The Integration of the Inaugural Graduates of the McMaster University Physician Assistant Education Program Into Ontario's Health Care System - Report November, 2010

E-mail Print

physician-assistant-program

Chan Kulatunga-Moruzi, M.Sc., Ph.D.
Educational Scientist, Physician Assistant Education Program, Assistant Professor, Department of Family Medicine, Faculty of Health Sciences, McMaster University, Michael G. DeGroote Centre for Learning and Discovery, rm. 2209, Hamilton ON, L8N 3Z5. Email: This e-mail address is being protected from spambots. You need JavaScript enabled to view it Phone (905) 525-9140 x 21584 Fax (905) 527-0131.

This research was supported by a grant from the Ontario Health Human Resources Research Network. I would like to thank John Cunnington, Nancy Weller, Kareen McCaughan and Danielle Laffan with the McMaster PAEP for their help with data collection and Ivy Bourgeault at the University of Ottawa for her support.

Synopsis

In September 2008, McMaster University admitted its inaugural Physician Assistant class into the first civilian physician assistant program in Ontario, with this cohort graduating in November 2010. As pioneers of this new profession, these graduates have secured positions in a variety of health care settings throughout Ontario, working collaboratively under the supervision of a physician. This report presents the results of an Employer Survey of this inaugural class. The online survey gathered a variety of information including where the graduates are working, initial assessment of employers’ perceptions of graduates’ preparedness and competence, as well as preliminary data on their perceived impact on the delivery of care at each site. The results indicate that supervising physicians were knowledgeable about physician assistants. Many of them had worked with a physician assistant in the past and were familiar with the training provided by the program as well as the scope of practice of physician assistants in Canada. Although the “Physician Supervision Grant” was less of a factor in their decision to hire a new physician assistant graduate, the “Provision of Physician Assistant Salary and Benefit Grant” was highly influential. Graduates were perceived to be well trained as indicated on the CanMeds and National Competency Profile ratings. Supervisory trends indicate that the large majority of physician assistants are supervised by more than one supervising physician and that the number of supervisory hours decreases with time. Supervising physicians identified “improving quality of patient care” as the primary motivating factor for hiring a physician assistant. Consistent with their motivations for employing a physician assistant, “improving quality of patient care” was also cited as the area in which physician assistants are having the most impact. In general, supervising physicians felt that the employment of a physician assistant had more impact on patient outcomes than personal outcomes. All data including free response comments indicate very positive outcomes.

Abstract

Background: In September 2008, McMaster University admitted its inaugural Physician Assistant class into the first civilian physician assistant program in Ontario, with this cohort graduating in November 2010. As pioneers of this new profession, these graduates have secured positions in a variety of health care settings throughout Ontario, working collaboratively under the supervision of physicians. All employers of this cohort participated in the “Physician Assistant Graduate Grant Initiative” financially supported by MOHLTC. The grants were provided to encourage the employment of the new graduates and to offset the cost of employing a physician assistant full-time for a period of  two years. A one-time physician supervision grant was also provided to the supervising physician(s) to compensate for their supervisory duties.

Purpose: This report presents the results of an Employer Survey of this inaugural class. The survey was designed to gather a variety of information including where the
graduates are working, initial assessment of employers’ perceptions of graduates’ preparedness and competence, as well as to gather preliminary data on their impact on
the delivery of care at each site.

Methods: In February 2011, 4 to 6 months following the commencement of the graduates’ employment, the supervising physicians (employers) of the inaugural Physician Assistant Education Program class were invited to complete an employer survey. The survey was conducted online and 22 responses were received reflecting the performance of 20 graduates. Completed surveys were downloaded, tabulated and statistically analyzed to uncover potential patterns.

