Enhancing service productivity in acute care inpatient settings using a collaborative clinical education model.

Key Words: Clinical education, Collaborative, Cooperative, Productivity.

The exposure of physical therapy students to acute inpatient settings (eg, general medicine, general surgery) is often difficult to achieve. These service areas are also plagued by frequent and chronic staff shortages. It is not surprising, therefore, that it is difficult to find enough clinical placements for students in these areas. When staff shortages exist, shortages of clinical placements may increase due to refusal to accept students because of concerns about institutional productivity. Such shortages create enormous difficulties for academic programs trying to plan clinical learning experiences. Creative methods of maximizing teaching resources are needed if students are to receive a well-rounded education prior to licensure.

The teaching model frequently used in physical therapy clinical education is the one student to one Clinical Instructor (CI) ratio (1:1 model). This model may limit the maximization of teaching resources and has not been shown to provide the most effective strategy for the development of clinical competence. This teaching model typically requires a CI to balance his or her caseload and administrative responsibilities with those of student supervision, evaluation, and teaching. This responsibility creates much anxiety on the part of the both the CI and the student. For the CI, anxiety comes in part from administrative pressure to maintain productivity and quality of patient care.[1] Maintaining this productivity and quality of care often comes at the expense of the student's learning experience.[2] For the student, anxiety comes in part from not having adequate opportunities for supervised practice, feedback, and reflection. Excessive reliance on the 1:1 teaching model may also create major limitations for academic programs trying to place students in a variety of clinical settings.

One model that has been demonstrated to maintain productivity, provide high-quality learning experiences, and increase the number of placements in the clinical setting is the collaborative or cooperative teaching model.2,3 In this model, two (or more) students are assigned to one CI. The students manage the care of the CI's caseload under the CI's direct supervision. Collaborative teaching refers to situations in which the interactions among students are not based on fixed teacher/learner roles.4 The participants are viewed as equal partners in the learning process, with the responsibility for learning placed on the students.4 Ladyshewsky and Healey[5] have described this model in detail. In their experience, they have found that this model encourages peer-based teaching and mutual problem solving. The CI functions as a facilitator, supervisor, and coach, with the students providing the bulk of patient care.[5] This model of teaching also decreases the reliance of students on the CI.[2-5]

This collaborative teaching strategy has been demonstrated to be effective in promoting learning in the clinical setting.[2,3,6-11] As students develop their ability to work together, they are better able to contribute and to use the contributions of their peer(s) to enhance their learning.[12] Further, as individual team members begin to learn collaboratively, they become empowered to promote additional learning among their peer group.[12] This peer-based learning strategy can lead to higher levels of clinical competence.[8] As Neufeld and Barrows[9] have stated, collaborative teaching provides a forum for students to identify their own emotional reactions and skills in listening, providing criticism, and pooling the collective experiences of the group in finding answers to a problem.

The impact of the collaborative teaching model on institutional productivity has yet to be fully studied. Disadvantages of this model such as increased time for evaluation, setting of objectives, managing different learning styles, differences in rapport between the CI and the students, and the need to share scarce resources can potentially increase the work load for the CI and influence the overall productivity of the supervisor-student team. Unfortunately, there are few evaluations of productivity when using multiple-student-single-instructor models.

Most evaluations looking at the impact of students on clinical agency output have used 1:1 models of student supervision and have focused on costs to facilities" and not "measures of productivity."[13-21] There have, however, been a handful of studies examing the productivity side of clinical education. Lopopolo[22] found that students produced a net financial benefit per day by increasing the number of patient visits per day. Leiken et al[23] examined productivity using occupational therapy, physical therapy, and radiology technology students in one large metropolitan hospital. The number of treatments provided was used as the productivity measure.[23] In two studies,[23,24] Leiken and colleagues found that the students contributed to the facility's capacity to provide patient care. Kaplow[25] concluded in her review of a physical therapy internship program in Quebec, Ontario, Canada, that the service provided by students outweighed the cost, in time spent, by CIs to supervise and coordinate the program. Ladyshewsky et al26 also found that physical therapy students did not produce negative effects on overall departmental productivity in an outpatient orthopedic setting. Instead, actual staffing levels, case mix, amount of meeting time, and length of waiting lists were the factors more likely to influence patient throughput.[26] Coulson et al[27] found that senior physical therapy students in an outpatient setting increased the clinic's net earnings by $216.77 per day and throughput by 3.25 patients per day. Graham et al[28] also demonstrated that longer placements (eg, 5 weeks) produced greater productivity and efficiency than did shorter placements. The student-CI team's productivity is measured by mean number of patients treated per day, mean revenue generated per day, and mean number of treatment units provided per day) in longer placements (5 weeks) was also substantially higher than when the therapist worked alone.28

The basis for this study was to evaluate the patient care productivity of the collaborative teaching model in acute care inpatient settings. A 2:1 model (two students to one CI) was utilized. Differences in patient care productivity between the CI alone and the CI with two students were investigated.

