Since an appropriate instrument that would gather desired data from the population to be studied was not located, a composite instrument was developed for the study. The instrument consisted of 69 questions and was organized into 4 sections. The instrument was developed to capture the propensity of physical therapists to adopt EBP, characteristics of the social system in the workplace of physical therapists, personal characteristics of physical therapists, and selected demographic variables of physical therapists. The central measure of the study was an 8 item scale measuring the propensity to adopt EBP which was adapted from a psychometric instrument by Green, Gorenflo, and Wyszewianski 26. (Table 1.) A histogram was examined to ensure that the dependent variable, the propensity to adopt evidence based practice, was normally distributed. The histogram was judged to satisfy the assumption of normality. The scales for two related variables practicality and, nonconformity which collected data about personal characteristics were also adapted from the psychometric instrument. Each item in section 1 was scored using a Likert style 6 point scale where 1 = strongly disagree and 6 = strongly agree. Section 2 gathered the individual physical therapist characteristics of the social system in the workplace using the shortened version of the Dimensions of the Learning Organization Questionnaire (DLOQ). 2728 Each item in section 2 was scored using a six point Likert style scale 1= almost never and 6 = almost always. Additional personal characteristics were collected in section 3 using the shortened version of the Self-directed Learning Readiness Scale for Nurse Education (SDLRSNE). 29Each item in section 3 was scored using a Likert style 6 point scale identical to the scale in section 1. Section 4 collected demographic information.

Section 1 was adapted from the psychometric instrument 26to reflect the practice world of physical therapists. The section measured propensity to adopt EBP, practicality and nonconformity. An expert panel of five physical therapists modified the section for clarity and natural language of physical therapists and reviewed the section for validation of content validity for physical therapists. The experts had experience in a variety of practice settings for physical therapists and were chosen because they understood the practice of physical therapy.

Section 2 measured physical therapists' perception of the characteristics of the social system in the workplace using the shortened version of the Dimensions of the Learning Organization (DLOQ).2728 Evidence of the short form of the DLOQ's construct validity was established by Yang, Watkins and Marsick.28

Section 3 measured physical therapists' self-assessment as self-directed learners using a shortened version of the Self-directed Learning Readiness Scale for Nurse Education (SDLRSNE). 29 Fifty two items in 3 clusters: self-management, desire for learning and self-control were identified by a panel of experts in self-directed learning using the Delphi survey method as representing characteristics of a self-directed learner. 29 The instrument for this study used a shortened version of SDLRSNE including only the items that loaded at .45 criterion or higher on three factor analysis.

Table 2 depicts the distribution and reliability of scale variables for the dependent variable, propensity to adopt EBP and the independent variables, personal characteristics and characteristics of the social system in the work place. The scale mean and scale standard deviation were derived. Coefficient alpha was assessed to measure how well the items in each construct measured a single unidimensional variable. The coefficient alpha for individual variables ranged from .56 to .89.

The potential participants consisted of 3,897 physical therapists who were licensed by the state of Georgia in the United States of America and were listed on a roster purchased from the state of Georgia. The goal was to create a sample large enough to generalize to the population of all physical therapists licensed by the state of Georgia. Based on Chassan's 30 recommendation and the advice of a statistician we selected 20 participants per varible as a desireable measure which would result in 660 potential participants. Twice that number were randomly selected from the list of potential participants to allow for nonrespondent problems and to ensure the stability of our analysis. 1320 names of potential participants were randomly selected using a computer generated software package BCC Software. 939 of the participants' questionnaires were returned for a response rate of 73%. 831 of the participants' questionnaires were useable and became the basis for the study.

Each therapist received an explanatory cover letter, survey instrument, a notice of Research Information for the Participants from the University of Georgia's Institutional Review Board, an incentive to participate, and a postage-paid return envelope. A follow-up reminder postcard was mailed 10 days later to all nonrespondents. Thirty-five days later all nonrespondents were contacted again by mail and sent another survey instrument, a different cover letter, Research Participants Information Sheet, another incentive to participate and a postage-paid return envelope. 31

SPSS 11.5 was used for all data analysis. Descriptive statistics, Pearson's correlation, a Spearman correlation, an independent t-test and stepwise multiple regression were conducted using SPSS statistical software package.

Lifelong self directed learning is essential in the context of a rapidly growing and changing body of knowledge in health care. 3233 The study confirms that desire for learning, a component of self-directed learning, accounts for a modest proportion of the variance in the propensity to adopt EBP. The information seeking behavior common to both self-directed learning and EBP may account for the association. A fully self-directed learner is one who has learned to think critically. 2934It is the critical reflection on practice that forms the basis for questions that lead to locating the best current research evidence.35 Developing a question also requires other behaviors identified as characteristics of the self-directed learner in the literature: initiative, 1929 self-discipline, 1933and the ability to develop and set goals. 19 Initiative and self-discipline were consistent with the behaviors described in the self-management scale, which emerged as a minimum predictor of the propensity to adopt EBP. Goal setting was addressed in two of the items in the self-control scale, which emerged as a minimum predictor of the propensity to adopt EBP.

