 

LIBRARY
Michigan State
University

 

 

 

PLACE ll RETURN BOX to movothin mum your record.

TO AVOID FINES mum on or baton duo duo.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

moismmmwomlm

Wm:

DETERMINANTS OF THE LOCATION OF THE
FREESI'AN DIN G AMBULATORY CARE CENTER

BY

Elizabeth Gail Lowell-Smith

A DISSERTATION

Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY

Department of Geography

1994

ABSTRACT

DETERMINANTS OF THE LOCATION OF THE
FREESTANDIN G AMBULATORY CARE CENTER

By

Elizabeth Gail Lowell-Smith

The freestanding ambulatory care center (FACC) is defined as any health
care facility which is open for extended hours (usually evenings and / or
weekends) and provides medical care on a no-appointment necessary basis.
FACCs, also known as walk-in clinics or urgent care centers, were originally
designed to provide care for minor injuries and illnesses in lieu of crowded
and expensive hospital emergency rooms. Today, many FACCs are offering
primary as well as urgent care. This dissertation had three research objectives:
to determine the locational patterns of PACCs at a variety of scales and the
reasons for these location patterns, and to understand the place of the FACC in
the American health care system.

This study focuses on the location of FACCs at three general scales:
regions, metropolitan statistical areas and the site level. For the regional and
MSA levels, secondary data on FACC locations were obtained for the years
1985 (regional only) and 1993 (all scales). At the regional scale, it was found
that although states in the South and West still contained a majority of the
FACCs, their relative dominance had declined between 1985 and 1993, while
the North Central region gained FACCs. Only the Northeast continued to lag
in the number of FACCs. In general, FACCs dispersed nationwide

between 1985 and 1993. 95% of PACCs in 1993 were found in metropolitan
areas and in three of the four Census regions, central city locations were more
prevalent that suburban locations for the clinics.

Analyses were conducted on FACC location for the 100 largest MSAs.
The location of FACCs in central cities was found to be related to measures of
demand and to the hospital beds/ population ratio. FACCs located in
suburban counties were correlated with employment structure and the
physician/ population ratio. Suburban FACCs showed relatively weaker
relationships to population variables than did central city PACCs.

Data for the analysis at the site level were obtained through a mail
survey of FACCs in Michigan. The response rate was 15% (33 surveys). The
survey gave information on clinic characteristics, operations and services. The
survey also provided data on competition and the interconnections between
FACCs and other health care providers. It is concluded that, at the regional
and metropolitan scales, FACCs show location patterns which are similar to
other services and health care providers. Future research at smaller scales may

reveal important differences.

ACKNOWLEDGENIENT'S

I would like to thank my advisor, Bruce Pigozzi, for his support and
advice on this dissertation and throughout my program at MSU. I would also
like to thank the members of my committee, Gary Manson, Assefa Mehretu,
Richard Groop, and Sharmistha Bagchi-Sen for their assistance. Thanks are
likewise due to the Geography department secretaries, Harriet Ashbay,
Marilyn Bria, Sharon Ruggles, and Judy Slate and to computer specialist Mike
Lipsey for their support and help during the last three years. I appreciate the
assistance of the Geography department in facilitating the completion of my
degree by providing assistantships, research money, and teaching
opportunities. Finally, I would like to acknowledge Ms. Molly Fisher and Mr.
Art Auer of NAFAC for providing the data on FACCs used in this study.

This dissertation is dedicated to my husband, Hugh S. Smith, and to my

mother, Gail B. Giles. I will always be grateful for their love and support.

iv

TABLE OF CONTENTS

LIST OF TABLES ............................................. vii
LIST OF FIGURES ............................................ ix
CHAPTER]. INTRODUCTION ................................. 1
The Freestanding Ambulatory Care Center .................... 1
Theoretical Perspectives on FACC Location ................... 5
Research Objectives .................................... 18
CHAPTER II. RESEARCH HYPOI‘HESES ........................ 21
General Hypotheses .................................... 21
Specific Hypotheses ................................... 26
CHAPTER III. DATA AND METHODS ......................... 32
Data ............................................... 32
Methods ............................................ 34
Descriptive Analysis ................................ 35
Regional Scale ................................... 35
Metropolitan Scale ................................ 37
Statistical Analysis .................................. 38
The Survey ....................................... 43
CHAPTER IV. RESULTS ..................................... 46
Descriptive Analysis ................................... 46
Regional Scale .................................... 46
Metropolitan Scale .................................. 57
Statistical Analysis ..................................... 62
Central Cities ...................................... 64
Central Counties ................................... 69
Other Counties ..................................... 77
Survey of Michigan FACCs .............................. 84
Location .......................................... 85
Characteristics ..................................... 89
Relationships ...................................... 95

CHAPTER V. CONCLUSIONS ............................... 102

S ........................................... 102

Future Research ..................................... 108
LIST OF REFERENCES ...................................... 113
Appendix A. MSAs Used in Analysis ........................... 120
Appendix B. Reconstituted Regression Coefﬁcients ................. 123
Appendix C. Sample Survey of Michigan FACCs .................. 124
Appendix D. Summary of Survey Results ........................ 130

vi

LIST OF TABLES

Table 1 Number and Proportion of FACCs by Region 1985 and 1993 . . . . 49
Table 2 Leading States as Locations of FACCs 1985 and 1993 ......... 50
Table 3 Leading Metropolitan Areas as Locations of FACCs 1993 ...... 52
Table 4 Location Quotients for Census Regions 1985 and 1993 ......... 53
Table 5 Regional Population and Employment Growth Rates: 1980-1985

and 1985-1990. ...................................... 53
Table 6 Proportion of FACC Excess and Deficit States by Region ....... 58
Table 7 Location Quotients for Intrametropolitan Locations by Region . . . 61
Table 8 Number and Proportion of FACCs by Intrametropolitan

Location and Region .................................. 61
Table 9 Variables used in MSA Level Analysis .................... 63
Table 10 Spearman Correlation Coefficients for FACC / POP Ratio for

Central Cities (N = 106) ............................... 65
Table 11 Bivariate Regression Results for Central Cities .............. 66
Table 12 Reconstituted Regression Coefficients for Central Cities ....... 68

Table 13 Spearman Correlation Coefficients for FACC / POP Ratio for
Central Counties Without Central Cities (N = 102) ........... 70

Table 14 Reconstituted Regression Coefficients for Central Counties
Without Central Cities ................................ 72

vii

Table 15 Spearman Correlation Coefﬁcients for FACC / POP Ratio for

Central Counties With Central Cities (N =111) .............. 74
Table 16 Bivariate Regression Results for Central Counties With

Central Cities ...................................... 75
Table 17 Reconstituted Regression Coefficients for Central Counties With

Cities ............................................ 76
Table 18 Spearman Correlation Coefﬁcients for FACC / POP Ratio for

Other Counties (N = 75) .............................. 78

Table 19 Bivariate Regression Results for Other Counties ............. 80
Table 20 Reconstituted Regression Coefﬁcients for Other Counties ...... 81
Table 21 Importance Scores for Location Factors for Metropolitan Area

and Site Locations ................................... 88
Table 22 Employment in FACCs ................................ 92
Table 23 Proportion of Clinics Providing Certain Services. ............ 94
Table 24 Variables Included in Regression Analysis ................. 96

LIST OF FIGURES

Figure 1 Location of Freestanding Ambulatory Care Centers by State

1985 ............................................. 47
Figure 2 Location of Freestanding Ambulatory Care Centers by State

1993 ............................................. 48
Figure 3 FACC Excess and Deficit States 1985 ..................... 55
Figure 4 FACC Excess and Deficit States 1993 ..................... 56
Figure 5 Lorenz Curves for 1985 and 1993 ........................ 59

ix

CHAPTER I
INTRODUCTION

The Freestanding Ambulatory Care Center

Freestanding Ambulatory Care Centers (FACCs) are a relatively new
phenomenon of the health care landscape in the US. In many ways, however,
FACCs may be seen as part of an evolution of the health care industry since
they are by no means entirely separate from traditional care givers such as
physicians and hospitals. The concept of freestanding ambulatory care centers
actually began with the ﬁrst freestanding emergency center (FEC) which
opened in Newark, Delaware in 1973. Called the Newark Emergency Center,
Inc., this facility was developed to provide immediate medical care when
patients' private physicians or hospital emergency rooms were not readily
available or conveniently located (Hone, 1985). The focus of the FEC has been
on convenience, cost efficiency, and quality health care. The FEC was
originally marketed as an alternative to the long waiting lines and high prices
of hospital emergency rooms, although the concept has since evolved to
emphasize primary and minor urgent care rather than service for major

emergency cases (Rylko-Bauer, 1988).

2
Research concerning freestanding ambulatory care centers has been

hampered to some extent by the multiplicity of names given to this source of
ambulatory medical care, and the lack of a single accepted definition of what
these centers are and what services they provide. Since the development of
the first FEC in Delaware, these centers have been frequently divided into at
least two subcategories (Friend and Shiver, 1985): the term freestanding
emergency center (FEC) is used to refer to those centers open 24 hours a day,
seven days a week, and which are equipped to handle all types of emergency
cases and are usually integrated into the local EMS (emergency medical
service) system. The terms freestanding ambulatory care center (FACC),
urgent care center (UCC), convenience clinic or walk-in clinic are used to
describe those centers open between 8 and 14 hours a day, seven days a week
(weekend hours vary) and which provide immediate medical care on a no-
appointment necessary basis for minor injuries and illnesses. Most FACCs also
provide primary care, although some authors separate urgent care centers
(UCCs) which offer episodic treatment for injuries and illnesses from primary
care clinics which promote continuity of care (Rylko-Bauer, 1988). In addition,
although referred to as freestanding, not all FACCs are truly freestanding in
the sense that they may be part of a strip mall or even a hospital wing.
However, most of these centers are in separate buildings (N AFAC, 1990). For
the purposes of this dissertation, a FACC is defined as any health care facility
which has extended hours (usually evenings and / or weekends) and does not

3
require appointments. Therefore, I will use the term FACC to refer to all

urgent and primary care centers, as well as to all FECs.

The purpose of this study is to determine the locational patterns of the
FACC, the reasons for the observed patterns, and to understand the position of
FACCs within the health care delivery system which currently exists in the
US. FACCs were chosen for study as they are a member of the fastest
growing sector of the economy - the service sector, they are one of several
modern health care providers which are currently challenging the traditional
health care system (Ermann and Gabel, 1985; Relman, 1991), and because the
growth and presence of FACCs has certain implications for the delivery of
health care from both the patient's and the provider's view points.

Few studies have been done on specific FACC locations. Hoskins (1983)
notes that FACCs in Missouri tended to concentrate in the St. Louis SMSA
(none were in the city of St. Louis itself). In Illinois in 1983, 23 FACCs were
identified, mostly located in middle to upper income suburbs in the
northeastern part of the state (Illinois Medical journal, 1983). The freestanding
ambulatory care center has experienced the most growth in Sunbelt (southern
and western) states such as California, Texas, and Florida and in the suburbs
rather than in central cities (Ermann and Gabel, 1985; Powills, 1985; Mullinax,
1980). These locational trends are not surprising given the nation—wide
movement of population and industry toward the Sunbelt and suburbs (Kim,
1987; Wheeler and Muller, 1986). In 1985, California, Florida, and Texas had

the largest number of FACCs, although Ohio, Michigan, Illinois, and

4
Massachusetts, all "Snowbelt" states, also had high numbers of FACCs

(Powills, 1985). Since 1985, there has been FACC growth in states such as
Georgia, Arizona, Washington, and Oregon.

Some authors have identified factors important to FACC site selection,
although no actual research has yet been conducted to determine if FACCs
really locate according to these factors. The current research is intended in
part to fill this gap in our knowledge of the FACC and its location patterns.
The FACC, although a health care supplier, is often considered to be similar to
retail firms, particularly consumer oriented retail and service firms such as
grocery stores, shopping centers, and branch banks. This similarity to
retail / service stores is borne out in most discussions of FACC site location
strategies (Latour, et al., 1985; Hosking, et al., 1985; Williams and Stukenberg,
1985). Factors that most FACCs consider in making their locational choices are
similar to those considered by consumer oriented retail and service firms:
population size and density, population growth, age and gender distributions,
income, home ownership, education, traffic routes, competition, and site
characteristics such as the building size and style, zoning regulations, access,
and parking facilities. However, it must be noted once again that research to
date has not considered whether or not FACCs actually use these guidelines to
choose sites.