Method

Subjects

Eight CIs employed in six different hospital-based acute inpatient settings and 16 physical therapy students were involved in the study. Five of the CIs had previously supervised students in a 2:1 placement and had received training in managing this type of supervisory arrangement from a workshop program regularly offered by the University of Toronto (Toronto, Ontario, Canada). Five CIs had supervised 7 or more students during the course of their career. The remaining three CIs had supervised 0, 1, and 3 students, respectively. All CIs involved in this study volunteered to supervise 2 students at the same time. The CIs were advised to turn over the majority of their patients to the 2 students and to spend their time in student supervision.

The 16 students who were involved in the study had completed their third year of academic study at the University of Toronto. The University of Toronto Program in Physical Therapy is a 4-year program leading to the Bachelor of Science (Physical Therapy) degree. By the end of the third year, students in this program have completed all studies in cardiorespiratory, neurology, and orthopedic sciences. They have also completed 4 weeks of full-time clinical practice, which occurred at the end of their second year of study. Of the 16 students, only 1 student had a previous 2:1 placement experience. The student pairs were randomly assigned to their placements by the Academic Coordinator of Clinical Education (ACCE) using the university's computerized placement matching system. The paired assignments were then reviewed by the ACCE, and changes were made if there were discrepancies in academic performance (eg, one student with an average grade of 60% and the other student with an average grade of 85%) and previous academic preparation (eg, one student holding a higher degree and the other student possessing only a high-school diploma). The students' cumulative grade point average was used as an academic performance measure. The number of years of pre--physical therapy higher education was used as a measure of previous academic preparation. in all cases, an equitable match was attained. For all of the students, this was their first acute inpatient experience in cardiorespiratory therapy.

Procedure

Workload Measurement System (WMS) data were collected and compared at two different points in time: once during the student placement and once during the control period. The WMS captures time spent in patient care, student supervision/teaching, and administration. Detailed descriptions of the various measures used in the system are presented in Table 1. The student placement was a 5-week full-time placement. The control period was a 1-week full-time period during which the CI worked independently. The data from this week were then multiplied by 5 so that a comparison could be made with the SP data. In Kaplow's study of physical therapy inter6n productivity, she outlined several considerations for ensuring reliable and valid data.[25] Data should come from the same institution with and without students, and the case mix across sampling intervals should be approximately the same.[25] This study met the first requirement by using data from the same institutions with and without students. The same CIs were also involved in both components of the study (control period and student placement).

[TABULAR DATA 1 OMITTED]

The strategy that I used to ensure case-mix consistency across the control period and the student placement was to capture only 1 week of CI data during the control period. The data were collected from 2 to 4 weeks before or after the actual student placement. A sufficient interval between the control period and the student placement was necessary to ensure that there were no influencing effects on the CI's productivity (eg, preparing for the student before or refamiliarizing oneself with the caseload after the students' departure). The control period had to be close enough to the student placement, however, to ensure that case mix had not changed substantially. I assumed that capturing 5 full weeks of CI data for the control period (with the 2- to 4-week grace period before or after the student placement) could increase the chance that the work unit characteristics would differ from the actual student placement, thus affecting the nature of the comparison. Ladyshewsky et al[26] have described several of these cofactors (eg, changes in staffing, case mix, meeting time, patient volume) that can have an impact on prod0uctivity measures.

Neither the students nor the CIs were aware of this study. This blinding was done to ensure that their reporting would be honest and typical of their regular activity rather than being influenced by their knowledge of being in an experiment. The 2:1 teaching model is part of the normal clinical education program at the University of Toronto, and the students and CIs believed they were engaged in a standard clinical placement. The students and CIs did complete an evaluation form asking them for qualitative information about the placement experience. The director of each physical therapy service, however, was sent a letter inviting him or her to participate in the study. If the director agreed to participate in the study, I met with the director to discuss the scope and objectives of the study and the methods for ensuring confidentiality. Following each student placement and control period, I met with the director and collected the appropriate data.