Results and Discussion: The results indicate that supervising physicians were knowledgeable about physician assistants. Many of them had worked with a physician
assistant in the past and were familiar with the training provided by the program as well as the scope of practice of physician assistants in Canada. Although the “Physician Supervision Grant” was less of a factor in their decision to hire a new physician assistant graduate, the “Provision of Physician Assistant Salary and Benefit Grant” funding was highly influential. Graduates were perceived to be well trained as indicated on the CanMeds profile and National Competency Profile ratings. Supervisory trends indicate that the large majority of physician assistants are supervised by more than one supervising physician and that the number of supervisory hours decreases with time. Although employers had several reasons for hiring a physician assistant, improving quality of patient care was identified as the primary reason. Consistent with the supervising physicians’ reasons for employing a Physician Assistant, improving quality of patient care was also cited as the area of most impact. In general, supervising physicians  felt  that  the  employment  of  a  physician  assistant  had  more  impact  on  patient  outcomes  than  personal  outcomes.  All data,  including  free  response  comments,   indicate  very  positive  outcomes.

McMaster Physician Assistant Graduates: Initial Data and Analyses

Background

The role of Physician Assistant was announced in May 2006 as part the government’s health human resources strategy, HealthForceOntario. The aim of this strategy was to improve patient care, provide more timely access to care and improve patient satisfaction by ensuring the appropriate number and mix of health care providers to Ontarians. The Physician Assistant Initiative was introduced to the Ontario health care system through multi-phase demonstration projects.

In January, 2007, the first phase began with the introduction of physician assistants, nurse practitioners and acute care nurse specialists joining 6 hospital emergency departments. The results of this initial project were generally positive indicating that the introduction of these provider roles had a positive impact on health care delivery,
including shorter wait times, shorter stays in the emergency departments and fewer patients leaving without being seen.

The second phase entailed the introduction of more than 50 physician assistants into hospitals, community health centres, diabetes care centres and long-term care settings across the province. In each of these settings, physician assistants worked collaboratively with other providers. The positive results of this phase led to the
expansion of the demonstration projects, introducing approximately 20 additional physician assistants into emergency departments and approximately 20 physician
assistants in family health teams.

The physician assistants participating in the demonstration projects were either trained in the Canadian Military, through an accredited American Physician Assistant Program or were internationally trained physicians who participated in an assessment process and 14-16 week integration program to orient them to the physician assistant profession.

As part of the physician assistant initiative, the Ministry of Health and Long-Term Care (MOHLTC) gave all Ontario medical schools an opportunity to submit expressions of interest to establish physician assistant education programs. Proposals were submitted by McMaster University and by the University of Toronto (in collaboration with the Northern Ontario School of Medicine and the Michener Institute for Applied Health Sciences) to develop physician assistant education programs.

In September 2008, McMaster University admitted its inaugural physician assistant class into the first civilian physician assistant program in Ontario. The program, which runs 24 months, focuses on the biological and clinical sciences underpinning health care in the first year, followed by a year of clinical placements in each of five clerkships, Medicine,

Emergency Medicine, Family Medicine, Psychiatry and Surgery. In November 2010, the inaugural class of 21 students graduated with a Bachelor of Health Sciences (Physician Assistant) degree. As pioneers of this new profession, these graduates have secured positions in a variety of health care settings throughout Ontario, working collaboratively under the supervision of a physician.

All employers of this first McMaster cohort participated in the “Physician Assistant Graduate Grant Initiative” financially supported by MOHLTC. The grants were provided
to encourage the employment of the new graduates and to offset the cost of employing a physician assistant full-time for a period of two years. Placement opportunities were focused in high priority areas to help reduce wait times and improve access to patient care. One hundred and fifty applications for funding were received from potential employers in a variety of geographical and clinical settings. Of these, over 100 sites were approved as possibilities to fund, with funding going only to the sites that successfully recruited a graduate. The sites approved for funding were based on employment setting and demonstrated need. Rural sites, sites in which there are
shortages of providers and sites with staffing challenges were given priority and funded at a higher level. Grants awarded ranged from $46,000 to $92,000, with the majority (15 sites) qualifying for the lower level funding of $46,000.

