Physical Therapist Productivity Using a Collaborative Clinical Education Model Within an Acute Care Setting: A Longitudinal Study

Abstract

Introduction. 

Collaborative models in clinical education are increasing in popularity, but concerns exist over the potential impact on patient productivity of clinical instructors. Little empirical research has examined the impact of patient productivity in the models of clinical education in the acute care setting. The objective of this study is to examine the productivity impact of physical therapists (PTs) participating in both the 1:1 and 2:1 models of clinical education over a 3-year time period in an acute care hospital setting.

Methods. 

This study utilized a longitudinal, retrospective design. Productivity data of 20 PTs was examined over a 3-year timeframe (8,981 days of productivity). Productivity of each PT was cross-referenced with their clinical education schedules to evaluate patient productivity while serving as a clinical instructor in a 1:1 or 2:1 clinical education model.

Results. 

PTs serving as clinical instructors in a 2:1 model (mean 21.53 units per day; 95% confidence interval (CI), 21.15–21.91) achieved a significantly greater productivity (P < .001) than those serving in a 1:1 model (17.05 units per day, 95% CI, 16.91–17.18). Within-group differences demonstrated a significant increase (P < .001) of units billed per day with the addition of each student, both in the transition from 0 to 1 student (0.591 units per day; 95% CI, 0.18–0.99) and from 1 to 2 students (4.75 units per day, 95% CI, 4.21–5.29) for therapists who participate in the 2:1 model, and for those who participate in a 1:1 model of clinical education (0.82 units per day; 95% CI, 0.51–1.13).

Discussion and Conclusions. 

PT productivity can increase during both a 1:1 and 2:1 model of clinical education in the acute care setting. Formal training of clinical instructors and continued mentorship of clinical education faculty is necessary to create an institutional culture that optimizes physical therapy clinical education.

INTRODUCTION

Clinical education is an integral component of physical therapist (PT) education. The academic preparation of effective health care professionals is dependent upon a sound foundation of clinical practice under the guidance and mentorship of clinical faculty. According to the Commission on Accreditation in Physical Therapy Education (CAPTE), on average, clinical education makes up 20.4% of the total credit hours, and almost a third (29%) of the enrollment weeks in the entire PT education curriculum.¹ Other authors have reported up to 44.9% of the PT education curriculum being devoted to clinical education.² Although there has been a progression of degree preparation from a baccalaureate to masters and now to the current entry-level professional doctoral degree in PT education, there has been no significant change in the method by which we provide clinical education. Other than increasing the number of hours that the students spend in clinical education and the form that those hours take in any given program, the traditional model of 1 clinical instructor to 1 student (1:1) is still the primary clinical education model utilized. With the growing number of PT educational programs nationally and the continued adherence to the traditional 1:1 model, academic institutions are facing increased difficulty in securing sufficient clinical education placements to meet the needs of all their students.

A collaborative model of clinical education involving 2 students with 1 clinical instructor (2:1) has been suggested by authors on several occasions.^3,4 One of the concerns that is identified anecdotally in the clinical education community regarding the 2:1 model is whether or not students will receive a quality clinical experience. The line of questioning and dialogue has continued in a recent systematic review that found inconclusive evidence regarding what constitutes best practice for PT clinical education.² Some early studies on this clinical model have suggested positive outcomes from using a 2:1 ratio. Several qualitative research studies, for example, have found that the 2:1 model has had positive outcomes in regard to student learning due to opportunity for collaborative input, discussion, and support.^4–6 Another earlier review held that students who were in the early stages of their education preferred the 2:1 model, while those in the later stages preferred 1:1 models.⁶ Regardless, as cited in the review by McCallum et al, there is a lack of evidence for best practice in terms of the various practices in physical therapy clinical education (eg, yearlong internship, self-contained, health system-based, clinician-paid, and required residency). These authors suggested that clinical education is often not a priority at the clinical site level and that advancing the education culture at both the site and clinical instructor levels should be considered.²

Several studies have noted that financial concerns are barriers to clinical education.^7–10 In light of this, some sought to examine the effect of clinical education on the productivity in clinical facilities across disciplines and the continuum of care.^11,12 These studies identified trends of productivity increases when students were involved in clinical care compared to times without students. However, these studies varied greatly in the disciplines examined as well as the models of education. Additional studies directly related to physical therapy are limited to only using the traditional 1:1 model of clinical education.^9,10,12–14 Only 1 of these studies examined changes in the acute care setting, but again utilized a 1:1 clinical education model.¹⁴ Cost analysis studies have also been performed to examine economic impact of clinical education on physical therapy departments.^11–14 However, the findings of these studies are limited in their generalizability due to the fact that they were completed in the 1980s, when there were definitive differences in both the health care climate and professional practice of physical therapy.

