ORGANIZATIONAL CONFIGURATIONS AND STRATEGIES RELATED TO FINANCIAL PERFORMANCE IN MEDICAL GROUP PRACTICES: A TEST OF PORTER’S GENERIC STRATEGIES by TODD B. SMITH S. Robert Hernandez, DrPH, Committee Chair Richard M. Shewchuk, PhD Jeffery H. Burkhardt, PhD Thomas L. Powers, PhD Douglas J. Ayers, PhD A DISSERTATION Submitted to the graduate faculty of The University of Alabama at Birmingham, in partial fulfillment of the requirements for the degree of Doctor of Philosophy. BIRMINGHAM, ALABAMA 2011
158
Embed
ORGANIZATIONAL CONFIGURATIONS AND … CONFIGURATIONS AND STRATEGIES RELATED TO FINANCIAL PERFORMANCE IN MEDICAL GROUP PRACTICES: A TEST OF PORTER’S GENERIC STRATEGIES by TODD B.
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
ORGANIZATIONAL CONFIGURATIONS AND STRATEGIES RELATED TO FINANCIAL PERFORMANCE IN MEDICAL GROUP PRACTICES: A TEST OF
PORTER’S GENERIC STRATEGIES
by
TODD B. SMITH
S. Robert Hernandez, DrPH, Committee Chair
Richard M. Shewchuk, PhD Jeffery H. Burkhardt, PhD Thomas L. Powers, PhD Douglas J. Ayers, PhD
A DISSERTATION
Submitted to the graduate faculty of The University of Alabama at Birmingham, in partial fulfillment of the requirements for the degree of Doctor of Philosophy.
BIRMINGHAM, ALABAMA
2011
ii
Copyright by Todd B. Smith
2011
iii
ORGANIZATIONAL CONFIGURATIONS AND STRATEGIES RELATED TO FINANCIAL PERFORMANCE IN MEDICAL GROUP PRACTICES: A TEST OF
PORTER’S GENERIC STRATEGIES
TODD BRENTON SMITH
HEALTH SERVICES ADMINISTRATION
ABSTRACT
Research in the field of organizational configurations (OC) involves the formation
of groups of firms that are similar to each other on certain characteristics, and dissimilar
from other groups, and explores organizational performance differences between the
groups (Ketchen & Shook, 1996; Short, Payne, and Ketchen, 2008). However, OC is
replete with literature lacking in common key terms, measurement methods, and
specification of variables, and too few empirical articles with a strong theoretical basis
have been published (Short, Payne, & Ketchen, 2008). Porter’s (1980, 1985) generic
strategies are a specific typology within the field of OC that have been used extensively
in the literature. Through a review of the major academic literature databases, however,
only one empirical study (Payne, 2001) has been published that specifically uses Porter’s
generic strategies with a sample of medical group practices. Therefore, following a call
from Shortell and colleagues (2005) for theory-driven research regarding predictors of
high-performing versus low-performing medical groups, this paper used data from the
MGMA 2009 Cost Survey to explore the financial performance differences between
medical practices that were classified into one of several strategy groups, based on
Porter’s generic strategies. This study also used an inductive methodology, a cluster
analysis, to divide the sample into six distinct groups, which were then compared with the
iv
features and performance of the groups created with deductive methodology, Porter’s
generic strategies.
The findings of both the inductive and deductive methodologies suggest that
medical group practices using a differentiated strategy were the best performers, with
cost leaders, hybrids, and mixed strategy groups having similar performance levels.
Specifically, medical group practices that provided a greater number of ancillary services,
had at least one branch office, and spent more on advertising and furniture/equipment,
were more likely to have higher profit ratios than other practices. However, the inductive
technique indicated that having a low accounts receivable ratio, a cost leadership strategy,
may also be related to a high profit ratio. Overall though, while this paper partially
supports Porter’s generic strategies typology, the inductive methodology was relatively
more efficient in explaining the variation in the dependent variable based on the eight
strategy indicator measures developed within the study. The findings of this study
provide additional support for Porter’s generic strategies, as well the overall field of
organizational configuration research, and will advance the knowledge base within the
field of health care regarding the strategy-performance links within medical group
practices.
v
DEDICATION
I dedicate this dissertation to my wife and parents. My wife has endured many
years of weekends and evenings without a husband during my time as a doctoral student.
Her unwavering support and overlooking of my “honey-do” list for many years is very
much appreciated. Additionally, my parents have always been available to listen to the
ordeals of my academic endeavors and have been understanding about my absence from
many family affairs these past several years. I cannot overstate the appreciation I have
for my family’s support of, and interest in, my doctoral studies.
vi
ACKNOWLEDGEMENTS
I wish to thank Drs. Shewchuk and Haiyan Qu for their extensive work with me
over many months regarding the methodology section of this dissertation. Their help was
invaluable and I learned so much from each of them. I also wish to thank my brother and
sister-in-law, Bryan and Misty Smith, for their review and feedback of the manuscript in
regards to the overall flow and grammar.
The University of Alabama at Birmingham (UAB) and my supervisors have been
instrumental regarding my doctoral studies during my employment at UAB. The tuition
reimbursement from UAB and the support of my supervisors was vital in the successful
completion of my studies. Specifically, Dr. David Kimberlin and Mr. Rich Pierce have
provided me with the occasional flexibility in my work schedule as I completed my
doctoral studies.
Mr. David Gans and MGMA were extremely gracious in providing me with the
data for this study and I will forever be thankful to them for their assistance.
Additionally, I want to also thank Dr. Hernandez for providing me with the link between
UAB and MGMA, as well as his guidance in the initial phases of my research, including
the identification of an applicable theory and subsequent theory development.
Finally, I wish to thank all of my committee for their continued support and many
other UAB faculty members who have guided me through my doctoral studies. As a
part-time student, I have taken somewhat longer than normal to complete my studies and
dissertation. However, the faculty and my committee have never wavered in their
support of my efforts to complete this degree.
vii
TABLE OF CONTENTS
Page
ABSTRACT........................................................................................................................iii DEDICATION ..................................................................................................................... v ACKNOWLEDGEMENTS ................................................................................................ vi LIST OF TABLES .............................................................................................................. ix LIST OF FIGURES ............................................................................................................ xi CHAPTER 1: INTRODUCTION ........................................................................................ 1 Purpose and Research Questions ............................................................................. 6 CHAPTER 2: LITERATURE REVIEW ............................................................................. 7 Foundations of Organizational Configuration Research ......................................... 7 Organizational Configuration Research ................................................................ 15 Porter’s Generic Strategies - Typology ................................................................. 25 The Outcome of a Successful Strategy - Performance .......................................... 33 Medical Group Practices ........................................................................................ 38 CHAPTER 3: METHODS ................................................................................................. 44 Research Questions ................................................................................................ 44 Study Population, Variables, and Operational Definitions .................................... 46 Statistical Methods in Organizational Configuration Research............................. 64 Hypotheses ............................................................................................................. 67 CHAPTER 4: RESULTS AND FINDINGS ..................................................................... 73 Preliminary Data Cleaning and Preparation for Data Analysis ............................. 73 General Practice Demographics ............................................................................ 78 Descriptive Analyses for the Variables Included in the Model ............................. 80 Grouping the Practices ........................................................................................... 85 Empirical Analyses ................................................................................................ 87 Hypothesis 2 – Performance Differences ............................................................ 100
Hypothesis 3 – Performance Differences Between the Theoretical and Empirical Models ................................................................................................. 102
viii
Table of Contents (continued) CHAPTER 5: SUMMARY AND CONCLUSIONS ....................................................... 104 Discussion of Study Findings .............................................................................. 104 Limitations ........................................................................................................... 106 Future Research ................................................................................................... 111 Final Conclusion .................................................................................................. 113 LIST OF REFERENCES ................................................................................................. 115 APPENDICES ................................................................................................................. 128 A Total RVUs and Total Number of FTE Physicians ...................................... 128 B Practices Grouped By Size ........................................................................... 129 C Comparison of Practices Included and Excluded ......................................... 130
D Descriptive Statistics and Discussion of Selected Variables ........................ 131
E Winsorization................................................................................................ 144 F Imputation ..................................................................................................... 145 G Post-Hoc Analysis – Hypothesis 1 ............................................................... 146 H IRB Approval ............................................................................................... 147
ix
LIST OF TABLES
Table
1 Measures of the Independent and Dependent Variables........................................ 72
2 Descriptive Statistics – Total Number of FTE Physicians .................................... 76
3 Practice Type (N = 1,413) ..................................................................................... 77
4 Legal Organization Type (N = 1,413) ................................................................... 78
Wagner, 2001; Karnani, 1984; D. Miller, 1992; D. Miller & Friesen, 1986b; Murray,
1988; Phillips, Chang, & Buzzell, 1983; White, 1986) have introduced another category
to Porter’s generic strategies, the hybrid group, which includes strong characteristics of
both the cost leadership and differentiator constructs. The hybrid strategy is distinct from
the mixed strategy in that firms with a mixed strategy do not exhibit strong characteristics
of either a cost leader or differentiator.
Through searches of several major literature databases, no published works in the
academic literature have been found that specifically use Porter’s generic strategies
model to explore the financial performance differences that may exist between medical
group practices. However, through a review of the academic literature related to Porter’s
generic strategies in other industries, and literature related specifically to medical group
practices and health care in general, this author believes it will be possible to derive
5
strategic characteristics found within medical group practices to develop groups that are
similar to those described within Porter’s generic strategies. Further, using literature
from numerous fields as a guide and data from medical group practices, this study will
build and test a model based on Porter’s generic strategies.
From a review of the literature on strategic typologies and organizational
configurations, this author theorizes that successful medical group practices will typically
conform to one of Porter’s pure generic strategies or the hybrid strategy. The proposition
is that medical group practices employing an organizational configuration strategy based
upon one of these pure generic strategies, or the hybrid strategy, will exhibit superior
financial performance compared with those medical groups practices that demonstrate the
characteristics of a mixed strategy (Powers & Hahn, 2004).
As there is a dearth of research examining the performance of medical group
practices, Shortell and colleagues (2005) have called for theory-driven research regarding
predictors of high-performing versus low-performing medical groups. Through a review
of the major academic literature databases, however, no research specifically using
Porter’s generic strategies regarding the organizational configurations, strategies, or
financial performance variations in medical group practices, and only one empirical
research work related to studies of medical group strategies using an objective financial
performance measure (Payne, 2006), has been discovered. Thus, this study will initially
lead to a further understanding of how one may categorize medical group practices by the
generic strategies they employ. Additionally, this study will advance the body of
knowledge regarding financial performance differences between medical group practices
based on their organizational configuration type and strategy. Finally, as Porter’s generic
6
strategies are known as a generalizable typology, this research study will lead to a greater
understanding of the strategy-performance link in the general field of business and further
the literature in regards to methods, terms, and measurements in the field of
configurational research.
To guide the theory development, the operationalization of the variables, and the
specification of the model within this study, the business and health care academic
literature in the fields of strategic management and configurational research will be
utilized. To validate the theoretical model, the study will employ both empirical (i.e.,
cluster analysis) and theoretical (i.e., Porter’s generic strategies) classification methods to
categorize the medical group practices within the sample. Subsequently, the study will
compare the characteristics and performance differences that may be found between the
groups that will be created from the empirical and theoretical classification methods.
Purpose & Research Questions
1. Identify specific organizational configurations and strategies used in medical group practices that can be linked specifically to one of Porter’s generic strategies.
2. Using a deductive methodology, do medical groups that exhibit the characteristics of one of Porter’s pure generic strategies (i.e., target scope and then differentiation or cost leadership) perform better financially than medical groups that use a hybrid strategy or groups that exhibit a mixed strategy (stuck in the middle)?
3. Using an inductive methodology (cluster analysis), do specific groups of organizations with similar characteristics perform better financially than others?
4. Comparing the inductive and deductive methodology, will one methodology lead to the formation of groups that will better predict the financial performance of certain medical group practices?
7
CHAPTER 2
LITERATURE REVIEW
The Foundations of Organizational Configuration Research
Research regarding organizational configurations (OC) has foundations in the
fields of both organizational theory and strategic management, as well as other popular
and historical fields of management. The configurational stream of research has
furthered our understanding of organizational configurations and performance differences
between organizations through a parsimonious method of grouping firms by similar
characteristics and then studying the different characteristics and performance differences
between the groups. The key concept of configurational research is that individual firms
can be viewed as clusters of firms with similar practices and strategies, and that key
organizational features and strategies can shape a firm’s performance, with performance
often measured as firm’s return on assets (Short, Ketchen, Palmer, & Hult, 2007).
Configurational research, which typically focuses on the similar patterns within
identified groups of firms as well as the performance outcome differences between the
different groups of firms, has many underpinnings from other fields of management
research. Research in the field of strategic management typically posits that specific
strategies can be matched to the structure and environment of a given firm and proper
deployment of these strategies can lead to superior performance compared with firms
using other strategies. In contrast, in the field of organizational theory, researchers
typically focus on the design and behavior of firms, sometimes through the use of
strategic choice and decision-making models, but without a central focus on an outcome
8
based on performance (Fiss, 2007). In the fields of structural contingency and industrial
organization, the focus is typically related to how environmental factors guide a firm’s
strategic decisions and impact a firm’s financial performance. Organizational
configuration research combines facets from all of the aforementioned fields of
management research in an effort to understand the performance differences between
firms belonging to distinct groups based on common attributes (Dess, Newport, &
Rasheed, 1993; Ketchen et al., 1997).
