questions before she encountered the woman based on expectations of the ongoing pregnancy. On the other hand a nurse in this setting would not be prepared for an encounter with a male patient suffering from a karate accident.
However, in both domains an early step in data organisation, which will enable problem identification, is to use the first recognised cues. These may be few and indefinite but they can activate a tentative, diagnostic hypothesis that may explain what is observed. It seems that the effectiveness of this step varies with experience of the person for example beginning practitioner or expert and is not specific to any practice field.
Diagnostic hypothesis activation follows cue recognition and is often a conscious, logical, critical-thinking manoeuvre although it can occur without conscious effort. Because of the tentative nature of early hypotheses it is important to avoid the manipulation of data to fit an inaccurate or imprecise diagnosis (Carnevali 1984) Subsequent cues may present a challenge to rethink the hypothesis - although research has shown a strong tendency for clinicians to ignore these cues Carnevali, 1984: Cox, 1993).
This phase of using data as the basis for generating explanations is a crucial one. Clearly, problem labels of diagnostic classifications that are not considered cannot be tested. Cox (1993) and Carnevali (1984) articulate this by simply saying that clinicians make right diagnosis because they think of them! The success of and the proficiency in the remainder of the assessment-diagnostic process rests on the quality of this first and recurrent step in the process.
After hypothesis generation, the process must begin refining (convergence) and revisiting (divergent) processes involved in evaluating the hypothesis. All possibilities generated must be included in this evaluative phase. The final stage of this process is to select a diagnostic classification as precisely as the available data will allow. This classification coupled the data then becomes the foundation for decisions about prognosis, goals, treatment plans and activities (Carnevali 1984).
The use of probability statements is commonplace in medicine with these statements having been derived from systematic observations on many individuals over a substantial period. However, one may argue that the notions of probabilities, especially when they are expressed in a mathematical form are meaningless in nursing clinical judgements. It is true that probability statements are best derived from systematic observations from many individuals, as the medical profession has done. It is also true that nursing lacks a sufficient empirical data base for precise expression of probabilistic relationships at this time. Nevertheless, it is argued by students of clinical decision making, that both the diagnosis and prognosis rest on at least the informal assignment of probabilities to clinical data (Tanner 1989)
Practising in the Combined Domains
As stated before, the combined domains of biomedicine and nursing used by nurses greatly affect clinical reasoning. To illustrate this the following example put forth by Tanner (1984) has been reiterated. This incident has its foundation in the biomedical domain when the nurse learns the signs and symptoms of hypoglycaemia, secondary to short acting insulin. The indications include changes in behaviour, irritability, cold clammy skin, tremulousness and diaphoresis. Students also learn that the causes of such a reaction are delayed or omitted meals, excessive exercise or insulin overdose. So, if this patient displays behavioural change, the nurse (based on this knowledge) would hypothesise a hypoglycaemic reaction.
The cue of behavioural change was an indicator but this may also signify other diagnoses. Therefore other cues are gathered and as these are recognised as positive for the original diagnosis so is it strengthened. If other factors were present such as the missed meal then the diagnosis becomes almost certain. In reality the only cue that will give a positive diagnosis is a blood sugar level. The diagnosis itself however, was confirmed through accumulation and informal revision of the probability hypothesis with each piece of data.
This example shows the nurse's dual domains of clinical reasoning and highlights two points. First, the recognition of probabilistic relationships between cues and diagnosis, which increases the thoroughness of data collected and so improves the diagnostic accuracy. Secondly, this example emphasises that the nature of the data available to the nurse both in the biomedical domain and especially in the nursing domain, is often unreliable. Nurses must rely almost entirely on their own perceptual process, and need to be skilled in the use of their senses because most often the precise measurement and instruments are not available. Therefore, the nurse's task is shown to be cognitively quite complex.
The combined domains also have the potential to cause conflict for the nurse practitioner. One of the nurse's most controversial roles and one of the hardest decisions she/ he will take is that to become a patient advocate when the patient does not want the medical treatment. In this situation conflict arises between the practice domains' of nursing. In the biomedical domain of clinical reasoning the nurse may feel the treatment decision is vindicated but in the nursing domain from a helping perspective it is not. While, in some cases, working in two domains may lead to personal conflict, the difficulty explaining and teaching the process is also compounded by these factors (Conrick 1994).