5
Theoretical Perspectives on FACC Location

Since the FACC is a part of the service sector as well as a health care

provider, the theory of FACC location is appropriately drawn from the factors
inﬂuencing the location of other services, retail functions, and medical
practitioners. Theories of service sector location can be considered at the
regional, metropolitan, and site levels. This study concentrates on the regional
and metropolitan scales, although the survey of FACCs will address some site
selection factors.

At the regional scale, the factor of importance is the presence of overall
economic and population growth. The growth of the South and West in terms
of population and employment has been a widely noted national trend, as
economic divergence among regions has occurred, particularly since the 19705.
As part of the process of economic restructuring, manufacturing firms have
relocated to the South and West in search of cheaper, nonunionized labor
(Noyelle, 1983). Studies show that corporations of both public and private
firms are increasingly headquartered in the South and West (Holloway and
Wheeler, 1991; Wheeler, 1990; Kim, 1987). The South and West have also been
increasing their share of health care professionals (DeVise, 1973; Pyle, 1989)
and other for-proﬁt health care providers, such as the freestanding ambulatory
surgery center (FASC) have shown Sunbelt preferences (Lowell-Smith, 1993a).
These trends give reason to expect FACCs to be more concentrated in the
Sunbelt than other areas of the country, particularly the Northeast.

6
Metropolitan areas have dominated nonmetropolitan areas in terms of

economic and population growth, and suburbs have prospered in relation to
central cities. Metropolitan locations have long attracted health care
professionals, particularly physicians. Studies show that physicians
concentrate in urban rather than rural areas for a variety of reasons including
population size, higher income potential, the proximity of hospitals and
medical schools, the presence of other medical professionals, and the cultural
and service amenities of urban areas (Gordon et al., 1992; Knaap and
Blohowiak, 1989; Rosenberg, 1984). Specialists (e.g., surgeons) are even more
likely than general practitioners to locate in urban areas because specialists
have greater population thresholds and closer ties to hospitals and to each
other (Ernst and Yett, 1985; Gober and Gordon, 1980). Therefore, despite
obvious shortages of care experienced in many nomnetropolitan areas (Koska,
1991; Joseph and Phillips, 1984), it is likely that a majority of physicians and
others have opened their FACCs in metropolitan sites.

Within metropolitan areas, the suburbs have been the dominant location
for population and businesses since the 19505. There are several reasons for
the movement of the population to the suburbs, among them the development
of the highway system, which made it possible to move quickly and easily
from residences in the suburbs to central city jobs (and later to jobs in other
suburbs) (Muller, 1986). Other factors which inﬂuenced the suburbanization
trend included the growth in automobile ownership, the desire for home

ownership and the amenities of the suburbs, and the perception of an

7
undesirable quality of life in the central city. Businesses and retail centers

followed the residential population to the suburbs as the CBD lost its centrality
due to the accessibility to other locations provided by the highway (Muller,
1986). Eventually, suburbs in the US. accounted for the majority of jobs and
retail sales while the CBD lost jobs, money, and people (Kellerman, 1985;
Hartshom and Muller, 1989).

Not only have population and consumer-oriented industries moved to the
suburbs, but so have manufacturing and office based firms (Kim, 1987;
Noyelle, 1983). Cheaper land and labor, space available for expansion, and the
improved environment of the suburbs have all contributed to this move
(Daniels, 1982). The health care field has not been immune to the
suburbanization trend as physicians and dentists have been migrating out of
traditional central city office locations to suburban sites (Lowell-Smith, 1993b;
Mattingly, 1991; Pyle, 1989; Dorsey, 1969). The trend for hospitals is less clear,
since newly built for-profit hospitals may locate in the suburbs (Gray, 1986),
but most hospitals remain in central cities. However, it is no surprise that
FACCs are expected to be found more frequently in the suburbs than in
central cities.

Also at the metropolitan scale, several factors have been found to be
important in determining the location of services, including health care
providers. These factors include the location and characteristics of the

population (demand), agglomeration economies (urbanization, localization,

8
economies of scale), competition, and the personal desires of the location

decision makers (which can also inﬂuence regional scale locations).

Most services, particularly retail services, target specific populations. The
FACC is no exception as it searches for metropolitan populations with a high
proportion of young people, people in the middle-upper income groups (i.e.,
have the ability to pay for services and / or are insured), and people who are
educated and relatively mobile (Miller, 1985; Williams and Stukenberg, 1985;
Yunker, et al., 1985).

Agglomeration economies are also identifiable at the metropolitan scale.
Reasons for firm agglomeration have been discussed by several researchers
(O'hUallachain and Reid, 1991 ; Brown, 1989; O'hUallachain, 1989; Daniels,
1985). These reasons include urbanization economies (e.g., the availability of
public utilities and transport networks), localization economies (beneﬁts of
locating near other firms e. g., information spillovers, shared labor pool, the
reduction of uncertainty) and to take advantage of consumer multipurpose
shopping patterns. O'hUallachain (1989) showed that for hospitals and other
health services, the level of urbanization was more important for growth than
localization economies. However, localization economies do affect the location
of health care providers as physicians, like other services, have been shown to
prefer locating near other each other and medical professionals as well as near
hospitals to facilitate consultations and to allow patients to easily access

different specialists (Knaap and Blohowiak, 1989; Rosenberg, 1984).

9
In the case of the FACC, it is expected that urbanization economies will of

course be important in the location of these facilities. Localization economies,
however, are unlikely to be important as FACCs are not expected to cluster
together to share patients or other services since they are in direct competition
with each other. Yet, localization economies in the form of locating near
physicians' ofﬁces and hospitals may occur.

Economies of scale, often considered part of the agglomeration economies,
are defined as benefits accruing to a ﬁrm because of its size and they may be
either internal or external. Internal economies of scale have to do with the size
of the ﬁrm or facility (physical size or number of employees, staff). External
economies of scale usually refer to the degree of centralization (number of
independent owners) within the industry. In retailing, for example, the trend
has been toward fewer, larger stores (Price and Blair, 1989). Hospitals and
individual medical practitioners have also felt the pull of scale economies. In
recent years, studies have shown that small hospitals, often in rural or low
socioeconomic status areas, have closed while larger, more specialized
hospitals have grown (McLafferty, 1986; Joseph and Phillips, 1984). Physicians
and dentists, likewise, have found it increasingly preferable to locate in groups
to share ofﬁce costs, staff, and equipment and to "ﬁll in" for each other on sick
leave or vacation (Lowell-Smith, 1993b; Shumsky, et al., 1986). FACCs are
expected to beneﬁt from economies of scale in the sense that chain clinics
should be able to expand their market areas faster and farther than

independent clinics. It would also be reasonable to expect larger clinics to be

10
more proﬁtable than smaller clinics, at least, up to a certain size where

diseconomies of scale may occur.

The tendency for agglomeration will of course be tempered somewhat by
the type and intensity of competition encountered by the service or retail ﬁrm
(in this case, by the FACC). The location of service ﬁrms is at least partially
the result of the interaction between the need for agglomeration (a centripetal
force) and the desire to be relatively distant from the competition (a centrifugal
force). FACCs are certainly aware of the effects of competition as it is
suggested that most walk-in clinics locate at least 3 miles from hospitals and
four miles from each other (Parsons, 1987).

One location factor which has been given more attention in the industrial
location literature than in the location of tertiary activities is the inﬂuence of
personal factors on the location decision (Chapman and Walker, 1991). This is
particularly true at certain scales, such as the state and metropolitan levels. In
other words, the choice of a state or city for ﬁrm location may be based on
nothing more than personal ties to the area or a preference for a certain
climate. This has been shown to be the case with physicians who often locate
in their home states, cities of similar size and character to their home cities, or
in states/ cities where they went to medical school (Ernst and Yett, 1985;
Joseph and Phillips, 1984; Watson, 1980).

Although the factors inﬂuencing the location of services and traditional
health care providers are relevant to FACC location, the particular nature of

the FACC (i.e., a normally for-proﬁt health care delivery system emphasizing

11

convenience and cost-reduction) means that there are other important factors
which have inﬂuenced the development and location of the centers. Rylko-
Bauer (1988) discusses general reasons behind the success of FACCs from both
patients' and providers' perspectives. For example, there have been changes in
the physician-patient relationship as society has become more mobile and
people have become more knowledgeable about their own health. Mobility
reduces the likelihood of a person forming strong ties with any particular
physician, while the increase in personal health knowledge means that people
are less likely to rely solely on physicians for their medical and health care
needs. The FACC which offers episodic care (as well as continuous care for
those who desire it) beneﬁts from trends which reduce patients' relationships
with ofﬁce-based private physicians. In addition, Rylko-Bauer (1988) notes the
growth of convenience as a "cultural value." FACCs offer immediate, no-
appointrnent necessary medical care at times when physicians are unavailable
and hospitals inconvenient. Surveys of FACC patients repeatedly cite
"convenient location" and "convenient hours" as reasons for the patients' use of
the centers (Rizos, et al., 1990; Miller, et al., 1985; Yunker et al., 1985).

One signiﬁcant trend in the US. health care delivery system which has
aided the growth of the FACC has been the rise of "commercial" or "for-proﬁt"
health care suppliers with their emphasis on reducing the soaring costs of
medical care while making the delivery of medical care a proﬁtable business
(Berliner and Burlage, 1987). The movement toward franchising in urgent care

has earned urgent care centers nicknames like "McDoctor," "Doc-in-a-Box," and

12
"Seven-11 medicine." For physicians faced with a doctor oversupply in many
areas and the high .coss of opening their own practice, the chance to begin
their careers on salary at an FACC is attractive (Berliner and Burlage, 1987;
Friedman, 1983; Weiss, 1982).

The ownership patterns of FACCs are also important to their location
strategies, as well as to their operational and service characteristics. In 1983, a
survey sponsored by the ambulatory care trade organization (then called the
National Association for Freestanding Emergency Care (N AFEC), now called
NAFAC), showed that 73% of FACCs were owned by physicians. A 1986
survey by the ACEP showed that the number of FACCs owned by a single
physician had declined while chain membership rose, as did hospital
participation in the FACC market (Rylko-Bauer, 1988). Given the trend toward
corporatization in health care, the rise of chain and corporate-owned FACCs
and FECs was to be expected. However, a recent survey produced by the
trade organization, N AFAC, showed that non-physician corporate ownership
of FACCs had actually declined between 1984 and 1990, while the number of
physician and hospital owned FACCs had again risen (N AFAC, 1990). Direct
ownership of FACCs is only one of many possible relationships hospitals and
physicians may have with the centers. For example, hospital involvement in
FACCs may be very high (ownership or management) or very low (hospital
owns facility but does not manage or ﬁnance services) (Burns and Ferber,
1985). However, for the purposes of this dissertation, only ownership and is

consequent location strategies are considered in detail.

13

Although FACCs did not originate directly from a hospital or physician
practice, they can nevertheless be seen as an evolution of these traditional
providers. This evolution, spawned by factors such as patient desires, the
need for cost control, and technological and organizational changes within the
health care industry iself, has made it necessary for traditional providers to
either become owners and operators of FACCs or to risk losing patiens and
income to the newcomers. The rise in hospital and physician ownership of
FACCs obviously indicates their decision.

Hospital-owned FACCs often have a dual purpose. They may be used to
alleviate emergency room crowding and offer a social service by providing
emergency room alternatives to areas distant from the hospital iself. Hospital-
owned FACCs may also be used to generate income from paying ambulatory
patiens and to act as feeders (through referrals) for inpatiens (Friend and
Shiver, 1985; Rylko-Bauer, 1988). Physicians, both as individuals and as
groups, often form their own FACCs merely by extending their ofﬁce hours
(nighs and weekends) and by accepting walk-in patiens, although they can of
course set up an F ACC while keeping private practices separate. Physicians
who are extending private practices into FACCs, often do so to be more
competitive by fulﬁlling perceived patient desires for convenience and greater
service options (Born, 1986; Stitham, 1986). Physicians and physician groups
who open an FACC which is more than an extension of their private practices,
do. so because FACCs are seen as good business and employment
opportunities in the face of sky-rocketing malpractice and other operating coss

14
(Friend and Shiver, 1985). FACCs allow physicians to share these coss,

particularly physician-employees who are often paid on salary and are covered
by the parent organization. In addition, it must be remembered that there are

(non-physician) corporate-owned FACCs which are presumably run much like
other consumer oriented service ﬁrms and where physicians are employees.