Instrumentation 

Standardized WMS data were used to measure productivity in these acute inpatient settings. These data are already available in usable format through the Canadian Management Information Systems group created through the merger of the Canadian Management Information Systems (MIS) Project and the Canadian National Hospital Productivity Improvement Program.[29] The WMS data that were utilized in this study are described in Table 1. One weighted unit of time is equivalent to approximately 1 minute of Staff time.[29] These data were then manipulated to create five specific productivity indicators. These productivity indicators are also described in Table 1.

Data Analysis

Once all the WMS data were collected, I calculated the productivity indicators. Each productivity indicator for the student placement included the WMS data for the CI and the students. The productivity indicators for each of the eight teams were then grouped together by period (control period and student placement) My averaged. By averaging each of the productivity indicators across the eight groups, one is able to make comparisons between the control period and the student placement.

The data collected during the student placement were analyzed in two different ways: (1) in their pure form and (2) with the student component reduced by a cofactor of 0.6. It was important to factor down the students' WMS data contributions because it would be incorrect to assume that 1 minute bf student time was equivalent to 1 minute of CI time. I selected a reduction factor of 0.6 because these students were expected to manage a two-thirds caseload by the end of the placement, as per guidelines established by the University of Toronto clinical education program. This factoring had the net effect of reducing the students' patient care productivity contributions by 400/o and served as an efficiency/effectiveness adjustment to the productivity indicator.

Results

Amount of Care Provided per Worked Hour

Table 2 illustrates the mean difference in patient care productivity during the control period and the student placement. The range across the eight facilities for the control period and the student placement is also provided. The total amount of direct care weighted units per worked hour (DCWU/WH) and patient care weighted units per worked hour (PCWU/WH) provided increased by 1060/o and 990/o from the control period to the student placement. When the students' contribution to DCWU/WH and PCWU/WH was reduced by a factor of 0.6, there was still a net increase in the teams' productivity of 47% and 43%.

[TABULAR DATA 2 OMITTED]

Number of Patients Treated per Worked Hour

Table 2 also illustrates the mean difference in the numbers of patients treated during the control period and the student placement. The range across the eight facilities for the control period and the student placement is also provided. The attendance days per worked hour (ADAY/WH) increased by 340/o from the control period to the student placement. That is, the two student-CI teams were able to treat 34% more patients as compared with what the CIs normally managed. This particular calculation did not include a factoring component of 0.6, given that it is measuring actual numbers of patients treated per worked hour.

Amount of Care Provided per Patient

Table 2 further illustrates the mean difference in the amount of care provided to each patient during the control period and the student placement. The range across the eight facilities for the control period and the student placement is also provided. The amount of direct care weighted units per attendance day (DCWU/ADAY) and patient care weighted units per attendance day (PCWU/ADAY) increased by 60% and 39% from the control period to the student placement. When the students' contribution was reduced by a factor of 0.6, there was still a net increase in the amount of DCWU/ADAY of 12%. With respect to PCWU/ADAY, there was a net decrease of 1%.

Amount of Clinical Instructor Time Spent in Non-Patient Care Activities

Clinical Instructors spent approximately the same amount of time managing their administrative responsibilities during the course of the student placement as they did during the control period. During the 5-week control period, CIs spent, on average, 27.04 hours on administrative duties. During the 5-week student placement, CIs spent, on average, 25.26 hours on administrative duties. In addition to this administrative time, the CIs were also able to spend, on average, 27.58 hours on didactic teaching activities during the student placement. During the student placement, therefore, the CIs spent, on average, a total of 52.84 hours on administration and didactic teaching.

Discussion

This study demonstrates that students placed in an acute inpatient setting are able to make a positive service contribution and that this contribution can exceed the unit standard when two students are assigned to one CI. One of the main reasons these productivity results were obtained was that the CI was able to delegate most of his or her direct responsibility for "hands-on" care to the two students. This delegation provides time for the CI to supervise the two students while mg any detrimental effects on productivity. Expert treatment can be provided by the CI when necessary.

It is true that if the CI also managed a large caseload in conjunction with the caseload of die two students, the productivity of the unit could theoretically increase even further (providing there were enough patients). This increased productivity, however, could have a detrimental effect on the quality of learning and supervision. This outcome was evidenced in an earlier study by Ladyshewsky,[2] which demonstrated that the quality of supervision and learning decreased (as reported by students) when their CI's personal caseload increased.