Previous studies on the collaborative model have largely used qualitative methodologies to assess the clinical model. These have included use of interviews and questionnaires,⁴ comparisons of varying constructs of the collaborative model,¹⁵ and feedback from observations and journal entries from clinical instructors and students.⁵ The results of these qualitative approaches have provided input for discussion and recommendations regarding the advantages, disadvantages, and perceptions of the competing collaborative models. Although these published commentaries and qualitative studies have added value to the examination of the impact of models of clinical education, a review of the literature indicates that there has been little research examining productivity using quantitative methods or the impact on patient outcomes. No studies have been conducted that analyze productivity changes over a large time period; that includes both clinical education models (1:1, 2:1) and all levels of students from many universities. The published studies evaluating the productivity of the PT serving as a clinical instructor in a 2:1 model have been limited in scope to either small sample sizes of therapists or students, or brief data collection timeframes.^5,15–19 With the exception of the most recent study of the United States Army-Baylor Doctoral Program,¹⁶ all are over 10 years old. In the last 10 years, 100% of PT education programs have progressed to a professional doctorate. Further, a recent systematic review focusing on quality enhancement in PT education identified insufficient evidence and lack of methodological rigor in research addressing clinical education models.²

The implementation of a collaborative clinical education model has been the subject of intense discussion for several decades. The collaborative model, in various formats, has been used in other health care professions such as nursing and medicine. Investigations into the collaborative model have been ongoing for several decades, resulting in mostly positive outcomes.^{4–6,15,16,20,21} Most of this research, however, has been limited in breadth of data analyzed, is somewhat dated, and of a lower hierarchical level of evidence. Many qualitative and descriptive research articles have provided good insight into the perceptions and attitudes toward the use of the 2:1 model, as well as the model's benefits and barriers. Research has been conducted globally with consistent results reported.^15,17,20 Much qualitative research has found acceptable levels of interaction, engagement, and learning by clinical instructors and students.^4,5 A more recent study supported the positive effect on the quality and benefits of collaborative clinical education models, stating that these models afforded the opportunity for a high level of mentorship and instruction with appropriate levels of individualized attention. It also noted little or no variability in productivity or efficiency with collaborative models, indicating no negative effect on productivity. This study, however, was conducted at academic medical centers associated with the investigators, limiting generalizability of findings.¹⁶

The present study aims to overcome limitations in the literature by using a larger number of clinical instructors and students to evaluate the productivity of clinical instructors over a 3-year period when they have no students (0:1), 1 student (1:1), and 2 students (2:1). The primary aim of this study, therefore, is to examine the productivity of PTs participating in both the 1:1 and 2:1 model of clinical education over a 3-year time period in an acute care setting. The authors hypothesize that PT productivity will increase in both the 1:1 and 2:1 models of clinical education when a longer time period of analysis is examined, and that PTs conducting 2:1 models of clinical education will have increased productivity. The results of the study are expected to have important implications for the culture of clinical education. If efficiency and patient productivity can be maintained or enhanced while using the collaborative model of clinical education, this model could gain support by the clinical community, thus ensuring more student experiences in the acute care setting and addressing the increased demand for available clinical education experiences by PT education programs.

METHODS

This study was a longitudinal design involving a retrospective review of PT productivity over a 3-year timeframe for services provided at Orlando Regional Medical Center (ORMC) in Orlando, Florida. ORMC is a part of Orlando Health, a comprehensive private, not-for-profit health care network. This PT practice was chosen as a convenience sample due to its location and widespread support of clinical education with numerous PT education programs.