The ownership patterns of the FACCs are expected to impact their location
strategies. Hospital-owned FACCs, because of the possible duality of their
mission (social service versus proﬁt generator) will likely exhibit two location
patterns. Those clinics located in the central city are most likely to be
performing the "social service" function (i.e., providing emergency room
alternatives). Those hospital-owned PACCs located in the suburbs are likely to
be the "proﬁt-makers," competing with private physicians for the middle-upper
income groups and the insured and providing patiens for the hospital through
referrals. Physician—owned FACCs, likewise should also be found more often
in the suburbs since 1) they are often formed out of existing physicians' ofﬁces
and 2) they are in the suburbs also for proﬁt. Also likewise, corporate-owned
FACCs, looking for maximum proﬁt are likely to choose suburban locations
over central city or nonmetropolitan areas.

Alternative systems of health care delivery, such as the FACC, have several
implications for both patiens and traditional health care providers like
physicians and hospitals. For patiens, the FACC has been beneﬁcial in the
sense that it offers spatially and temporally convenient health care, often at

lower coss (in particular, coss lower than those of hospital emergency rooms).

15
However, the fact that FACCs are most often run for proﬁt (even non-proﬁt

institutions may use FACCs as for-proﬁt income generators) has raised
questions about the actual accessibility of the clinics and about possible long-
term effecs on the availability of health care for the poor and uninsured. For
example, FACCs (and other for-proﬁt providers) have often been accused of
"cream skimming" the best patiens (the wealthy and insured) and leaving
public facilities to support the indigent and uninsured without the income
generated by the wealthy and insured patiens (Berliner and Burlage, 1987).
However, although it is difﬁcult to understand or. measure the impacs of one
health care institution on another, there is little evidence that "cream
skimming" has occurred or that the presence of FACCs forces hospitals to close
their emergency departmens (Rylko-Bauer 1988; Gray 1986). Finally, patiens
who do use FACCs must face the issue of continuity of care. Although many
walk-in clinics would like to promote continuous care, the fact that patiens are
usually seen by one of several part-time physicians at the clinic would make
continuous care difﬁcult. In addition, the no-appointment necessary policy of
FACCs is rather antithetical to the notion of continuous care since patiens are
presumably more concerned with convenience than with which physician they
visit. Questions then arise concerning the nature of health care that we prefer.
Will the FACC form of health care which is more episodic than preventative
prevail?

From the point of view of (traditional) providers such as physicians and

hospitals, FACCs are seen as both opportunities for business and employment

16
and as competition. As discussed above, FACCs are often extensions of a

physician's (or group of physicians') private practice(s); FACCs may also be
used by hospitals as patient feeders and as alternatives to overcrowded
emergency rooms. The involvement of hospitals in the FACC movement has
in fact lessened fears of hospitals closing their emergency rooms or removing
charity services. However, FACCs, especially when owned by competitors
(i.e., other hospitals, physicians, or corporations), may force traditional
providers into competitive stances. For example, physicians competing with
convenience clinics have had to begin offering Saturday and evening hours,
and in some cases lower coss (Born, 1986; Stitham, 1986).

Competition from FACCs in the wealthier suburban areas may even be
used to push traditional providers into unserved areas such as the inner cities
or nonmetropolitan areas, similarly to the way that competition among
physicians in metropolitan areas has caused some movement of physicians to
small towns and rural areas (Phelps, 1992). However, given the needs and
desires of these professionals (i.e., proximity to hospitals and each other, urban
amenities, proximity to certain population groups), this movement seems
unlikely.

For physicians, FACCs (as well as other providers which employ
physicians) hold further implications besides competition for patiens.
Physicians are often employees in FACCs, and as such they face a possible loss
of autonomy or status (Lowell-Smith, 1994). Several authors discuss the

physician's loss of autonomy as the physician becomes "proletarianized" by

17

working as employees rather than as private practitioners (McKinlay and
Stoeckle 1988; McKinlay and Arches, 1986). They deﬁne "proletarianization" as
the process by which,

"an occupational category is divested of control over certain

prerogatives relating to the location, content, and essentiality

of is task activities, thereby subordinating it to the broader

requiremens of production under advanced capitalism. "
For physicians, the movement towards "proletarianization" has occurred for
several reasons. The authors cite increased specialization (more likely to be
practiced in group and corporate settings), increased knowledge on the part of
consumers (i.e., less unquestioning trust in physicians), increased technology
and the demand by consumers for that technology (increased coss beyond the
scope of the individual practitioner), and increased malpractice premiums
(again costly to individual practitioners) as mechanism which have propelled
physicians in the direction of salaried employment. In addition, the
bureaucracy of third party payers is thought by some physicians to be the
third party's "secret weapon" (Grumet, 1989). That is, slow and inefﬁcient
insurance companies and government agencies delay reimbursement of
physicians for their own proﬁt, once again making salaried status for the
physician attractive (also, Yunker, et al., 1986). Forphysicians, alternative
health care delivery systems such as FACCs has mixed blessings. For some,
ﬁxed salary and employment will be a welcome relief of medical school debs

(since they would be getting immediate income) and a way to avoid the coss

of setting up a new practice. However, in exchange for regular hours and

18

employee beneﬁs, physicians will lose some of their autonomy and status as
professionals. They may face reduced income, decreased authority, and less
choice of location and service activities.

While HMOs and other providers of health care will exercise more control
over physicians than in the past, it is uncertain whether these new provider
patterns will produce "proletarianization" since physicians will continue to
maintain their knowledge and expertise and to have legal righs and
responsibilities protected by state laws (O'Connor and Lanning, 1992; Derber,
1984). Therefore, it is unlikely that physicians will ever be "proletarianized" to
the same extent that, say, crafsmen were in the early part of this century.
However, as long as other parties (whether they be corporations, insurance
companies, or the US. government) exert control over the health-related
decisions of physicians (e. g., the length of hospital stay, or the suitability of a
lab test), there will always be some questions concerning the quality and

quantity of care patiens receive.

Research Objectives

Most previous studies of FACCs have been descriptive and have placed
emphasis on the growth and character of the FACC, as well as on patient-
oriented issues such as accessibility and utilization. Few studies have
considered the location of FACCs, and those that have were either limited in
scale (e. g., one city), merely involved state totals, or have been based on

anecdotal evidence. The current research hopes to ﬁll the gap in our

19
knowledge of the geographical patterns of FACCs and the reasons behind the

observed patterns. The location of FACCs is analyzed at several scales, the
regional, inter and intrametropolitan, and the "local" or site level.

In many ways, FACCs may be seen as particularly rebellious children of
traditional parens (i.e. physicians and hospitals). FACCs challenge and
compete with their "parens," while at the same time are not completely
independent of them. Even FACCs controlled by "ousiders" (e.g.,
corporations) can not be operated without input from physicians and hospitals.
It is this relationship between FACCs and traditional providers, this evolution
of health care in response to modern needs and desires, that at least in part
makes them such an interesting study. In addition, in these times of health
care reform, FACCs seem an appropriate setting for implementing the part of
the plan concerned with offering primary (GP) level care at convenient
(neighborhood level) locations.

The current research has two primary goals: 1) to determine the geographic
patterns of FACCs at a variety of scales and the reasons behind these
loCational patterns and 2) to gain a better understanding of the relationship
between FACCs and other health care providers and other economic sectors.
At the regional and metropolitan scales, available secondary data is used to
obtain the necessary information. A survey of Michigan FACCs is also
conducted to obtain more detailed analyses involving speciﬁc FACCs.

The dissertation is organized into four remaining sections. The next

section (Chapter II) discusses the general and speciﬁc hypotheses relating to

20
FACC location patterns at various scales. Data and methods are discussed in

Chapter III. Chapter IV presents the resuls of the analyses and Chapter V

contains a summary of the resuls and directions for future research.

CHAPTER II
RESEARCH HYPOTHESES

General Hypotheses

There are four general factors which are assumed to underlie the location
decision of FACCs at all scales of analysis: the demand for FACC services, the
economic structure (as it relates to economic growth and agglomeration
economies), competition, and FACC ownership patterns. The last two factors,
competition and FACC ownership, may actually be considered together as
they interact to produce FACC location patterns, particularly at the
metropolitan and site levels.

Each of these four factors can be broken down into more speciﬁc pars.
Demand for FACC services, for example, is basically concerned with
population size, growth, and population characteristics such as demographics,
income, mobility, and socioeconomic status. Although FACC research has not
considered the speciﬁc effecs of these factors on FACC location, population
and is characteristics have been studied in relationship to the location patterns
of other retail and service functions. The positive relationship between income
and demand for goods/ services is a well-established economic principle.

Studies have found that the location of such activities as shopping centers,

21

22
grocery stores, and branch banks is related to the presence of middle-upper

income groups, higher socioeconomic status groups, and low unemployment
rates (Morrill, 1987; Hall, 1983; Olsen and Lord, 1979). In addition, the location
of health care providers, in particular, physicians, has been correlated with
various population attributes such as population size, age distributions, race,
and income. Physicians are often attracted to areas with large populations,
high median incomes, high education levels, older populations (GPs) and
younger populations (OB / GYN s, interniss) (Knaap and Blohowiak, 1989;
Guzick and Jahiel, 1976; Kaplan and Leinhardt, 1973; Elesh and Schollaert,
1972).

In addition to considering the location decisions of retail/ service ﬁrms,
much research has concentrated on consumer shopping patterns and how.
these can be used by businesses to determine the location of their stores.
Although the literature on consumer shopping preferences is beyond the scope
of this research, it is appropriate to point out that many studies have found
relationships between population characteristics and the choice of
stores/ services. Businesses can then use this information to determine who
will likely patronize their stores/ services. For health services, it has been
found that variables such as age, income, education, family status, and race
inﬂuence how often people visit physicians and hospitals and what they desire
from these providers when they do visit them. The elderly, women, families
with children, the educated, and the middle-upper income groups all use
physician services frequently (Phelps, 1992). Lower income groups tend to use

23
hospitals for both emergency and primary medical care, and to use both

physician and dental services less than higher income groups due to inability
to pay fees, the lack of insurance coverage, and (often) the spatial
unavailability of physician/ dental services (Bohland, 1984; O'Mullane and
Robinson, 1977; Morrill, et al., 1970). Thus for FACCs, attractive populations
will be those with higher incomes and socioeconomic status.

Population characteristics also inﬂuence the expectations of
consumers/patiens when they choose a store or service. For example,
researchers have found that the choice of a pharmacy depends on age as the
elderly are less concerned with distance or convenience than with service and
quality (Shannon, et al., 1985; Nickel and Wertheimer, 1979). Likewise, in
considering the choice of a physician, the young and middle aged, single-
parent families, and the middle-upper income groups are often concerned with
physician promptness, convenience, waiting times, and facility distance from
home or work (Gesler and Meade, 1988; Nordstrom and Steinke, 1987). These
resuls suggest that FACCs, with their emphasis on providing convenient
health care, will attract young to middle aged patiens more so than the
elderly.

Based on these resuls concerning the use of health care services by various
population groups, it is hypothesized that the location of FACCs will be
related to the following measures of demand: population size, growth, density,
age distribution, income, the unemployment rate, family status, and mobility.

All of these factors have been found to be important in the location decisions

24
of retail and service ﬁrms, including health care providers. In addition,

studies of FACC patiens indicate that they have certain characteristics. For
example, the typical FACC patient tends to be young (one study found that
85% of patiens were less than 50 years old), educated, in the middle to upper
income groups, and mobile (Miller, et al., 1985; Rylko-Bauer, 1985; Yunker, et
al., 1985). Therefore, FACCs should locate in order to be close to these
populations.

The second general factor considered to be important in FACC location is
the economic structure of the area, particularly as it relates to economic
(employment) growth and agglomeration economies (advantages of locating
near other businesses). The main issue of concern here is which areas are
experiencing economic growth. In the US. today, economic growth is
associated with the service sector, especially business and professional services.
Rather than manufacturing, business and professional services are currently
being touted as producers of economic growth (Goe, 1990; Hansen, 1990;
Harrington and Lombard, 1989). In addition to providing employment and
income for an area, services are also relevant because they are important
linkages for other businesses, both manufacturing and other service ﬁrms.
FACCs are businesses and as such need access to business services such as
accountans, lawyers, and consultans. Finally, the other necessary linkage for
FACCs, that connection to other health care providers, is likely to be entranced
as physicians, in particular, congregate in areas of economic and population

growth.