Structural Contingency Theory - Categorizing Firms and Their Strategy
Structural contingency theory is a major foundation of OC. Weber (1947) was
one of the first authors to describe various configurations of organizations and ascribe
specific structural differences between, and attributes to, these organizations. As a
historian and sociologist though, Weber’s works did not lead to practical managerial or
academic implications within the various configurations he found. In later years though,
Woodward (1958) took Weber’s work a step further by identifying numerous categories
of production firms with differing strategies, finding that small batch production
companies focused on meeting the customer’s need while large batch production
companies focused on efficiency. The focus of Woodward’s (1958) organizational work
was on the type and quantity of products produced, and how the technology enabling the
production affects various aspects of the structure of the organization. Woodward (1958)
also used empirical methods to describe the performance differences between specific
firms and their peers on a number of structural dimensions.
9
Many other structural contingency theorists have laid the foundation for the
current research field of organizational configurations. Burns and Stalker (1961) studied
the differences between firms with either a mechanistic or organic structure, determining
that within dynamic economic sectors, firms with an organic structure would perform
better. Lawrence and Lorsh (1967) studied the differences between six organizations and
the impact the environment had on their structures and performance, finding generally
that firms able to achieve both high integration and differentiation will perform better
than other firms. In a study of hospital operating rooms, Galbraith (1973) used
contingency theory to explore how firms should organize in uncertain environments,
finding that there was no one best way to organize, but not all ways of organizing are
equally effective (thus exploring the issue of equifinality, which will be discussed later).
Filley and Aldag (1978) created a taxonomy of organizations based on contingency
theory and identified patterns within a number of firms, creating craft, promotion, and
administrative groups. Subsequent contingency theory research by Filley and Aldag
(1980) determined that the product selection varied between firms with an efficiency
strategy and firms with a “made-to-order” strategy.
In summary, structural contingency theorists typically view the unit of analysis as
the firm, as do OC theorists. However, unlike OC theorists, those in the field of
structural contingency typically limit their research to structural and situational concepts
related to firms. Building on structural contingency theory, however, organizational
configuration theorists study organizations as open-systems and in a multi-dimensional,
non-linear context. Thus, OC theorists believe organizations determine their adaptive
needs to an environment, rather than the environment determining the characteristics of
10
an organization, and situations of equifinality often exist in the view of OC theorists
(Meyer, Tsui, & Hinings, 1993).
Organizational Ecology
The widely cited organizational ecology works of Hannan and Freemen (1977,
1984) have also laid the foundation for organizational configuration theory. These
authors state that firms will develop certain structural characteristics allowing them to
adapt to their environment in order to become or remain successful. Thus, certain firms
through collective, rational action will succeed while others will not. The debate in the
organizational configuration versus ecology literature relates to whether the structure
leads to strategy or vice-versa. Organizational configuration theorists emphasize
strategic choice and thus tend to downplay the structural elements, while organizational
ecologists posit that the structure or environment lead to the strategy of an organization
and certain organizations will fail because they are unable to adapt to a certain
environment (Ketchen, Combs, Russell, Shook, Dean, et al., 1997).
Industrial Organization Theory
Organizational configuration theory also has strong links to industrial organization
theory. Concepts from industrial organization (IO) theory lead one to believe that macro
industry-wide factors are the primary drivers of performance of an individual firm (Bain,
1956; Barney, 1986a; Mason, 1939; Thornhill & White, 2007). This view takes the
external environment of a firm and explores these environmental factors as they relate to
a firm’s internal performance. The traditional model in industrial economics is the
11
structure-conduct-performance paradigm: firm performance depends on its conduct
related to pricing, research and development, and investments, and firm conduct is based
on an industry’s structure – concentration levels, barriers to entry, and degree of product
differentiation (Spanos, Zaralis, & Lioukas, 2004). Thus, the success of the industry as a
whole is thought to be primarily influenced by “barriers to entry, the number and relative
size of firms, the existence and degree of product differentiation in the industry, and the
overall elasticity of demand for the industry” (Barney, 1986b, p. 792).
In the IO view of business, organizations tend to position themselves within an
industry so that they can enhance their productivity and ultimately their profitability
(Parnell, 2006). However, the industrial organization theory of the firm cannot explain
the wide variations in firm performance within a specific industry (Parnell, 2006). The
original work regarding Porter’s five-forces model (1980), which deals with a firm’s
opportunities and threats, is based on the IO economics view.
A number of researchers (Caves & Porter, 1977; Porter, 1980) have studied the
opportunities and threats that exist for a firm in a competitive environment using
industrial organization theory. However, this type of analysis typically assumes that a
firm’s resources are similar to those of its competitors (Barney, 1991). Thus, the focus in
IO theory is more on the attractiveness of a given industry, rather than on the individual
strategies that a firm may use to improve its performance in comparison with its
competitors.
In industrial organization theory, specific industries are often seen as more attractive
than others based upon certain environmental structural characteristics such as the
competitive intensity, barriers to entry, bargaining power of suppliers and buyers, and the
12
threat of substitutions within a given industry (Bishop & Megicks, 2002). Researchers
Xu, Cavusgil, & White, 2006). This method has the researcher define an ideal type with
an empirical profile and then each case is compared with the ideal profile. Using this
method, along with an appropriate theoretical basis, researchers may test the fit of various
profiles. However, this method lacks in the ability to determine whether any causal
relationships exist between the independent variables and which variables may be related
to the performance differences between groups. Additionally, the deviation scores’
method is less than idyllic as it requires a great deal of researcher subjectivity when
creating the ideal profile (Fiss, 2007).
To counter these methods, Fiss (2007, p. 1183) says that a “set-theoretic approach
using Boolean algebra to determine which combinations of organizational characteristics
combine to result in the outcome in question” is the best method for empirical research
using the organizational configuration theory. With this method, the researcher develops
a “truth table” listing all possible organizational configurations and then a determination
is made as to whether each outcome leads to the desired outcome. However, no
published articles using this methodology were found in the empirical literature of
organizational configurations.
67
Although the previous discussion of statistical methodology in OC demonstrates that
a number of methods have been utilized in the literature to study performance differences
between group, cluster analysis is still the primary method used in organizational
configuration research (Short, et al., 2008). However, research using the cluster analysis
with only inductive methods may receive serious criticism for lacking a theoretical basis.
Additionally, in a meta-analysis of 40 empirical configurational studies, Ketchen and
colleagues (1997) found no difference in effect size between the inductive and deductive
methods. However, these researchers did find that studies using a broader range of
organizational configurations yielded stronger effect sizes than those studies with a more
narrow range. Ketchen and colleagues (1997) also found unequivocal evidence from the
40 articles reviewed that configurational research methods were able to predict the
performance of a firm. In summary, as clustering analysis is a widely used method in OC
research, and to further the explanatory power of the results for both the medical group
population as well as the overall research stream, this study will utilize a clustering
algorithm based initially on a deductive methodology using Porter’s generic strategies,
followed by a clustering algorithm using a purely inductive methodology.
Phase I: Deductive, Theoretically Based Model/Typology
1. Using a deductive methodology, do medical groups that exhibit the characteristics of
one of Porter’s pure generic strategies (i.e., target scope and then differentiation or
cost leadership) perform better financially than medical groups using a hybrid
strategy or mixed strategy?
68
Broad-focused:
H1a: On average, broad-focused (multispecialty) medical group practices exhibiting a
pure generic strategy (cost leader or differentiated) will have higher levels of financial
performance compared with groups exhibiting a hybrid or a mixed strategy.
Narrow-focused:
H1b: On average, narrow-focused (single specialty) medical group practices exhibiting a
pure generic strategy (cost leader or differentiated) will have higher levels of financial
performance compared with groups exhibiting a hybrid or a mixed strategy.
Phase II: Inductive Taxonomy – An Empirical Model (Cluster Analysis)
2. Using an inductive methodology, do specific groups of organizations with similar
characteristics perform better financially than others?
Broad-focused:
H2a1: Discrete clusters will not form based on a cluster analysis technique using the
same variables used in the deductive analysis for broad-focused medical groups.
H2b1: Clusters that emerge from the cluster analysis technique (inductive technique) will
not exhibit significant performance differences between groups of broad-focused medical
groups.
69
Narrow-focused:
H2a2: Discrete clusters will not form based on a cluster analysis technique using the
same variables used in the deductive analysis for narrow-focused medical groups.
H2b2: Clusters that emerge from the cluster analysis technique (inductive technique) will
not exhibit significant performance differences between groups of narrow-focused
medical groups.
Phase III – Comparing Taxonomy and Typology Models (Goodness of Fit test)
3. Will the inductive or deductive methodology lead to groups that will better predict the
financial performance of certain medical practice groups?
Broad-focused:
H3a: Using a goodness of fit test, the variation in the performance between the broad-
focused groups using the inductive approach will be greater than that of the deductive
approach.
Narrow-focused:
H3a: Using a goodness of fit test, the variation in the performance between the narrow-
focused groups using the inductive approach will be greater than that of the deductive
approach.
70
Figure 1. Hypothesis Model: Differentiator, Cost Leader, Hybrid, and Mixed Groups – Broad Focused Medical Groups (Multispecialty)
BROAD FOCUS: MULTISPECIALTY
COST LEADER
DIFFERENTIATOR
FINANCIAL PERFORMANCE
HYBRID
MIXED
71
Figure 2. Hypothesis Model: Differentiator, Cost Leader, Hybrid, and Mixed Groups – Narrow Focused Medical Groups (Single Specialty)
NARROW FOCUS: SINGLE SPECIALTY
COST LEADER
DIFFERENTIATOR
FINANCIAL PERFORMANCE
HYBRID
MIXED
72
Table 1
Measures of the Independent and Dependent Variables
Variables Included in the Model as Measures of Strategy Type Independent Variables
• Market Scope Strategy Variables: Focused vs. Broad - Type of medical group practice (single or multispecialty)
Variables and Measures of the Differentiation Strategy 1. MARKETING (Advertising Intensity): promotion &
marketing costs / total general operating costs
2. SPECIALTY SERVICES (Providing Specialty Products/Services): Number of ancillary / supplementary services provided
3. IMAGE: furniture and equipment costs + furniture and equipment depreciation/ total general operating costs
4. BRANCHES (Range of Market Segments): number of
branch or satellite clinics
Variables and Measure of the Cost Leadership Strategy
5. EFFICIENCY (Production and Operations Efficiency): a. Total operational costs / total RVUs b. Total Number of FTE Physicians / total RVUs
6. COST CONTROL (Control of Costs): – physical size: gross
square footage of all practice facilities / total RVUs
7. OVERHEAD (Tight Control of Overhead Costs) – front office support staff FTEs / total RVUs
8. ACCOUNTS RECEIVALBE (Tight Control of Marginal Accounts) – total accounts receivable > 120 days (in dollars)
Dependent Variable 9. PROFIT RATIO - total medical revenue after operating
costs / total number of FTE physicians
Note: Variables are from the MGMA 2009 Survey (Based on 2008 Data)
73
CHAPTER 4
RESULTS AND FINDINGS
The 2008 MGMA Cost Survey included a number of variables from a total of 1,797
medical group practices. This chapter will include a discussion regarding the initial
review of the applicable variables from the dataset, a summary of the descriptive statistics
for the applicable variables, an explanation regarding necessary modifications of the
originally proposed methods, the results from the inferential analysis (i.e., an ANOVA
and a cluster analysis), and the results of the hypothesis testing. The preliminary review
of the applicable variables in the MGMA dataset revealed numerous issues including
duplicate cases, numerous cases with missing data, and wide variability. Along with
these issues, several variables that were included in the original model had to be replaced
with alternative variables. Due to these factors, which will be discussed briefly in this
chapter and in greater detail in the appendix, the final analyses of the MGMA data
included only 1,413 of the 1,797 practices included in the complete dataset.
Preliminary Data Cleaning and Preparation for Data Analysis
Duplicate Cases and Missing Data
An initial review of the specific variables included in the originally proposed model,
as well as several other variables, led to the identification of nine cases that were obvious
duplicates in the dataset. These practices were eliminated from the dataset, leaving 1,788
cases for further analyses. For subsequent analyses that describe the data using all
practices, the duplicated practice data were excluded.
74
Normalizing the Variables
As practice size and volume will obviously vary between the practices, the data
analysis plan included methods to normalize many of the variables in the model. In the
original model, the total physician RVUs for each practice was to be used to normalize
three of the four cost leadership strategy indicator variables (i.e., RVUs was proposed as
a denominator for the strategy indicator ratios that were to be created using these three
variables). However, only 684 of the 1,788 original cases reported a value for total
physician RVUs in their practice. Thus, after a review of the data for possible alternatives
for the RVU variable, a decision was made to include the total number of FTE physicians
in each practice as a proxy for the total physician RVUs variable. Total number of FTE
physicians was also selected as it had been proposed in the original methods to normalize
the dependent variable, the mean profit ratio for each practice. The correlation between
these two variables, total physician RVUs and total number of FTE physicians,
demonstrated a significant (p < 0.01), yet weak (r = 0.185) relationship (see Appendix
A). However, there were no other variables present within the MGMA dataset that were
intuitively related to physician practice size or volume, other than several variables that
also included a large number of cases with missing data.
The variable total general operating costs within each practice was initially
proposed as the variable that would be used to normalize two of the four differentiator
strategy variables, advertising intensity and providing specialty products/services.
