Clinical Decision-making in Speech-Language Pathology Graduate Students: Quantitative Findings

Clinical decision-making, a product of critical thinking, is defined as a “contextual, continuous, and evolving process,” where data are “gathered, interpreted, and evaluated” in order to make an evidence-based decision (Tiffen, Corbridge, & Slimmer, 2014, p. 401). Practicing clinicians are continually engaged in the decision-making process as they perform differential diagnosis and provide treatment to persons with communication disorders. Indeed, clinicians’ decision-making skills are the foundation for the development and implementation of high quality clinical care (American Speech-Language-Hearing Association [ASHA], 2005).

Although the assumption has been that degree-seeking students will develop clinical decision-making abilities on their own, through coursework and clinical experiences, (Arum & Roksa, 2010; Crebbin, Beasley, & Watters, 2013), research suggests more explicit instruction is needed for novice clinicians to develop these skills (Ginsberg, Friberg & Visconti, 2016). While 99% of faculty endorse teaching critical thinking as an important goal of university education, a survey of public and private colleges and universities found that as few as 9% of instructors felt they taught critical thinking on a regular basis (Abrami, et al., 2015; DeAngelo, 2009). In response, Finn (2011) suggested requiring the critical thinking to be taught in professional training programs, declaring it a core skills of 21st century education.  As Abrambi et al. explained, students are more likely to learn how to think critically when they are taught the skills directly.

Literature Review

Experts and Novices: Knowing More, Knowing Differently

Research has long identified differences in clinical-decision making between experts and novices (Coderre, Mandin, Harasym, & Fick, 2003; Crebbin, et al., 2013; Ginsberg, et al., 2016; Norman, 2005; Tschikota, 1993).  Evans and Gadd (1989) asserted that experts in a given field of study not only know more in a given situation, but they also know differently; having internalized strategies to manage and evaluate information. Novices, on the other hand, require more exposure and explicit training to master thinking and action strategies. Tschikota (1993) published a clinical decision-making study of 19 nursing students. The findings showed that these novices, senior diploma nursing students, assigned equal importance to all pieces of data and made decisions based on factual information rather than hypotheses. Tschikota found, due to limited experience, novice nurses in the study processed information serially and in small amounts, relying on theories instead of schemas or patterns to help them select and use data in making decisions.

Another recognizable difference between a novice and an expert is the faster speed and greater fluidity of thinking; a result of pattern recognition that draws on previously stored schematic representations.  Experts use these patterns to make clinical decisions, eliminating the need to analyze each step and component as novices often do (Coderre, et al., 2003; Crebbin, et al., 2013). Further, advanced decision-making processes allow clinicians to work more efficiently in fast-paced clinical environments (Crebbin, et al., 2013). 

In the field of speech-language pathology, studies of novice clinicians suggest similar trends. Hill, Davidson, and Theodoros (2012) found that novice speech-language pathology  students demonstrated reflective skills focused on the process and content of clinical experiences; few students showed characteristics of deeper, more critical reflectors, such as considering the patient perspective and noting changes in their own perspectives. Ginsberg and colleagues (2016) employed a qualitative methodology to explore the thought processes of 15 SLPs with at least five years of experience and 15 novice speech-language pathology graduate students. The aim of the study was to identify the thinking strategies, or heuristics, used in diagnostic reasoning between these two groups. The results showed that experienced clinicians were more likely to engage in higher-order planning of specific assessment hierarchies, develop contingency plans for the assessment process, and make connections between the assessment process and treatment planning. These processes show that the experienced clinicians in the study were able to prepare for evaluations more efficiently and foresee how the evaluations would impact future treatment goals. The experienced clinicians engaged in the diagnostic process more holistically, using past experiences to prepare and implement the evaluation process. According to Ginsberg and colleagues, modeling connections between assessment data and implications for treatment is crucial to the development of diagnostic reasoning skills. With this insight, training programs can begin to implement intentional teaching practices that will foster development of prototypes and schemas for graduate students.