25
The third and fourth factors related to FACC location, competition and

ownership patterns are themselves interrelated, since the importance of
competition will be affected by who (or what institution) owns the FACCs. As
previously discussed, ownership patterns inﬂuence FACC location by
inﬂuencing market strategy. Hospital-owned FACCs may be fulﬁlling one of
two purposes: those located in the central city are likely to exist for "social
service" purposes, that is, to be branches of the hospital where care is provided
for those who cannot easily reach the hospital. Those hospital-owned FACCs
located in suburbs are likely to be there for proﬁt - to take advantage of
paying, insured patiens, to compete with physicians and to provide referral
patiens for the hospital. Physician and corporate owned FACCs are also
likely to be located in the suburbs since most physician ofﬁce are already there
and the opportunity for success is presumably greater than in central cities.
Competition among FACCs is likely to be most intense in the suburbs. It is
expected that when choosing a particular site, FACCs put a certain distance
between themselves and each other, and other providers. Only at the site
level can competition patterns be speciﬁcally determined. At other scales, the
measures of "competition," physicians and hospitals, may also be considered as
measuring to some extent F ACC ownership patterns.

Although the ﬁrst factor, demand, receives the most attention in this
discussion, the actual importance of the four factors may in fact vary with the
scale of analysis. For example, FACCs located in central cities may consider

population characteristics (the type of population they wish to serve) more

26
important than "beating" the competition (e.g., physicians or even other
FACCs). FACCs locating in suburban counties may consider the economic
structure (growth, employment, etc.) of the area of top importance since it is
an indicator of their potential success and the availability of the necessary
service linkages. The separation, where possible, between central cities and
suburban areas in the analysis is used to further understand these
relationships and the varying importance of the four factors thought to be

related to FACC location.

Speciﬁc Hypotheses

The relationship between FACC location and the four factors (demand,
economic structure, competition, and ownership) may be tested at a variety of
scales. The importance of the four factors will vary with scale, and not all
factors will be tested at each scale of analysis. The ﬁrst analyses, at the
regional, state, and metropolitan scale, will be primarily descriptive. Three
hypotheses concern the location of FACCs at the these scales. First, FACCs
will be found more in the South and West than in either the North Central or
Northeast regions. The next two hypotheses are that FACCs will be found
more often in metropolitan than nonmetropolitan areas and within
metropolitan areas, suburban locations will dominate over central city
locations. The reasons for these location patterns relate to two of the four
factors expected to inﬂuence FACC location: the location of population

(demand) and economic growth. The fourth factor (ownership patterns) can

27
also be tested to some extent at this scale since a dominance of suburban

locations also implies a dominance of particular ownership patterns (e.g.,
physicians, corporations). A dominance of central city locations would imply
that there are more hospital owned FACCs than other types. The resuls of
this scale of the analysis will then be used to propose further analyses at
smaller scales.

Statistical analyses will be performed at the metropolitan (central city and
county) scale. At the metropolitan level, several relationships can be
postulated regarding the determinans of FACC location. For this analysis,
metropolitan areas are divided into central cities and suburban counties
(central counties and other metropolitan counties). Population and is
characteristics comprise the ﬁrst factor related to FACC location. These
characteristics can be divided into two broad categories: general demand,
measured by population size, density, growth, and the age distribution, and
socioeconomic status measured by other characteristics such as income, long-
term residens, education, and family status. Positive relationships are
expected between FACC location and population size, growth rate, the number
of people between the ages of 20 and 45, and the number of children between
the ages of 0 and 5. A negative relationship is expected between FACC
location and the number of elderly. Socioeconomic characteristics of the
population will be measured by income, the unemployment rate, the number
of female-headed families, education levels, and the number of long-term

residens in the area. Positive relationships are expected with income,

28
education, and the number of long-term residens in the area while negative

relationships are expected between FACC location, and the unemployment
rate, and the number of female-headed families (usually lower socioeconomic
status - see Jones and Kodras, 1990). Although these relationships are
expected to remain constant for all pars of the metropolitan area (cities and
suburbs), differences are expected regarding the importance of the variables
(and consequently the general factors) related to FACC location. In central
cities, population variables should be more important in determining FACC
location than in suburban counties. This is because FACCs in central cities are
likely to be there to serve demand rather than as proﬁt makers which would
necessarily increase the importance of such factors as the economic structure
(see below).

At the metrOpolitan scale, speciﬁc hypotheses can be tested concerning the
second general factor to inﬂuence FACC location: economic structure
(economic growth and agglomeration economies). Positive relationships are
expected between FACC location and employment patterns, particularly
employment in the service sector, business services, health services and retail
services. This is not to say that it is important to have residents employed in
these sectors (since they may actually work ouside the study area), but rather
that the employment offered in the area where the FACCs are located is in
these service sectors. It is a measure of business structure, not residential
employment. In metropolitan areas experiencing the decline of manufacturing

activities, there should be a negative relationship between FACC location and

29

manufacturing employment. However, since manufacturing is not
synonymous with economic decline nationwide, the relationship with FACC
location is not easily predictable. This factor (economic structure) is expected
to be somewhat more important for FACCs locating in central and other
counties (where FACCs are looking for proﬁs) than for those FACCs in the
central cities. Unfortunately, the nature of the available data will not allow for
direct comparison of economic structure variables between central cities and
counties, however, the relative importance of economic versus population
variables for the counties can be determined.

In addition to population attributes, and the economic structure of the area,
competition and FACC ownership patterns will interact to inﬂuence FACC
location. A positive correlation is expected between FACC and hospital
location in central cities since it is expected that most central city FACCs will
be "branches" of the hospitals designed to alleviate emergency room crowding
and provide care for central city patiens who cannot easily reach the hospitals.
In suburbs, positive relationships are expected between FACC location and
both physicians and hospitals. These positive statistical correlations should
exist since 1) both hospitals and physicians use suburban FACCs for proﬁt-
making, 2) there should be other ties between these health care providers
through referrals and employment, and 3) many of the FACCs are owned by
these physicians and hospitals. Although the data on physicians and hospital

beds do measure the presence of competitors, the relationships with FACCs at

30
the metropolitan scale are not expected to be negative. Only at the site level

will FACCs really separate themselves from the competition.

Of the four general factors which effect the location of FACCs, competition
and ownership patterns are two which cannot be completely tested based on
available secondary data at the regional or metropolitan levels. A survey of
FACCs in Michigan was designed to obtain speciﬁc information concerning the
ownership and competition patterns of the clinics, as well as on other factors
of interest such as linkages between FACCs and other businesses.

The survey allowed for the testing of "subsidiary" hypotheses regarding
FACCs such as differences between FACCs in terms of services, age, and
economies of scale. For example, differences are expected between city and
suburban FACCs in terms of the services offered. Central city FACCs
(expected to be predominantly hospital-owned) would be more likely to
concentrate on services such as drug and alcohol abuse treatment, psychiatric
care, and, of course, urgent care. For FACCs located in suburbs, the service
emphasis is expected to be on emergency (urgent) care and on "specialized"
services such as gynecology/ obstetrics and pediatrics. In the suburbs, FACCs
may take the place of hospitals which are spatially more distant than in central
cities. Differences may also occur in the services offered by various ownership
types (hospitals, physicians, etc.). Hospital-owned FACCs may involve
themselves in services perceived to be of need to central city (and suburban)
populations. Physician or corporate-owned FACCs may include services

perceived to be proﬁtable (but not necessary, such as diet control or physical

31
therapy), although they will also offer needed services such as urgent care. It

may also be hypothesized that there will be differences in FACC age by
ownership type. It is likely that hospital-owned FACCs are newer than other
FACCs since hospitals have been opening FACCs in response to competition
from physician and corporate-owned FACCs (NAFAC, 1990). Still another
hypothesis which could be tested using the data obtained from the survey is
the idea that larger clinics are more proﬁtable (based on economies of scale).
In general, the survey will be particularly useful in fulﬁlling one of the major
objectives of this dissertation: understanding the place of the FACC in the

American health care system.

CHAPTER III
DATA AND METHODS

Data

In order to test the hypotheses discussed above, certain types of data are
needed. Data on the location of FACCs are, of course, required. The location
of FACCs is the dependent variable which the analyses are attempting to
explain. The 1993 data on FACCs were obtained from the National
Association for Ambulatory Care (N AFAC). NAFAC is a trade organization
formed in 1981 and represens all ambulatory care centers, their owners,
managers, medical directors, and all other related staff. All ambulatory care
centers in the US, as of January 1, 1993, are listed in the data set. The list
includes the name, address, and member name for each freestanding
ambulatory care center. The list contained no other information on the centers.
Since more than one person at the same FACC could be a member of NAFAC,
the list had to be edited and all duplicates removed. In addition, the list
included a few ambulatory surgery centers, chiropractors, and other such
facilities which are not a part of the present study. After editing the data, the
list contained just over 3800 FACCs which were eligible to be included in this

study. It must also be noted that due to the nature of the list, it is impossible

32

33
to tell whether or not some of the centers are truly FACCs. For example, some

centers included on the list may require appointmens or be run by public
health departmens, in which case they would not be considered FACCs.
Unfortunately, the list is the only such data set available and must be used
despite is shortcomings. In addition to the 1993 data, the number of FACCs
in each state for 1985 was also available through another publication (Powills
1985). In this article, Powills has obtained data on the number of FACCs per
state in 1985 from information collected by the American College of
Emergency Physicians. The 1985 data were used to analyze changes in the
location and concentration of FACCs at the regional and state levels which
may have occurred between 1985 and 1993.

The four general factors expected to be related to F ACC location mean
that data should be collected on demand factors (population and is
characteristics), economic structure (employment), competition, and ownership
patterns. For the ﬁrst factor, demand, population data are readily available.
All data on population and is characteristics used in the statistical analyses
were obtained from the 1990 US. Census. These data include population size,
density, growth rates, age breakdowns, education, female-headed families,
long-term residens, income, and unemployment. Data on population and
employment at the regional scale were obtained from the Statistical Abstract of
the United States (US. Department of Commerce, 1992).

For the second factor, economic structure, there were fewer data options,

particularly for cities. Data on employment patterns (for counties only) were

34
obtained from the 1990 County Business Patterns (US. Department of

Commerce, 1990). The "demand" factors of income and unemployment may

‘ also be used to represent the economic structure of an area. The third factor,
competition, was measured by the number of physicians and hospital beds in
an area. Physician data were obtained from the American Medical Association
(1992) and data on hospital beds were from the American Hospital Association
(1991). The ideal situation for the fourth factor, ownership patterns of FACCs,
would be to know who (or what institution) owns each FACC on the NAFAC
list. However, this information is not available through secondary data
sources. To some extent, the data on physicians and hospital beds can be
substituted for actual ownership data since physician-owned FACCs are
expected to occur where there are more physicians (e.g., suburbs), while
hospital-owned FACCs would often be located where there is a greater
presence of hospitals (e.g., central cities). In addition, the survey of Michigan
FACCs was useful in obtaining speciﬁc information on competition and
ownership patterns at the site level. All analyses (except the survey) included

all 50 states and the District of Columbia.

Methods

The study of the four factors and their relationship to FACC location
requires that the analyses be conducted at several scales. These scales are the
regional (including state) level, the metropolitan level and the site level. The
relationships between FACC location and the four factors can not be studied

35
equally well at all scales due to the nature of the relationships. For example,

competition is not expected to be relevant to the regional or state level location
of the FACC. The analyses at each scale are designed to understand FACC
location patterns based on factors thought to be relevant at that scale. For
regions, states, and all US. metropolitan areas, the analysis is descriptive and
only two factors, demand and economic structure, are examined relative to
FACC location. Statistical analyses are used to examine FACC location
patterns in central cities, central counties, and other metropolitan counties. At
these scales, all four factors, demand, economic structure, competition, and
ownership patterns become relevant to FACC location. Finally, a survey of
Michigan FACCs looks at FACC location at the site level. At this scale,

competition and ownership are examined more closely.