However, to maintain consistency within the model to the greatest degree possible, total
number of FTE physicians was replaced as the normalizing variable for both advertising
intensity and providing of specialty products/services within each practice. The inclusion
75
of the total number of FTE physicians variable to normalize many of the strategy
indicator variables also led to the elimination of this variable as one of the cost leadership
indicator variables, as it could not be used as both a distinct strategy indicator variable as
well as a variable to normalize many of the other strategy indicator variables.
As discussed in greater detail below, two of the remaining strategy indicator
variables - total ancillary services provided by each practice and total number of
branches within each practice - were recoded as binary variables, thus negating the need
for normalization of these two variables. The last of the eight indicator variables,
accounts receivable greater than 120 days (in dollars), was normalized independently of
the total number of FTE physicians in each practice. For this variable, the total accounts
receivable value (in dollars) reported by each practice was used to normalize each
practice’s value for their accounts receivable greater than 120 days. Table 10 provides
the revised strategy indicator measures, as well as the outcome variable, and the specific
variables used to create each new strategy indicator measure.
Full-time Equivalent Physicians - Inclusion of Only Cases with 1 – 35 FTE physicians
Within the original overall sample (n = 1,788), after the duplicate cases were
removed, 13 cases included a missing value for the total number of FTE physicians
variable and 3 cases were deemed as extreme outliers (> 1, 000 total number of FTE
physicians). Overall, the mean of the total number of FTE physicians for the remaining
1,772 practices was 17.21. One hundred and forty-eight (8.4%) practices had less than
1.0 physician FTEs and 203 (11.5%) practices had greater than 35 FTEs, after the 3
outlier cases were eliminated (see Appendix B). The comparison of the variability of the
76
total number of FTE physicians between the two groups was stark, with a standard
deviation of 7.56 for the group including only the practices with between 1 – 35 total
FTE physicians (n = 1,421), 47.35 for the group with all 1,772 cases, and 108.54 for the
group with only those practices with more than 35 FTE physicians (with the three cases
with extreme outliers and missing values excluded). Thus, to create a more
homogeneous sample for our inferential analysis, only those practices that had between 1
and 35 total FTE physicians (n =1,421 or 80.2% of the overall sample) were included in
the subsequent data analyses (see Appendix B for the frequencies within selected size
groups of practices).
Table 2
Descriptive Statistics – Total Number of FTE Physicians
Total Number of FTE Physicians N Min Max M SD All Practices 1,772 .00 968 17.21 47.35 Practices with 1 – 35 FTEs 1,421 1.00 35 7.38 7.56 Practices with > 35 FTEs 203 35.40 968 98.18 108.54 Note: The three extreme outlier cases were excluded from this analysis.
Total General Operating Costs and Profit Variables
Two of the 1,421 remaining practices did not report a value for their practice’s total
general operating costs, two other cases reported a negative value, and three reported a
value of zero. These seven cases were eliminated from further analysis, leaving 1,414
cases for further analyses. One practice did not report a value for the model’s outcome
variable, the profit of each practice (i.e., medical revenue less operating expenses). This
single case was also eliminated from further analyses, leaving 1,413 cases as the final
sample size for the subsequent descriptive analyses and the inferential analyses.
77
Single specialty vs. Multispecialty Practices
Of the 1,413 practices that remained, 81.7% (1,154) were single specialty practices,
17.1% (242) were multispecialty practices, and 1.2% (17) reported “other” as their
practice type. The initial data analysis plan included grouping the cases within the
sample by either single or multispecialty groups, which would relate to Porter’s broad
and narrow target market scope construct. However, with the relatively small number of
cases represented as multispecialty practices (n = 242), and the proposed methods that
would lead to the division of the multispecialty practices into four separate groups, it was
determined that the small sample size for the multispecialty practices would be
inadequate for a separate inferential analysis. Therefore, the methods were altered to
combine both the single and multispecialty practices (n = 1,413) into a single sample and
eliminate the separate analyses using each these practice types.
Table 3
Practice Type (N = 1,413)
Practice Type Frequency Percent Single Specialty 1,154 81.7 Multispecialty 242 17.1 Other 17 1.2
As is evident from the preceding pages in this chapter, a significant amount of
effort ensued during the data cleaning and preparation stage of the analysis. The
following sections of this chapter will detail the results from the descriptive and
inferential analyses that were completed after the initial data cleaning and preparation.
78
General Practice Demographics
Regarding the legal organization of the practices (N = 1,413), 27.6% (390) reported
their practices as not-for-profit; 50.3% (709) were for-profit businesses, LLCs,
partnerships, professional corporations, sole proprietorships, or other; and 22.2% (314) of
the practices did respond to this question.
Table 4
Legal Organization Type (N = 1,413)
Legal Organization Type Frequency Percent Not-for-profit corporation/foundation 390 27.6 For-profit entity 709 50.3 Other 22 1.6 Missing 314 22.2
Fifty percent (709) of the practices were owned by an integrated system or
hospital, 46.4% (655) of the practices were owned by physicians, and the remaining 3.5%
(49) reported another type of ownership.
Table 5
Majority Owner (N = 1,413)
Majority Owner Frequency Percent IDS or hospital 709 50.2 Physicians 655 46.4 Other 49 3.5
79
The population category of the metropolitan areas for the practices was split
relatively evenly within the sample (n = 1,413), with 18.4% (260) residing within a
nonmetropolitan area of fewer than 50,000 residents, 23.4% (331) residing within a
metropolitan area between 50,000 – 250,000 residents, 19.5% (276) residing within a
metropolitan area between 250,001 – 1,000,000 residents, 16.1% (227) residing within a
metropolitan area greater than 1,000,000 residents, and 22.6% (319) of the practices did
not respond to this question.
Table 6
Population category (N = 1,413)
Population Category Frequency Percent Nonmetropolitan (fewer than 50,000) 260 18.4 Metropolitan (50,000 to 250,000) 331 23.4 Metropolitan (250,001 to 1,000,000) 276 19.5 Metropolitan (more than 1,000,000) 227 16.1 Missing 319 22.6
Overall, the cases included in the model did not vary significantly from the excluded
cases based on legal organization type (p = 0.196) and majority owner type (p = 0.57).
However, when comparing those practices included in the final analysis with all practices
within the MGMA dataset, those included were significantly (p = 0.006) more likely to
reside in smaller metropolitan areas. Additionally, the effect size was also very small (<
0.01) for each of these three variables (see Appendix C for the ANOVA and Effect Size
results).
80
Descriptive Analyses for the Variables Included in the Model
The model, after elimination of the second variable to measure one dimension of
production and operational efficiency (total number of FTE physicians), now includes
four variables that indicate whether or not a practice exhibits the characteristics of a cost
leader and four variables that indicate the designation of a practice as a differentiator, in
addition to the dependent variable, the profit ratio. However, many of the practices did
not respond to one or more of the questions that were included as variables in our model.
The number of missing responses varied from 100 – 500 for each of the strategy indicator
measures. The descriptive statistics for the variables, prior to normalization, are detailed
below. A further discussion of each variable included in the model, including the
descriptive statistics for each variable, is included in Appendix D.
Table 7
Descriptive Statistics – All Variables Included in Model (N = 1,413)
Variables N Min Max M SD
Promotion and marketing (1) 995 $0 $873,594 $36,562 $68,471
Furniture and equipment (3a) 880 $0 $1,357,256 $48,896 $121,625
Furniture and equipment dep (3b) 926 $(167,081) $6,501,692 $102,734 $305,583
Total general operating cost (5) 1,413 $2,160 $79,274,718 $1,860,000 $3,629,000
Total square feet (6) 877 587 500,000 18,605 27,059
Total FTEs of Support Staff (7) 1,317 .01 346.88 32.53 44.69
AR > 120 (8) 1,294 .00 $10,632,000 $326,341 $724,400
Total accounts receivable (8) 1,315 0 $61,581,000 $1,585,979 $3,083,723
Profit (9) 1,413 $(16,733,495) $38,127,495 $2,860,000 $4,359,000 Note: The number in parentheses after each subheading represents the strategy dimension measured by each variable, as detailed in Table 10.
81
After the initial review of the data, two of the indicator variables, total number of
ancillary services and the number of reported branches for each practice, were recoded
into binary variables. Binary recoding was necessary as numerous practices reported no
ancillary services provided by their practice and/or no additional branches beyond their
primary practice site. Thus, a binary variable to measure these two dimensions of
differentiation was deemed more practical. The frequencies for each of these binary
variables are detailed in the tables below, and further details regarding these two
variables are included in Appendix D.
Table 8
Specialty Services (N= 1,413)
Practices Reporting One or More Ancillary Services (2) Frequency Percent No 820 58.0 Yes 593 42.0 Note: The number in parentheses represents the strategy dimension measured by this variable, as indicated in Table 10. Table 9 Branches for Each Practice (N = 1,413)
Branches (4) Frequency Percent No Branches 948 67.1 1 or More Branches 465 32.9 Note: The number in parentheses represents the strategy dimension measured by this variable, as indicated in Table 10.
82
Table 10 Measures for Each of the Strategy Indicator Variables & the Outcome Variable (Profit Ratio) Independent (Strategy Indicator) Variables & the Outcome Variable Differentiation Variables
1. MARKETING - Advertising Intensity: promotion & marketing costs / total number of FTE physicians
2. SPECIALTY SERVICES - Providing Specialty Products/Services: the provision
of ancillary services (binary variable – yes/no)
3. IMAGE - Image: furniture and equipment costs (3a) + furniture and equipment
depreciation (3b) / total number of FTE physicians
4. BRANCHES - Range of Market Segments: number of branch or satellite clinics
(binary variable – yes/no)
Cost Leadership Measures (Variables) 5. EFFICIENCY - Production and Operations Efficiency: Total general operating
costs / total number of FTE physicians
6. COST CONTROL - Control of Costs: – physical size: gross square footage of all
practice facilities / total number of FTE physicians
7. OVERHEAD - Control of Overhead Costs – front office support staff FTEs / total number of FTE physicians
8. ACCOUNTS RECEIVABLE - Control of Marginal Accounts – total accounts receivable > 120 days / total accounts receivable
Outcome Variable 9. PROFIT - Profit (total medical revenue after operating costs) / total number of
FTE physicians
83
Creation of Ratios - Winsorized Ratios & Outliers
Once the data cleaning and descriptive analyses were finalized, the ratios for the
applicable variables were computed. However, as discussed previously, the significant
variability with many of the variables, as well as numerous outliers, presented possible
issues. Nenide and colleagues (Nenide, Pricer, & Camp, 2003) have challenged the
validity in predicting firm performance through traditional methods when using financial
ratio data calculations. These authors contend that many researchers who utilize ratio
data to predict firm performance do not test for sample assumptions (e.g., normal
distribution, negative denominators, and outlier influence) and do not address potential
errors within the data (Nenide, et al., 2003). To reduce the effect of these issues, Nenide
and colleagues propose the use of the Winsorizing technique. Winsorizing the data is
accomplished by deeming values beyond a certain threshold for a given variable as
outliers. These outlier values are then replaced with a value that is one point above the
outlier threshold value.
Thus, based on recommendations from Nenide and colleagues (Nenide, et al., 2003),
each of the six non-binary strategy indicator variables were Winsorized. Specifically, an
outlier threshold value was generated by adding the value of the 75th percentile for each
variable to the interquartile range (IQR) value multiplied by 1.5 (i.e., 75th percentile +
(IQR)*1.5). In addition to the previous reviews of the data to correct for data errors,
Winsorizing the data led to the inclusion of many cases that might have otherwise been
eliminated (i.e., the extreme outliers) and dramatically reduced the variability of the
strategy indicator ratios in our model. A comparison of the values of the Winsorized and
84
non-Winsorized variables, as well as the changes in variability for each strategy indicator
ratio variable, is included in Appendix E.
Imputation of Values
As mentioned previously, many of the strategy indicator variables had numerous
cases with missing data, including one variable with 536 (38%) missing values. Only
711 cases would have remained for further analysis if a simple listwise deletion of the
cases was utilized. A review of the patterns of missing data ensued via SPSS and the
missing data were verified as missing completely at random. Several methods for
overcoming the missing data issue were reviewed, including mean substitution,
indicator/dummy variables adjustment, and imputation methods. Overall, multiple
imputation was found to be the best alternative. “Multiple imputation allows pooling of
the parameter estimates to obtain an improved parameter estimate,” thus incorporating
the uncertainty into the standard errors, which is an improvement over single imputation
methods (Acock, 2005, pp. 1,019). Compared with others method that would eliminate
cases with missing data (e.g., listwise deletion), multiple imputation enabled the inclusion
of many more cases in the final analyses. Five maximum likelihood multiple imputation
methods followed, and the method of providing the best fit for the variables was used to
impute the missing values of each six strategy indicator ratios, as well as the two binary
indicator variables. Imputation of these values enabled all 1,413 practices to remain in
the sample for further analyses. Residuals from the imputation process are included in
Appendix F.
85
Grouping the Practices
Differentiator Construct and Cost Leader Constructs
The steps in the data cleaning and data preparation process detailed above ultimately
left 1,413 cases for further analyses and led to the creation of eight strategy indicator
variables, four for cost leadership and four for differentiation. Two of these variables
were binary (i.e. yes/no) and the six other strategy indicator variables were ratios. The
individual values for each of the strategy variables were then reviewed, and using the
Spanos (2004) methodology, a cutoff point for each variable was computed. Thus, for
each of the eight strategy indicator variables, approximately one-third of the sample was
designated as a differentiator for each four measures of differentiation and approximately
one-third of the sample was designated as a cost leader for each of the four measures of
cost leadership. Subsequently, the four variables in each dimension, cost leadership and
differentiation, were summed together to create two separate composite strategy variables
for each practice. Initially, the overall composite values from each dimension led to the
designation of each practice as a cost leader, a differentiator, or both. Next, those
practices deemed as both a cost leader and differentiator were designated as hybrids (and
thus eliminated from the cost leader and differentiator groups), and those practices not
exhibiting the cost leader or differentiator traits were labeled as having a mixed strategy.