Much work has been undertaken in an attempt to explain and teach Clinical Decision Making in nursing and other practice disciplines. Many attempts have been made to use paper simulation and more recently computer simulation to teach the process. Mattingly, (1991) described clinical reasoning as largely tacit and highly imaginative. He also sees it as a deeply phenomenological mode of thinking because one can only draw meaning from such complex reality through careful analysis of narrative subjective material. It is also argued that clinical reasoning involves more than the ability to offer explicit reasons that justify clinical decisions because it is also based on implicit understanding and habitual knowledge gained through experience. These factors reveal a very complex cognitive process and one that is critical for the student to grasp and become skilled in its use. Cost-effective, safe nursing requires accurate clinical judgement and decisions from the nursing domain as a basis for treatment, as surely as in any other health discipline.
A fact often overlooked is that the client actually hires health professionals, either directly or indirectly for their expertise in both diagnosis and treatment. Distinctive discipline specific perspective and expertise is therefore crucial. Nursing has attempted to simplify and clarify the clinical decision making process and many frameworks have been articulated. However, because of the complexity of the process, the task of teaching students clinical decision making remains a substantial challenge.
Frameworks for Decision Making in the Nursing Domain
The trend in nursing practice, especially in North America and to a lesser extent Australia, has been towards the formation of a taxonomy for clinical decision making. It is difficult however to find the term that covers the entire process. Dick (1991) is probably correct when he says that there does not seem to be a general term in common usage that covers the processes of planning, problem solving and decision making. He uses the term "problem solving" to cover all three processes.
In nursing education circles, a common working definition for problem solving is that it is the process used to resolve or answer a proposed question or achieve an answer to a client need (Pinell and de Maneses, 1986). Roy (1980) describes problem solving as helping within the nursing process, whereas Orem (1985) discusses problem solving as the means by which clients strive for self care.
In Australia, the United States of America, Canada and the United Kingdom, clinicians resist the rigid use of frameworks such as the Nursing Process Model regarding them as unrepresentative of the process by which experts make decisions. However, theorists and many educators see a great value in such frameworks for guiding the novice and student towards an understanding of the process underpinning problem solving. In many quarters, it is argued that this framework is the way of the future with nursing information systems likely to incorporate these types of nursing frameworks. Indeed much work has been carried out in the United States towards this end.
However, it is the express purpose of nursing to identify and solve client problems in the nursing domain, whether it is by using a problem solving approach, problem statements or the Nursing Process Models. All these approaches are valid methods for problem solving in nursing education and clinical practice. Although practitioners have been slow to embrace the Nursing Process Model in their practice, this has not been the case with theorists and educators. According to Klaassens (1992) most of the models and strategies described the literature use the nursing process as a guide or reinforcement According to Koch and McGovern (1993), since the introduction of the Nursing Process, difficulties have been encountered by both nurse educators teaching this subject matter and practitioners trying to rationalise the process. An identified lack of knowledge of the diagnostic task, and subsequent recommendation that educators address the issue, were made by Kelly (1986) and supported by many others (Grier 1981; Field 1983). These authors speculate on the cognitive strategies used by nurses. Field (1983) saw community health nurses frequently collecting data, but failing to examine the relationships. Consequently, they were seen to make poor judgements or they did not arrive at appropriate conclusions. However, Grier (1981) thought that the nurses might have lacked formal education in the necessary cognitive strategies. It is pertinent then to discuss the methods used in nursing education and how well these approaches prepare students in the skills of clinical decision making.
Teaching Methods for the Classroom
There are still few schools using the more liberating structures of contemporary curricula such as Problem Based or student centred and integrated learning. This then leaves students struggling to grasp the concepts of problem solving with process taught in didactic sessions with practice being gained on clinical placement If students have not grasped the concepts of the problem solving in a controlled setting (the university ) it is unreasonable to expect the application of these concepts on clinical practice. Finding the most appropriate teaching strategy which encourages deep learning is the challenge facing nursing educators.
Outcomes of learning are dependent on the teaching approaches utilised with (Ramsden 1992) describing these approaches as the fusion of two parts- the "How" and the "What". The "what" is seen as "actively trying to understand or passively trying to reproduce", while the how refers to the way in which the student structures the task.
A student can structure or organise a task by embracing the entire task and in so doing maintain the structure and use a holistic approach. However, it is also possible for the student to dissect the task into component parts, distorting the structure and focusing on the segments instead of the whole task. This Ramsden (1992) refers to as an atomistic approach to learning. The meaning derived from the task is also divided into two branches - surface learning and deep learning. Surface learning is described by Biggs (1989:10) in the following manner:
Knowing facts and how to carry out operations might be part of the means for understanding and interpreting the world, but the quantitative conception stops at the facts and skills. A quantitative change in knowledge does not change understanding. Rote learning scientific formulae may be one of the things scientists do, but it is not the way scientists think (Biggs 1989, p.10)
When discussing surface learning Ramsden describes it saying that "surface learning at its best is about quantity without quality" (1992, pp 45). On the other hand, deep learning is seen as gaining an understanding and focuses on meaning. It is this deep learning coupled with an holistic structure that Marton and Saljo (1984) regard as the only way to understand learning materials.