Descriptive Analysis
Regional Scale

Analyses for the regional, state, and metropolitan areas (for the entire
US.) are primarily descriptive and are used to understand general locational
trends of the FACC. The regions of the US. are determined by the US.
Census (Northeast, North Central, South, and West). At this scale, it is useful
to describe the location and distribution patterns of FACCs and to determine
any changes in these patterns between the two periods for which data is
available (1985 and 1993). For this purpose, FACC data were tabulated for

regions, states, and metropolitan areas. In addition, location quotiens were

36
calculated for the four regions, each of the 50 states and the District of

Columbia, and for metropolitan versus nonmetropolitan areas. The location
quotient is used to indicate the relative degree of locational concentration of
FACCs in a region, state, or metropolitan/nonmetropolitan areas.
The location quotient was calculated as follows:

LQ' = (FACC.‘/Popi')/(X FACCHXPopi‘)
where LQ.t = location quotient for area i for time period t; FACC,t = number of
FACCs in area i for time period t; and Pop,t = population in area i for time
period t. For this study, the location quotient compares the regional, state, or
metropolitan proportion of FACCs divided by the regional, state, or
metropolitan proportion of the population to the TOTAL number of FACCs in
the US. (or in the metropolitan U.S.) divided by the TOTAL US. population
(or metropolitan US. population). A location quotient greater than one
indicates that a region, state or metropolitan area has more than is expected
share of FACCs based on is share of the population while a location quotient
less than one means that the area has less than is expected share of FACCs
compared to the area's share of the population.

Along with the location quotient analysis, changes in the regional level of
concentration of FACCs were examined by using the Lorenz curve and gini
coefﬁciens. The Lorenz curve is used to plot the relationship between the
cumulative percentages of population and FACCs across the four Census
regions. The distance between the diagonal (which represens an equal
distribution of the cumulative proportions of FACCs and population) and the

37
Lorenz curve indicates the degree of spatial concentration of FACCs. The

farther away from the diagonal line the Lorenz curve is, the greater the degree
of concentration of FACCs. The gini coefﬁcient is a numerical representation
of the degree of concentration; it is a ratio of the area between the Lorenz
curve and the diagonal to the entire area beneath the diagonal. The coefﬁcient
falls between zero and 100, with higher values meaning greater concentration.
The tabulations of FACCs, location quotiens, and Lorenz curve/gini
coefﬁciens lead to an overall description of FACC location at the regional and
state levels. In addition to this description, an attempt is made to analyze two
of the four general factors expected to inﬂuence FACC location at the regional
and state scales. These two factors are demand and economic structure.
Regional FACC locations (and changes in location between 1985 and 1993) are
compared to regional population and economic (i.e., non-farm employment)

growth rates.

Metropolitan Scale

The descriptive part of the analysis continued with an examination of
FACC locations at the intrametropolitan scale for the entire US. FACCs were
divided according to their location in one of these four areas: the central city of
a metropolitan statistical area (MSA), the central county of the MSA, the
surrounding metropolitan counties, or a nonmetropolitan county. This
division separates the FACCs into central city and metropolitan-suburban sites

and allows for a description of intrametropolitan location patterns.

38
Metropolitan and nonmetropolitan location quotiens were also calculated (see

above). Demand, economic growth, and to some extent, ownership patterns of
FACCs can be analyzed by looking at the intrametropolitan location of FACCs.
Suburban locations imply that FACCs are searching for economically and
socially prosperous areas, and indicate a greater degree of physician—ownership
of FACCs. Central city locations, on the other hand, suggest that FACCs are
more service than proﬁt oriented and that hospitals are greatly involved in

FACC ownership and operation.

Statistical Analysis

All four factors (demand, economic structure, competition, and ownership
patterns) are expected to play a role in the location of FACCs in cities and
suburban counties. The statistical methods used in this analysis ascertain
which variables (factors) are important to F ACC location and which variables
are relatively more or less important to FACCs in cities versus central counties
versus other suburban counties. This portion of the analysis included the
largest 100 MSAs nationwide plus the largest MSA in each state not
represented in the top 100 MSAs, for a total of 111 MSAs (Appendix A). Each
MSA was divided into three pars: the central city, the central county, and all
other metropolitan counties. The division of MSAs into central cities and
suburbs allows for maximum variability in the independent variables and
allow for comparisons to determine the importance of the factors to the

different FACC locations.

39
The dependent variable is deﬁned as: the number of FACCs per 100,000

people (FACC/ POP * 100,000) in the central city, the central county (central
county with central city and without central city), and all other metropolitan
counties (aggregated for one value for each MSA). This ratio was chosen as
the dependent variable to take into account the obvious effect of population on
FACC location. The Spearman rank order correlation coefﬁcient between
FACCs and population was .73 in central cities, .68 in central counties (without
cities), and .77 in other counties. A regression analysis showed that for central
cities, population accounted for 69% (R2=.69) of the variation in FACCs.
Population accounted for 82% of the variation in central county FACCs
(R2=.82), and 61% of other county FACCs (R2=.61).

Most of the same independent variables (particularly those concerned with
population attributes) were used in all models. However, due to data
availability, the model for central cities included no data to measure the
employment structure of the city. Data on physicians were also not available
for cities. In addition, the fact that data on employment structure and
physicians were only available for counties (i.e., the central city could not be
separated for these variables) meant that central county data had to be
examined both with the central city (as one unit) and without the central city.
The model without the central city included only population variables and the
hospital beds / population ratio (similar to the model for central cities alone).
The model for central counties including the central city included all variables

(similar to the model for other counties).

40
Twelve independent variables are in the models for central cities and

central counties without the cities, while the models for the entire central
county and for other metropolitan counties include 19 variables. The majority
of variables represent the ﬁrst factor, population (demand for FACCs). Six
employment measures were used to represent the second factor (economic
structure), while two variables (physicians and hospitals) represent
competition/ ownership patterns. The large number of variables meant that
the potentially complex relationships needed to be ﬁrst analyzed in a
simpliﬁed form. To accomplish this, several bivariate statistical methods were
used. First, all independent variables were plotted against the dependent
variable to check for the linearity of the relationships. In addition, Spearman
rank-order correlation coefﬁciens were calculated for all variables. Spearman
correlation coefﬁciens were chosen rather than the Pearson method since all of
the variables were right-skewed (not normally distributed). Based on the
resuls for these analyses, it was determined that for the most part, the
relationships were either linear and/ or weak. In addition to these simple
statistics, bivariate regression analyses were performed for each of the
independent variables. Finally, all independent variables were used in a single
model (one for cities, central counties (with and without cities), and one for
other counties).

The general model: Y: DENSITY + POPGRO + AGE2044 + AGEOS - AGE65

+ INC-r- FEMALE +EDUC +RES - UNEMP + RETAIL + SERVICE + BUS +
FIRE+MFG+HEALTH+PHYS+HOSP

41
Y: dependent variable (see above)

INC: per capita AND median family income by county (or Cent. City)
DENSITY=population density

POPGRO=population growth rate

AGE2044=percent population age 20 to 44

AGE05=percent population age 0 to 5

AGE65=percent population age 65+

FEMALE=percent of female headed households with children 18 and under
EDUC=percent of population over 25 with bachelor's degree

RES=percent of population over age 5 who lived in the same house in 1985
UNEMP=unemployment rate

RETAIL=percent of total employment in retail

SERVICE=percent of total employment in services

BUS=percent of total employment in business services

FIRE=percent of total employment in ﬁnance, insurance, and real estate (FIRE)
MFG=percent of total employment in manufacturing

HEALTH=percent of total employment in health services

PHY S=physician/ population ratio

HOSP=hospital beds/ population ratio

As could be expected, there was high multicollinearity between the

independent variables in this model. Therefore, in order to remove the
problem of multicollinearity, a Principal Componens Analysis (PCA) was

42
performed. Since the purpose of the PCA is to reorganize the independent

variables to remove multicollinearity, no componens were dropped from the
analysis. Therefore, the number of componens equaled the number of
variables in the analyses. Componens were rotated using the varimax
rotation procedure. The componens produced by PCA are orthogonal to each
other and thus can be used as independent variables in the regression analysis.
However, the component loadings are often difﬁcult to interpret. Therefore, a
regression equation in which componens are independent variables may be of
little interpretive value. In order to overcome this problem, reconstituted
partial regression coefﬁcients were obtained following the example of Riddel
(1970) (Appendix B). First, the regression equation is calculated using the
component scores for all of the componens (12 in the case of central cities and
central counties without cities, 19 in the case Of combined central cities and
central counties and in the case of other metropolitan counties). The
dependent variable (FACC/ POP) is also standardized to make interpretation
consistent. Next, matrix multiplication is used to multiply the matrix of
component score coefﬁciens (calculated by SYST AT) and the vector of
standardized regression coefﬁciens obtained from the regression of component
scores on the standardized dependent variable. This multiplication yields the
reconstituted regression coefﬁcient for each of the original variables. Since
component scores are standardized, reconstituted regression coefﬁciens are
directly comparable. However, interpretation of reconstituted regression

coefﬁciens is not completely straightforward since they represent the inﬂuence

43
of the relevant original variables, incorporating their joint effect with other

independent variables with which they are collinear (Johnston, 1991). In this
analysis, the reconstituted regression coefﬁciens allow for a between-scale
comparison of the importance of variables (i.e., which variables are more
important in explaining FACC location in central cities versus central counties,
etc.). Within-scale comparisons are more difﬁcult since the nature of the data
means that there are far more demand variables than variables representing
competition or economic structure. Therefore, the importance of demand
compared to either competition or economic structure may be over stated

using reconstituted regression coefﬁciens.

The Survey
The ﬁnal part of this study concerns obtaining information on FACC

location which is not available through secondary data. At this "site" level,
speciﬁc information on the effects of the third and fourth factors (competition
and ownership) can be obtained. Also at this level, the relationships between
FACCs and other health care providers can be determined. To obtain these
data, a survey of all 245 Michigan PACCs was conducted (Appendix C). The
FACCs were referred to as "clinics" and "ambulatory care centers" since the
term FACC is not familiar to all FACC employees. The survey was sent to the
clinics by mail and was addressed to the clinic owner/ manager (although that
person did not necessarily answer the survey). In implementing the survey,

the suggestions of Dillman (1978) were followed when possible, including

44
preparing the cover letter, identifying the survey, and assembling the mail-out

package. One follow-up mailing (a postcard reminder) was sent to all clinics,
two weeks after the initial mailing of the survey.

The survey was composed of three pars: questions concerning the location
of the FACC and the reasons for the location choice at both the metropolitan
and site levels, questions concerning the FACC's operational and service
characteristics, and questions concerning the relationships between FACCs and
other businesses and other health care providers (including competition). In
the ﬁrst section, repondens were asked questions on the speciﬁc location of
their clinic (inside city limis, in a suburb, or nonmetropolitan area), the type
of building (freestanding, medical building, strip mall, etc), and on the parking
and trafﬁc patterns found near their clinic. They were also asked to state the
importance of several factors (population characteristics, labor availability,
competition, legal/ cost factors, accessibility, etc) in the location decision of the
FACC. These questions were the only ones which required the respondent to
have more than a factual knowledge of the F ACC. However, since most
respondents were not the clinic owners or mangers, they were likely not
involved in the location decision of the FACC. Therefore, these two questions
cannot be considered very representative of the actual decision-making process
for F ACCs.

The second section was designed to get speciﬁc information on the clinics.
Respondens were asked about clinic ownership, chain membership, age of the

clinic, operating hours, clinic size (in square feet), services offered, annual

45
revenues, proﬁs, and expenses, Medicare/ Medicaid policy, and average

monthly patient census. Through these questions, information was obtained
on one of the four inﬂuential factors - ownership patterns. The hypothesis that
ownership affects FACC location decisions could then be tested at this local
scale. In addition, this section allowed for the development of further
hypotheses concerning the FACC, including questions on FACC services, age,
and economies of scale.

The third section asked such questions as "how often does your clinic
receive and give referrals from/ to local physicians and hospitals?" "Does your
clinic have a contract with an HMO/PPO?" "Do your physicians also work
elsewhere?" The respondens were also asked about the clinic's employment
patterns and where the clinic received supplies and services. All of these
questions were designed to be able to better understand the relationship
between FACCs and other providers and businesses. Agglomeration
economies and linkages can be looked at with this information. The third
section also asked respondens to estimate the size of their market area and to
rate other health care providers (physicians, hospitals, other FACCs) on the
intensity of their competition with their clinics. They were also asked to
estimate the distance in miles between the clinic and their nearest competitors.

