Investigating Assessment of Clinical Performance: Feasibility of Adopting a New Tool
Student clinical performance assessment has been instrumental to physical therapy (PT) education for many years and has evolved to include several tools: Clinical Performance Instrument (CPI), PTMacs, Clinical Internship Evaluation Tool, and Performance Assessment System (PAS) among others. Historically, the profession encouraged a single tool to lessen variability, but over time, increased variability has occurred. As variability has increased, validity and reliability have been investigated for some developed tools. The purpose of this study was to investigate feasibility and concurrent validity of the PAS.
Physical therapists serving as clinical instructors (CI) were recruited to complete the PAS and CPI during midterm and final evaluations of a terminal clinical education (CE) experience. Students of participating CIs were also recruited. A follow-up survey was completed with participants rating usefulness, ease of use, ease of learning, level of satisfaction, and ability to rate performance. Time to complete the PAS and open-ended responses were gathered. Ratings on identical or similar categories on the CPI and PAS were correlated using Spearman Rho; survey data was evaluated with descriptive statistics.
A total of 27 CIs and 20 students participated; 87 assessments were analyzed. Correlational analysis shows strong correlations (rs > .75, p < .01) between the PAS and CPI on several items including safety, procedural interventions, teaching and patient education, examination, administrative/management, and documentation. Time in minutes to complete the PAS compared to the CPI was decreased by half. All survey respondents indicated the PAS was easy to use and assessed strengths and deficits in clinical performance. Open-ended responses regarding the PAS highlighted less redundancy and improved ease of use.
Conclusions/Relevance to the conference theme:
Developing a diverse and culturally competent physical therapy workforce, ready to provide optimal patient care, requires measuring competency with valid and reliable clinical performance assessment tools. The PAS demonstrates fair to strong correlations on many of the clinical performance categories related to psychomotor and cognitive domains. Less correlated are categories falling under affective behaviors. The weaker correlations in the affective domain could be due to the general difficulty with assessing students in this domain rather than a problem with the tool. Academic programs considering changing tools should consider conducting similar processes to make informed changes and enhance training of students and CIs on tool use. The PT profession is primed for greater and deeper discussion on student performance assessment and larger studies need to be conducted. Continued variability and expansion in the number of performance assessment tools used in PT education hinders the professionÕs ability to identify best practice in CE. Clinicians and future employers must weigh in on the essential elements that should be measured. It is time to not alter, but to recreate.