This study provides empirical evidence that the higher the degree of education physical therapists obtain the more likely they are to demonstrate the propensity to adopt EBP. This finding is consistent with previous research by Warren and Pierson 36who found that physical therapy students with a master's degree demonstrated a more positive attitude toward research than baccalaureate students.

Age and years licensed as a physical therapist were negatively correlated with the propensity to adopt EBP. When we examined age and years licensed with the predictor highest degree held, we extrapolated that physical therapists who were older and have been licensed longer were less likely to demonstrate the propensity to adopt EBP because they may not understand how to access the scientific literature. Time spent in direct patient care was negatively correlated with the propensity to adopt EBP. The negative correlation is noteworthy. Based on the findings of this study, physical therapists who are providing direct patient care value clinical experience and authority over scientific studies when making clinical decisions. This suggests direct patient care continues to be based on clinical experience and authority. Reaching the provider of direct patient care is imperative if the paradigm is going to shift to practice based on "the integration of the best research evidence with clinical experience and patient values." 7

Numerous authors have cited lack of time as a barrier to the adoption of EBP. 13738 It is therefore not surprising that those physical therapists who reported that they agreed that evidence-based guidelines and scientific studies can be used to make clinical decisions in the day-to-day practice of physical therapy without interfering with productivity or the smooth and orderly flow of patients 21 demonstrated the propensity to adopt EBP. This finding suggests that research evidence needs to be easy to access, time efficient, and relevant to practice so that productivity and the smooth orderly flow of patients will not be adversely affected. Linking evidence to workflow so that physical therapists can act on it in a time efficient manner is essential. 39Research evidence needs to be available at the point of care so that clinical questions can be answered quickly. 40

The study demonstrated there was a relationship between physical therapists who view themselves as engaging in clinical practices that are out of step with how others in the local community provide care or what opinion leaders recommend, nonconformist, 21 and the propensity to adopt EBP. We speculated that as physical therapists have moved toward autonomous practice in the United States and thereby perceived themselves as having the capability, ability, and responsibility to exercise professional judgment, they concomitantly demonstrated a propensity to value research evidence over experience and authority when making clinical decisions. EBP may require a physical therapist to depart from the usual and ordinary care that has been provided by other physical therapists in the facility in order to provide the right treatment for the right patient based on the best current research available. 41

The variables desire for learning, practicality, and highest degree combined account for a moderate proportion of the observed variance in the propensity to adopt EBP. This combination of variables out performed any single variable predicting the propensity to adopt EBP. This finding reveals the fact that a combination of factors works synergistically to foster the propensity to adoption of EBP.

Despite the theoretical integrity of these predictors we have limited explanatory power. The predictor variables employed in this study were selected based on a review of the literature. However, as with any study we studied only a finite number of variables. There were many variables that could have been considered including: skills in critically appraising the literature, computer access, knowledge of online databases, amount of time available to research evidence, presence of weekly case studies, or presence of a journal club in the facility or the community.

Another limitation deals with the sample. Because the sample frame only covered physical therapists licensed by the state of Georgia in the United States of America generalization to other groups must be handled by logical rather than statistical inference.

This study has a number of implications for the practice of physical therapy and other health care practices. In this study desire for learning, a component of self-directed learning, accounts for a modest proportion of the variance in the propensity to adopt EBP. Adults learn from experience and from problems in their everyday life. ^{19 Each day questions arise in the practice of health care that go unanswered. 35 It is in the day-to-day practice that continuing education should be taking place. Cervero, an expert in continuing education of medical professionals, has proposed that if "physicians are going to make good clinical judgments they need to learn from their experience in the swamp of practice."42The same is true for all health care practitioners. EBP harks back to the natural teaching method described by Osler. 43 The patient is the source of the health care practitioner's inquiry. To maximize the benefit of self-directed learning health care practitioners will need to reflect on practice and develop a process of questioning. It is this critical reflection on practice that forms the basis for questions that lead to locating the best current research evidence on interventions. 35 Reflecting on practice illustrates to members of the health care team how the ordinary challenges of work can be converted into opportunities for learning. 24 To further foster a desire for learning and encourage the use of EBP opportunities should be made available to provide health care practitioners the leisure to "reflect on their reflection in action" 42 in clinical decision making in the presence of colleagues. In other words there should be a forum where there is critical debate of the treatment interventions based on the best current research evidence available. These communities of practice 44 can take the form of weekly journal clubs or case studies. Exploiting the problems that arise in day-to-day practice of the health care practitioners by reflecting on practice, converting the patient problems to questions and reflecting on actions with colleagues develops habits of practice that are consistent with facilitating self-directed learning. 384546}

Practicality accounted for a moderate proportion of the variance in the propensity to adopt EBP among physical therapists. It is imperative that the health care professions, adult educators, change agents and department managers investigate ways to make EBP easy to access, time efficient, and relevant to practice so that productivity and the smooth orderly flow of patients will not be adversely affected. Research evidence needs to be linked to workflow so that health care practitioners can act on it in a time efficient manner. 39The best current research evidence for treatment interventions needs to be readily available electronically in the same place and at the same time that a health care professional is making a clinical decision so that the right treatment intervention is provided for the right patient in a timely manner.