Due to a low response rate for the survey, primarily nonparametric statisties

(e.g., Chi-Square and Mann-Whitney U tess) were used in the analysis.

CHAPI'ERIV

RESULTS

Descriptive Analysis
Regional Scale

There were 1074 FACCs in the US. in 1985 (according to the American
College of Emergency Physicians (Powills 1985)) (Figure 1), While the number
of FACCs totaled 3845 in 1993 (Figure 2). In 1985, the South contained the
greatest proportion of FACCs (Table 1). Although the South still contained the
largest proportion of FACCs in 1993, is share of the clinics had declined since
1985, while the shares for the West and North Central region increased. In
1985, 60% of the leading ten states in the number of FACCs were located in
the South and West, while this proportion declined to 50% in 1993 (Table 2).
The number of FACCs per person (or per 1,000,000 people) for each state
indicates a slightly different relationship in that many Northeastern and
Midwestern states with small populations actually have large p
FACC/ population ratios. These relationships are similar to those exhibited

using the location quotient technique (see below), therefore, they were not

mapped separately here.

mwma 35m sn— mumuﬁmU ounU bop—canines wﬁvﬁﬂmomuh mo Garcons..—

 

mama. madam kn— mumuﬁmnv muanv bOuagﬁg meﬁGQ—wmounm no COMM—MUG..—
~ 3:me

nUU<um uO gas—Z
con 8...

cos...

 

 

 

 

 

49
Table 1 Number and Proportion of FACCs by Region 1985 and 1993

 

 

 

1985 1993 '
Region Number Percent Region Number Percent
South 436 41 South 1349 35
West 258 24 West 1038 27
North 240 22 North 997 26
Central Central
Northeast 140 13 Northeast 461 12

 

Source: 1985 data compiled by author from Powills (1985); 1993 data compiled
by author from list provided by NAFAC.

50
Table 2 Leading States as Locations of FACCs 1985 and 1993

 

 

1985 1993

State Number State Number
California 132 California 594
Florida 128 Florida 345
Texas 105 Texas 290
Ohio 60 Illinois 250
Michigan 56 Michigan 220
Illinois 53 Ohio 153
Massachusets 42 Colorado 110
Georgia 38 New York 100
Washington 31 Virginia 99
Colorado 28 Massachusets 96

 

Source: 1985 data compiled by author from Powills (1985); 1993 data compiled
by author from list provided by NAFAC.

51
A majority of the leading (top 15) metropolitan areas in the number of

FACCs were located in the South and West in 1993 (Table 3). Californian
metropolitan areas dominated the leading metropolitan areas list. The North
Central region has three metropolitan areas in the top 15, while the Northeast
has only one.

The distribution of FACCs across the United States is correlated with the
state population distribution (the Spearman rank-order correlation coefﬁcient
(Spearman rho) was .85 in 1985 and .89 in 1993). In order to better assess the
relationship between population and the location of FACCs at the regional
and state levels, location quotiens were calculated. In 1985, the lecation
quotiens were greater than one for the South and West, while they were less
than one for the Northeast and North Central Regions (based on both 1985 and
1990 population couns). In 1993, the location quotiens for all regions except
the Northeast were greater than one (Table 4). Therefore, only the Northeast
did not have is expected share of FACCs based on is share of population.
However, the location quotient for the South was almost equal to one,
indicating that the proportion of FACCs in the South is not more than could
be expected based on is share of the population. These resuls suggest that
although FACCs in 1985 appeared to be concentrated in the South and West,
they have begun to disperse throughout the country since then. The states of
California, Florida, and Texas remain dominant, but the Southern and Western

regions as a whole do not appear to be attracting FACCs to the exclusion of

52
Table 3 Leading Metropolitan Areas as Locations of FACCS 1993

 

MSA/PMSA NUMBER

 

of FACCS
Chicago 198
Los Angeles- 196
Long Beach
Detroit 133
Houston 81
Tampa-St. 75
Petersburg
Atlanta 72
Boston 69
Denver 69
Dallas 59
Washington 58
DC.
Anaheim-Santa 57
Ana
Cleveland 53
San Jose 47
San Diego 46
Riverside-San 44

Bernadino

 

53
Table 4 Location Quotiens for Census Regions 1985 and 1993

 

 

1985 1993
Region LQ Region LQ
South 1.19 South 1.01
West 1.20 West 1.30
North central .904 North Central 1.07
Northeast .622 Northeast .583

 

Source: 1985 data computed by author from Powills (1985); 1993 data
computed by author from list provided by N AFAC.

Table 5 Regional Population and Employment Growth Rates: 1980-1985

 

 

 

and 1985-1990.
Population Employment
Region 1980-1985 1985-1990 1980-1985 1985-1990
South .082 .049 .111 .137
West .108 .104 .109 .171
North Central -.0007 .014 .015

126
Northeast .015 .019 .071 .065

 

54
other regions. The Northeast, however, does appear to be the least favorable

location for walk-in clinics.

Table 5 gives data on the regional population, employment, and income
growth rates between 1980 and 1990. These data indicate that the South and
West have experienced greater growth in population and employment than the
Northeast and North central regions. However, population growth rates in the
South and West did decline between 1985 and 1990 (compared to 1980-1985),
while the Northeast and Midwest experienced an increase in population
growth rates between the two periods. In addition, the North Central region
experienced the gmatest growth in both p0pulation and employment between
the 1980-1985 and 1985-1990 periods. The North Central region also
experienced the greatest growth in FACCS between 1985 and 1993. These data
indicate that FACCS are strongly tied to economic and population growth.

An indication of the level of concentration of FACCS at the state and
regional levels can be determined by classifying states as excess states (i.e.,
those with a location quotient greater than one, based on national FACC and
population totals) and deﬁcit states (those with a location quotient less than
one, based on national FACC and population totals). In 1985, there was a total
of 19 FACC excess states, while in 1993 excess states numbered 22. A higher
proportion of the states in the West were in the excess category in both 1985
and 1993 than in the other regions, particularly the Northeast and North
Central (Figures 3 and 4 and Table 6). The Northeast and West regions both
experienced a decline in excess states, while the South gained 3 and the North

 

meme cream nomen— ccn ontuxm 004m
m 8:me

Amwoﬁv cans—um 59G dud—u E .5551 kn— Bﬂm—HOU "0858

ems 3 or!
on; 2 oo;
oo _ 3 S. o s.

Sioosooo D

ecwzoso c0580..

 

 

was MEN—m «momma Urns mmouxm UU<nm
¢ mud—mum

new 3 an.“
an; 2 8; g

and on and D

c5530 5:33

 

 

57
Central gained 2. These results indicate a decrease in concentration at the

regional scale. This conclusion is also supported by an analysis using the
Lorenz curve and gini coefﬁciens (Figure 5). The gini coefﬁcient is lower for
1993 than for 1985, indicating an decrease in the spatial concentration of

freestanding ambulatory care centers during the time period.

Metropolitan Scale

A second hypothesis was that FACCS will be located primarily in
metropolitan areas. This analysis was performed only for the 1993 data. In
1993, 95% of all freestanding ambulatory care centers were located in
metropolitan areas. By comparison, metropolitan areas accounted for only 78%
of US. population in 1990. Location quotiens indicated that metropolitan
areas in all regions had more than their expected share of FACCS (Table 7).
To compare FACC locations with other health care providers, data show that
only 78% of all community hospital beds in the United States were located in
metropolitan areas in 1990 (American Hospital Association 1991). However,
metropolitan locations did account for 88% of all non-federal physicians in
1990 (American Medical Association 1992).

The third hypothesis stated that more FACCS would be found in suburban
locations than in central cities. Table 8 provides data on the proportion of
FACCS in suburban and central city locations. Only in the case of the
Northeast region did the data support the hypothesis. In the other three

regions, central cities accounted for more FACCS than suburbs. A similar

58
Table 6 Proportion of FACC Excess and Deﬁcit States by Region

 

 

1985 1993
Region LQ>1 LQ<1 Region LQ>1 LQ<1
South 29 71 South 47 53
West 62 38 West 54 46
North 25 75 North 42 58
Central Central
Northeast 33 67 Northeast 22 78

 

Cumulative Proportion of FACCS

100

80

60

4O

20

59

 

 

 

 

 

i l I l
Gini Coefﬁcients:
_ / ...
1985 = 11 /
/
_ 1993 = 8 _
/
_ / _
/
/
/
1985
— - - 1993
l l i l
o 20 40 so so 100

Cumulative Proportion of Population

Figure 5
Lorenz Curves for 1985 and 1993

 

60
result has been found for freestanding ambulatory surgery centers (FASCs)

(Lowell-Smith 1993a). These resuls may merely be a function of the
deﬁnitions of "city" and "suburb" which may vary by regions. It could be that
Northeastern cities have well-deﬁned limis while Southern and Western cities
have larger, more ﬂexible boundaries. Therefore, more Northeastern FACCs
have addresses that actually place them ouside the cities, while FACCs in
other regions (which may actually be in "suburban" areas) have addresses
inside the cities.

The resuls of these broad scale analyses have shown that FACCs are
related to two of the proposed factors (demand (population) and economic
structure). FACCs seem to prefer areas of economic and population growth
(South and West). This relationship can be further tested at the metropolitan
scale by looking at F ACC locations in counties and cities and comparing these
locations to p0pulation and employment factors. Also, the dominance of
central city locations may mean that, at least in the South, West, and North
Central regions, there are more hospital-owned FACCs than physician or
corporate owned FACCs since hospitals are most likely to locate their FACCs
in central cities. In addition, many FACCs may be part of a chain or have a
parent organization. This centralized corporate structure may dictate that
member FACCs locate relatively near to the parent/ headquarters which, as a
corporation, is likely to be iself located in the central city (Brown, 1981).
These hypothesis also can be further tested, particularly at the site level.

 

61
Table 7 Location Quotiens for Intrametropolitan Locations by Region

 

 

Region Metropolitan Nonmetro
LO LO
South 1.35 .145
West 1.11 .412
North Central 1.32 .193
Northeast 1.10 .276

 

 

Table 8 Number and Proportion of FACCs by Intrametropolitan Location
and Region

 

Central City Suburban Nonmetropolitan

 

Region Number Percent Number Percent Number percent

 

South 823 61 472 35 54 4
West 509 49 467 45 62

North 548 55 399 40 50 5
Central

North- 152 33 295 64 14 3
east

 

62
Statistical Analysis

The regional scale resuls suggested that two factors, demand and
economic growth, were important determinans of FACC location patterns. At
the intrametropolitan scale, the importance of demand and economic structure,
as well as of the other factors (competition and ownership), on FACC location
can be further tested. This section discusses the resuls of the statistical
analyses on FACCs in MSAs.

Each of the 100 largest MSAs (plus the largest MSA in any state not
included in the top 100 for a total of 111 MSAs) was divided into three pars:
central city, central county (excluding the central city), and all other counties
(aggregated into one "other county" for each MSA). Not all MSAs had a
deﬁnable central city (as in the case of Nassau-Suffolk, NY), or a central
county (as in Washington DC), and several MSAs contained only one county
(e.g., Los Angeles). Therefore, there were a total of 106 central cities, 102
central counties (in both central county analyses), and 75 other counties
included in the analysis. Obviously, MSAs are not equal in size (boundaries).
This may affect the density variable, but it is not expected to affect any other
variables in the analysis. These MSAs accounted for 77% of all FACCs in 1993.

Table 9 liss the variables used in the analyses. In this section, the resuls
for central cities will be discussed ﬁrst, then those for central counties, and
ﬁnally the resuls for the other counties will be discussed.