A Continuum of Development

In medical literature, the highest outcome of refined critical thinking, having the ability to make clinical decisions, is often described as a continuum (Arocha & Patel, 1995; Banning, 2008; Crebbin, et al., 2013). At one end of the continuum, novice clinicians rely most heavily on a slow, analytical and deductive approach for making decisions because of their lack of experience. At the other end, experienced clinicians have the ability to recognize similarities and familiar patterns in a fast and frugal process, requiring little mental energy and less time (Crebbin, et al., 2013).  Furze and colleagues (2005) described a gradual developmental process of clinical reasoning among students. Physical therapy students with beginner level clinical reasoning skills demonstrated a focus on self, compartmentalized thinking, and limited acceptance of responsibility. At this early stage in development, students viewed  each piece of information about a patient as being disconnected from other pieces making it difficult to select and synthesize important data.  Over time, students in the Furze et al. study began to incorporate information from the patient into their clinical reasoning. In their last semester of study, they demonstrated dynamic patient interaction and integrated situational awareness. At this stage, the students began relating to the patient’s background, integrating patient information into care plans, and modifying their interventions around the patient’s needs. Furze et al. found students in this advanced stage of the development process were flexible and able to change directions during evaluations based on interactions with patients, suggesting a higher level of clinical decision-making abilities. Students at this stage of training began to see the client and his plan of care more holistically, moving away from their initial view of clients as segmented parts and pieces. Similar patterns of development have been identified by researchers in the fields of medicine and nursing (Arocha & Patel, 1995; Banning, 2008). 

These findings have implications for teaching clinical decision-making skills to novice clinicians. Research literature suggested that exposure to knowledge, skills and strategies is not sufficient for the development of clinical decision-making (Crebbin, et al., 2013; Norman, 2005). It requires a variety of clinical experiences and “the opportunity for deliberate practice with multiple examples and feedback, to facilitate effective transfer of basic concepts” (Norman, 2015, p. 425). This suggests the necessity of hands-on clinical experiences to foster the development of students’ clinical decision-making skills. 

Purpose of the Study

Based on the previously discussed research, students in health professions demonstrate differences in clinical decision-making as compared to experts (Crebbin, et al., 2013; Ginsberg, et al., 2016).  These skills appear to develop over time (Arocha & Patel, 1995; Banning, 2008; Crebbin, et al., 2013; Furze, et al., 2015). It is proposed that these changes occur as a result of hands-on clinical experiences (Crebbin, et al., 2013; Norman, 2005). Yet some researchers contend that the development of clinical decision-making skills requires direct instruction in critical thinking (Abrami, et al., 2011; Finn, 2011). 

The current study serves as an exploratory examination of the types of changes in clinical decision-making specific to (a) formulation of hypothesis, (b) selection of appropriate evaluation instruments, (c) diagnosis, and (d) recommendations for therapy that occurred in a group of graduate speech-language pathology students, enrolled in academic and clinical coursework and practica, without direct instruction in critical thinking. Results will serve to guide future research in students’ development of clinical decision-making skills in graduate speech-language pathology programs. 

This study examines the following research questions:

1. Do speech-language pathology graduate students demonstrate changes in diagnostic clinical decision–making as they gain clinical experiences? 
2. If so, which specific diagnostic clinical decision-making skills evidence change?   

It was hypothesized that clinical decision-making skills would change over the course of study as a result of didactic academic coursework and supervised clinical practica experiences; without direct instruction in critical thinking. It was expected that students would require less time to complete case studies. It was further hypothesized that students would demonstrate changes in clinical decision-making across the four diagnostic skills examined. 

Method

Participants

Recruitment of participants was conducted with approval of the Internal Review Board (IRB) at James Madison University (JMU). Participants were recruited from a convenience sample consisting of 22 first year students enrolled in a five-semester speech-language pathology graduate program at JMU; the program is accredited by the American Speech-Language-Hearing Association (ASHA). A total of 11 students (50%) participated in this study following submission of informed consent. Participation was voluntary. Participants had the opportunity to win a gift card in the amount of $25 for completion of each case, but no other compensation was offered. In accordance with the JMU IRB protocol, the researcher was not permitted to recruit students. A research assistant, a peer of the participants, conducted recruitment activities, which may have been a factor in the limited participation. Additionally, it is suspected that the commitment to participate over three semesters may have limited participation. 

While 11 participants started the study, only eight participants (73%) completed both the survey and case study portions of the study at all three stages and were included in data analysis. Of the three participants who did not complete the study, one completed only the initial case and survey and two additional participants failed to complete the final case and/or survey. 