Common Teaching Methods Used in Nursing
In nursing education it is expected that students will be able to combine technical skills and knowledge and apply these to the clinical setting. This knowledge includes decision making frameworks to underpin nursing care plans and therefore the nursing care that students give while on clinical practice. Students require an understanding of the processes and must have the ability to apply the concepts and to implement safe plans while in the clinical area. To enable this, it is the deep holistic approach to learning which must be implemented. Four specific teaching methods have been used in nursing to achieve this - the lecture, the seminar, guided design and clinical teaching (de Tournay and Thompson 1982). These could be termed the traditional methods of teaching. There are also those processes of contemporary curricula design which are based on clinical decision making skills and processes which are beginning to gain greater prominence. Then, there are also the methods termed nontraditional, for example, multimedia and Computer Based Learning (CBL), which are becoming increasingly acceptable as teaching and learning tools.
The aim in teaching is to enable the students to relate to the material in a purposeful way and to understand and apply knowledge. The teacher in doing this will facilitate a deep learning approach to the materials. But, when looking at effective teaching, the lecturer should not be scrutinised alone. Effective teaching and the strategies used, are tied very much to the curriculum base and philosophy that underpin the curriculum processes. Do the curriculum base and philosophy of the school allow students to actively find knowledge, interpret their results and test hypotheses against reality? Is it flexible enough to allow cooperative and/or individual inquiry? Does the learning environment support and enable understanding and retention of factual knowledge as espoused in the deep holistic approach to learning?
The deep holistic approach to learning necessarily considers areas such as keeping students interested and motivated in the learning process. This can be accomplished in different ways but with varying degrees of success again depending on the curriculum base and the teaching strategies. As optimal learning must be the goal of instruction, high levels of learning, interest and motivation must be stimulated and maintained. Interest is necessary for arousing curiosity and holding attention while motivation stirs learners to persevere in the pursuit of a goal and gives them direction in learning.
There has been much discussion in academia about the need to specify goals, aims and objectives for courses of all kinds. The importance of ensuring that course aims are consistent with general goals and that the more specific objectives are consistent with the course aims has also been stressed. Contemporary curricula also demand recognition of the learner goals and expectations and objectives. So, for all objectives to be satisfied the correct teaching method must be implemented. Other considerations, when selecting a teaching methodology are entry level, prior learning and the background of the student. Culture, language and age of the student will also affect the effectiveness of one's approach to the teaching task. Cultural differences may dictate the way in which the student learns and the language barrier, the ability of the student to understand what is being said. The age of the student may set an expectation for the learning experience and if ageing affects the adults' capacity to learn, the rate of learning may also be affected. A teacher must be aware of these student differences and has to work hard to keep students motivated and interested and to find strategies which will enable deep holistic learning to take place.
Motivation is complex and consists of both intrinsic and extrinsic forms. Unilateral control and assessment extrinsically motivate students to learn and work as do external rewards and punishments, (Heron 1981:81) regards this as the wrong sort of motivation which can "breed intellectual alienation". It is the intrinsic motivation, born from the student's own interest and thirst for knowledge, the competition with oneself to attain excellence and the fertile discussion and debate with one's self and peers, which the teacher must work to achieve.
Students who use the surface approaches to learning, fostered by the traditional approaches to education are seen to be more extrinsically motivated while the deep approaches foster intrinsic motivation. Very often it is the approach to study which is at the crux of the students "inability to study" rather than a lack of motivation. However in nursing, although the trend is changing, the traditional methods with their extrinsic motivation abound. In these situations the teacher needs to use strategies which will encourage students to participate and become active in their own learning encouraging intrinsic motivation.
Of the traditional methods, the lecture is the most readily recognised teaching method and is perhaps the oldest strategy for teaching large groups. Many teachers favour lecturing because it is time and resource efficient and takes less time than discovery or experiential learning. Lecturing does give the teacher control over the content to be learned but the lecturer must be well prepared and the lecture well delivered to be successful. deTornay and Thompson (1987) relate that there is evidence showing other methods of teaching to be more effective than expository techniques when achieving higher cognition and attitudinal objectives. The classical lecturing situation creates an environment which has the potential for students to become the passive receivers of information. Ramsden (1992) sees students in a lecture as dependent on the lecturer and says that this is the exact opposite to the environment necessary for deep learning to occur.