Even though this particular study utilized Canadian work-load measurement data, the results are generalizable to most health care institutions at which the physical therapy salary component of the departmental budget is relatively constant (eg, the physical therapists are paid their salary whether students are present or not). What was measured in this particular study was the activity of the CI and the students when providing patient care.

Although this work-load measurement system does not attempt to measure every minute of a physical therapist's time, it does give a good measure of the amount of care provided. The WMS is part of standard hospital recording in Canada, so most physical therapists are well versed in the reporting method. This method enhances the reliability of their reporting. Because students are learning how to report their WMS data, there is a greater possibility of incorrect reporting. Fortunately, CIs are expected to review these data, and all students receive a 2-hour training session on the Canadian WMS as part of their administrative education in the second year of their program. Both of these safeguards served as important reliability checks. As mentioned earlier, CIs and students were not aware that these data were being collected. Hence, any reporting bias should have been minimized.

The fact that the amount of DCWU/WH and PCWU/WH increased is not surprising, given that there were more bodies" available to provide patient care. The increase in these values is a function of the mathematics involved in deriving these productivity measures. Only the paid worked hours of the CI were included in the denominator. The numerator, however, included the DCWU and PCWU of all three team members. The reason for calculating the measures in this way is that the students' service contributions to the agency are a benefit. The agency does not pay for them. It is, therefore, only appropriate to consider the cost of the CI's time. What was important to determine was whether the students' contributions to productivity during the student placement were sufficient to equal or outweigh the normal productivity of the CI during the control period.

The increase in ADAY/WH demonstrated that the team was able to manage their caseload responsibilities as well as treat additional patients. For those agencies that bill their patients on a per-attendance basis, the potential for increased revenue is evident. One of the reasons productivity increased in this case was not because there was an unlimited supply of patients for the students to treat. Most acute inpatient units have fixed bed complements and therefore a fixed number of patients for treatment. Students were able to treat patients from other physical therapists as the students' skill and competence increased. This factor accounted for the increase in the number of attendances per worked hour.

The amount of direct and overall patient care provided per attendance DCWU/ADAY and PCWU/ADAY) was also found to increase and is a direct measure of the amount of time, on average, spent with each patient. In many ways, these work-load statistics offer a measure of quality, if one assumes more time with a patient is better. Because the single caseload of the CI was divided between two students, each student was able to spend a proportionately greater amount of time with each patient. This is why an increase in the amount of care per patient was seen. With factoring, however, PCWU/ADAY decreased by 1%. By itself, this decrease is insignificant because it is an average measure and reflects the variance in performance among the eight 2:1 teams. This decrease does, however, demonstrate the effect of incorporating the students' efficiency on productivity.

Although beyond the scope of this study, it would have been interesting to measure the impact of students on overall work unit productivity. one would expect to see, at least, a break-even outcome, with students offsetting the decrease in the number of patients seen by other physical therapy staff. In addition, one would expect to see either an increase in administrative time by the other physical therapists (using the time made available by students treating their patients) or an increase in the amount of time these physical therapists spent treating their remaining patients. Both findings have been reported in earlier work undertaken by Kaplow[25] and by Ladyshewsky et al,[26] who found that the influence of students did not have a detrimental effect on overall work unit productivity.

The absorption of the CI's caseload by the two students provided the CI with time to plan and engage in clinical teaching as well as time to maintain his or her administrative responsibilities. This teaching time is important if the benefits of a cooperative learning strategy, which were described earlier, are going to be realized. The CI needs time to structure the learning environment such that students can learn cooperatively. The fact that the amount of CI administrative time remained relatively consistent between the control period and the student placement period suggests that students utilized the expertise of one another to solve problems. This possibility further reinforces the concept that this model has the potential to decrease the reliance of students on CIS.[2-5]

From a cost-accounting perspective, the additional time that was made available for teaching should not be interpreted as a cost to the agency. Although this "teaching time" costs the agency in the sense that this professional physical therapist could be engaging in some other activity, it ignores the fact that the only reason this time has become available is because the students are providing a free service to the hospital. I believe that this student service contribution should be interpreted as a bayback or benefit to the facility. The principle that the investigator is trying to elucidate is that one cannot isolate clinical teaching costs from productivity. Both must be considered as part of the full cost-benefit equation.

One of the critical elements behind the success of this placement model is that the students had completed their clinical studies and had already completed a full-time 4-week placement during their second-year program. Hence, they were able to make a useful contribution to the service. These same results might not have occurred in a clinical teaching situation using junior students.