Because of a lack in male enrollment, all eight participants were female; a representative demographic of the communication sciences and disorders field. At the initial time of the study, participants completed 27 credit hours of graduate level coursework with grade point averages (GPAs) ranging from 3.30 to 3.94 on a four-point scale (M = 3.66, SD = .19). Directly related to the study, students had completed three graduate courses (nine credit hours) in child language disorders, child phonological disorders and phonetics. At the start of the study, the total number of clinical hours completed by the participants in the university clinic ranged from zero to 61 (M = 45, SD = 20.5), and the number of child assessment hours ranged from zero to 30 (M = 13, SD = 9.9).

Measures 

Case Simulations. Participants were required to complete a total of three web-based pediatric case simulations created for this study.  One simulated case was of a child with an articulation disorder, one with a phonological disorder, and the other with typical speech and language. The web-based case simulations were developed, housed and accessed by participants through DecisionSim, a hosted, secure software service that allows for the creation of multi-media simulations. Each student participating in this study was assigned a unique username and password to the DecisionSim website. Once logged into the case simulation, students were given a comprehensive case history of a patient: the reason for referral, background information, a video example of connected speech, and test results. DecisionSim recorded the number of steps and time in seconds required to complete each case. 

Participants were then asked to provide text-based input on a series of questions related to the simulation. The prompts were as follows:    

1. Identify the three most important pieces of information (free response). 
2. Identify initial impressions (free response). 
3. Select tests to administer to the patient (multiple choice). 
4. Select a diagnosis (multiple choice). 
5. Make recommendation for treatment (yes, no). 
6. Determine referrals needed (multiple choice). 
7. Create treatment goals (free response).  

These prompts were identified from requisite knowledge and skills in the area of evaluation as outlined in Standard V-B of ASHA’s 2014 Standards for Certificate of Clinical Competence in Speech-Language Pathology (Council for Clinical Certification in Audiology and Speech-Language Pathology of the American Speech-Language-Hearing Association, 2013). 

The case simulations were parallel in complexity and structure. All three cases were built employing identical branching and node structures. That is, all cases presented information and prompts in identical manner. A post hoc analysis of overall student accuracy and seconds per step to complete cases suggested equivalency of cases (M = 42, SD = 7.22, M = 153, SD = 31; respectively). To determine accuracy of responses, five clinical educators with at least five years of clinical experience, completed each of the cases online and provided answers to each prompt. A research assistant compiled the responses. The researcher and clinical educators met as a group to discuss each set of responses. Correct responses for each of the cases were determined by group-consensus method; that is all clinical educators were in agreement in order for a response to be deemed correct.  

Online Survey. The students who participated in this study were also required to complete an online survey providing the following information: their academic status (GPA), number of completed direct clinical hours, coursework completed, and confidence levels in their knowledge and skills in the area of assessment outlined in ASHA’s 2014 Standards for Certificate in Speech-Language Clinical Competence in Speech-Language Pathology.  The survey required a four-point Likert scale response indicating level of confidence (i.e., self-efficacy) in each of the 21 clinical skills presented.  Participants completed the survey at the initiation of each case simulation, for a total of three times.  The purpose of the online survey was to probe for factors that might account for changes in clinical decision-making.  

Research Design

Participants completed two measures (case simulation and online survey) at three stages in their five-semester graduate program. The first stage occurred after completion of the first semester of coursework, which included graduate level courses in phonological disorders, child language disorders, and the diagnostic process. The students had also completed  their first clinical assignment in the university clinic with an average of 45 clinical hours (SD = 20.5). The second set of data was obtained after the completion of three semesters of graduate study. At this point in the program, students had completed 79% of coursework and three practica at the university clinic. At this stage, participants reported a mean GPA of 3.60 (SD =.18) and ranged in clinical hours from 54 to 171 (M = 114, SD = 38). The third set of data was completed after the fourth semester of graduate coursework; this stage followed completion of all coursework, as well as, an off-campus clinical practica (three or four days per week). At this time, students reported an average GPA of 3.70 (SD =.14) and ranged from 115 to 356 clinical hours (M = 273, SD = 74.60). This time frame was determined to be the final data point because the researchers were concerned that students, enrolled full-time in off-campus placements, distanced from the program and close to graduation would not elect to participate in the final phase of this project if it were extended to the end of the final semester. 

The order of case simulation completion (i.e., a child with an articulation disorder, one with a phonological disorder, and the other with typical speech and language) was randomized across participants. Participants were instructed to complete the DecisionSim simulated case and online survey at a time and location convenient to them within a given two-week period.