The retrospective review included productivity as the dependent variable, which was measured in 20-minute units of billable PT services (units) provided each day over 3 full calendar years and involving full-time PTs at ORMC. Thus, a “unit” represented a billable service that could denote various physical therapy practices, such as a therapeutic exercise or patient evaluation. All clinical instructors at the hospital had at least 1 year of clinical experience and 6 months experience of working in the facility at the time of data collection. During the timeframe of this study, the productivity data collected involved a total of 20 PTs. The units billed during each calendar day by each PT were recorded into a database. This productivity database was cross-referenced with the clinical education calendar to determine the days in which the PT was not serving as a clinical instructor, and those in which he or she served PT students in either a 1:1 or 2:1 clinical education model.

The productivity units in a given day were included for all days in the 3-year period (calendar year), during which PT practice standardized to an 8-hour work day. Days with reduced productivity due to meetings, patient rounds, onsite continuing education, or other nonclinical-related activities were not eliminated from the study. Thus, all days in which the PT was paid as an employee, outside of personal time off, were included. PT productivity was excluded from the study if it involved physical therapist assistant (PTA) students in either a 1:1 or 2:1 model. Potential bias was also controlled through inclusion of all PT students, regardless of clinical length, clinical experience, and institution.

Data Analysis

Descriptive and inferential statistics were generated and analyzed with IBM SPSS Statistics for Windows, version 22 (Armonk, New York).²² Productivity for the given dates was analyzed based on the PT serving as a clinical instructor for 0, 1, or 2 students each day over the 3-year period. Baseline productivity was assessed between the groups of clinical instructors by dichotomizing the productivity of those therapists who do versus do not participate in the 2:1 model of clinical education, and by comparing the group differences with an analysis of variance (ANOVA). Mean group differences for productivity of PT services for clinical instructors while supervising 0, 1, and 2 students were compared using an ANOVA, with further investigation through a post hoc test. In addition, to further corroborate the results and control for potential differences at baseline productivity (not serving as a clinical instructor), an analysis of covariance (ANCOVA) was performed to assess productivity differences among the groups. Within-group differences were then assessed through repeated measures ANOVA to evaluate productivity changes based on the addition of each clinical education student. Level of significance was set at P < .05 for all analyses.

RESULTS

Demographic Information About Clinical Education Facility

The clinical education facility is an 808-bed acute care hospital in an urban setting specializing in trauma, critical care, emergency care, cardiology, orthopedics, and neurology. PT students attended each of these settings during the period over which data were assessed. The standard productivity expectation for full-time PTs is 16 billed units in an 8-hour work day. The clinical education facility affiliates with 15 universities for PT clinical education. During the 3-year period in which the productivity data was assessed for this study, 196 students were educated at the hospital during 45 total clinical education time periods, and all 15 PT institutions were represented. Forty-seven of the students (24%) were on their first clinical experience, 97 (49.5%) were on an intermediate clinical experience, and 52 (26.5%) were on their terminal clinical experience. Four students included in the data were remediation students sent out of sequence by request of contracted institutions. None of the remedial students had been in the facility prior to the clinical experience. The range of the length of the clinical education experiences was 6 to 15 weeks. During the 3-year data collection time period, clinical education programs from the universities cancelled 43 rotations. Facility demographics are outlined in Table 1.

Table 1: Clinical Education Facility

Demographic Data of Physical Therapists

Data from 20 full-time PTs were collected in this study. All 20 (100%) PTs were American Physical Therapy Association (APTA) credentialed clinical instructors, with 25% (5/20) being advanced APTA credentialed clinical instructors. The mean years of experience of the PTs was 5.25 years (SD 4.99). The 2:1 PTs possessed a mean 8.4 years of experience, while the 1:1 PTs had a mean 4.2 years of experience. Sixty-percent of the PTs possessed entry-level Doctor of Physical Therapy (DPT) degrees, with an additional 30% possessing DPT degrees earned in the transitional model. PT demographic data are reported on Table 2.

Table 2: Physical Therapist Characteristics

Productivity

There was a total of 8,951 days of productivity assessed in this study, with an overall mean of 17.23 units (SD 3.8) billed per day. When analyzed by days with a PT student in clinical education, PT productivity was 16.17 (SD 3.27) with no student (0:1 model), 17.05 (SD 3.35) with 1 student (1:1 model), and 21.53 (SD 5.61) with 2 students (2:1 model). Descriptive data about therapist productivity are provided in Table 3.