In addition to funding to support physician assistant salary, a one-time physician supervision grant of $10,000 was provided to the supervising physician, or group of
supervising physicians. This grant, issued 6 months after the commencement of physician assistant employment, was intended to support the time and effort required
to integrate the physician assistants into the clinical setting.

Introduction

In February 2011, the supervising physicians (employers) of the inaugural Physician Assistant Education Program from McMaster University were invited to complete an
employer survey. The survey was designed to gather a variety of information including where the graduates are working, supervisory trends, initial assessment of employers’ perceptions of graduates’ preparedness and competence, preliminary data on their perceived impact on the delivery of care at each site as well as general satisfaction with graduates’ performance. The survey was conducted online and 22 responses were received reflecting the performance of 20 graduates. One graduate moved overseas and was not employed at the time of survey. Reported below are highlights of the main findings. Graduates had been employed for 4-6 months at the time of survey.

This report presents the results of the Employer Survey for the inaugural McMaster University Physician Assistant Education Program. Since some response distributions
may be skewed due to one or a few extreme scores, the median, which is a more appropriate measure of central tendency under such conditions, is provided along with
the mean. The mode, reflecting the most frequently occurring response is also provided along with the standard deviation to provide a measure of variability.

1. Recruitment Trends: Employers and Placement of the Graduates.

Many of the employers (46%) knew the employee prior to employment, either as a preceptor in a clerkship rotation (23%) or in an elective (23%). The majority of
supervising physicians (59%) had worked with a physician assistant in some capacity in the past.

To further understand the employers’ background with respect to physician assistants, employers were asked to rate their familiarity with the training provided by the
McMaster University Physician Assistant Education Program as well as their familiarity with the scope of practice of physician assistants in Canada on a 9-point scale
(1 = completely unfamiliar, 9 = extremely familiar). Familiarity ratings of the training provided by the Physician Assistant Education Program at McMaster University were
high (median 6.0, standard deviation 2.0). Familiarity rating of the scope of practice of physician assistants in Canada was also high (median 7.0, standard deviation 1.9).

Although where the graduates are employed is based on which sites applied for a grant through the “Physician Assistant Graduate Grant Initiative,” there were over 100
opportunities available to the graduates to choose from. The graduates are working in a variety of disciplines; however, most of the graduates are working either in emergency rooms (40%) or in internal medicine (30%). See Table 1.

table1-2

 

The median and modal physician assistants salary reported in the survey was $75 000. Salaries ranged from $70 000 to $85 000 (standard deviation 3 500). Many factors such as clinical discipline, geographical location, characteristics of the location and the clinical background and experience of individual physician assistants are likely to play a role in the salary distribution. Further analysis with subsequent graduates will allow for identifying which of these factors have the most impact on physician assistant salary.

2. PA Task Profile

In order to understand how the physician assistants are being utilized, supervising physicians were asked to rank order the frequency with which their physician assistant
performs various tasks (with 1 being the most frequently performed). The options provided in the survey were:

  • Taking medical histories and performing physical exams
  • Ordering and interpreting tests
  • Performing clinical procedures
  • Developing treatment plans
  • Chronic care management
  • Counselling (e.g. addiction, smoking cessation, lifestyle)
  • other, please specify

Table 2 provides summary statistics. The mean and median ranks indicate that the most frequently performed task is “taking medical histories and performing physical exams.” “Counselling” is the least frequently performed task. Although there is some variability in the ranks assigned to the various tasks, Friedmans rank order test was statistically significant, χ2(5, N = 17), p < .0001. Post-hoc analyses using Wilcoxan Sign Rank test indicate statistically significant differences between many of the pair-wise comparisons (all post-hoc comparisons below Table 2 are significant at p < .05). Kendall’s W = 0.47, indicating considerable agreement among supervising physicians in the rank order of the tasks performed by physician assistants. Other tasks performed by physician assistants listed by supervising physicians include teaching (4.5%), acute patient care management (4.5%), and filling of forms (4.5%).

