University of Michigan Health Systems
Program and Operations Analysis
Physician’s Assistant Staffing and Best Practices Analysis in Urology and Otolaryngology
Final Report
To:Marc Moote, MS, PA-C, Chief Physician’s Assistant, UMHS,
Ian Perry, Management Engineer, Program and Operations Analysis, UMHS,
Andreea Duma, Management Engineer Fellow, Program and Operations Analysis, UMHS,
Prof. Mark Van Oyen, Ph.D., IOE 481 Professor, Department of Industrial and Operations Engineering,
Mary Duck, Industrial Engineer Expert & Lean Coach, Program and Operations Analysis, UMHS,
From:IOE 481 Project Team #9
Karina Hudak
Rachel Katz
Erica Segre
Erin Winn
Date:December 13th, 2016
1
TABLE OF CONTENTS
EXECUTIVE SUMMARY
Background
Methods
Findings and Conclusions
Recommendations
INTRODUCTION
BACKGROUND
Key Issues
Goals and Objectives
Project Scope
METHODS
Literature Review
Observations
Surveys
Beeper Study
Analysis of MiChart Scheduling Data
Microsoft Excel Tabulation and Visualization of wRVU Data
DATA ANALYSIS
Findings and Conclusions
Literature Review
Observations
Survey
Beeper Study
Analysis of MiChart Scheduling Data
Excel Tabulation and Visualization of wRVU Data
SUMMARY OF FINDINGS AND CONCLUSIONS
Literature Review
Observations
Survey
Beeper Study
Analysis of MiChart Scheduling Data
Microsoft Excel Tabulation and Visualization of wRVU Data
Consolidated Data Summary
RECOMMENDATIONS
EXPECTED IMPACT
REFERENCES
Appendices
APPENDIX A: Survey Questions
APPENDIX B: Survey Results
APPENDIX C: Data Adjustments
APPENDIX D: Beeper Study Forms
APPENDIX E: Figures per PA
LIST OF FIGURES
Figure ES-1:Comparison of Metrics per PA
Figure 1: Percentage of Direct and Indirect Patient Care and Miscellaneous tasks.
Figure 2: Top 10 tasks performed by Otolaryngology clinic.
Figure 3: Top 10 tasks performed by Urology clinic as shown in a Pareto chart.
Figure 4: Top Direct Patient Care tasks performed by Otolaryngology clinic.
Figure 5: Top Direct Patient Care tasks performed by Urology clinic.
Figure 6: Top 10 Indirect Patient Care tasks performed by Otolaryngology clinic.
Figure 7: Top 10 Indirect Patient Care tasks performed by Urology clinic
Figure 8: Top Miscellaneous tasks performed by Otolaryngology clinic
Figure 9: Top Miscellaneous tasks performed by Urology clinic.
Figure 10: Percentage of Appointment Type for the Otolaryngology clinic.
Figure 11: Percentage of Appointment Type for the Urology Clinic.
Figure 12: Percentage of Clinic Type Division for the Otolaryngology clinic.
Figure 13: Percentage of Clinic Type Division for the Urology Clinic.
Figure 14: Summary table of wRVU data for each of the PAs in both clinics
Figure 15: Comprehensive quantitative data for all PAs in both clinics
Figure A-1: Percentages of tasks by PA in both clinics
Figure A-2: Top 10 overall tasks per PA
Figure A-3: Top 10 tasks per clinic divided by PA
Figure A-4: Percentages of appointment tasks by PA in both clinics
Figure A-5: Percentage of clinic type division per PA
1
EXECUTIVE SUMMARY
In a variable environment such as the University of Michigan Health System (UMHS), Physician Assistants (PAs) greatly improve performance due to their high adaptability and low cost [1]. The Chief Physician Assistant would like to know what scheduling and environmental factors influence the high productivity levels demonstrated by the UMHS Otolaryngology and Urology clinics.
To assess the factors influencing these high productivity levels, the Chief Physician Assistant asked a team of the University of Michigan Industrial and Operations Engineering (IOE) Department 481 class to conduct a beeper study to understand the workload profile of PAs and compare this to the wRVU, work relative value units, data provided by UMHS. The team also collected MiChart appointment data, distributed surveys, and conducted observations at the clinics to develop recommendations regarding the high productivity levels.
Background
The wRVU is a transformation of clinical work into a measurable unit developed by the Centers for Medicare and Medicaid. UMHS has developed a unique methodology that provides a comprehensive view of PA work effort, including “shared” and “bundled” services that are not captured by traditional wRVU reporting. Using this new metric, Otolaryngology and Urology appear to be two of the most efficient clinics within UMHS; further study may reveal actionable best practices that can be extended to other services.
Four issues motivated this project. First, there were no documented best practices for PAs at UMHS. Second, previously unavailable PA-specific wRVU data had not yet been compared to the PA workload profile to understand how each activity impacts performance. Additionally, there was no analysis of clinic utilization with respect to proportion of appointment types. “Shared” and “Independent” indicate the level of physician participation in a patient visit: a Shared Visit requires direct patient care from both the supervising physician and the PA, whereas an Independent Visit requires direct patient care from only the PA. Finally, there was limited understanding of qualitative environmental factors or clinic-specific habits that aid productivity.
This project seeks to mitigate these issues and inform solutions by observing PA practices in the two clinics, establishing a workload profile, determining clinic utilization and distribution of appointment types, and tying the developed workload profile with the related wRVU data.
Methods
The team performed 6 types of tasks to evaluate and assess the Otolaryngology and Urology clinics’ high productivity levels.
●Conducted a literature review. The team conducted a literature review of 4 source types: past IOE 481 projects, government sources, previous Programs and Operations Analysis (POA) studies, and peer-reviewed journals.
●Conducted observations. The team conducted observations in the Otolaryngology and Urology clinics to understand the tasks performed and to edit the data collection sheet.
●Distributed surveys. The team created and sent out a survey to each of the 7 PAs to collect qualitative data regarding their perception of workload and productivity.
●Conducted beeper study. The team conducted a beeper study to collect data on the tasks performed by the 7 PAs in both clinics over a 10 workday period. Data was collected in the Otolaryngology clinic from November 1st until December 2nd, and in the Urology clinic from November 11th until December 7th.
●Analyzed MiChart scheduling data. The team obtained the MiChart Scheduling data for January-June 2016. With this data, the team identified the percentage of appointments labeled “Canceled,” “Completed,” “Left Without Seen,” and “No-Show” for each clinic.
●Analyzed wRVU data. The team obtained wRVU data from July 2015 to June 2016 for related Microsoft Excel analysis and used this data set to assess PA productivity levels.
Findings and Conclusions
The literature review was helpful in execution of the beeper study data collection sheet, but found no previous attempts to document best practice among outpatient PAs.
While observing each clinic, the team noted that multitasking and different styles of preparation enabled these PAs to be more productive when interacting with future patients. The style of preparation plays a role in productivity. Some PAs prefer to prepare for patient visits by brainstorming solutions beforehand while others brainstorm solutions during the patient visit. The second style results in a larger portion of time spent in Direct Patient Care, but does not necessarily correlate to an effective use of time during the workday.
From the survey results, the team concluded that clinic preparation, support staff, and scheduling influence PA productivity. Survey responses indicated that 71% of PAs perform indirect patient care tasks outside of work hours and reported that this preparation time increased their productivity in clinic. 71% of PAs indicated that they delegate some indirect patient care tasks to other care team members. PAs reported that well-trained and effective nurses and MAs are essential for good clinic performance.
From the beeper study data, the team found that the PAs in Otolaryngology and Urology spend 32.1% and 31.7% in Direct Patient Care, respectively. These percentages were comparable to past IOE studies when PAs were feeling underutilized, such as the Radiation Oncology Project in 2011 with 32% of time spend in Direct Patient Care [2]. This result demonstrates that time spent in Direct Patient Care is not correlated to high clinic productivity.
In both clinics, the most frequently performed Indirect Patient Care task was Dictation - typing in MiChart. This finding, along with the qualitative survey finding that 71% of PAs perform indirect patient care tasks outside of work hours, lead to the conclusion that both Otolaryngology and Urology PAs spend a large amount of time in dictation and paperwork.
After reviewing the MiChart Scheduling Data, the team found no correlation to high wRVU data.
Figure ES-1 presents the quantitative results from all data sets and clearly demonstrates that the only actionable correlation to high wRVU data is a high percentage of Independent Visits.
Figure ES-1:Comparison of Metrics per PA
Consolidated Team 9 data, varied sample size
A portion of appointments are missing an assigned “Shared Visit” (SV) , “Independent Visit” (IV), or “Directed Visit” (DV) charge modifier, meaning that either there is only a modifier mapping to an American Medical Association (AMA) designated circumstance or there is no charge modifier.
Recommendations
From the findings and conclusions, the team has identified that best practice exists in that Otolaryngology and Urology clinics. The following practices can be implemented to improve PA productivity across clinics:
●Increase percentage of independent visits for PAs
●Complete clinic prep outside of business hours to streamline Direct Patient Care
●Ensure effective clinic prep prior to seeing patients
●Delegate Indirect Patient Care to support staff
The team recommends that the Otolaryngology and Urology clinics pursue the following actions for improvement and further study:
●Investigate opportunities to allocate clinic prep time in business hours
●Consider compensation options for PAs working outside business hours
●Assess charge modifier workflow to assign SV, IV, and DV to every charge
●Perform further research on the use of clinic support staff
●Assess whether implementing PA scribes would be cost effective
INTRODUCTION
In the University of Michigan Health System (UMHS), the introduction of Physician Assistants (PAs) was found to be crucial due to their adaptability and lower cost, in addition to the variability of the health system [1]. Given their value to the health system, it is important for PAs to be fully utilized. The Physician Assistants in the University of Michigan Otolaryngology and Urology clinics have higher productivity levels of patient care than most other clinics, as shown through the work relative value unit (wRVU) , or measurement of work effort. Therefore, the Chief Physician Assistant would like to know what factors influence their high productivity levels. To achieve this, the Chief Physician Assistant asked a team of the University of Michigan Industrial and Operations Engineering (IOE) 481 class to conduct a beeper study to understand the workload profile of PAs and compare this data to the wRVU data provided by UMHS. Additionally, the team collected appointment information from MiChart, as well as conducted observations and interviews at the clinics regarding the PAs’ daily activities.
Based on the data collected, the team has determined what scheduling and environmental factors affect the high productivity levels of the Otolaryngology and Urology PAs. This report summarizes the team’s observation, data collection, and data analysis on the performance data of the PAs of the Otolaryngology and Urology clinics to determine and provide recommendations regarding if and which best practices exist in these clinics. The results of this project may in the future inform other clinics looking to optimize PA utilization.
BACKGROUND
The Chief Physician Assistant of UMHS identified the Otolaryngology and Urology clinics to have high wRVU data relative to many other outpatient clinics. The wRVU is a transformation of clinical work into a measurable unit, developed by the Centers for Medicare and Medicaid, and defined informally by the National Institutes of Health as “[a] numerical unit that can be added together to create a simple measure of volume, divided per clinician in aggregate to examine productivity per provider” [3]. It is defined more formally in the Federal Register [4].
Measuring PA performance using traditional wRVU data may miss a significant portion of PA work effort, depending upon specialty, role, and function of the PA, notably with regards to PA contribution to direct and indirect patient care, and “Shared,” “Independent,” or “Directed” visits, none of which are captured in traditional wRVU reporting. Direct Patient Care is defined by tasks involving face-to-face interaction with a patient. Indirect Patient Care is defined by tasks pertaining to patient, but do not require face-to-face interaction. “Shared,” “Independent,” and “Directed” indicate the level of physician participation in a patient visit: a Shared Visit requires direct patient care from both the supervising physician and the PA, a Directed Visit requires direct patient care from the PA and consultation from the physician, whereas an Independent Visit requires direct patient care from only the PA. Recently, UMHS has developed a unique methodology that provides a more comprehensive view of PA work effort, including these “shared” and “bundled” services that are not captured by traditional wRVU reporting. Using this new wRVU metric, Otolaryngology and Urology appear to be two of the most efficient clinics within UMHS; further study may reveal actionable best practices that can be extended to other services. The Chief Physician Assistant assumed that there is a direct correlation between the amount of Direct Patient Care that a PA performs and wRVU values. This assumption suggests that it is ideal for PAs to do as much direct patient care as possible.
From 2009 to the present, many studies have been undertaken due to PAs reporting to UMHS that they were feeling underutilized, one of which being the Radiation Oncology utilization study performed in 2011. This study found that the PAs in the Radiation Oncology clinic spent 46% of their time performing indirect patient care tasks. Of this time, 18% was spent in dictation or charting. This finding led to the conclusion that the PAs were underutilized [2]. Though all previously performed studies found valuable data related to the workload of the PAs, these studies addressed problems specific to these clinics. This project pushed further and looked into the factors affecting the high wRVU data so that the factors can be better understood.
Key Issues
The project was motivated by the following key issues:
●No documented best practices for PAs at UMHS
●Previously unavailable PA-specific wRVU data that is not yet compared with PA workload profile
●No analysis of shared vs. independent visits or clinic scheduling data
●Limited understanding of clinic-specific environmental factors that aid productivity
Goals and Objectives
To determine why the PAs in the Otolaryngology and Urology clinics have much higher wRVU data than all other clinics, the project team has achieved the following tasks:
●Observed PA practices in Otolaryngology and Urology clinics
●Determined workload profile in Otolaryngology and Urology clinics
●Determined clinic schedule utilization and division of PA appointments between shared and independent visits
●Tied workload profile findings of PAs and physicians with related wRVU data
Using this information, the project team developed recommendations to:
●Establish best practices for PAs in Otolaryngology and Urology clinics
●Implement found best practices in other clinics to improve wRVU data
●Investigate areas for further study
Project Scope
The project focused primarily on developing a workload profile of the 7 available outpatient PAs in the Otolaryngology and Urology clinics, specifically in the division between direct and indirect patient care. It also compared this workload profile with clinic utilization data extracted from MiChart. This comparative analysis included the newly available wRVU data and determined which aspects of these two clinics allow for high wRVU data. The wRVU and MiChart Scheduling data analyses will also include an additional Otolaryngology PA currently on maternity leave and unavailable for the observation and beeper study portions. The project has also started documentation of environmental factors specific to Otolaryngology and Urology that result in high wRVU data, but does not include any formal documentation of these processes (e.g. value stream maps, process maps, etc.)
This study focused only on activities performed by the PAs in the Otolaryngology and Urology clinics. Any tasks that are not performed by the PAs were not documented in the workload profile; that is, the team did not establish a workload profile of physicians, nurse practitioners, nurses, orderlies, or support staff to contrast with other clinics. The participation of these agents and any mention of their activities only shows their portion of work that is directly contributing to the PA work process.