Table 3: Productivity for All Clinical Instructors

A further assessment was conducted to determine if there are baseline differences between clinical instructors who used the 2:1 model versus those who did not use the 2:1 model. The baseline ANOVA found statistical significance between the groups (F = 18.81, P < 0.001), with clinical instructors not using the 2:1 model having a higher baseline mean (16.81, SD = 3.38, 95% confidence interval (CI), 16.7–16.9) than those using the 2:1 model (16.42, SD = 3.05, 95% CI, 16.3–16.6). Results of this comparison are in Table 4.

Table 4: Overall Baseline Productivity: All CIs when Practicing With no Students - Units per 8-hour Day

A 1-way ANOVA examining betweengroup differences revealed significant differences between the PT productivity when using the different clinical education models (0, 1, 2 students)(F_2,8978 = 685.4, P < .001). The post hoc test identified statistically significant differences (P < .001) in all pairwise comparisons. Group comparisons of clinical instructors working with 0, 1, and 2 students are presented in Table 5. A follow-up ANCOVA was also performed to address concerns of the baseline productivity serving as a covariate that may influence group comparisons when students are utilized in patient care. The ANCOVA revealed statistically significant differences (F = 559.1, P < .001, partial eta squared = .163) when controlling for baseline productivity of the PTs.

Table 5: Clinical Instruction Group Comparisons - Productivity in Units per 8-hour Day

PTs were then grouped based on whether or not they participate in the collaborative model (2:1) of clinical education and assessed for within-group differences in productivity using a 1-way repeated-measures ANOVA. The ANOVA considered 851 days in which the clinical instructor supervised 2 students, and compared means with the same number of days instructors served 1 and 0 students. The mean number of units billed by PTs who participate in the 2:1 model of clinical education for this sample of 851 days each was 16.77 (SD = 3.15) without a student, 17.36 (SD = 3.61) with 1 student, and 21.52 (SD = 5.61) with 2 students. The repeated measures ANOVA indicated significant changes (F = 305.6, P < .001). There was a statistically significant increase in productivity with each student added in clinical education (P < .05). Increase from 0 to 1 student resulted in a mean difference of .591 units, change from 1 to 2 students resulted in a 4.16 unit increase, and change from 0 to 2 students resulted in a mean increase of 4.75 units (95% CI, 4.21–5.29). Data for the repeated-measures ANOVA for clinical instructors working with 0, 1, and 2 students is presented in Table 6.

Table 6: Productivity of Clinical Instructors: Changes in Units Billed per Day when Serving as CI

The mean number of units billed by PTs who do not participate in the 2:1 model of clinical education was 15.95 (SD = 4.3) without a student, and 16.77 (SD = 3.20) with 1 student. The repeated measures ANOVA indicated a significant change from 0 students to 1 student (F = 26.88, P < .001). There was a significant increase in productivity (P < .001) when a student was added in clinical education, with a mean increase of .818 units (95% CI, .508–1.13). Data for the repeated-measures ANOVA assessing within-group differences for clinical instructors who only do the 1:1 model are presented in Table 6.

DISCUSSION

The primary aim of this study was achieved through the analysis of the large sample of PT productivity over a 3-year timeframe. This study examined 8,981 days of PT productivity, of which, 3,254 days included the involvement of at least 1 physical therapy student. This study found statistically significant differences in PT productivity in the acute care setting while serving in any capacity as a clinical instructor. Considering the analyses displayed an average increase between 1.05 and 5.53 units of physical therapy per day, the results demonstrate a significant impact in productivity. These results will be of high interest to both clinical sites and academic institutions. When comparing the mean productivity differences between the groups of clinical instructors while controlling for baseline productivity, a statistically significant difference was observed between each model (0, 1, 2 students). In addition, within-group differences were observed with statistical significance for the clinical instructors when there was an addition of a student at any level.