Again using the methods of Spanos and colleagues (2004), a best fit for dividing the
sample roughly into thirds was derived for the two composite strategy indicator variables,
cost leader and differentiator. The results indicated that 23.4% (n = 330) of the sample
were initially deemed as exhibiting the traits of a differentiator and 37.2% (n = 525) of
After the practices had been categorized into the cost leader and differentiator
constructs, those practices that were deemed as both costs leaders and differentiators
were identified as hybrid practices (2.5%; n = 36). Additionally, those practices deemed
as neither cost leaders nor differentiators (42%; n = 594) were labeled as having a mixed
strategy. The remaining groups were those exhibiting only one of the initial two traits -
cost leadership (34.6%; n = 489) or differentiation (20.8%; n = 294).
Table 12
Final Categorization of All Practices (N = 1,413)
Strategy Group Frequency Percent Stuck In The Middle 594 42.0 Differentiator 294 20.8 Low Cost Leader 489 34.6 Hybrid 36 2.5
87
Empirical Analyses
Hypothesis 1 - Theoretical Groupings (Typology)
After the sample had been divided into four distinct groups based on the exhibited
organizational configuration strategy of each practice, an ANOVA was used to test the
theoretical performance hypothesis. To account for the previously discussed combination
of both the multispecialty and single specialty groups into a single group, Hypothesis 1
was revised as follows:
H1: On average, medical group practices exhibiting a pure generic strategy (cost
leader or differentiator) will have higher levels of financial performance as compared
with groups exhibiting a hybrid or a mixed strategy.
Overall, we find support for part of Hypothesis 1. The ANOVA produced a
significant (p < .001) finding of a difference between the mean profit ratios of the four
groups. The Sheffe post hoc analysis (see the Appendix G) led us to reject the null
hypothesis and find support for our research hypothesis, thus determining that the cases
categorized within the differentiator group produced a significantly (p < 0.05) higher
mean profit ratios than the other three groups. However, Hypothesis 1 was not fully
supported, as the cost leader group was not found to be significantly different, based on
the mean profit ratio, from the hybrid or the mixed strategy group. Also, it should be
noted that the hybrid group, those practices exhibiting the traits of both a cost leader and
differentiator strategy, was quite small (n=36) when compared with the other three
groups, which were relatively similar in size. The mean and standard deviation for each
of the four groups are listed in the table below. The large standard deviation values
88
relative to the group means, especially for the differentiator group, demonstrate the high
degree of variability for many of the indicator measures in this model.
Table 13
Descriptive Statistics for the Four Groups (N = 1,413)
Groups M SD
Stuck In The Middle 218,884 310,921 Differentiator 585,928 836,923 Low Cost Leader 254,077 284,039 Hybrid 352,835 230,110 Table 14 One-Way Analyses of Variance for Effects of Group Membership on Profit Ratio (N = 1,413)
Source df SS MS F Between Groups 3 2.891E13 9.637E12 44.697*** Within Groups 1409 3.038E14 2.156E11 Total 1412 3.327E14
***p < .001.
Figure 3. Mean Profit Ratio of Practices by Organizational Configuration Type (N =
Differentiator Dimension. The figure below represents all six groups, but only includes
the four indicator variables for the dimension measuring differentiation. In this figure,
Group 6 is clearly depicted as the group most resembling a group of practices with the
traits of a differentiator. Group 3 is also depicted as the near polar opposite of Group 6,
as Group 3 ranks below average in all four measures of differentiation, and last for two of
these measures. The remaining four groups do not exhibit patterns that would clearly
identify them as a differentiator or exclude them from the cost leadership category.
Figure 11. Strategy Dimension - Differentiator
1.0
0.8
0.6
0.4
0.2
0.0 Marketing Image Specialty
Services Branches
Differentiation Strategy Variables Group 1 Group 2 Group 3 Group 4 Group 5 Group 6
99
Cost Leader Dimension. Only the four measures of the cost leadership strategy are
depicted in the figure below, but all six groups are simultaneously included in this figure.
Group 3 clearly stands out as the cost leader, ranking highest of all six groups in three of
the four measures of cost leadership, and Group 6 clearly stands out as nearly a mirror
image of Group 3, ranking the lowest of the six groups on three of the three costs
leadership strategy variables. The other four groups are much less well defined with
regards to the costs leadership dimension.
Figure 12. Strategy Dimension - Costs Leadership
1.0
0.8
0.6
0.4
0.2
0.0 Efficiency Cost
Control Accounts
Receivable Overhead
Cost Leader Strategy Variables Group 1 Group 2 Group 3 Group 4 Group 5 Group 6
100
Hypothesis 2 – Performance Differences
Hypothesis 2b utilized an ANOVA to test for performance differences between the
six groups. The analysis of the results incorporated an evaluation of both the magnitude
and range of the F values to determine if there were significant performance differences
between the groups, as well as whether there was significant discrimination between the
clusters.
H2b: Clusters that emerge from the cluster analysis technique (inductive
technique) will not exhibit significant performance differences between groups of medical
group practices.
Hypothesis 2b was not supported, as significant (p < 0.001) differences between the
mean profit ratios of the groups were found (see Table 17 below). The mean profit per
physician within each group ranged from $105,288 to $725,732 (see Table 16 below).
Group 6 has the highest profit ratio (profit per physician in each practice), with nearly
twice the average profit of the next highest group. Groups 2 and 4 have the lowest profit
ratios.
As noted previously, Group 6 exhibits many of same characteristics as the
theoretically generated group of practices that were categorized as differentiator. Group
6 has high values for the four differentiation strategy variables, and low values for all but
one of the cost leadership strategy measures, Accounts Receivable. Group 3, the subset
of practices that are most analogous to the Cost Leadership group created from the
theoretical model, had the third highest profit ratio among the six groups.
Group 1, the group deemed most analogous to those practices with a stuck in the
middle strategy, had the second highest profit ratio amongst the six groups. This group
101
was also the largest of the six groups, with 419 practices, and was relatively similar in
size to the group categorized as the stuck in the middle group in the theoretical model.
The remaining three groups - Groups 2, 4, and 5 – exhibited a range of high to
low values on each of the eight strategy indicator variables, but none of these groups
corresponds very well with one of Porter’s groups or with a group exhibiting the hybrid
strategy, though they could be interpreted as practices employing the stuck in the middle
strategy. It is noteworthy that none of the six groups exhibited relatively high values on
all eight indicator variables, which would indicate a group of practices with the traits of a
hybrid.
Table 16
Mean Profit Ratio for Each Group (N = 1,413)
Group # N M ($/FTE)* SD
Group 1 419 369,121 244,545
Group 2 277 105,288 243,461
Group 3 231 367,902 309,609
Group 4 181 112,006 340,060
Group 5 151 254,157 228,482
Group 6 154 725,732 1,122,002
Total 1,413 310,846 485,402
*Profit Ratio = Profit / Total Number of FTE Physicians
102
Table 17 One-Way Analysis of Variance for Effects of Group Membership on the Profit Ratio (N = 1,413)
Source df SS MS F Between Groups 5 4.803E13 9.606E12 47.479*** Within Groups 1407 2.847E14 2.023E11 Total 1412 3.327E14
***p < .001.
Hypothesis 3 – Performance differences between the theoretical and empirical model
Hypothesis 3 is related to whether the inductive or deductive methodology would
lead to groups that will better predict the financial performance of certain medical group
practices. To test Hypothesis 3, the performance differences among all of the groups
were compared and a determination was made regarding which methodology generated
groups that were superior in explaining performance differences between the groups - the
empirical or theoretical model.
H3: Using a goodness of fit test, the variation in the performance between the
medical group practices using the inductive approach will be greater than that of the
deductive approach.
Hypothesis 3 was rejected as it was found that the inductive methodology was
relatively more efficient in explaining the variation in the dependent variable based on
the eight indicator measure. The Eta squared value for the inductive methodology, the
cluster analysis, indicates that approximately 14.4% of the variation in the profit ratio of
each practice can be explained by inclusion in one of the six groups. Using the deductive
103
methodology, only 8.7% of the variation in the profit ratio was accounted for by group
membership (see Table 18).
Table 18
Variation Explained by the Inductive and Deductive Methodologies
Methodology η2
Inductive .144 Deductive .087
104
CHAPTER 5
SUMMARY & CONCLUSION
Discussion of Study Findings
Overall, Porter’s generic strategies model was found to be useful in the
categorization of physician practices by the various organizational strategy configurations
they employ and in determining if differences in financial performance vary between the
groups. Using Porter’s model, the results indicate that practices conforming to a
differentiated organizational configuration strategy perform significantly better that those
practices with another organizational configuration. However, the analysis did not fully
support Porter’s theory, as the cost leadership group did not produce a significantly
higher mean profit ratio as compared with the other three groups. These results are
further supported by Payne’s (1998) dissertation, in which he used Porter’s generic
strategy labels for groups that formed following a cluster analysis technique, as his
differentiator group also performed better than the other groups.
Additional support of Porter’s theory can be derived from the small size of the
hybrid group in the analysis using the deductive methodology and the absence of this
group after using the inductive methodology. While certain authors (Buzzell &
Wiersema, 1981; Cross, 1999; Hambrick, 1981; Helms, et al., 1997; Hill, 1988;
Hlavacka, et al., 2001; Karnani, 1984; D. Miller, 1992; D. Miller & Friesen, 1986b;
Murray, 1988; Phillips, et al., 1983; White, 1986) have supported a hybrid group in
addition to Porter’s original groups, Porter does not acknowledge the presence of hybrid
groups in his theory. Rather, Porter believes that the differentiator and cost leader are
105
thought to be at opposite extremes, thus eliminating the ability for an organization to
perform well as both a differentiator and cost leader. The limited number (2.5%) of
medical group practices exhibiting the strategy traits of the hybrid classification supports
the difficulty of a practice employing the strategies of both a cost leader and a
differentiator. It may be beneficial to eliminate the hybrid category from future
theoretical models exploring the financial performance and strategies within medical
group practices as it does appear that in general, the strategies of cost leadership and
differentiation are at opposite extremes, and that few medical group practices attempt to
deploy both strategies simultaneously.
While the theoretical, a priori, classification model was posited to generate groups
that would better predict the financial performance of the medical groups, this was not
supported. Rather, the inductive classification scheme explained much more of the
variation in the profit ratio, nearly twice that of the inductive methodology, for each
practice. However, it is interesting that the inductive classification technique created
several groups that were very similar to the cost leader, differentiator, and mixed strategy
groups that were generated by the theoretical methodology. Further, the similarities in
the groups formed by each methodology provide validation for the a priori, theoretical
classification of the medical group practices using Porter’s generic strategies.
The composite strategy indicator variable that created the construct for the groups
noted as having a differentiated strategy included the number of branches, the dollars
spent on marketing and furniture/equipment, and the number of ancillary services for
each practice. The construct for the cost leadership strategy included a number of
measures related to efficiency, with practices designated as a cost leader having smaller
106
practice sizes (measured by square feet), fewer support staff, lower account receivable
ratios, and lower operating cost ratios. With the support from both the inductive and
deductive methodologies of the differentiator as the group with the highest mean profit
ratio, those overseeing the management of medical group practices may want to focus
their efforts on the differentiation strategies, and less on efficiency. However, in the
cluster analysis, the group with the highest mean profit ratio, as well as the first or second
highest rankings on the four differentiator measures, had the lowest accounts receivable
ratio (i.e., fewer dollars in accounts receivable greater than 120 days compared to the
total dollars in accounts receivable). This may indicate that medical group practices
should implement not only the strategies of differentiation, but should also focus on
lowering their total days in accounts receivable in order to achieve superior profit
margins.
Limitations
This study has numerous limitations that will be discussed in the following pages.
First, equifinality holds that multiple paths may be possible to reach the same outcome,
while strict OC theorists believe that there is a single optimal organizational
configuration. Next, the somewhat unique characteristics of the health care industry may
have somewhat confound the analysis regarding the cost leadership strategy, as demand
and payment issues are typically quite different in health care, as compared with other
industries. Additionally, the various types of physician specialties within the sample may
have obfuscated the analysis if certain specialties were more likely to pursue one strategy
over another (e.g., a cardiology practice may be more likely than a pediatric practice to
107
pursue the differentiator strategy). Other limitations include the exclusion of several
variables (e.g., total RVUs) proposed in the original methodology plan due to missing
data, the elimination of the separate analysis of the single and multi-specialty practice
types, and the inclusion of the hybrid group in the model, which was found to be
extremely small.
Equifinality, an open systems concept introduced by von Bertalanffy (1949), is a
concept which finds that multiple paths may be possible to reach the same outcome. In
OC, researchers use the concept of equifinality to denote that multiple types of OCs may
be employed to create high performing firms. Thus, strict configurational theories that
espouse that one specific OC form is better than another may be problematic in the view
of equifinality. Gresov and Drazin (1997) state that many researchers point to
equifinality when their results do not find specific performance differences between
groups. In Payne’s (2001) study of medical group practices, he employed the concept of
equifinality in a sub-optimal environment, finding that contextual situations in the
medical group environment led to conflicts when employing specific strategies, and that
multiple types of medical group configurations had superior performance. However, as
noted previously, Porter’s generic strategies model has been widely used in the literature
and strategy textbooks and offers a parsimonious model to explore the different practice
configurations that performed as well as those practices that attempted to differentiate
themselves from other medical group practices. Additionally, the data supports the
contention that practices with organizational configurations that are aligned with both the
cost leader and differentiator constructs, the hybrids, are a small segment of the sample.
