While many factors elicited during the admissions process have been used to predict academic success in physical therapy curricula (1-3), very little research has focused on factors that may predict clinical performance (4). A study on positive graduate professional performance traits (5) indicate that communication, active learning, professionalism, responsibility, ethical behavior and critical thinking are highly important. These traits happened to be areas measured on the standard Physical Therapy Central Application Service letters of recommendation (PTCAS-LORs). We examined whether data derived from standard PTCAS-LORs could discriminate between upper and lower clinical performers during full-time clinical experiences.
Upper clinical performers (UCPs) from 5 cohorts were identified as students with web-CPI scores >1.5 standard deviations above the normal distribution curve. Lower clinical performers (LCPs) from these cohorts were identified as students who failed clinical rotations or required remedial experiences or substantial intervention during clinical experiences. In most cases (86%), this equated to students who scored >1.5 standard deviations below the normal distribution of web-CPI scores. PTCAS-LOR categories of highly recommend, recommend, recommend with reservations, and cannot recommend were coded 1-4 from least favorable to most favorable for an overall score. Component ratings of commitment, interpersonal skills, communication skills, effective use of time, use of constructive feedback, ethical/professional behavior, responsibility, critical thinking, stress management, problem solving, and leadership were scored on a scale of 1 (poor) to 5 (excellent). Not observed responses were treated as blank responses. T-tests examined between-group differences, with significance at p=0.05. If normality failed, the Mann-Whitney U test and Yates continuity correction were used. Qualitative analysis of PTCAS-LORs for conceptual themes is ongoing.
All PTCAS-LORs submitted for each student were examined, yielding a total of 55 PTCAS-LOR for UCPs (n=14) and 49 PTCAS-LOR for LCPs (n= 14). Analysis revealed that UCP had higher overall scores than LCP (p=0.041). Component analysis indicated that interpersonal skills (p=0.021) and leadership (p=0.004) were higher in the UCP group. The average total of component scores, number of PTCAS-LORs submitted, number of not observed ratings, type of reviewer, and length of time the reviewer knew the applicant were not different between the groups. Preliminary qualitative analysis appears to compliment these quantitative findings.
Conclusions/Relevance to the conference theme:
Problematic clinical performance is a major concern for both faculty and clinical educators, often requiring a substantial investment of resources for intervention and remediation. Our analysis indicates that, on average, UCPs tended to receive more overall ratings of highly recommend on PTCAS-LOR, whereas LCPs tended to receive more overall ratings of recommend. UCPs also tended to demonstrate higher scores in interpersonal skills and leadership when compared to LCPs. This preliminary analysis suggests that careful consideration of PTCAS-LORs during the admissions process may be a useful means of discriminating between potential UCPs and LCPs. Future analyses should more closely examine the role of this and other admissions data in identifying students who are potentially at risk of facing significant challenges in the clinical setting.