 

63

Table 9 Variables used in MSA Level Analysis

 

Number of FACCs/100,000
people (dependent)

Percent of Population Aged 0 to 5
Percent of Population Aged 20 to 44
Percent of Population Aged 65 and Over

Population Growth Rate 1980 to 1990
Population Density
Percent of Pop. Over 25 with Bachelor's

Percent of Families Headed by Women

Median Family Income

Per Capita Income

Percent of Population Living in
Same House for the Last Five
Years

Unemployment Rate
Hospital Beds/ Population Ratio
Physician/ Population Ratio"

Percent Employment in
Manufacturing‘

Percent Employment in Retail"
Percent Employment in Services"
Percent Employment in Business
Services*

Percent Employment in Health
Services"

Percent Employment in FIRE"

 

* These variables not available for use in the analysis on central cities

Central Cities

The Spearman rank correlation coefﬁciens show that the FACC / POP ratio
in central cities is only weakly correlated to each of the independent variables
(Table 10). The strongest relationships were with density, the hospital
beds / population ratio, the unemployment rate, and the number of long-term
residens in the city. The relationships with density and long-term residens
were negative rather than positive as hypothesized. Similarly, bivariate
regression analyses revealed that only these four independent variables were
signiﬁcant regressors on the FACC/ POP ratio (Table 11). For the most part,
these relationships were as hypothesized. Central city FACCs are attracted to
areas with higher employment rates, education levels, and a lower proportion
of female-headed families (usually lower socioeconomic status). The
unexpected negative relationship with long-term residens may reﬂect the idea
that FACCs are actually looking for more mobile populations (less likely to
have a regular source of health care) or it may simply be a reﬂection of the
population of these large central cities (i.e., these cities tend to have fewer
long-term residens). The variable, density, may represent to some extent the
socioeconomic status of the cities (less densely populated = "newer" cities, less
crowding, perhaps , better services), rather than simply demand. This would
explain the negative relationship with the FACC / POP ratio. However, this
relationship may simply be a function of "overbounded" versus
"underbounded" cities where cities with more ﬂexible boundaries (less dense)

also have more FACCs with city addresses.

 

65

Table 10 Spearman Correlation Coefﬁciens for the FACC/ POP Ratio for
Central Cities (N: 106)

 

 

Variable Coeff.
Density -.351
Hosp/ Pop .314
Unemployment -.234

Lg. term Residens -.197
Fem. Head. Pam. -.152

 

Education .124
Age 65+ .073
Growth .059
Age 0-5 .052
Med. Inc. .044
Per Capita Inc. .039

Age 20-44 -.022

 

66
Table 11 Bivariate Regression Resuls for Central Cities

 

 

Variable Std. Coef. t p

Density -.329 -3.55 .001
Hosp/ Pop .401 4.47 .000
Unemployment -.239 -2.51 .014
Lg. term Residens -.220 -2.30 .024
Fem. Head. Fam. -.142 -1.46 .146
Education .114 1.17 .244
Med. Fam. Inc. -.056 -.567 .572
Age 0-5 -.045 -.461 .646
Age 20-44 .043 .439 .661
Age 65+ .043 .434 .665
Per Capita Inc. .040 .404 .687
Growth .004 .039 .969

 

 

67
FACCs are also positively correlated with hospital beds, as expected.

These resuls for the most part support the regional scale resuls which
suggested a relationship between FACC location, demand and (for cities) the
location of hospitals.

The model containing all twelve independent variables suffered in part due
to multicollinearity. To remove these multicollinearity problems, principal
componens regression was performed and reconstituted regression coefﬁciens
were calculated. All independent variables were included in the principal
componens analysis and componens were rotated using the varimax rotation
procedure. Component scores for all 12 componens were regressed on the
standardized dependent variable (FACC/ POP ratio). Reconstituted regression
coefﬁciens were then calculated using the component score coefﬁcient matrix
and the standardized regression coefﬁciens from the regression equation.
Table 12 shows the reconstituted regression coefﬁciens for the central cities.
These coefﬁciens show that the growth rate, the percentage of people ages 20
to 44, the hospital beds/ population ratio, and density are relatively more
important variables in explaining the location of FACCs in cities.

It is concluded then that FACCs located in central cities show the greatest
relationships to the population growth rate, density (the negative relationship
suggesting that density is here a measure of the socioeconomic condition of the
city) and to hospital beds (presumably an ownership measure and / or a
measure of agglomeration economies). FACCs also show less strong

relationships with other demand and socioeconomic variables such as the

 

68
Table 12 Reconstituted Regression Coefﬁciens for Central Cities

 

 

 

Variable Coeff.
Growth .601
Age 20-44 .424
Hosp/ Pop .320 .
Density -.318
Unemployment -.187
Med. Fam. Inc. -.136
Age 0—5 -.116
Per Capita Inc. -.103
Lg. term Residens -.094
Education .076
Age 65+ -.061

Fem. Head. Farn. .057

 

69
percentage of people aged 20 to 44 (the users of FACCs), the percentage of

children, and the unemployment rate. The bivariate regression equations and
the reconstituted regression coefﬁciens both indicate a negative relationship
between FACCs and income which was unexpected. This suggess that cities
with relatively more FACCs are not necessarily those with "wealthier"
populations. In fact, the service orientation expected of central city FACCs may
mean that income as a measure of demand or ability to pay is less important

than it might be for proﬁt-oriented FACCs.

Central Counties

The central counties analysis has two pars: one involves an examination of
the data for central counties without the central city, while the other involves
data for the entire central county, including the central city. The relationships
for central counties without the central city are discussed ﬁrst.

Spearman rank correlation coefﬁciens were used to test the strength of
the relationships between the FACC/ POP ratio and the independent variables
(n=12 for model without central city). The relationships were much weaker
than those exhibited in central cities alone, with no correlation exceeding .21
(Table 13). Likewise, in all bivariate regression analyses, no independent
variables were signiﬁcant and most adjusted R2 values for these equations
were .000.

A principal componens analysis was performed on all 12 independent

variables. The componens were rotated using the varimax rotation procedure.

70

Table 13 Spearman Correlation Coefﬁciens for FACC / POP Ratio for Central
Counties Without Central Cities (N=102)

 

 

 

Variable Coeff.
Education .205
Density .163
Lg. term Residens -.163
Growth .151
Fem. Head. Fam. .128
Per Capita Inc. .093
Hosp / Pop .063
Age 20-44 .061
Age 65+ -.(B6
Age 0—5 .035
Med. Fam. Inc. .035

Unemployment .018

 

71

The component scores were used in a regression equation with the
standardized dependent variable. Reconstituted regression coefﬁciens were
calculated using the component score coefﬁciens and the standardized
regression coefﬁciens for each component. The reconstituted regression
coefﬁciens are shown in Table 14. The number of long-term residens, the
unemployment rate, and median family income were the relatively more
important variables in explaining FACC location. Interestingly, two of the
most important "demand" variables, unemployment rate and income, are also
the variables which can be used to some extent to represent economic
structure. However, it can be seen that almost all of the variables were less
important than they were for central cities.

The lack of relationships between the FACC / POP ratio and the
independent variables may be due to the strong relationship to population (i.e.,
only the presence of demand is important at this scale), or it may be because
the relevant independent variables were simply not included here. Another
possible explanation is that central county FACCs take into account
characteristics of much smaller areas than the entire county, perhaps towns
and cities within the county or perhaps census tracts. On the other hand, it
may be that the characteristics and inﬂuence of the central city are so great
that FACCs do not look to the characteristics of the central county alone to
make location decisions. The sharp decline in the importance of the
population variables compared to their relative signiﬁcance in central cities

was as expected. This implies that, as hypothesized, central county FACCs are

 

72
Table 14 Reconstituted Regression Coefﬁciens for Central Counties Without

 

 

Central Cities
Variable Coeff.
Lg. term Residens -.175
Unemployment -.139
Med. Fam. Inc. -.110
Density .096
Growth .077
Education .068
Hosp/ Pop -.055
Per Capita Inc. -.054
Fem. Head. Fam. .046
Age 65+ .009
Age 20-44 0%

Age 0-5 -.001

 

73

choosing their locations for other reasons, perhaps economic growth, rather
than to serve a particular population.

The same analyses were conducted for entire central counties (including
the central city). Spearman correlation coefﬁciens are given in Table 15.
Median family income, the proportion of children, manufacturing employment,
and the hospital beds / population ratio showed the strongest correlations. The
bivariate regression analyses had similar resuls (Table 16). Also contrary to
expectations, the population variables have assumed greater importance than
the economic structure (employment) variables, with the exception of the
negative relationship to manufacturing employment. Apparently, the
locational relationships of central city FACCs (i.e., concern with population
variables) dominate over those of central counties (which alone showed no
relationship to the population variables other than, of course, population size).
Also of interest is the signiﬁcant positive relationship between the FACC/ POP
ratio and the hospital beds/ population ratio. This relationship was expected,
although since it was not signiﬁcant for central counties alone, this signiﬁcant
relationship is likely due to the influence of the central cities.

As with central counties alone, a Principal Componens Analysis and
regression were performed on the variables for the combined central counties
and cities. Reconstituted regression coefﬁciens were calculated (Table 17).
The most important variable is the hospital beds / population ratio (presumably
due to the inﬂuence of this variable on central city FACCs). Interestingly, the

age variables, education, and female-headed families assume greater

74

Table 15 Spearman Correlation Coefﬁciens for FACC / POP Ratio for Central
Counties With Central Cities (N=111)

 

 

Variable Coeff.
Med. Fam. Inc. -.306
Age 0-5 .300
Manufact. -.281
Hosp/ Pop .259
Per Capita Inc. -.246
Lg. term Residens -.224
Age 65+ -.217
Education .201
Business .191
Density -.187
Services .164
Age 20-44 .150
Phys/ Pop .115
Growth -.103
Unemployment -.103
Fem. Head. Fam. -.095
FIRE .067
Health -.048

Retail -.008

 

Table 16 Bivariate Regression Resuls for Central Counties With

 

 

Central Cities

Variable Std. Coeff. t p
Age 0-5 .284 2.97 .004
Med. Fam. Inc. -.244 -2.51 .014
Manufact. -.201 -2.05 .043
Density -.197 -.201 .048
Hosp/ Pop .196 2.00 .048
Education .194 1.98 .176
Per Capita Inc. -.181 -1.84 .069
Lg. term Residens -.168 -1.70 .092
Unemployment -.151 -1.53 .130
Age 20-44 .143 1.44 .138
Fem. Head. Fam. -.135 -1.36 .176
Age 65+ -.132 -1.33 .187
Services .125 1.26 .212
Business .113 1.14 .258
Health -.113 -1.14 .257
Phys/ Pop -.066 -.664 .508
FIRE .072 .717 .475
Growth -.016 -.162 .872
Retail .007 .068 .946

 

76

Table 17 Reconstituted Regression Coefﬁciens for Central Counties With
Cities

 

 

Variable Coeff.
Hosp/ Pop .588
Education .449
Age 65+ -.416
Age 0-5 .398

Fem. Head. Pam. -.352
Lg. term Residens -.216

 

Med. Pam. Inc. -.205
Health -.186
Per Capita Inc. -.144
Manufacturing -.134
FIRE .133
Growth -.081
Unemployment .069
Retail -.066
Services -.055
Business .042
Density .037
Phys/ Pop -.032

Age 20-44 .031

 

77
importance for the combined model than they did for either the central cities

or central counties alone. Once again, the employment variables rank lower in
importance than the population variables. However, this may be a result of
the fact that there were far more population variables (many of which were
interrelated) than there were employment variables.

The combined resuls for the central cities, central counties alone, and
central counties with cities suggest that the hypothesis regarding the
importance of the factors, in particular demand versus economic structure, was
basically correct. In the central cities, population variables are important,
especially measures of socioeconomic status. In central counties alone, these
relationships did not appear. In the combined model, both population
variables and economic structure variables were important, reﬂecting the
relative contributions of the central cities and the central counties. This
hypothesis can be further tested by looking at the FACC / POP ratio
relationship to population and economic variables for other metropolitan

counties.

Other Counties

The same analyses were conducted for the other (non-central) counties in
the MSAs. Spearman rank correlation coefﬁciens indicated that the
relationships between the FACC / POP ratio and the independent variables
were stronger than they were for either the central cities or the central counties

(Table 18). Of particular interest is the fact that the four highest correlations

78

Table 18 Spearman Correlation Coefﬁcients for FACC / POP Ratio for Other
Counties (N=75)

 

 

Variable Coeff.
Manufact. -.430
FIRE .389
Service .341
Business .320
Education .291
Density .287
Med/ Fam. Inc. .259
Lg. term Residens -.255
Phys/ Pop .248
Growth .237
Per Capita Inc. .237
Age 20-44 .173
Age 65+ -.164
Age 0-5 .137
Unemployment -.126
Retail .093
Fem. Head. Fam. -.073
Health .026

Hosp/ Pop .003

 

79

are with measures of employment (representing the economic structure factor)
and that the relatiOnships were as expected: more FACCs in areas with more
employment in the services, especially employment in FIRE and business
services. The bivariate regression analyses tell a slightly different story (Table
19). In these equations, manufacturing employment, services employment, the
median family income, and the physician/ population ratio were each
signiﬁcant.