While these hybrid practices did not exhibit higher profit ratios than the other three
108
groups, the small sample size for this group limited a discussion about the profit ratios
found in this group.
The lower average profit ratio found in the cost leadership group may be due to
the somewhat unique traits of the health care industry. For example, most health care
consumers are somewhat insulated by medical practice costs through third-party payers.
The higher profit ratios in practices exhibiting the traits of a differentiator may indicate
that consumers are drawn to those practices that are able to differentiate themselves from
other practices, rather than the lower costs that might be present in practices with a cost
leadership strategy. Additionally, most consumers do not demand health care as with
traditional commodities. Health care is typically viewed as something required when ill,
and thus is very dissimilar to a service or product that one would purchase in a store.
Therefore, the price of the service, or searching for a lower price, may not be as relevant
for consumers of the services provided by a medical group practice.
There is another plausible explanation for the Hypothesis 1 finding that medical
groups exhibiting differentiator behavior experienced significantly higher mean profit
ratios than the other three groups. Since most medical groups in this sample were single
specialty, it is possible that the differentiator group with the higher profit ratios was
comprised of specialties which generate higher fees, such as cardiology or cardiovascular
surgery. The cost leadership group may have been comprised of specialty groups which
generate lower fees such as pediatrics and family practice. In this case, specialties which
generate higher fees may have slack resources. These slack resources may allow these
groups to pursue differentiator strategies such as investing in marketing or spending more
on furniture and equipment. However, groups which traditionally generate lower fees
109
(e.g., pediatric practices) may have fewer resources and be restricted to cost leadership
activities such as controlling overhead costs, lowering square footage per FTE physician,
and hiring fewer front office personnel.
The variable total RVUs produced by each practice was also excluded from the
analysis due to the limited number of practices that responded to this question. However,
as this variable is widely used to measure a practice’s productivity, an inferential analysis
using this variable to normalize many of the indicator variables would be beneficial.
Additionally, the variable measuring the total number of FTE physicians in each practice
was used as a proxy for total RVUs in each practice. However, these two variables
demonstrated a relatively low correlation value, which may indicate that the decision to
replace the total RVUs variable with the total FTE physicians’ variable may be
questionable.
As discussed in the results section, and in much greater detail in the appendices,
problems with the MGMA data were plentiful. Variables with missing data and/or
obvious errors confounded the data analysis to a certain degree and necessitated an
alteration of the originally planned methods, as well as the imputation of the values of
many of the variables for a number of practices. The analysis provides strong support for
changes in the data collection methods for this instrument, including a computer-based
survey that would include stipulations regarding specific questions which must be
answered by the respondent, as well as parameters based on pre-defined thresholds for
many of the questions (e.g., the respondent would not be able to include a value of
greater than 1,000 total FTE physicians in a single practice). Further, MGMA may want
110
to refine the instrument to clarify the instructions for specific variables that were found to
have a large number of extreme, dubious, or missing values.
Another limitation of this study includes the lack of an analysis of any potential
differences that may be present in the average profit ratios in each of the various types of
single specialty practices. This study did include a process to normalize the profit of each
practice by dividing the profit value by the total number of physicians in each practice
(i.e., profit / total number of FTE physicians). However, the analysis did not include
grouping the practices by specialty type (e.g., cardiology practices, pediatrics practices,
anesthesia practices, etc.) and then determining if the average profit per physician was
different across the various types of specialty groups. It would be useful if further
research using this sample included a review of whether the ratios of specific specialty
types of practices varied significantly among the groups created through Porter’s theory
and the groups created through the cluster analysis. For example, one might expect that
the average profit per physician of a cardiology practice would be significantly higher
than a pediatric practice. Thus, the group designated as differentiators might have a
higher percentage of certain types of specialty practices with higher average profit ratios
than other specialty groups. While it is not posited that the specialty group type drives the
strategy of a practice, one might expect that certain specialty practices with a higher
average profit per physician may have greater resources, which may enable those high
profit specialty practices to employ the differentiation strategy.
111
Future Research
Further analysis of this data should include a comparison of the overall sample,
including the excluded cases (those practices with less than 1 FTE physician or greater
than 35 FTE physicians), to determine if any differences in the outcome might vary based
on the excluded practices. Including the very small and large practices in future studies,
in addition to the medium-sized practices, may be beneficial in furthering the knowledge
base regarding how organizational strategies impact profit ratios, as well in determining
if differences exist between the three groups. The practices excluded from the analysis
were relatively small in number, and either very large or small practices. However, it
would be beneficial to explore whether or not any differences may occur in the outcome
if these practices are included in the analysis. Also, an exploration of the differences in
the outcome between the three groups – small, medium, and large – may aid in our
understanding of the organizational strategies used by different sizes of practices.
An analysis of the ratios of various types of specialty practices (e.g., cardiology,
pediatrics, etc.) may lead to a better understanding of the organizational strategies used
by different types of medical group practices and whether certain types of specialties are
more likely to pursue the cost leadership or differentiator strategy.
Rather than analyzing a single year of MGMA data, as was completed with this
study, a longitudinal study would with the same medical group practices may lead to
details about whether or not medical group practices may change their strategies over
time and in which practices this may occur (e.g., Leask and Parker’s 1997 study of firm-
level performance in the pharmaceutical industry over time). A longitudinal study may
112
also provide evidence regarding whether specific performance factors may predict which
firms will alter their strategies over time.
Due to the small sample size of the multispecialty practices, both the single and
multispecialty groups were combined for this study. While the small number of
multispecialty practices may limit the applicability of any findings from a study of only
this group of practices, it still may be beneficial to perform the inferential analysis
separately and compare the results of the two separate samples. This separate analysis
could provide support for practices choosing either the single or multispecialty practice
type and would lead to details regarding any differences in the strategies used by each of
the different practice types.
Further research related to how the accounts receivable ratio impacts a highly
profitable medical group practice would be beneficial as well. While Porter (1985)
specifically stated that avoidance of marginal customer accounts is an indicator of a cost
leadership strategy and MGMA (MGMA, 2006) found that the better performing medical
practices had a lower percentage of total accounts receivable greater than 120, the
outcome of this study provided a bit of a conundrum. The best performing group created
from the inductive cluster analysis technique had the highest values for the each of the
four differentiator variables and most closely resembled the differentiator group from the
theoretical model. However, this group also had the highest ranking for the accounts
receivable variable (the lowest ratio of A/R < 120 days to Total A/R), which indicated a
cost leadership strategy, and relatively low rankings on the other three cost leadership
variables. Thus, while certain organization configuration strategies related to
differentiation were found to be positively related to a medical group practice’s
113
profitability, maintaining a low A/R ratio, and not necessarily the other traits of a cost
leader, may also play an important role the profitability of medical group practices.
Although Porter did not include a hybrid group in his original theory, as discussed
previously, many researchers have found evidence that this group does exist in many
populations. However, the inclusion of the hybrid group in the theoretical model slightly
limited the power of the analysis, as those included in the hybrid group were excluded
from other groups (i.e., if the hybrid classification were eliminated, practices previously
assigned to the hybrid group would be classified as another group). As the hybrid group
was quite small in this study, the inclusion of this group in future studies with medical
group practices may limit the applicability of any inferential analyses.
Final Conclusion
Porter’s generic strategies have been used for nearly 30 years to categorize firms by
the strategies they employ and study the performance differences between the different
groups. In the field of health care, this author found no previous empirical research
specifically using Porter’s generic strategies as a theoretical grouping model for medical
group practices and little empirical research related to financial performance and strategy
differences between physician practices. Therefore, this paper will be a contribution to
the field of strategic management, organizational configurations, and medical group
practice management as it elucidates the various strategies medical group practices
employ and compares the performance levels of firms employing different organizational
configurations. It also supports Porter’s generic strategy typology in general as it
demonstrated performance differences between medical group practices grouped by
114
Porter’s typology, with the differentiator group performing significantly better than those
medical group practices conforming to another organizational configuration type. This
paper also supports the exclusion of the hybrid group in further studies of medical group
practices. For practitioners and physician practice managers, this research provides
valuable information in regards to the specific strategies and organizational
configurations that are typically associated with the most highly profitable medical group
practices. Overall, this research will provide a contribution to the literature related to
Porter’s generic strategies and the field of organizational configurations, as well as to the
field of health care and the specific strategy-performance links within medical group
practices.
115
REFERENCES
Abouleish, A. E., Prough, D. S., Whitten, C. W., Zornow, M. H., Lockhart, A., Conlay, L. A., et al. (2002). Comparing Clinical Productivity of Anesthesiology Groups. Anesthesiology, 97(3), 608-615.
Acock, A. C. (2005). Working with missing values. Journal of Marriage and Family, 67(4), 1012-1028.
Akan, O., Allen, R., S. , Helms, M., M. , & Spralls, S. A. I. (2006). Critical tactics for implementing Porter's generic strategies. [Generic strategies; specific tactics; empirical; detailed definitions of each strategy; practitioner]. The Journal of Business Strategy, 27(1), 43.
Alexandraki, I., Palacio, C., House, J., Catalano, C., & Mooradian, A. (2009). Resource-Based Relative Value Scale Analysis Between Teaching and Nonteaching Hospitalist Services. The Health Care Manager, 28(1), 81.
Allen, R. S., & Helms, M. M. (2002). Employee perceptions of the relationship between strategy, rewards and organizational performance. [Generic strategies; empirical; employee/managerial reward related to strategy and firm performance]. Journal of Business Strategies, 19(2), 115.
Allen, R. S., & Helms, M. M. (2006). Linking strategic practices and organizational performance to Porter's generic strategies. [Generic strategies; empirical; tactics leading to performance]. Business Process Management Journal, 12(4), 433.
Allen, R. S., Helms, M. M., Takeda, M. B., & White, C. S. (2007). Porter's Generic Strategies: An Exploratory Study of Their Use in Japan. [Generic strategies; empirical; international (Japan); survey only on attitudes of strategy (no performance outcome variable)]. Journal of Business Strategies, 24(1), 69.
Astley, W. G., & Van deVen, A. H. (1983). Central Perspectives and Debates in Organization Theory. Administrative Science Quarterly, 28(2), 245-273.
Bacharach, S. B. (1989). Organizational Theories: Some Criteria For Evaluation. Academy of Management. The Academy of Management Review, 14(4), 496.
Bain, J. S. (1956). Barriers to new competition, their character and consequences in manufacturing industries. Cambridge,: Harvard University Press.
Baliga, B., & Johnson, B. (1986). Analysis of an industry in transition. Health care strategic management, 4(12), 5.
Barnard, C. I. (1938). The functions of the executive. Cambridge, Mass.,: Harvard University Press.
116
Barney, J. B. (1986a). Types of Competition and the Theory of Strategy: Toward an Integrative Framework. The Academy of Management Review, 11(4), 791-800.
Barney, J. B. (1986b). Types of competition and the theory of strategy: Toward an integrative framework. Academy of Management Review, 791-800.
Barney, J. B. (1991). Firm Resources and Sustained Competitive Advantage. [Competitive advantage, sustained; firm resources; RBV theory]. Journal of Management, 17(1), 99.
Barney, J. B., & Hoskisson, R. E. (1990). Strategic Groups: Untested Assertions and Research Proposals. Managerial and Decision Economics, 11(3), 187-198.
Bharadwaj, S. G., Varadarajan, P. R., & Fahy, J. (1993). Sustainable competitive advantage in service industries: A. [Competitive advantage; theoretical and prescriptive - great article with lots of strategy tactics, possible operationalization of variables]. Journal of Marketing, 57(4), 83.
Bierly, P., & Chakrabarti, A. (1996). Generic Knowledge Strategies in the U.S. Pharmaceutical Industry. Strategic Management Journal, 17, 123-135.
Bishop, P., & Megicks, P. (2002). Competitive strategy and firm size in the estate agency industry. Journal of Small Business and Enterprise Development, 9(2).
Blair, J. D., & Buesseler, J. A. (1998). Competitive forces in the medical group industry: a stakeholder perspective. [Stakeholders - competitive advantage; healthcare - physician practices]. Health Care Management Review, v23(n2), p7(21).
Borch, O. J., Huse, M., & Senneseth, K. (1999). Resource Configuration, Competitive Strategies, and Corporate Entrepreneurship An Empirical Examination of Small Firms.(Statistical Data Included). [Porter - small firms]. Entrepreneurship: Theory and Practice, 24(1), 49.
Bowman, C., & Faulker, D. O. (1997). Competitive and Corporate Strategy.
Burns, T., & Stalker, G. M. (1961). The management of innovation. [London]: Tavistock Publications.
Buzzell, R. D., & Wiersema, F. D. (1981). Successful Share-Building Strategies. Harvard Business Review, 59(1), 135.
Campbell-Hunt, C. (2000). What Have We Learned about Generic Competitive Strategy? A Meta-Analysis. [Generic Strategy - Meta-analysis]. Strategic Management Journal, 21(2), 127-154.
Casalino, L. P., Devers, K. J., Lake, T. K., Reed, M., & Stoddard, J. J. (2003). Benefits of and Barriers to Large Medical Group Practice in the United States. Arch Intern Med, 163(16), 1958-1964.