The data reported in this article are mean data. There were definite differences in the data among the CI-student teams, the ranges of which are shown in Table 2. Not every team demonstrated the same changes in productivity. This finding reflects the difficulty in implementing controls in this type of study, in which there may be different student performance capabilities, different case mixes among acute inpatient settings, different patient acuity levels, different institutional systems in operation, and differences in how the CIs spend their time teaching and supervising. For example, in one particular group in this study, a thoracic surgeon went on holiday in the middle of the placement, which had a dramatic effect on the productivity of the team.

Another important limitation of this study is that it did not measure the quality of patient care provided by the students. This is not an issue exclusive to 2:1 models of clinical supervision. It is an issue for any model of clinical teaching. What is interesting about this particular model of clinical supervision is that it has the potential for improving the quality of care provided by students during a placement. If the CI properly delegates the majority of his or her clients to the students, the CI is in a much better position to oversee the quality of care provided by the students, a finding reinforced in the study by Emery and Nallette.[3] This finding is in direct contrast to the 1:1 situation, in which the CI is constantly struggling to meet the demands of his or her personal caseload of 65% to 75%, for example, with the 25% to 40% caseload of the student. Another important factor to consider about this teaching model, outside of productivity, is that it has the potential to increase the number of acute inpatient placements for university programs.

There was a net positive influence in productivity across the eight facilities. There will always be instances in which a placement may result in a decrease in productivity because of problems related to patient availability, supervision, or student performance. These same issues are not exclusive to 2:1 models of clinical teaching and can be extended to the 1:1 model. Over the course of time, however, one would hope that the facility benefits overall from its student program. ideally, further investigations with larger samples would add to the research. It was difficult to find enough CIs who were willing to supervise two students at the same time. The general shortage of acute inpatient cardiorespiratory placements in itself also led to a less-than-ideal sample size.

Repeating this type of analysis in different speciality areas would also determine whether these same productivity benefits occur. Investigations using this type of data analysis on placements of different durations should also be undertaken, as Coulson et al[27] demonstrated that this factor also has an effect on levels of productivity.

This type of productivity analysis varies somewhat from the type of cost-benefit analyses undertaken in the literature. Most cost-benefit studies in the past have looked at very few multiple measures. The impact of students on institutions was often assessed by counting changes in numbers of patients seen or in numbers of treatments provided. Tracking the amount of time CIs spent teaching students and then calculating the cost of this supervision (without considering the service benefit accrued to the agency by the students) has also been undertaken. Other studies have examined changes in the amount of billed revenue when students are placed in an institution.

With the advent of computerized management information systems, much more accurate work-load and productivity information can be collected and analyzed in tandem. Time spent in patient care interventions, administration, and teaching can be easily collected and tied into cost estimations. This approach provides a much more reliable and valid method of cost-benefit analysis. The impact on an individual CI's productivity or the productivity of an entire work unit can also be assessed using these management information systems.

Conclusion

This study has demonstrated that a cooperative model of clinical education using two students assigned to one CI can positively influence a service unit's productivity. In this study, acute inpatient services were examined. The implications of these results should interest ACCEs, CIs, and administrators because all are concerned about the impact of students on institutional productivity. There is clearly much more room for other investigations of this sort. For example, does one get similar results using a different model of work-load measurement? How does placement length influence productivity, and do you get similar types of results with more senior or junior students? What are some of the effects on patient care quality when you have multiple students and one CI? Finally, do you get the same type of effects in different specialty areas?

In spite of these additional research questions, the physical therapy profession can use the current information gleaned from this study to reexamine the way in which it delivers its clinical education programming. Current resource constraints, institutional staff shortages, deficiencies in the number of clinical placements, and the changing nature of our profession require that we reexamine the way we prepare our students.