In the 1950s, Knowles began to identify strategies that would work best for adult learners. He found that social interactions are relevant to learning for these students. He also believed that adult learners were selfdirected in their learning which was based on self identified needs. Therefore, he proposed concepts of adult learning that were congruent with the Gestalt Theories (cited in Knowles 1950) and Dewey's work (1920). Subsequently, seminars combined with lectures became an accepted format and these are still used extensively in universities. These gave a new dimension of lecture and small discussion group to the educational scene. It also gave students time to interact with each other and their lecturer. This approach to learning heralded the more student centred approach to learning and has been used in universities for many years.
Despite these developments, Wong (1979) revealed that many students who received classroom lectures are still unable to relate that knowledge to the clinical practice setting. Therefore, experiences that provided opportunity for students to scrutinise realistic patient situations and problems are necessary. These situations are set up in the laboratory setting or another non-threatening and ethically safe environment. To this end guided design, a process orientated teaching strategy, is becoming more frequently used in nursing education.
Guided design considers the fact that learning reams of facts and memorising pre-established answers will not help the student to deal with "real life" situations and complex problem solving. Wales and Stager (1977) developed the concept of guided design which they describe as an organised problem solving format, integrating the teaching of subject matter while developing decision-making skills. In 1979, and Hageman developed this technique further as a strategy for use in nursing education. They see this method as utilising many learning theories and say that students are actively involved in their own learning, they receive immediate feedback and simultaneously develop both cognitive and process skills. Guided design does not simply transmit knowledge but encourages the move to the realm of deep learning. This could be argued as the basis for the contemporary curriculum approaches.
Regarded as a specialised field, clinical teaching differs in many ways from classroom teaching. Clinical teaching provides the student with the opportunity to experience but remain supernumerary in the "real situation" with "real clients". They have the time and opportunity to observe the clinical setting without direct responsibility and to be socialised into their profession gradually. It also allows the student to integrate and apply those principles, theories and concepts learned in the classroom or laboratory setting and to reflect on the processes.
The clinical experience must be well planned and integrate the concepts of relevance, be interesting and motivational. Balla (1990a), describes the situation in medical education where the knowledge taught in the pre-clinical years is often not integrated into the clinical ones. He says that students perceive the two areas as unrelated and have problems applying conceptual knowledge to clinical problems. Ramsden (1992 p 164) refers to this as the familiar "concomitants of surface - atomistic approaches to learning". However, a holistic deep approach to on-campus learning coupled with integration and properly managed clinical experience has the potential to become the complete learning experience.
Much of the traditional learning has taken place in large group settings, but there is a plethora written to suggest that each student learns at an individual rate (Briggs 1977: Dick and Carey 1985: deTorney, 1982, White and Ewan 1991). Many authors have also described the learning environment and have advanced criteria they regard as enhancing learning. Hunkins for instance, describes significance coupled with high interest as one criterion, in which significance "refers to the essentialness of the content to be learned" (1980 pp 221). He also asserts that principles of learning and motivation that suggest holding student interest are critical in making learning more productive. Bevis (1982) sees high student involvement as an exciting, vibrant, alive environment for learning. Many educationists also acknowledge that previous learning experiences and student interest in the task will determine a student's approach to learning. Ramsden in 1992 added to this saying that deep learning approaches, related to students' interest in the task for its own sake, is associated with a welldeveloped knowledge base in that field of study.
Ramsden (1992) also relates that in all these areas, the use of computers in education has many advantages and a great deal to offer. When applying the use of technology to learning Laurillard (1987, 1988) says that theoretically computers should provide an intensive and relevant encounter with learning matter, they should be individualised and self-paced, allow immediate access to large data bases, test understanding, and provide guidance when mistakes or misconceptions are noted. If the courseware is designed and written to support these specifications then the potential exists for high level interaction and deep holistic learning.
"Teaching" clinical decision making and problem solving while encouraging learning is sometimes seen as a complex undertaking even though all students have been engaged in problem solving behaviours from their preschool years. It may be that instead of teaching the complexities of problem solving educators are really surfacing and embellishing a talent which already exists in their students. After having achieved this they move on to assist the learner to capture the specifics of techniques used in clinical decision making by nurses. However, "teachers" in nursing do have varied outcomes to achieve and are faced with a range of learners' ages, the individuals past learning and life experiences, their individual approach to solving problems and the dynamic nature of the content to be covered.
There are many ways of approaching the surfacing of such a complex and individualised cognitive skill but this it difficult to imagine that any one teaching method could be sufficient for instructing all students and for their gaining some level of competence in clinical decision making.
Last modified on: