PYA Principal Tynan Olechny and Consulting Manager Zach Doolin recently presented, “Beware of Benchmarks: Use of Survey Data in Determining FMV,” as part of NACVA’s Online Winter Summit.

1. December 8, 2014
Tynan P. Olechny, MBA/MPH, CVA
Zach K. Doolin, Macc
December 8, 2014

2. December 8, 2014
DISCLAIMER
All rights reserved. No part of this work covered by the copyrights herein may be reproduced or copied in any form or by any
means—graphically, electronically, or mechanically, including photocopying, audio/video recording, or information storage and
retrieval of any kind—without the express written permission of the CTI, NACVA ,and the presenter.
The information contained in this presentation is only intended for general purposes.
It is designed to provide authoritative and accurate information about the subject covered. It is sold with the understanding that the
copyright holder is not engaged in rendering legal, accounting, or other professional service or advice. If legal or other expert advice
is required, the services of an appropriate professional person should be sought.
The material may not be applicable or suitable for the reader’s specific needs or circumstances. Readers/viewers may not use this
information as a substitute for consultation with qualified professionals in the subject matter presented here.
Although information contained in this publication has been carefully compiled from sources believed to be reliable, the accuracy of
the information is not guaranteed. It is neither intended nor should it be construed as either legal, accounting, and/or tax advice, nor
as an opinion provided by the Consultants’ Training Institute (CTI), National Association of Certified Valuators and Analysts
(NACVA), the Institute of Business Appraisers (IBA), the presenter, or the presenter’s firm.
The authors specifically disclaim any personal liability, loss, or risk incurred as a consequence of the use, either directly or indirectly,
of any information or advice given in these materials. The instructor’s opinion may not reflect those of the CTI, NACVA, its policies,
other instructors, or materials.
Each occurrence and the facts of each occurrence are different. Changes in facts and/or policy terms may result in conclusions
different than those stated herein. It is not intended to reflect the opinions or positions of the authors and instructors in relation to
any specific case, but rather to be illustrative for educational purposes. The user is cautioned that this course is not all inclusive.
© 2014—1997 NACVA • 5217 South State Street, Suite 400 • Salt Lake City, UT, 84107—ALL RIGHTS RESERVED.
The Consultants' Training Institute (CTI) is registered with the National Association of State Boards of Accountancy (NASBA) as a sponsor of continuing professional
education on the National Registry of CPE Sponsors. State boards of accountancy have final authority on the acceptance of individual courses for CPE credit. Complaints
regarding registered sponsors may be submitted through its web site: learningmarket.org.

3. December 8, 2014
How to Obtain CPE Credit
for This Webinar:
The Presentation will include periodic online Polling Questions/Codes to
assess continuous participation and to determine the program’s
effectiveness.
You MUST respond to all polling questions (live) or complete a quiz
(recorded) in order to receive CPE credit.
If you view the webinar with a Smartphone or Tablet you may NOT be able to
answer polling questions that are required for obtaining CPE credit.
If you are viewing this presentation as a recorded webinar it will not qualify
for NASBA QAS CPE credit. You can however, obtain CPE credit per the
instructions included with your recorded webinar purchase.
© 2014 National Association of Certified Valuators and Analysts

4. December 8, 2014
Why Benchmark?
• Benchmarking is the process of measuring and comparing an
organization’s performance to national, regional, or industry
averages
 Provides quantitative data to support informed decision-making
• Provides insight into
 How a business is operating compared to peers
 How individuals/entities are producing and/or being compensated
compared to peers
 How efficiently staff or processes are working
• Can represent a starting point for determining fair market
value

5. December 8, 2014
What Can Be Benchmarked?
• Production
– Procedures - Patients - RVUs
– Cases - Weeks Worked - wRVUs
– Visits - ASA Units
• Compensation/Benefits
– Compensation per wRVU
– Compensation to collections ratio
– Benefits as a percentage of salary

6. December 8, 2014
What Can Be Benchmarked?
• Overhead
– Numerous expense categories (i.e., staff, space, supplies, etc.)
• Staffing Complement
• Efficiency Ratios
– Days in A/R
– Gross/Adjusted Collection %
• Collections
– Collections per wRVU
• Payer Mix
• Net Income or Loss

7. December 8, 2014
Healthcare Benchmarking Surveys
• Medical Group Management Association (“MGMA”)
o Physician Compensation & Production Survey
o Management Compensation Survey
o Cost Survey
o Academic Practice Compensation and Production
o Medical Directorship and On-Call Compensation
o Physician Placement Starting Salary
o Individual Specialty Surveys (e.g., Anesthesia, etc.)
 Historically represented small, single specialty medical groups
and had the largest sample size
 Shift to larger, hospital-owned, and multispecialty groups

8. December 8, 2014
Healthcare Benchmarking Surveys
(Cont.)
• American Medical Group Association (“AMGA”)
o Medical Group Compensation & Financial Survey
 Generally mid-to large multispecialty medical groups
 Not as many expense category benchmarks available
• Sullivan, Cotter & Associates (“SullivanCotter”)
o Physician Compensation & Productivity Survey
o Physician On-Call Pay Survey
 Generally represents larger organizations, including hospital
based systems and academic groups

9. December 8, 2014
Healthcare Benchmarking Surveys
(Cont.)
• Hospital & Healthcare Compensation Service (“HHCS”)
o Physician Salary & Benefits Report
 Primarily represents hospitals, with approximately 15% of
responses from group practices
 Evenly split between teaching and non-teaching facilities
• TowersWatson
o Health Care Clinical and Professional Compensation Survey
Report
 Includes in-depth data for non-MD positions including nursing,
clinical research, laboratory, technical services, etc.
 Generally larger organizations, hospitals

10. December 8, 2014
Use of Multiple Benchmark Surveys
in Determining FMV
• Dependence upon only one of many different surveys
can result in significantly different conclusions
 Data comes from a variety of sources; not always
comparable
• Regulatory guidance encourages the use of numerous
salary survey sources when determining FMV
compensation
 Relying on only one salary survey can create regulatory
compliance issues
• “But, MGMA says…”

11. December 8, 2014
Regulatory Issues
• Specifically, with respect to FMV determination, the
Phase III rule of the Stark Law states that “a reference
to multiple objective, independently published salary
surveys remains a prudent practice for evaluating fair
market value”
• Documentation of the use of multiple objective
independently published salary surveys is beneficial in
case an organization’s physician compensation
practices are challenged

12. December 8, 2014
Potential Compensation
Consequences - Example
• Consider a hospital looking to recruit a hematology/oncology
physician. Based on a detailed production and compensation
analysis, the hospital has decided to pay at the 90th percentile.
Survey data at the 90th percentile for the Southern region:
MGMA = $864,561 SullivanCotter = $596,205
AMGA = $693,954 Average = $718,240
Relying on only one survey results in drastically different results
than an average of all three
• Consider specific facts and circumstances when deciding which
surveys to utilize
• The importance in considering the number of respondents

13. December 8, 2014
Other Possible Pitfalls When
Benchmarking Using Survey Data
• Variance in participant data points across multiple
surveys
 Use a weighted average instead of a straight average when
sample size and reported compensation data vary significantly
for the same specialty
• Selecting the wrong type of survey respondent
 Utilize surveys that include the closest representation of
population to which data is being compared
o Academic versus private practice
o Hospital-employed versus non-employed

14. December 8, 2014
Other Possible Pitfalls When
Benchmarking Using Survey Data (Cont.)
• Compensation increases should be in overall
compensation, not compensation per unit of
production
− Higher quartiles of production do not necessarily mean
higher compensation per wRVU quartiles.
• Example:
− Hospitalist annual wRVUs = 5,747 (MGMA 75th %tile) and
is paid comp/wRVU at the 75th %tile or $81.01/wRVU
− Compensation approximates $466,000 which exceeds the
90th %tile of annual compensation for a hospitalist
($363,000 per MGMA).

15. December 8, 2014
Comparing Survey Data
• Normally the median should be used for comparisons
across specialties and surveys
 Since the median is the midpoint of all data, it is not subject to
the distortion which can occur with the average when extremely
high or low values are present
• The mean should be used in cases where benchmarks
include categories that should total to 100% (e.g., payer
mix)
 The median and other percentiles will not total due to statistical
distributions of each sample

16. December 8, 2014
Comparing Survey Data (Cont.)
• To gain a better understanding of the variability among
survey respondents, one can review the often
overlooked standard deviation, which measures the
extent of variability within the dataset
 A standard deviation that is similar to the mean indicates that the
data points are disperse and there is weak central tendency
 A standard deviation less than a third of the mean indicates that
the data points are clustered tightly around the mean and there
is strong central tendency

17. December 8, 2014
Additional Factors to Consider
• Geographic variances
 By region, state, etc.
• Physician’s level of experience
• Practice ownership
 Private practice vs. hospital-owned
• Inclusion of all surveys in the analysis

18. December 8, 2014
Additional Factors to Consider
(Cont.)
• Understanding survey definitions
 Total compensation in one survey may not include the same
components as total compensation in another survey
 Categorization of expenses from MGMA Cost Survey
 Classification of states into regions
 Definition of revenue
 Specialty categorization

19. December 8, 2014
Geographic Variances
• Example – effect of regional variance on compensation:
Region matched to California:
Region matched to New York:
Regional Compensation Survey Data, OB/GYN
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey 484 $300,806
2013 SullivanCotter Physician Compensation and Productivity Survey 731 $296,614
2014 HHCS Physician Salary & Benefits Report 70 $275,400
2014 AMGA Medical Group Compensation and Financial Survey 775 $338,668
2014 TW Healthcare Clinical and Professional Compensation Survey Report ISD ISD
Average of Surveys, Rounded $302,872
Regional Compensation Survey Data, OB/GYN
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey 439 $267,909
2013 SullivanCotter Physician Compensation and Productivity Survey 467 $250,000
2014 HHCS Physician Salary & Benefits Report 207 $262,581
2014 AMGA Medical Group Compensation and Financial Survey 403 $267,528
2014 TW Healthcare Clinical and Professional Compensation Survey Report 57 $284,600
Average of Surveys, Rounded $266,524

20. December 8, 2014
Physician Experience
• Example – effect of physician experience variance on
compensation:
National Compensation Survey Data, Gastroenterology
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey
All Survey Respondents 860 $510,671
1 to 2 Years in Specialty 25 $450,000
3 to 7 Years in Specialty 74 $487,729
8 to 12 Years in Specialty 65 $519,763
13 to 17 Years in Specialty 50 $551,396

21. December 8, 2014
Inclusion of Surveys
• Example – effect of averages excluding certain surveys:
National Compensation Survey Data, Cardiology: Invasive/Interventional
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey 815 $560,000
2013 SullivanCotter Physician Compensation and Productivity Survey 544 $530,450
2014 HHCS Physician Salary & Benefits Report 395 $413,599
2014 AMGA Medical Group Compensation and Financial Survey 543 $544,733
Average of Surveys, Rounded $512,196
National Compensation Survey Data, Cardiology: Invasive/Interventional
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey 815 $560,000
2013 SullivanCotter Physician Compensation and Productivity Survey 544 $530,450
2014 AMGA Medical Group Compensation and Financial Survey 543 $544,733
Average of Surveys, Rounded $545,061

22. December 8, 2014
Practice Ownership
• Example – effect of practice ownership on
compensation:
National Compensation Survey Data, Family Medicine (without OB)
Description Data Points Median
2014 MGMA Physician Compensation and Production Survey
All Survey Respondents 5,983 $211,452
Physician Owned 1,229 $225,143
Hospital/IDS Owned 4,521 $209,965
Other Majority Owner 233 $189,349

23. December 8, 2014
• Example – effect of practice ownership on overhead costs:
Hospital/IDS Owned:
Physician Owned:
Practice Ownership (Cont.)
Operating Cost Data, Family Medicine
Description Data Points Median
2014 MGMA Cost Survey
Information technology expense 11 $22,149
Medical and surgical supply expense 11 $12,189
Building and occupancy expense 11 $74,878
Malpractice insurance expense 11 $9,763
Total Operating Cost 11 $467,783
Operating Cost Data, Family Medicine
Description Data Points Median
2014 MGMA Cost Survey
Information technology expense 70 $9,963
Medical and surgical supply expense 93 $9,427
Building and occupancy expense 93 $53,018
Malpractice insurance expense 75 $6,884
Total Operating Cost 429 $364,852

24. December 8, 2014
Survey Trends Change Annually
• Do not be fooled by assuming one year is consistent
with another
• Trending analyses that compare compensation
benchmarks over multiple years are useful when
determining FMV for a multi-year agreement
• Increases in hospital-employed respondents over recent
years may impact certain metrics, such as overhead
costs
• Keep it in perspective – a benchmark with an n = 356 is
probably more reliable than one with an n = 28

25. December 8, 2014
Keys to Remember
• Benchmarks do not tell the whole story, but instead offer
insight
• Not all benchmarks are relevant, and not everything is
worthy of comparison
• Don’t take benchmarks at face value – use common
sense!
• Always step back and look at the relationships of the
analysis
• If wRVUs and collections are greater than the median and
expenses are at or below the median, why isn’t
compensation greater than the median? What is missing?

26. December 8, 2014
Questions?
Tynan Olechny, MBA/MPH, CVA
tolechny@pyapc.com
(404) 266-9876
Zach Doolin, MAcc
zdoolin@pyapc.com
(865) 673-0844