117
Caves, R. E. (1980). Industrial Organization, Corporate Strategy and Structure. [Industrial Organization Economics]. Journal of Economic Literature, 18(1), 64-92.
Caves, R. E., & Porter, M. E. (1977). From Entry Barriers to Mobility Barriers: Conjectural Decisions and Contrived Deterrence to New Competition. [Industrial Organization Economics]. The Quarterly Journal of Economics, 91(2), 241-262.
Chan-Olmsted, S., & Jamison, M. (2001). Rivalry through alliances::: Competitive Strategy in the Global Telecommunications Market. European Management Journal, 19(3), 317-331.
Chrisman, J. J., Hofer, C. W., & Boulton, W. R. (1988). Toward a System for Classifying Business Strategies. [Strategy Typology - Porter and Abbell; a revised, new typology; theoretical - great article; need to revisit]. The Academy of Management Review, 13(3), 413-428.
Cool, K. O., & Schendel, D. (1987). Strategic Group Formation and Performance: The Case of the U.S. Pharmaceutical Industry, 1963-1982. Management Science, 33(9), 1102-1124.
Cronshaw, M., Davis, E., & Kay, J. (1994). On Being Stuck in the Middle or Good Food Costs Less at Sainsbury's. [Generic Typology - Stuck in the Middle]. British Journal of Management, 5(1), 19-32.
Cross, L. (1999). Strategy drives marketing success. Graphic Arts Monthly, 71(2), 96.
Daft, R., & Lewin, A. (1993). Where are the theories for the" new" organizational forms? An editorial essay. Organization Science, 1.
Davis, G., & Marquis, C. (2005). Prospects for organization theory in the early twenty-first century: Institutional fields and mechanisms. Organization Science, 16(4), 332.
Day, G. S., & Wensley, R. (1988). Assessing Advantage: A Framework For Diagnosing Competitive. Journal of Marketing, 52(2), 1.
Dess, G. G., & Davis, P. S. (1984). Porter's (1980) Generic Strategies as Determinants of Strategic Group Membership and Organizational Performance. Academy of Management Journal, 27(3), 467-488.
Dess, G. G., Newport, S., & Rasheed, A. M. A. (1993). Configuration research in strategic management: key issues and suggestions. Journal of Management, v19(n4), p775(721).
Dess, G. G., & Robinson, R. B., Jr. (1984). Measuring Organizational Performance in the Absence of Objective Measures: The Case of the Privately-Held Firm and Conglomerate Business Unit. Strategic Management Journal, 5(3), 265-273.
118
Doty, D. H., Glick, W. H., & Huber, G. P. (1993). Fit, equifinality, and organizational effectiveness: A test. [Configurational Theory]. Academy of Management Journal, 36(6), 1196.
Dranove, D., Peteraf, M., & Shanley, M. (1998). Do strategic groups exist? an economic framework for analysis. Strategic Management Journal, 19(11), 1029-1044.
Ebeler, J., Neuman, T., & Cubanski, J. (2011). Medicare Policy. The Kaiser Family Foundation, 1 - 24.
Fiegenbaum, A., McGee, J., & Thomas, H. (1988). Exploring The Linkage Between Strategic Groups And Competition. International Studies of Management & Organization, 18(1), 6.
Fiegenbaum, A., & Thomas, H. (1990). Strategic Groups and Performance: The U.S. Insurance Industry, 1970-84. Strategic Management Journal, 11(3), 197-215.
Filley, A., & Aldag, R. (1980). Organizational growth and types: Lessons from small institutions. Research in organizational behavior, 2, 279-320.
Filley, A. C., & Ramon, J. A. (1978). Characteristics and Measurement of an Organizational Typology. The Academy of Management Journal, 21(4), 578-591.
Fiss, P. C. (2007). A Set-Theoretic Approach to Organizational Configurations. Academy of Management. The Academy of Management Review, 32(4), 1180.
Fleming, N. S., Becker, E. R., Culler, S., Cheng, D., McCorkle, R., & Ballard, D. J. (2009). Financial performance of primary care physician practices prior to electronic health record implementation. Baylor University Medical Center Proceedings, 22(2), 112-118.
Forte, M., Hoffman, J. J., Lamont, B. T., & Brockmann, E. N. (2000). Organizational form and environment: an analysis of between-form and within-form responses to environmental change. Strategic Management Journal, 21(7), 753-773.
Galbraith, C., & Schendel, D. (1983). An Empirical Analysis of Strategy Types. Strategic Management Journal, 4(2), 153-173.
Galbraith, J. R. (1973). Designing complex organizations. Reading, Mass.,: Addison-Wesley Pub. Co.
Gapenski, L. C. (2005). Healthcare finance : an introduction to accounting and financial management (3rd ed.). Chicago: AUPHA/HAP.
Greenwood, R., & Hinings, C. R. (1993). Understanding strategic change: The contribution of archetypes. Academy of Management Journal, 36(5), 1052.
119
Gresov, C., & Drazin, R. (1997). Equifinality: Functional equivalence in organization design. Academy of Management. The Academy of Management Review, 22(2), 403.
Hambrick, D. C. (1981). Strategic Awareness within Top Management Teams. Strategic Management Journal, 2(3), 263-279.
Hambrick, D. C. (1982). Environmental Scanning and Organizational Strategy. Strategic Management Journal, 3(2), 159-174.
Hambrick, D. C. (1983). High Profit Strategies in Mature Capital Goods Industries: A Contingency Approach. The Academy of Management Journal, 26(4), 687-707.
Hambrick, D. C. (1984). Taxonomic Approaches to Studying Strategy: Some Conceptual and Methodological Issues. Journal of Management, 10(1), 27-41.
Hannan, M. T., & Freeman, J. (1977). The Population Ecology of Organizations. The American Journal of Sociology, 82(5), 929-964.
Hannan, M. T., & Freeman, J. (1984). Structural Inertia and Organizational Change. American Sociological Review, 49(2), 149-164.
Hatten, K. J., & Hatten, M. L. (1987). Strategic Groups, Asymmetrical Mobility Barriers and Contestability. Strategic Management Journal, 8(4), 329-342.
Hatten, K. J., & Schendel, D. E. (1977). Heterogeneity within an industry: firm conduct in the U.S. brewing industry, 1952-71. The Journal of Industrial Economics, 26(2), 97.
Hawes, J. M., & Crittenden, W. F. (1984). A Taxonomy of Competitive Retailing Strategies. Strategic Management Journal, 5(3), 275-287.
Helms, M. M., Dibrell, C., & Wright, P. (1997). Competitive strategies and business performance: evidence from the adhesives and sealants industry. Management Decision, 35(9), 689.
Hill, C. W. L. (1988). Differentiation Versus Low Cost or Differentiation and Low Cost: A Contingency Framework. [Porter - Cost, Differentiation; contingency approach]. Academy of Management Review, 13(3), 401-412.
Hing, E., & Burt, C. W. (2007). Office-based medical practices: methods and estimates from the national ambulatory medical care survey. Adv Data(383), 1-15.
Hing, E., & Burt, C. W. (2008). Characteristics of office-based physicians and their medical practices: United States, 2005-2006. Vital and health statistics. Series 13, Data from the National Health Survey(166), 1.
120
Hlavacka, S., Bacharova, L., Rusnakova, V., & Wagner, R. (2001). Performance implications of Porter's generic strategies in Slovak hospitals. Journal of Management in Medicine, 15(1), 44-66.
Hofer, C. W., & Schendel, D. (1978). Strategy formulation : analytical concepts. St. Paul: West Pub. Co.
Hoskisson, R. E., Hitt, M. A., Wan, W. P., & Yiu, D. (1999). Theory and research in strategic management: swings of a pendulum (Yearly Review of Management). [Strategic management - review]. Journal of Management, 25(3), 417(414).
Hult, G., Ketchen, D., Cavusgil, S., & Calantone, R. (2006). Knowledge as a strategic resource in supply chains. Journal of Operations Management, 24(5), 458-475.
Hunt, M. S. (1972). Competition in the major home appliance industry, 1960-1970 : a thesis.
Hunt, S. D. (1997). Resource-advantage theory: An evolutionary theory of competitive firm behavior? Journal of Economic Issues, 31(1), 59.
Hyatt, L. (2001). A simple guide to strategy. Nursing Homes, 50(1), 12.
Inamdar, N. S. (2007). Examining the Scope of Multibusiness Health Care Firms: Implications for Strategy and Financial Performance. Health Services Research, 42(4), 1691-1717.
Ittner, C. D., Larcker, D. F., & Pizzini, M. (2007). Performance-based compensation in member-owned firms: An examination of medical group practices. Journal of Accounting and Economics, 44(3), 300-327.
Jensen, M. (1983). Organization theory and methodology. Accounting Review, 319-339.
Jiang, H. J., Bernard, F., & James, W. B. (2006). Sustaining and Improving Hospital Performance: The Effects of Organizational and Market Factors. Health Care Management Review, 31(3), 188.
Johnson, S., & Newton, W. (2002). Resource-based relative value units: a primer for academic family physicians. FAMILY MEDICINE, 34(3), 172-176.
Kang, S.-C., Morris, S. S., & Snell, S. A. (2007). Relational Archetypes, Organizational Learning, and Value Creation: Extending the Human Resouce Architecture. Academy of Management Review, 32(1), 236-256.
Ketchen, D. J., Combs, J. G., Russell, C. J., Shook, C., Dean, M. A., Runge, J., et al. (1997). Organizational configurations and performance: A meta-analysis. Academy of Management Journal, 40(1), 223.
Ketchen, D. J., Combs, J. G., Russell, C. J., Shook, C., & et al. (1997). Organizational configurations and performance: A meta-analysis. Academy of Management Journal, 40(1), 223.
Ketchen, D. J., & Shook, C. L. (1996). The Application of Cluster Analysis in Strategic Management Research: An Analysis and Critique. Strategic Management Journal, 17(6), 441-458.
Ketchen, D. J., Thomas, J. B., & Snow, C. C. (1993). Organizational Configurations and Performance: A Comparison of Theoretical Approaches. Academy of Management Journal, 36(6), 1278.
Kikulis, L. M., Slack, T., & Hinings, C. R. (1995). SECTOR-SPECIFIC PATTERNS OF ORGANIZATIONAL DESIGN CHANGE. Journal of Management Studies, 32(1), 67-100.
Kumar, K., Subramanian, R., & Yauger, C. (1997). Pure versus hybrid: performance implications of Porter's generic strategies. Health Care Management Review, v22(n4), p47(14).
Langland-Orban, B., Gapenski, L., & Vogel, W. (1996). Differences in characteristics of hospitals with sustained high and sustained low profitability. Hospital & health services administration, 41(3), 385.
Lawrence, P. R., & Lorsch, J. W. (1967). Differentiation and Integration in Complex Organizations. Administrative Science Quarterly, 12(1), 1-47.
Leask, G., & Parker, D. (2007). Strategic groups, competitive groups and performance within the U.K. pharmaceutical industry: Improving our understanding of the competitive process. Strategic Management Journal, 28(7), 723-745.
Lin, H.-C., Chen, C.-S., Liu, T.-C., & Lee, H.-C. (2006). Differences in practice income between solo and group practice physicians. [solo vs. group practice]. Health Policy, 79(2-3), 296-305.
Locke, R., & Srinivasan, M. (2008). Attitudes Toward Pay-for-Performance Initiatives Among Primary Care Osteopathic Physicians in Small Group Practices. JAOA: Journal of the American Osteopathic Association, 108(1), 21.
March, J., Simon, H., & Guetzkow, H. (1958). Organizations: John Wiley & Sons Inc.
Mark, T. L. (1999). Analysis of the rationale for, and consequences of, nonprofit and for-profit ownership conversions. Health Services Research, 34(1), 83(81).
122
Mason, E. S. (1939). Price and Production Policies of Large-Scale Enterprise. [Industrial Organization Theory - historical]. The American Economic Review, 29(1), 61-74.
Maxwell, S., & Zuckerman, S. (2007). Impact of Resource-Based Practice Expenses on the Medicare Physician Volume. Health Care Financing Review, 29(2), 65.
McDonagh, K., J. , & Umbdenstock, R., J. . (2006). Hospital Governing Boards: A Study of Their Effectiveness in Relation to Organizational Performance. Journal of Healthcare Management, 51(6), 377.
Meyer, A. D., Tsui, A. S., & Hinings, C. R. (1993). Configurational approaches to organizational analysis. Academy of Management Journal, 36(6), 1175.
MGMA. (2006). Performance and Practices of Successful Medical Groups: 2006 Report Based on 2005 Data. Englewood, CO: Medical Group Management
Association (MGMA).
Miles, R. E., & Snow, C. C. (1978). Organizational strategy, structure, and process. New York: McGraw-Hill.
Miller, A., & Dess, G. G. (1993). Assessing Porter's (1980) model in terms of its generalizability, accuracy and simplicity. The Journal of Management Studies, 30(4), 553.
Miller, D. (1986). Configurations of Strategy and Structure: Towards a Synthesis. Strategic Management Journal, 7(3), 233-249.
Miller, D. (1988). Relating Porter's Business Strategies To Environment And St. Academy of Management Journal, 31(2), 280.
Miller, D. (1992). The Generic Strategy Trap. [Generic strategies - mixed strategies; theory]. The Journal of Business Strategy, 13(1), 37.
Miller, D., & Friesen, P. H. (1978). Archetypes of Strategy Formulation. Management Science, 24(9), 921-933.
Miller, D., & Friesen, P. H. (1986a). Porter's (1980) Generic Strategies and Performance: An Empirical Examination with American Data -- Part II: Performance Implications. Organization Studies, 7(3), 255.