References

[1] Bohannon RW. Statistical analysis of productivity in one physical therapy department. Phys Ther. 1987;21:1553-1557. [2] Ladyshewsky R. Clinical teaching and the 2:1 student-to-clinical instructor ratio. Journal of Physical Therapy Education. 1993;7:31-35. [3] Emery M, Nallette E. Student staffed clinics and creative clinical education during the times of constraint. Clinical Management in Physical Therapy. 1986;6(2):9-10. [4] McDonald BA, Larson CO, Danscreau DF, et al. Cooperative dyads: impact on text learning and transfer. Contemporary Educational Psychology. 1985;10:369-377. [5] Ladyshewsky RK, Healey E. The 2:1 Teaching Model in Clinical Education: A Manual for Clinical Instructors. Toronto, Ontario, Canada: Department of Rehabilitation Medicine, Division of Physical Therapy, University of Toronto; 1990. [6] Tiberius R, Gaiptman B. The supervisor-student ratio: 1:1 versus 1:2. Canadian Journal of Occupational Therapy. 1985;52:179-183, [7] Abercrombie MLJ. Anatomy of judgement. Harmondsworth, England: Penguin Publishing Co Ltd; 1969. [8] DeClute J, Ladyshewsky RK. Enhancing clinical competence using a collaborative clinical education model. Phys Ther. 1993;73:683-689. [9] Neufeld VR, Barrows HS. The McMaster philosophy: an approach to medical education. Journal of Medical Education. 1974;49: 1040-1050. [10] Glendon K, Utrich D. Using cooperative learning strategies. Nurse Educator. 1992; 1 7: 37-40. [11] Barrows HS. The scope of clinical education. Journal of Medical Education. 1986;61: 23-33. [12] Irby D. Clinical teaching and the clinical teacher. Journal of Medical Education. 1986; 61:35-45. [13] Ramsden EL, Fischer WT. Cost allocation for physical therapy in a teaching hospital. Phys Ther. 1970;50:660-664. [14] Rabkin MT. Reducing the costs of medical education. Health Affairs. 1986;5:97-104. [15] Hammersberg SS. A cost/benefit study of clinical education in selected allied health programs. J Allied Health. 1982;1:35- 41. [16] Chung YI, Spelbring LM, Boisonneau R. A cost benefit analysis of fieldwork education in occupational therapy. Inquiry. 1980;17:187-191. [17] Mackinnon JR, Page GG. An analysis and comparison of the educational costs of clinical placements for occupational therapy, physical therapy, speech pathology and audiology students. J Allied Health. 1986;15:225-237. [18] Page GG, Mackinnon JR. Cost of clinical instructors time in clinical education-physical therapy students. Phys Ther. 1987;67:238-243. [19] Porter RE, Kincaid CB. Financial aspects of clinical education to facilities. Phys Ther. 1977; 57:905-909. [20] Kling DR, Bulgrin JA. Clinical education costs and benefits: application of a fiscal analysis model. J Allied Health, 1987; 16:135-147. [21] Pobojewski TR. Case study: cost/benefit analysis of clinical education. J. Allied Health. Summer 1978:192-198. [22] Lopopolo RB. Financial model to determine the effect of clinical education on physical therapy departments. Phys Ther. 1984;64: 1396-1402. [23] Leiken AM, Stern S, Baines RE. The effect of clinical education programs on hospital production. Inquiry. 1983;20:88-92. [24] Leiken AM. Method to determine the effect of clinical education on production in a health care facility. Phys Ther. 1983;63:56-59. [25] Kaplow M. Cost Benefit of Physiotherapy Internship in Quebec. Montreal, Quebec, Canada: McGill University; 1980. Thesis. [26] Ladyshewsky RK, Bird N, Finney J. The impact on departmental productivity during physical therapy student placements: an investigation of outpatient physical therapy services. Physiotherapy Canada. 1994:45;94-98. [27] Coulson E, Woeckel D, Copenhaver R, et al. Effects of clinical education on the productivity of private practice facilities. Journal of Physical Therapy Education. 1991;5:29-32. [28] Craham CL, Catlin PA, Morgan J, Martin E. Comparison of 1-day-per-week, 1-week, and 5-week clinical education experiences. journal of Physical Therapy Education. 1991;5: 18-23. [29] Government of Canada. Physiotherapy Workload Measurement System National Hospital Productivity Improvement Program, Health and Welfare Canada Statistics Canada Health Division, 1987-1988 Edition. Ottawa, Ontario, Canada: Minister of Supply and Services; 1988.

RK Ladyshewsky, MHSc, BMR(PT), is Lecturer, School of Physiotherapy, Curtin University of Technology, Perth, Western Australia, Australia. He was Academic Coordinator of Clinical Education, Division of Physical Therapy, University of Toronto, Toronto, Ontario, Canada, at the time of this study. Address all correspondence to Mr Ladyshewsky at School of Physiotherapy, Curtin University of Technology, Selby St, Shenton Park, Western Australia 6008, Australia.

This article was submitted May 18, 1994, and was accepted February 2, 1995.