As noted, the group differences in productivity were statistically significant on all levels. On average, the productivity of the group with 1 student was 0.37 units greater than the group without a student. Although this may not be a substantial increase displaying a large clinical impact, it does deflect the notion that clinical education has a negative impact on PT productivity. In addition, this difference was demonstrated in an analysis of over 3 years of data. Of even greater impact, however, is the finding that PTs treating in the 2:1 model had a mean productivity that was 4.48 units greater than the group with 1 student, and 4.85 units greater than the group with no students. This data is further supported in the within-group analyses performed. The 2:1 clinical instructors increased their productivity by an average of 0.59 units (95% CI, 0.18–0.99) when moving from 0 to 1 student, and increased by an average of 4.16 units (95% CI, 3.6–4.7) when moving from 1 to 2 students. The clinical instructors not utilizing the 2:1 model saw an average increase of 0.82 units (95% CI, 0.51–1.12) when moving from no students to 1 student in clinical education.

While the changes in productivity are reached on a statistically significant level over the 3-year period, the clinical significance warrants further discussion. The primary dependent variable in this study was productivity in terms of units of therapy services billed per day. A change of less than 1 unit may not be clinically impactful. However, one can more readily determine that the changes seen in productivity in 1:1 or 2:1 models are clinically significant. The standard of practice at this facility is 16 units per day. The results of this study suggest that moving from 1 to 2 students equated to reaching over 125% of the productivity standard. One additional consideration in regard to this study is that the units were not examined on a level to define the exact service and reimbursement amount. A unit not only included daily interventions, such as therapeutic exercise, but also more substantially reimbursable services such as physical therapy evaluation.

It should also be noted that the present study included all days that the PT was practicing, not just days spent treating patients for an entire day. It did not eliminate therapist days of billing that involved nonclinical activities to account for the typical activities and responsibilities of PTs in the acute care setting (eg, patient rounds, training and development, quality assurance meetings, etc). Thus, the authors feel the results are externally valid. More important are the considerations involved with the demographics of the PT students during the timeframe of data analysis in this study. The clinical site is affiliated with numerous institutions, both inside and outside of its residential state. The duration of the clinicals ranged from 6 to 15 weeks. Contrary to other studies, the clinical site involved nearly an equal distribution of first and final affiliation students (24.0% and 26.5%, respectively). This distribution is rather unique considering other published studies have been limited to more experienced cohorts or those from only 1 institution.^15,16,18 Thus, the authors believe that the productivity data is not skewed because of perceived ability of students due to their experience or academic training. Lastly, although the data were not dichotomized based on practice area within the acute care hospital, the clinical instructors represented all areas within the acute care level-1 trauma center (critical care, cardiovascular, orthopedics, neurology, oncology, step-down, and intensive care).

Examining the empirical data longitudinally over a 3-year period yields greater reliability and validity of results, rather than using data for a single internship period or over a shorter time span. Studies that gathered data from using 1 or 2 internship periods, a small sample of clinical instructors, and students at the same level or from the same academic institution could create a bias in favor of positive results. Again, most of those studies did not examine productivity data. Utilizing data from 20 clinical instructors at varying stages of clinical practice who were supervising 196 students at varying levels of clinical rotation from 29 academic sites around the nation increases validity of the results following the sample of 8,981 days of productivity in the data analysis.

The sample involved in this study included productivity from clinical instructors who both do not participate in the 2:1 model and do participate in the 2:1 model. Surprisingly, at baseline, even though they possessed slightly less clinical experience, the clinical instructors who do not participate in the 2:1 model possessed a greater productivity than those who do participate in the 2:1 model. Although statistical significance was reached in this analysis, the differences in mean units (16.81 versus 16.42) may not be clinically significant. A follow-up assessment of the groups comparing productivity when both involved with 1 student, however, demonstrated no statistical significance between the groups. Furthermore, the addition of the ANCOVA to control for baseline differences yielded no impact to the results and substantiated the significance of group differences.

LIMITATIONS

Through the inclusion of productivity data from a large sample involving students at various levels from numerous institutions working in varied settings within the acute care hospital, the investigators are confident in the generalizability of this study's effects to a broad clinical education population in the acute care hospital setting. However, like all empirical studies, the current study is not without limitations. First, the clinical site presented in this study has a strong commitment to clinical education. As evident in its demographics, the hospital has a large contingency of PTs serving as clinical instructors, with 100% of them being certified by APTA. In addition to this training, it can be calculated that the average instructor included in the study was serving as a clinical instructor for 163 days per year (45% of the year). With such a level of experience, formalized training, and a large contingency of colleagues in clinical education immediately accessible, the sample of clinicians in this study likely would produce a greater ability to manage their patient caseload than others. In addition, the study did not take into account whether the increased productivity in the 2:1 model might have been influenced by the years of practice of the instructors. Although the 2:1 clinical instructors possessed a lesser productivity at baseline (no students), they achieved greater productivity in 1:1 models than their counterparts. This data would support the notion that they are effective clinical instructors in any model of clinical education, and also resonate with the educational culture of the facility.