table-2

Although it would be interesting to determine the types of tasks performed by physician assistants as a function of discipline or other characteristics of the site (e.g. urban/rural), the small samples in this cohort does not allow for this type of analysis. Pooling of data across several years would be useful for such analyses. Certainly, the type of setting would be expected to dictate, to some extent, the tasks performed by physician assistants. Physicians in emergency departments, for example, are not likely to spend time counselling or engaging in chronic care.

3. Supervision Trends

In order to understand supervision trends, employers were asked questions regarding the number of physicians supervising the physician assistant at their site, the proportion of supervision time they were responsible for as well as the number of supervision hours in a given week. Respondents only provided information with regard to their own interactions with the physician assistant. The responses to the questions pertaining to supervision trends indicate that only 36% of the respondents reported that they were the only supervising physicians responsible for 100% of the supervision. The majority (63%) of respondents reported that they shared supervisory responsibility. Of those who shared supervision, 10% reported that they shared supervision with one other physician, 20% with two other physicians, 5% with three other physicians and 45% with four or more physicians. Thus, shared supervision with more than one supervising physician is the norm for this particular cohort. For those supervising physicians who shared supervision, the percent of the total time spent supervising ranged from 5% to 90% (median 41.5%, standard deviation 30.0%). No differences in supervision trends as a function of type of clinical setting were apparent from the data. Pooling of data across several years will allow for such analysis.

While it is difficult to interpret the data because some supervising physicians did not supervise in all time periods queried (1st month, 2nd - 4th month, 4th month and beyond) in the survey, there appears to be a decrease in the number of supervisory hours per week over time. See Table 3. In order to examine the trends more closely, the response options (< 3, 3‐5, 6‐8, 9‐11, 12‐14, 15+) were substituted with the mean number of hours for each of the options (1, 4, 7, 10, 13, 16) and entered into a repeated measures analysis of variance. The results of this analysis indicate a decreasing trend with time F(2, 36) = 9.2, p < .001. While there is a decrease from period 1 (1st month) to period 2 (2nd- 4th month), post hoc analysis using the Newman-Keuls test only indicated a statistically significant decrease in period 3 (4th month and beyond). See Figure accompanying Table 3 (all post-hoc comparisons below Table 3 are significant at p < .05).

Further analyses with a larger sample would be helpful in determining whether the downward trend over each of the time periods is significant. Examining the number of
supervisory hours in shorter time periods or simply asking for an estimate of the number of hours supervising might also be helpful in understanding the trends.

table-3

4. Motivational Factors for Employing a Physician Assistant

There are numerous potential ways in which the utilization of physician assistants can have an impact on the delivery of health care. In order to understand the reasons for hiring a physician assistant, supervising physicians were asked to rank order the reasons for employing a physician assistant (with 1 being your primary reason). The options provided in the survey were:

  • Increase the number of patients seen per day
  • Improve quality of patient care (e.g. ability to spend time on more complex cases).
  • Decrease patient wait times
  • Increase access to care (number of patients on roster)
  • Increase time allotment for learner supervision (e.g. clerks, residents)
  • Increase time allotment for professional development
  • Reduce workload
  • Increase pay
  • Improve work-personal life balance
  • Other (please specify)

Table 4 provides the mean and median ranks and the frequency data for each of the options. The primary reason supervising physicians cited for employing a physician
assistant was to “improve the quality of patient care” followed by “increase the number of patients seen per day.” “Increase supervising physician salary” was cited as the least motivating factor. Other reasons cited by supervising physicians for employing a physician assistant not provided as a response option were: restructuring issues (4.5%) and increase time for research (4.5%).