Miller, D., & Friesen, P. H. (1986b). Porter's (1980) Generic Strategies and Performance: An Empirical Examination with American Data Part I: Testing Porter. Organization Studies, 7(1), 37-55.
Miller, D., Friesen, P. H., & Mintzberg, H. (1984). Organizations : a quantum view. Englewood Cliffs, N.J.: Prentice-Hall.
123
Mintzberg, H. (1979). The structuring of organizations : a synthesis of the research. Englewood Cliffs, N.J.: Prentice-Hall.
Mintzberg, H. (1990). 5 STRATEGY FORMATION Schools of Thought. Perspectives on strategic management, 105.
Murray, A. I. (1988). A Contingency View of Porter's "Generic Strategies". Academy of Management Review, 13(3), 390-400.
Nayyar, P. R. (1993). On the measurement of competitive strategy: Evidence from a large multiproduct U.S. firm. Academy of Management Journal, 36(6), 1652.
Nenide, B., Pricer, R. W., & Camp, S. M. (2003). The Use of Financial Ratios for Research: Problems Associated With and Recommendations for Using Large Databases. Paper presented at the Fourth Annual Conference of the National Business and Economics Society, Virgin Islands.
Newman, H. (1978). Strategic groups and the structure-performance relationship. The Review of Economics and Statistics, 417-427.
Oster, S. (1982). Intraindustry structure and the ease of strategic change. The Review of Economics and Statistics, 376-383.
Parker, B., & Helms, M. M. (1992). Generic Strategies and Firm Performance in a Declining Industry. [Generic strategies; empirical; survey]. Management International Review, 32(1), 23.
Parnell, J. A. (2006). Generic strategies after two decades: a reconceptualization of competitive strategy. [Generic strategies - review of literature]. Management Decision, 44(8).
Payne, G. T. (2001). Strategy and Structure Configurations: An Examination of Fit and Performance. Texas Tech.
Payne, G. T. (2006). Examining Configurations and Firm Performance in a Suboptimal Equifinality Context. Organization Science, 17(6), 756.
Penrose, E. T. (1959). The theory of the growth of the firm. New York,: Wiley.
Pham, H. H., Devers, K. J., May, J. H., & Berenson, R. (2004). Financial Pressures Spur Physician Entrepreneurialism. Health Affairs, 23(2), 70.
Phillips, L. W., Chang, D. R., & Buzzell, R. D. (1983). Product Quality, Cost Position and Business Performance: A Test of Some Key Hypotheses. [Generic strategies; support of mixed]. Journal of Marketing, 47(2), 26-43.
Porter, M. E. (1979a). How competitive forces shape strategy. Harvard Business Review, 57(2), 137-145.
124
Porter, M. E. (1979b). The structure within industries and companies' performance. The Review of Economics and Statistics, 214-227.
Porter, M. E. (1980). Competitive strategy : techniques for analyzing industries and competitors. New York: Free Press.
Porter, M. E. (1981). The Contributions of Industrial Organization To Strategic Management. Academy of Management Review, 6(4), 609-620.
Porter, M. E. (1985). Competitive advantage: Creating and sustaining superior performance: Free Press.
Porter, M. E. (1996). What Is Strategy? Harvard Business Review, 74(6), 61-78.
Porter, M. E. (1997). New strategies for inner-city economic development. Economic Development Quarterly, v11(n1), p11(17).
Porter, M. E., & Teisberg, E. (2004). Redefining competition in health care. Harvard Business Review, 64-77.
Powers, T. L., & Hahn, W. (2004). Critical competitive methods, generic strategies, and firm performance. [Generic strategies - performance and competitive methods; empirical]. The International Journal of Bank Marketing, 22(1), 43.
Rangone, A. (1999). A resource-based approach to strategy analysis in small-medium sized enterprises. Small Business Economics, 12(3), 233.
Reger, R. K., & Huff, A. S. (1993). Strategic Groups: A Cognitive Perspective. Strategic Management Journal, 14(2), 103-123.
Relman, A. S. (2007). A second opinion: rescuing America's healthcare: a plan for universal coverage serving patients over profit: Public Affairs.
Reschovsky, J., Hadley, J., & Landon, B. (2006). Effects of compensation methods and physician group structure on physicians' perceived incentives to alter services to patients. Health Services Research, 41(4 Pt 1), 1200.
Robinson, J. C. (2001). Physician organization in California: Crisis and opportunity. Health Affairs, 20(4), 81.
Romano, M. (2001). United we do quite well, thank you. Modern Healthcare, 31(12), 34-34.
Sandino, T. (2007). Introducing the First Management Control Systems: Evidence from the Retail Sector. The Accounting Review, 82(1), 265.
125
Schappert, S. M., & Burt, C. W. (2006). Ambulatory care visits to physician offices, hospital outpatient departments, and emergency departments: United States, 2001-02. Vital and Health Statistics(159), 1-66.
Schoemaker, P. J. H. (1990). Strategy, Complexity and Economic Rent. Management Science, 36(10), 1178.
Segev, E. (1989). A Systematic Comparative Analysis and Synthesis of Two Business-Level Strategic Typologies. Strategic Management Journal, 10(5), 487-505.
Selznick, P. (1957). Leadership in administration; a sociological interpretation. Evanston, Ill.,: Row.
Short, J. C., Ketchen, D. J., Palmer, T. B., & Hult, G. T. M. (2007). Firm, strategic group, and industry influences on performance. Strategic Management Journal, 28(2), 147-167.
Short, J. C., Payne, G. T., & Ketchen, D. J. (2008). Research on Organizational Configurations: Past Accomplishments and Future Challenges. [Feature]. Journal of Management, 34(6), 1053-1079.
Shortell, S. M., Schmittdiel, J., Wang, M., Li, R., Gillies, R., Casalino, L., et al. (2005). An empirical assessment of high-performing medical groups: results from a national study. Medical Care Research and Review, 62(4), 407.
Shortell, S. M., & Zajac, E. J. (1990). Perceptual and Archival Measures of Miles and Snow's Strategic Types: A Comprehensive Assessment of Reliability and Validity. The Academy of Management Journal, 33(4), 817-832.
Sides, R. W. (2000). Cost analysis helps evaluate contract profitability. Healthcare Financial Management, 54(2), 63.
Sisko, A., Truffer, C., Smith, S., Keehan, S., Cylus, J., Poisal, J., et al. (2009). Health spending projections through 2018: Recession effects add uncertainty to the outlook. Health Affairs, 28(2), w346.
Spanos, Y. E., Zaralis, G., & Lioukas, S. (2004). Strategy and industry effects on profitability: evidence from Greece. Strategic Management Journal, 25(2), 139-165.
Spence, A. M. (1977). Entry, Capacity, Investment and Oligopolistic Pricing. [Industrial Organization Economics]. The Bell Journal of Economics, 8(2), 534-544.
Stonehouse, G., & Snowdon, B. (2007). Competitive Advantage Revisited: Michael Porter on Strategy and Competitiveness. Journal of Management Inquiry, 16(3), 256.
126
Sunshine, J., & Burkhardt, J. (2000). Radiology Groups' Workload in Relative Value Units and Factors Affecting It. Radiology, 214(3), 815.
Tennyson, D., & Fottler, M. (2000). Does system membership enhance financial performance in hospitals? Medical Care Research and Review, 57(1), 29.
Thibadoux, G. M., Scheidt, M., & Luckey, E. (2007). Accounting and Medicine: An Exploratory Investigation into Physicians' Attitudes Toward the Use of Standard Cost-Accounting Methods in Medicine. Journal of Business Ethics, 75(2), 137.
Thornhill, S., & White, R. E. (2007). Strategic purity: A multi-industry evaluation of pure vs. hybrid business strategies. [Generic strategies - mixed strategies; theory]. Strategic Management Journal, 28(5), 553-561.
Torgovicky, R., Goldberg, A., Shvarts, S., Dayan, Y. B., Onn, E., Levi, Y., et al. (2005). Application of Porter's Generic Strategies in Ambulatory Health Care: A Comparison of Managerial Perceptions in Two Israeli Sick Funds. [Porter - healthcare]. Health Care Management Review, 30(1), 17.
Truffer, C. J., Keehan, S., Smith, S., Cylus, J., Sisko, A., Poisal, J. A., et al. (2010). Health Spending Projections Through 2019: The Recession's Impact Continues. Health Aff, hlthaff.2009.1074.
von Bertalanffy, L. (1949). Problems of organic growth. Nature, 163(4135), 156-158.
Vonderheid, S., Pohl, J., Schafer, P., Forrest, K., Poole, M., Barkauskas, V., et al. (2004). Using FTE and RVU Performance Measures to Assess Financial Viability Of Academic Nurse-Managed Centers. Nursing Economics, 22(3), 124.
Vorhies, D., & Morgan, N. (2003). A configuration theory assessment of marketing organization fit with business strategy and its relationship with marketing performance. Journal of Marketing, 67(1), 100-115.
Wagner, B., & Digman, L. (1997). The relationships between generic and time-based strategies and performance. Journal of Managerial Issues, 9(3), 334.
Watt, J., Derzon, R., Renn, S., Schramm, C., Hahn, J., & Pillari, G. (1986). The comparative economic performance of investor-owned chain and not-for-profit hospitals. New England Journal of Medicine, 314(2), 89.
Weber, M., Henderson, A. M., & Parsons, T. (1947). The theory of social and economic organization. London: W. Hodge.
Westbrook, K. W. (1997). Assessing the strategy-performance paradigm: Implications for organizational adaptation in discontinuous environments for home health care providers. Health Marketing Quarterly, 14(4), 27 - 44.
127
White, R. E. (1986). Generic Business Strategies, Organizational Context and Performance: An Empirical Investigation. Strategic Management Journal, 7(3), 217-231.
Wiggins, R. R., & Ruefli, T. W. (1995). Necessary conditions for the predictive validity of strategic groups: Analysis without reliance on clustering techniques. Academy of Management Journal, 38(6), 1635.
Wirtz, B. W., Mathieu, A., & Schilke, O. (2007). Strategy in High-Velocity Environments. Long Range Planning, 40(3), 295-313.
Woodward, J. (1958). Management and technology. London: H. M. Stationery Office.
Wright, P. (1987). A Refinement of Porter's Strategies. [Porter - Theory]. Strategic Management Journal, 8(1), 93-101.
Xu, S., Cavusgil, S., & White, J. (2006). The impact of strategic fit among strategy, structure, and processes on multinational corporation performance: a multimethod assessment. Journal of International Marketing, 14(2), 1-31.
Zammuto, R. F. (1988). Organizational Adaptation: Some Implications of Organizational Ecology for Strategic Choice. The Journal of Management Studies, 25(2), 105.
128
APPENDIX A
Total RVUs and Total Number of FTE Physicians
Table A1 Correlation Between Total Number of FTE Physicians (N = 1,775) and Total RVUs (N = 684) Correlation Coefficients Pearson’s Correlation Kendall’s tau b Spearman’s rho Total Physician FTEs & Total RVUs
.185** .488** .582**
**Correlation is significant at the 0.01 level (2-tailed).
129
APPENDIX B
Practices Grouped By Size Table A2 Descriptive Statistics - Total Number of FTE Physicians By Groups – All Cases (N = 1,788)
Number of FTEs Frequency Percent 0 to .99 148 8.3 1 to 4.99 743 41.6 5 to 9.99 327 18.3 10 to 14.99 139 7.8 15 to 19.99 89 5.0 20 to 35 123 6.9 36 to 50 68 3.8 51 to 100 83 4.6 101 to 250 40 2.2 251 to 1,000 12 .7 Greater than 1,000 3 .2 Missing 13 .7
Table A3 Total Number of FTE Physicians By Groups – 1 Physician FTE to 35 Physician FTEs (N = 1,421) Number of FTEs Frequency Percent 1 to 4.99 743 52.3 5 to 9.99 327 23.0 10 to 14.99 139 9.8 15 to 19.99 89 6.3 20 to 35 123 8.7
130
APPENDIX C
Comparison of Practices Included and Excluded Table A4 Comparison of Practices Included and not Included in the Final Analysis Practice Characteristics Inc. In Final Analysis? Legal organization Majority owner Population category
No M 3.54 3.33 2.64 N 247 375 245 SD 1.50 1.60 1.04
Yes M 3.40 3.50 2.43 N 1099 1413 1094 SD 1.52 1.58 1.07
Total M 3.43 3.46 2.47 N 1346 1788 1339 SD 1.52 1.58 1.06
Table A5 One-Way Analysis of Variance Summary for Practice Characteristics Practice Characteristic df SS MS F Legal organization Between Groups 1 3.85 3.85 1.68
Within Groups 1,344 3089.22 2.30 Majority owner Between Groups 1 9.08 9.08 3.63
Within Groups 1,786 4463.56 2.50 Population category Between Groups 1 8.86 8.59 7.62**
Within Groups 1,337 1506.75 1.13
**p<0.01.
Table A6
Effect Size for Practice Demographic Variables Practice Characteristics η2 Legal organization .001 Majority owner .002 Population category .006
131
APPENDIX D
Descriptive Statistics and Discuss of Selected Variables
Descriptive Analysis for the Independent Variables - Differentiator Indicators
Marketing. The promotion and marketing costs reported by each practice were a
proxy for their advertising intensity and an indicator of whether or not the practice was
deemed a differentiator. Out of the sample of 1,413 practices, 418 respondents did not
provide a response to this question. Of those responding (n = 995) with the total dollars
spent by their practice on promotion or marketing, the practices had a mean of $36,562
and a range between $0 and $873,594. To account for varying sizes of the practices, this
variable was normalized with the total number of FTE physicians in each practice. Thus,
a new variable was created as a ratio to measure the advertising intensity in each practice,
with promotion and marketing costs as the numerator and the total number of FTE
physicians in each practice as the denominator.