Although it was not possible to assess the potential effect of APTA clinical instructor credentialing on the results (since all the instructors in this study were already APTAcredentialed at the basic level), there was a greater number of clinical instructors who perform 2:1 clinical education certified at the advanced level. Although this study did not seek to examine this potential influence, it suggests that there is likely an association with this level of credentialing and years of practice as a clinical instructor. In addition, it is difficult to determine the effect that the PT environment may have had on the results. The acute care setting for the hospital (ORMC) where the data were retrieved provides a flexible and supportive environment that allows the clinical instructors time to orient their students and gradually increase patient care productivity. A supportive environment has been identified as a key factor in the success of a clinical education program,^7,14,23 and when not present, is a stressor to the clinical education faculty.⁸

Another limitation to the study is that it did not include patient outcomes in the assessment. The addition of such data would undoubtedly add to the body of knowledge and discussion in regard to clinical education models. Although productivity concerns have been noted as 1 of the greatest barriers to clinical education for several years,^3,18 there have been few studies focusing on clinical outcomes. Two recent studies, however, focused on outcomes of patients being treated by students versus licensed PTs. Although neither found statistical significance, both identified trends of lesser outcomes in the student population that warrant further investigation.^24,25

Last, the current study did not include an assessment of qualitative measures for the students’ clinical experiences. This component would certainly add to the examination of actual learning quality within the clinical education models. It is possible that data gathered in smaller, less diverse acute care settings may have yielded different results, but if the collaborative model is implemented correctly, there should nevertheless be a positive outcome in clinical instructor productivity and in the efficiency of patient care.

Future Directions

In consideration of previous studies examining collaborative models in clinical education, it is apparent that adoption of the collaborative model as a standard of practice for PT clinical education will require a culture shift from all the stakeholders involved. Due to ongoing concerns of clinical education productivity, some authors have proposed changes to lengthen clinical education experiences to achieve greater productivity gains.^16,19,26 The findings in this study, however, based on overall mean productivity analyses, demonstrate that increased patient productivity can occur through either 1:1 or 2:1 models of clinical education in clinical rotations ranging from 6 to 15 weeks in the acute care setting. In consideration of the possibility of implementing the collaborative model as a standard of practice for PT clinical education, clinical education faculty must consider all factors contributing to an effective model. Stakeholder perceptions, misconceptions, and concerns should be addressed through the results of the current research.

In the literature, there is an underlying agreement that the collaborative model is an effective and valuable learning method, but the 2:1 model is not guaranteed to be successful for all clinics or all instructors. Clinic caseloads, administrative support, clinical settings, instructor training, and student preparedness are all factors to be addressed when considering a collaborative model of clinical education. As indicated in the discussion of this study's findings, limitations, and those of other relevant studies, the support of both the clinical sites and academic institutions are an essential component for creating a successful culture in clinical education.^{7–10,14,21,23}

CONCLUSION

The current study demonstrates the ability for PT productivity to increase in the acute care setting in both a 1:1 and 2:1 clinical education model. While the 2:1 model provided the greatest increase in productivity during the 3-year timeframe, the productivity of all PTs in the study significantly increased with each added student. As indicated in previous qualitative studies and reemphasized herein, a supportive atmosphere in the clinical setting with the appropriate level of communication and supervision is warranted for success in all models of clinical education. The ability to successfully implement the 2:1 model will be dependent upon the institutional support, training, and experience of the clinical instructor. This study adds to the body of knowledge in regard to clinical education by providing quantitative data to support the positive impact that clinical education can have on productivity in the clinical setting. More research is needed to examine clinical productivity of PTs in other settings of patient care not evaluated in this study, and future studies should also include strong consideration of outcomes of patients treated by students and clinicians in the 2:1 model of clinical education.