Although mean and median ranks provide some idea of which factors supervising physicians regard as important, there is significant variability and these findings must be interpreted as tentative. This is supported by Kendall’s W = 0.35. An examination of the frequency per rank provides a sense of this variability. Although Friedman’s rank order test was significant χ2 (8, N = 11) = 30.8, p < .0001, post-hoc pair-wise comparisons using Wilcoxan Sign Rank Tests indicates that the only statistically significant difference in rank was for “ increase pay.” (All post-hoc comparisons below Table 4 are significant at p < .05).

table-4

Grouping the survey response options provided into those reflecting patient outcomes (Increase the number of patients seen per day, Improve quality of patient care, Decrease patient wait times, Increase access to care) or personal outcomes (Increase time allotment for learner supervision, Increase time allotment for professional development, Reduce workload, Increase pay, Improve work-personal life balance), indicates that the reasons for employing a physician assistant are more centered around patient outcomes than personal outcomes. With the exception of reducing workload, the options pertaining to patient care were sited as being more significant motivating factors in employing a physician assistant. In interpreting these findings it is also important to keep in mind that many of the constructs of the reasons provided are inter-dependent and thus, the evaluations are not likely orthogonal.

5. Perceived Preliminary Outcomes of Physician Assistant Placement

In order to understand in which areas the physician assistants are perceived to be having the most impact, supervising physicians were asked to rank order the same
response options as above (with 1 being your primary reason). Table 5.1 provides the mean and median ranks and the frequency data for each of the response options.

As with the previous data on the reasons for hiring a physician assistant, although the mean and median ranks provide some idea of which areas are being most affected by physician assistants, there is significant variability and these findings must be interpreted as tentative. This is supported by Kendall’s W = 0.14, indicating considerable disagreement in ranking. Friedman’s rank order test approached statistical significance, χ2 (8, N= 12) = 13.2, p = .10. An examination of the frequency data for each
of the response options provides a sense of this variability.

table5-1

Although the graduates had only been employed for 4-6 months, many of the supervising physicians perceived an impact. The perception of which areas are being
affected by physician assistants is largely consistent with supervising physicians’ reason for employing a physician assistant. The average and median ranks indicate that
supervising physicians feel that physician assistants have the most impact on improving the quality of patient care followed by increasing the number of patients seen per day. Increase in supervising physician salary was cited as the factor with the least impact. See Table 5.2.

Also consistent with the reasons for employing a physician assistant, grouping survey response options into those reflecting patient outcomes or personal outcomes,
indicates greater perceived impact on patient outcomes.

table5-2

In order to gauge the overall impact of the physician assistants at each site, supervising physicians were asked to rate “the impact of the Physician Assistant on the way your site delivers health care” on a 9-point scale (1= extremely negative impact, 5 = no impact, 9 = extremely positive impact). The median impact rating was 8.0 (standard deviation 1.3), indicating a very positive impact. Below is a sample of the comments provided by supervising physicians to support their numeric ratings.

“Does Q/A, targets core population, leaves time to see more complex patients plus allows time for teaching.”

“has allowed more efficient delivery of care.”

“Sees lots of patients.”

“Too early to assess if affecting our Physician Initial Assessment Times.”

“Better access, good care.”

 6. Expectations and Perceptions of Preparedness of New Graduates

In order to assess employers' perceptions of the graduates' preparedness for the role of a physician extender, they were asked to rate their employee on a 9-point scale (1 = completely unprepared to 9 = completely prepared) on each of the CanMeds Competencies, Medical Expert, Communicator, Collaborator, Scholar, Health Advocate, Manager and Professional. The ratings indicate that employers felt that the graduates were well prepared for each of the roles. The median satisfaction ratings range from 7.0 (Medical Expert, Manager, Scholar) to 8.0 (Communicator, Collaborator, Health Advocate, Professional). A repeated measures analysis of variance indicated significant differences in perceptions of preparedness across the 7 competencies F(6,102) = 7.66, p < .00001. Employers felt that graduates were most prepared as a professional and least prepared as a manager. See Table 6.1 for summary statistics and specific comparisons (all post­‐hoc comparisons below Table 6.1 are significant at p < .05 or less).

