Physical therapy degree program

Background and Purpose. Physical therapy schools use a variety of admission criteria to select qualified candidates who will succeed academically and professionally. Outcome variables include scholastic achievement, clinical competence, and successful passage of the licensure examination. Identifying reliable and valid predictors presents significant challenges. The purpose of this study was to assess the ability of one set of admission criteria to a master's degree program in physical therapy to predict graduate program grade point average (GPA), applied skills as measured by the Clinical Performance Instrument (CPI), and national physical therapy licensure examination (NPTE) scores. Subjects. A retrospective multiple regression analysis focused on 121 students who earned an entry-level master's degree in physical therapy (MPT) at the College of St Catherine from 1998 to 2001. Methods. Admission criteria included overall and prerequisite preadmission GPAs, Graduate Record Examination (GRE) scores, and admission committee evaluations of multiple factors. Outcome variables were final cumulative MPT GPA (MPTGPA) and CPI and NPTE scores. Results. The MPT grades were best predicted by preadmission grades. With GRE scores and age, prior grades accounted for 37% of the variability in graduate grades. The MPTGPA was the best predictor of NPTE scores; however, with prior grades and GSE scores, it accounted for only 11% of NPTE variability. The CPI scores increased significantly across 4 placements over a 2-year period, but were not predictable above chance level by any of the present variables. Discussion and Conclusion. Academic outcomes of MPT grades can be moderately predicted using traditional admission criteria. However, licensure examination scores appear to be only weakly predictable. The total absence of predictability of clinical performance, as measured by the CPI, is problematic and demands closer evaluation. Restricted ranges in both the admission and outcome variables contributed to the limited predictive power of the present admission criteria.

Key Words: Admission criteria, Clinical Performance Instrument, National Physical Therapy Examination, Predicting academic performance, Predicting clinical performance.

INTRODUCTION

Competition for admission to Professional (entry-level) master's degree programs in physical therapy (MPT programs) has declined due to smaller numbers of applicants in the last few years.' Nonetheless, there is still a need to use discriminating selection criteria that can effectively identify candidates capable of successfully completing a rigorous academic program, displaying effective clinical performance and passing the national physical therapy !!censure examination (NPTE) at the outset of their professional careers. Surveys of admission criteria used in allied health professions, including physical therapy, confirm the strong reliance on previous scholastic achievement, including both overall preadmission grade point average (GPA) and GPA based on prerequisite science courses (most often biology, anatomy, physiology, chemistry, and physics).2-7 When academic performance in the physical therapy program is the dependent variablc in a forecasting study, it is logical to use as one's predictor the same type of behavior at the undergraduate level (overall and prerequisite GPAs). Indeed, in several studies,**' ' 2 prior GPA was found to be the single best predictor of physical therapy grades.

Unfortunately, physical therapy grade predictors have not been found to be equally reliable forecasters of clinical performance.13 Silver and Hodgson14 have recently reported similar constraints on predicting clinical performance in medical school. Attempts to predict clinical skills have introduced a quite different set of issues for admission committees, as well as for researchers investigating this question. The first difficulty is identifying admission measures that assess applicants' potential for developing skills necessary for clinical competence, although interviews, reference letters, and application essays have all been used with varying amounts of success.9,13,15

Noncognitive variables may be essential components of effective clinical performance and may be distinct attributes that work in concert with cognitive abilities to allow an individual to be an effective clinician.'*' Guffey et al'? explored the correlation between 8 noncognitive domains and performance on the NPTE of 57 physical therapy graduates, but they did not examine the correlation of these factors with clinical performance. The domains that were measured included academic positive self-concept, realistic self-appraisal, support of academic plans, leadership, long-range goals, ability to establish community ties, understanding of racism, and academic familiarity. Sisola's explored moral reasoning, an attribute that may reflect both noncognitive and cognitive processes, and found that moral reasoning scores on the Defining Issues Test (DIT) '9 correlated weakly (y=.28) with clinical performance. In her study, the DIT was completed by physical therapist students in the first week of the program, rather man as part of the admission process. She did not advocate using moral reasoning ability as a factor in the admission process. Morris and farmer(TM) suggested that nonacademlc factors, such as motivation, personality, learning style, and approach to learning, may be better predictors of clinical performance than is academic performance and that these areas warrant further investigation. ? different approach was proposed by Mann and Banasiak.2* They suggested that correlational analysis may be an inappropriate method to investigate the relationship between academic and clinical performance in occupational therapy students, given the homogeneity of student performance in these 2 areas. They urged researchers to examine the relationship between academic course work content and whether or not it directly relates to clinical performance. In addition, they suggested studying students who perform poorly in the clinic in an attempt to understand significant factors associated with poor clinical performance. While there is evidence in the literature to suggest the importance of identifying and measuring predictors of clinical performance, ways of doing so have not yet been defined.

A second equally critical factor in examining predictors of clinical performance has been the absence of a reliable and valid method of measuring clinical performance within the physical therapy curriculum. Although the American Physical Therapy Association (APTA) published its 24-itcm physical therapy Clinical Performance Instrument (CPI) in 1997,22 only recently have the first studies of its reliability and validity appeared/J'^' Based on analyses of early drafts of the CPI conducted by an APTA task force, the instrument appears to have high internal consistency and construct validity. The high level of homogeneity of individual items was supported by Cronbach alphas of .97. Construct and concurrent validity were supported by CPl scores that increased across clinical placements and days of clinical experience. It should be noted mat these validity measures were conservative, since scores on some items were found to be stable already by the end of the first clinical experience (eg, "Presents self In a professional manner"), whereas scores on other items iudexing specific skills such as screening, performing physical therapy examinations, and designing and performing physical therapy interventions were found to increase with experience, finally, the interrater reliability of trained physical therapy clinical instructors yielded intraclass correlation coefficients for individual items ranging from .21 to .76.??

Adams et al^ confirmed mat the CPI has high internal consistency and that it faithfully measures student progress toward entry-level competency across clinical exp??riences; however, they found that it does not significantly correlate win any other physical therapy outcome measures such as physical therapy GPA, licensing examination scores, or performance on a final practical examination (Objective Structured Clinical Examination). Thus, it remains to be determined if it is possible to predict performance on the CM and if CPl performance itself predicts or is correlated with other measures of physical therapist competence.