Specialty Services. The total number of ancillary services provided by each practice
was used as a proxy for whether or not a practice offered specialty products/services and
was an indicator of whether or not a practice was a labeled as a differentiator. Each
respondent answered “yes” if their practice offered one of 17 specific ancillary services
(e.g, general radiology, drug administration, durable medical equipment, etc.). A new
variable was computed to sum the total number of ancillary services available in each
practice. In the sample used for the final model (n = 1,413), 37.1% (524) of the cases
included missing values one or more of the 17 ancillary services.
132
Table A7
Specialty Services: Number of Practices Providing Ancillary Services – 17 Possible
Due to the findings in the descriptive analysis for the total ancillary services
available in each practice, a determination was made to recode this variable into a binary
134
variable, with 58% (820) of the practices reporting no ancillary services and 42% (593)
of the practices reporting one or more ancillary services. Comparing the final sample
with the original sample, the two samples were similar, with the original sample of 1,788
having 59.2% (1,059) of the practices reporting no ancillary services and 40.8% (729) of
the practices reporting one or more ancillary services.
Image. Each respondent was asked to provide the costs of general furniture and
equipment, as well as the depreciation costs of the general furniture and equipment, in
their practice. These values were used as a proxy for each practice’s “Image” and were
an indicator of whether or not a practice was included as a differentiator. The average
value for furniture and equipment for the practices was $48,896, with a range between $0
and $1,357,256. The average value for furniture and equipment depreciation was
$102,734, with a range between a negative $167,081 and a positive $6,501,692. Many
respondents (n = 533 and n=477, respectively) did not report a value for their general
furniture and equipment costs or the depreciation value for the general furniture and
equipment category.
The values for these two variables were summed together and created the numerator
for the “Image” indicator ratio variable. To account for the missing values with one or
the other variables in the numerator, it was necessary to address the missing data. Thus,
74 missing values for the furniture and equipment variable and 120 missing values for the
furniture and equipment and depreciation variable were recoded as a zero, resulting in a
total of 607 practices reporting a missing value for this variable. Once this was
completed, the “Image” variable was created by summing these two variables. The
resulting variable included 1,000 cases without missing data. The mean was $138,160
135
with a range of $0 to $6,634,036. To account for the varying sizes of each practice in the
sample, a ratio for this variable was created with the summed furniture and equipment
variables as the numerator and a denominator including the number of FTE physicians in
each practice.
Table A10
Descriptive Statistics for the Image Ratio Variable (Differentiator) (N=1,000)
Min Max M SD Image ($) 0 6,634,036 138,160 339,056
Branches. The total number of branches reported by each practice was a proxy for
their range of market segments and was an indicator or whether or not a practice would
be labeled as a differentiator. Of the initial 1,788 practices, 27.5% (492) of the
respondents did not report the number of branches for their practice. Thus, an analysis
was completed of the practices not reporting the number of branches compared with
those practices reporting the number of branches, based on the total number of FTE
physicians in each of these two groups. The intention of this analysis was to determine if
an assumption could be made that the practices not reporting a branch typically did not
have any braches beyond their primary location. The analysis excluded data from 13
respondents who did not include data regarding the total number of branches for their
practice, as well as the three cases previously identified as outliers (reporting > 1,000
total FTE physicians in their respective practices).
136
Table A11
Descriptive Statistics for Total Number of Branches (N=1,788)
Practices Reporting Number of Branches Valid or Missing Total Percent Valid 1,296 72.5 Missing 492 27.5
The results of this analysis indicated that those practices not reporting the number of
branches were significantly smaller in total number of physician FTEs as compared with
all practices, with a mean of 3.8 physicians per practice. Even those practices reporting
zero or one branch, reported a higher average total physician FTEs per practices as
compared with a missing values for the question regarding their number of branches. In
fact, the other practices significantly increased in number of FTE physicians when
grouped by number of branches. Due to these results, for those with missing data for the
total number of branches, a determination was made to recode this variable to zero for
their reported branches value. The tables below details the results of this analysis.
Table A12
Comparison of Total Physician FTEs vs. Number of Branches (N=1,772)
Total Number of Physicians in Each Practice
Number of Branches N M SD Min Max 0 644 10.0893 41.93252 .00 968.00 1 157 12.9248 26.22903 .00 222.00 2 118 13.2851 13.72947 .00 63.00 3 to 5 145 19.0412 18.81416 .00 138.51 5 or More 217 72.7080 92.11816 1.00 733.60 Missing 491 3.8011 8.04773 .00 97.67 Total 1772 17.2118 47.34691 .00 968.00 Note: All Cases, Excluding 13 with missing Total Physicians FTEs and 3 Outliers
137
Table A13 One-Way Analyses of Variance for Total Number of Physician FTE and Number of Branches (N=1,772)
Source Df SS MS F Between Groups 5 794486.73 158897.35 88.37*** Within Groups 1,766 3175616.90 1798.20 Total 1,771 3970103.62 *** p < 0.001
Once the practices with missing values for their number of branches were changed to
zero, the descriptive results from the sample included in the model (n = 1,413) indicate
that 67.1% (948) of the practices reported no branches (the primary facility/clinic did not
count as a branch), 26.4% (373) reported 1 – 5 branches, and 6.5% (92) reported 6 or
more branches. Due to this finding, a determination was made to recode the number of
branches for each practice into a binary variable, with the results indicating that 67.1%
(948) of the cases having no branches and 32.9% (465) of the branches having 1 or more
branches.
Table A14
Practices with No Missing Values for Number of Branches (N=1,413)
Number of Branches Frequency Percent 0 948 67.1 1 145 10.3 2 105 7.4 3 to 5 123 8.7 6 to 10 64 4.5 11 or more 28 2.0
138
Descriptive Analysis for the Independent Indicator Variables – Cost Leader
Production and Operations Efficiency – Cost Leader. The survey respondents
reported the total general operating costs for their practices. This value served as a proxy
for our production and operations efficiency variable, and was an indicator of whether a
practice was deemed a cost leader. As discussed earlier, two of practices did not report a
value for their practice’s total general operating costs, two other cases reported a negative
value, and three reported zero. These seven cases were eliminated from further analysis.
Of the 1,413 cases in our final sample, total general operating costs ranged from $2,160
to $79,274,718, with a mean of $1,860,000. To normalize this value, a ratio was created
using the total general operating costs as the numerator and each practice’s total number
of FTE physicians as the denominator.
Our initial model included a second ratio to measure of the production and
operation efficiency of each practice – total FTE physician in each practice divided by
total RVUs. As the variable for total FTE physicians in each practice has been used as
the denominator for many of our new construct’s variables, and our sample size of cases
including the total RVUs in each practice was small (n = 692 for the overall sample and
n=551 for those cases included in the final sample), this specific indicator variable was
eliminated from the final model.
Control of Costs – Cost Leader. Each respondent reported the value for gross
square footage of all of their practice facilities, which was a proxy for how well their
practice controlled costs as indicated by the physical size of each practice. This variable
was used as an indicator within the cost leader construct and was normalized by using the
original square footage as the numerator and total number of physician FTEs as the
139
denominator of a newly created ratio. Our initial descriptive analysis indicated 50
practices responded with “zero” for their square footage value and another 13 practices
responded with very low values (42 – 587). For this variable, the values for these 63
cases were recoded as missing data. Adding these two subsets to the originally missing
data for this variable, there were a total of 536 cases with missing data. Of the 877
practices reporting data, the average square footage was 18,605, with a range between
587 – 500,000 total square feet. As with the other four cost leader indicator variables, a
ratio was created with the square footage as the numerator and the total number of FTE
physicians as the denominator.
Table A15
Descriptive Statistics – Square Footage
Square Feet N Min Max Mean SD
Only Cases Inc. In Final Model (N=1,413)* 877 587 500,000 18,605 27,059 All Cases (N=1,772)** 1,053 587 1,646,582 41,909 97,322 * Not including cases with missing values ** Not including cases with missing values or outliers Tight Control of Overhead Costs – Cost Leader. Respondents were asked to provide
the total number of support staff employed by their practice, which was a proxy for how
well each practice controlled their overhead costs. This variable was used to determine
whether or not a practice was designated as a cost leader. Seventy-two practices reported
a zero as the value for the total number of support staff in their practice. As it was
determined that is was unlikely that a practice could operate without any support staff,
the total number of support staff variable was recoded as a zero for these practices. Of
the remaining 1,317 cases with reported values, the average total number of support staff
140
reported in each practice was 32.5, with a range between 0.01 and 346.88. As with the
other four cost leader indicator variables, a ratio was created with the total number of
support staff FTEs in each practice as the numerator and the total number of FTE
physicians as the denominator.
Table A16
Descriptive Statistics - Total Number of Support Staff FTEs
Total Number of Support Staff FTEs N Min Max M SD
Only Cases Inc. In the Final Model (N=1,413)* 1,317 .01 347 32.56 44.69 All Cases (N=1.772) 1,643 .01 9,545 81.07 304.51 * Not including cases with missing values ** Not including cases with missing values or outliers
141
Tight Control of Marginal Accounts – Cost Leader. In our analysis, a ratio for each
practice’s accounts receivable value was created (measured in dollars) by dividing the
total accounts receivable greater than 120 days by each practice’s overall accounts
receivable value. This ratio was an indicator of the practice’s tight control of marginal
costs and part of the construct that determined whether each practice was deemed a cost
leader. Sixteen cases with a negative value for this variable were recoded as a zero to
prevent potential issues with a negative numerator value for the subsequent ratio that will
be created for this variable.
For each practice’s accounts receivable greater than 120 days value, the average was
$326,340 with a range between $0 and $10,632,000. The variable measuring accounts
receivable greater than 120 days had 119 missing values. The total accounts receivable
for each practice was an average of $1,585,979, with a range between $0 and
$61,581,000. The variable measuring total accounts receivable had 98 missing values.
Once the ratio was computed, the values varied from 0 to 1.0, with an average of 0.2113,
and resulted in 121 practices with a missing value for their accounts receivable ratio.
AR > 120 Days ($) Only Cases Inc in Final Model* All Cases**
Total Accounts Receivable ($) Only Cases Inc In Final Model* All Cases** Accounts Receivable Ratio Only Cases Inc In Final Model* All Cases**
1,294 1,642
1,315 1,666
1,292 1,636
0.00
0
0 0
0.00 0.00
10,632,000
378,44,715
61,581,000 106,885,913
1.00 1.00
326,341 662,390
1,585,979 3,120,629
.2113 .2213
7.24E5 2.06E6
3,083,723 8,069,586
.14709 .17328
* Not including cases with missing values ** Not including cases with missing values or outliers
143
Descriptive Analysis for the Dependent/Outcome Variable – Profit.
In the inferential model, the dependent variable is a measure of the profit ratio for
each practice. This variable is measured by the total medical revenue after operating
costs for each practice divided by the total number of FTE physicians within each
practice. The mean profit for our sample is $2,860,000, with a range between a negative
$16,733,495 and a positive $38,127,495. As with many of the independent variables in
the model, this value was normalized by using the total number of FTE physicians in
each practice as the denominator for the profit ratio, and the total medical revenue after
operating costs as the numerator.
Table A18
Descriptive Statistics – Profit (Total Medical Revenue Less Operating Costs)
Profit N Min Max M SD
Only Cases Inc. In Final Model* 1413 $-16,733,495 $38,127,495 $2.86E6 $4.359E6 All Cases** 1783 $-49,607,581 $430,085,103 $6.08E6 $1.767E7 * Not including cases with missing values ** Not including cases with missing values or outliers
Imputation of Values - Residuals for Eight Indicator Variables via Cluster Analysis
Indicators Marketing Image Specialty Services Branches Efficiency Cost Control AR Overhead
Marketing .
Image 0.22 .
Specialty Services 0.21 0.53 .
Branches 0.20 0.64 2.19 .
Efficiency 0.08 0.52 1.50 0.36 .
Cost Control 0.00 0.00 0.44 2.45 1.89 .
AR 0.65 2.43 0.99 1.18 0.19 0.00 .
Overhead 0.05 0.79 1.04 0.00 2.44 0.00 0.93 .
146
APPENDIX G
Post-hoc Analysis – Hypothesis 1
Table XYZ
Post-Hoc Analysis for Hypothesis 1 - Scheffe
Porter’s Groups N 1 2
Stuck In The Middle 594 218,884 Low Cost Leader 489 254,077 Hybrid 36 352,835 Differentiator 294 585,928 Sig. .190 1.00
Table G2
Hypothesis 1 –Post Hoc Analyses - Bonferroni Comparisons for Strategy Groups
Comparisons
Mean
Difference($)
Std.
Error
95% CI Lower Bound
Upper Bound
Differentiator vs. Stuck In The Middle 367,044* 33,110 279,567 454,521 Differentiator vs. Cost Leader 331,851* 34,267 241,318 422384 Differentiator vs. Hybrid 233,093* 81,988 16,479 449,707 Cost Leader vs. Stuck in the Middle 35,193 28,352 -39,174 110,099 Hybrid vs. Stuck in the Middle 133,951 79,698 -76,611 344,513 Hybrid vs. Cost Leader 98,759 80,186 -113,091 310,609