table6-1

In order to determine whether the graduates have obtained the necessary knowledge, skills and competencies as defined by the National Competency Profile, employers were also asked to rate how satisfied they are with the graduates’ level of preparedness on specific competencies on a 9-point scale (1 = extremely dissatisfied to 9 = extremely satisfied). The ratings indicate that employers were very satisfied with each of the skills and competencies. Median scores ranged from 7.0 to 9.0. See Table 6.2

table6-2

7. Satisfaction with New Graduates.

In order to gauge the level of satisfaction with the quality of training provided by the McMaster University Physician Assistant Education Program, employers were asked two questions: How satisfied are you with this employee and how likely are you to hire another McMaster University graduate in the future on a 9-point scale (1 = extremely dissatisfied to 9 = extremely satisfied). Employers were also asked to provide comments about their numeric rating. The median rating regarding satisfaction with the PA they supervise was 9.0, indicating the highest level of satisfaction (mean = 8.1, mode = 9, standard deviation = 1.3). The median rating indicating the likelihood of hiring another McMaster University graduate was 8.0 (mean = 7.7, mode = 8, standard deviation = 1.6).

This score reflected two employers who gave a rating of 5, commenting that it was too early to assess. Provided below are a selection of representative comments.

“Doing an outstanding job. Helps us do our jobs well. Hard working very professional.”

“Very professional, dedicated and a fast learner. She will become an invaluable asset when training phase is complete.”

“Exceptional work ethic, knowledge base, and clinical skill.”

 8. Impact of Physician Assistant Hiring Incentives

In order to elucidate the role of the government grants in hiring the physician assistants, employers were asked to rate on a 9-point scale how influential (1= no influence at all, 9 = extremely influential) the “Provision of PA salary and Benefit Grant” and the “Provision of the Physician Supervision Grant” were in their decision to hire a physician assistant. They were also asked to rate on a 9-point scale (1= extremely unlikely, 9 = extremely likely) scale how likely they would have been to hire a physician assistant had there been no government funding and how likely they would be to hire another physician assistant if no government funding were provided?

Supervising physicians reported that the “Provision of PA Salary and Benefit Grant” was very influential in their decision to hire a physician assistant (median = 8.0, standard deviation = 1.5). The “Provision of the Physician Supervision Grant” was less of a factor in their decision to hire a physician assistant (median = 5.0, standard deviation = 2.5). Respondents reported that they were unlikely to have hired a physician assistant if the “PA Graduate Salary and Benefits Grant” had not been available (median = 3.0, standard deviation = 2.3). Respondents also reported that they would be unlikely to hire another physician assistant if no government funding was provided (median = 3.0, standard deviation = 2.7).

 9. Free Response Comments

In order to gather information not reflected in other parts of the survey, respondents were also asked for comments. Possible suggestions of comment topics were:
integration of PA into your setting, attitudes towards PA, Interprofessional relationships. Below is a sample of the comments provided:

“I think PAs are good for patient care and clarity of communication (has the time to review information with patient), decrease length of stay (able to help organize discharges in a more timely fashion) and improve communication with allied health care workers.”

“….fit well with multidisciplinary team in the emergency room.”

“Great addition for us, functions as hospitalist is main role I would say.”

Conclusions

The results of this survey, although preliminary, indicate that the inaugural class of Physician Assistants from McMaster University is functioning well as physician extenders in a variety of clinical settings. The graduates are working collaboratively; performing a variety of clinical tasks and are typically supervised by more than one supervising physician. They are perceived to be well prepared with the need for less physician supervision over time. The responses with respect to the perceptions of supervising physicians indicate that the graduates are meeting their expectations, improving access and quality of patient care.

Download Report