Once again, the model which contained all 19 independent was not
unexpectedly insigniﬁcant due to multicollinearity. Principal componens
analysis was used to take care of this multicollinearity problem. Componens
were rotated using the varimax rotation procedure. Component scores for all
19 components were used in a regression analysis with the standardized
dependent variable. Reconstituted regression coefﬁciens were calculated
using the component score coefﬁciens matrix and the standardized regression
coefﬁciens for each component (Table 20). At this scale, the variables
unemployment, retail employment, employment in FIRE, and the growth rate
are most important. Apparently, economic structure is more important than
demand in determining FACC location in outlying metropolitan counties.
Along with the employment variables, the unemployment and growth rates
can also be considered measures of economic prosperity (as well as measures
of demand). In addition, a comparison with the reconstituted coefﬁciens for

the combined central counties/ cities model shows that variables representing

80
Table 19 Bivariate Regression Resuls for Other Counties

 

 

Variable Std. Coeff. t

Manufacturing -.266 -2.36 .021
Med. Fam. Inc. .256 2.26 .027
Phys/ Pop .241 2.12 .037
Services .234 2.05 .044
Education .221 1.94 .056
FIRE .205 1.79 .077
Growth .204 1.78 .079
Unemployment -.146 -1.26 .212
Age 0-5 .130 1.13 .268
Density .119 1.02 .310
Lg. term Residens -.096 -.82 .843
Retail 095 .84 .418
Age 65+ .047 .41 .686
Business 035 .30 .765
Fem. Head. Pam. -.034 -.29 .770
Hosp/ Pop .029 .25 .802
Health .025 .21 .834
Age 20-44 .023 -.20 .843
Per Capita Inc. .000 .42 .689

 

81
Table 20 Reconstituted Regression Coefﬁciens for Other Counties

 

 

Variable Coeff.
Unemployment -.549
Retail -.514
FIRE .370
Growth .369
Fem. Head Fam. -.223
Health -.171
Med. Pam. Inc. .167
Per Capita Inc. .159
Age 65+ .143
Services .137
Business .127
Education -.113
Age 20-44 -.112
Age 0-5 .092
Lg. term Residens -.084
Density .069
Hosp/ Pop .046
Phys/ Pop .023

Manufacturing -.016

 

82
economic structure are more important for the "other" counties model for all

employment categories except health services.

It seems apparent that for FACCs in outer suburban counties, the
economic structure (economic growth and agglomeration economies) of the
area is most important in determining their location. Measures of demand
(other than population size, of course) are not as important to FACC locations.
In addition, the relationships between FACCs and the location of both
physicians and hospital beds were positive as expected, although the
relationship was signiﬁcant only for physicians.

Although the resuls of these analyses at the metropolitan scale were not as
strong as had been hoped, they do reveal important aspects of FACC location
and allow for suggestions as to the direction of further research at the next
scale. There are three general Conclusions which can be drawn from these
metropolitan scale analyses. First, when population variables are signiﬁcant
(e.g., in central cities), the relationships are, for the most part, as predicted;
that is, areas with higher levels of demand as measured by those most likely to
use FACCs and as measured by socioeconomic status are areas with more
FACCs. Second, the relative importance of the factors expected to inﬂuence
FACC location seem to vary with location. Population variables are important
to FACCs in central cities, presumably because those FACCs are in central
cities to serve the population and to provide alternative care to hospitals
(particularly, the emergency room). Economic structure variables,

representative of economic and employment growth and of agglomeration

83
economies, are relatively more important to FACCs locating in suburban

counties. This is presumably because of the proﬁt-maximizing focus of
suburban FACCs which causes them to concentrate on ﬁnding economically
prosperous areas. Third, all three areas had the expected relationships with
the "competition/ ownership" variables, hospitals and physicians, although not
all relationships were statistically signiﬁcant. In general, FACCs were
associated with more hospital beds in central cities and counties, and with
physicians in other counties (based on both bivariate analyses and
reconstituted regression coefﬁciens). This suppors the hypothesis that
hospital-owned FACCs would be found more often in central cities due to
their service orientation while suburban FACCs are more likely to be owned

by physicians.

 

84
Survey of Michigan FACCS

Both the descriptive (regional) analyses and the statistical (MSA) analyses
indicated that demand and economic structure are signiﬁcant factors in
determining FACC locational patterns. The MSA-level analyses also suggest
that the competition-ownership factors affect FACC location. In order to
further test the inﬂuence of these factors, primarily competition and
ownership, on FACC location, a mail survey was sent to Michigan FACCs. In
addition, the survey method allowed other hypotheses to be tested such as
those regarding differences in the services offered by suburban and central city
FACCs.

This section discusses the resuls of the survey of Michigan FACCs, the
site-level analysis of the determinans of FACC location. Of the 245 surveys
sent to Michigan FACCs, 17 were returned as undeliverable. Forty-one
surveys were completed and returned, although eight of these were not
FACCs (they required appointmens and / or had no evening or weekend
service). Therefore, 33 surveys out of a possible 220 were used in the analysis
(response rate of 15%). In order to statistically test the resuls of the surveys,
the responses were divided by three characteristics: by location - central city (n
= 13) and suburban (n = 17); by ownership - hospital owned (n = 17) and non-
hospital owned (n = 16); and by chain (n = 14) and non-chain (n = 19). Where
appropriate, Chi-square and Mann-Whitney U tess were used to test for

signiﬁcant differences. Most of the resuls discussed here are descriptive and

 

85
include all respondens. Only when the statistical tess were signiﬁcant are

they reported here. A summary of the resuls can be found in Appendix D.
The resuls are broken down into three sections: the ﬁrst discusses the
location of the FACCs and the reasons for their location (as given by the
respondens). The second section discusses the characteristics of the FACCs,
including ownership patterns, operations, and services. The last section
considers the relationship of the FACCs to other health care providers and

their competitors, and the linkages between FACCs and other businesses.

Location

The questions on clinic location address both the intrametropolitan (central
city versus suburbs) scale and the site level. This information can be used to
test the hypothesis discussed at the regional scale which states that FACCs
should have a preference for suburban areas since these areas have
experienced economic and population growth. In addition, the preference for
suburbs over central city or nonmetropolitan areas would also indicate that the
"proﬁt generation" motive is greaterlthan the "social service" motive as
discussed in the MSA level analyses. At the site level, locational information is
used to better understand what site characteristics the clinics considered
important and thus how location decisions are made.

In this survey, 39% of the clinics were located inside the city limis (central
city), 52% were suburban, and 9% were in nonmetropolitan areas. By

comparison, 40% of all 220 Michigan clinics were inside the city limis, 55%

86
were suburban and 5% were nonmetropolitan. The survey resuls appear to

be representative of Michigan FACCs in terms of the intrametropolitan location
of clinics. The dominance of suburbs as FACC locations was originally
hypothesized since presumably most FACCs are looking for economically and
socially proﬁtable areas. However, this result is in contrast to the overall
dominance of central cities as FACC locations as determined in the descriptive
analyses at the regional scale. This means only that Michigan FACCs seem to
prefer suburban locations, while those FACCs in other North Central states
more often chose central city sites. These resuls could suggest that there are
differences in the reasons for the existence of FACCs by state, in Michigan,
the proﬁt motive appears to dominate, while in other states, the "social service"
motive might be dominant. However, it must also be kept in mind that many
of the North Central states included in the analysis may have only small cities
or overbounded cities which would place FACCs within the city limits (central
city location), but which may actually be "suburban" in character. Therefore, it
is likely that the suburban preferences displayed by Michigan FACCs are more
representative of the true location decision and the true motive (proﬁt).
Questions about site level location choices also gave information regarding
the locational needs of the clinics and the characteristics considered important
for clinic success. Two-thirds of the clinics (64%) were located in freestanding
buildings, as expected. Twelve percent were in medical buildings and 9% each
were in hospitals and multi-use buildings. Only two clinics were located in

strip malls and none were in enclosed shopping centers. Parking was free at

87
100% of the clinics, and 70% of clinics had their own parking lot. Fifty-eight

percent were located on 4-lane highways, 30% on two lane roads, and 12% on

6-8 lane highways. Over two-thirds were NOT located at a major intersection.
These resuls are all similar to those reported by NAFAC (1990). In general, it
would appear that the typical FACC is oriented toward auto-owning suburban

and city dwellers. Physical inaccessibility in fact may be used as a deterrent to

‘13

the poor and uninsured rather than ﬁnancial barriers as most clinics do accept
Medicare and Medicaid (see below).

These next questions were used to indicate the importance of measures of
the four factors, demand, economic structure, competition, and to a lesser
extent, ownership, in the metropolitan and site scale location choices of the
clinics. Respondens were asked to evaluate the importance of a series of
reasons for the location of their clinics at both the metropolitan and site levels.
Reasons were ranked on a scale of 1 to 5 (with 1 being the most important
reason). Scores were calculated for the location factors by adding the number
of Is, 25, 3s, 4s, and 58 each reason. received, multiplying those numbers by 1,
2, 3, 4 and 5 respectively, adding the total "score," and dividing the score by
the number of respondens (Table 21). Due to the nature of the question, the
lower the score, the more important the reason. At the metropolitan level,
competition was considered the most important consideration, with population
characteristics (demographics, socioeconomics) coming in a close second (note
that when chain clinics and those with a parent organization were separated,

nearness to headquarters / parent ranked third behind competition and

88
Table 21 Importance Scores for Location Factors for Metropolitan Area and

 

 

Site Locations

Factor Metro Area Factor Site

Competition 1.8 Target Population 1.7

Population Chars. ' 1.9 Unserved by 1.8
Competition

Nearness to HQ 2.7 Accessibility of site 1.8

Personal Reasons 2.7 Site Characteristics 2.1

Legal/ Cost Factors 2.8 Business Structure 2.4
Around Site

Availability of 2.8 Legal/ Cost Factors 2.8

Labor

Nearness to Other 2.8
Health Care
Professionals

 

89
population). The fact that parent/ headquarter availability is important to the

respondens in chains (or with a parent) suggess the direct importance of the
ownership variable for these clinics.

At the site level, nearness to a target population, site unserved by
competition, and site accessibility were the top reasons for clinic location. The
only statistically signiﬁcant difference was between hospital-owned clinics and

non-hospital clinics. N on-hospital clinics ranked nearness to target population

544.424.]

somewhat higher than hospital-owned clinics (U=81.5, p=.032). The resuls do
support the theoretical hypotheses concerning the importance of two of the
four factors, demand (population) and competition. In addition, the responses
also suggest the importance of site level factors such as accessibility and site
characteristics. At both scales, measures of economic structure (labor
availability, site business structure, and nearness to other health care
providers) all ranked low in importance. However, as the actual location
decision was probably not made by the person answering the survey (in most
cases), these resuls cannot be considered as true guidelines to understanding

the FACC location decision process.

Characteristics

The questions on clinic characteristics gave information on two of the four
factors: ownership, and to a lesser extent, economic structure (in terms of
employment in FACCs as an agglomeration economy). Also, these questions

looked at additional hypotheses of FACC operation such as clinic age, services

90

and economies of scale. The questions concerning services, in particular, can
be used to understand the location choices of FACCs since FACCs (their
owners) will offer services based on a perception of need (social service)

and / or the possibilities of proﬁt.

A slight majority of the respondens were hospital owned (52%), 24% were
owned by a single physician, 21% were owned by multiple physician groups,
and 3% (1 clinic) were non-physician owned. The dominance of hospital-
owned FACCs indicates that there will be a dual purpose in the location
patterns: to provide a social service (in response to demand) and to generate a
proﬁt for the hospital. However, there were no statistically signiﬁcant
differences between FACC locations (city versus suburbs) and ownership
categories (hospital-owned versus non-hospital owned). The industry survey
conducted by N AFAC in 1990 shows hospital ownership to be 26% of the total,
with non-physician owned clinics accounting for 32% (NAFAC 1990). There
may be regional differences in FACCs ownership, with Michigan FACCs being
primarily owned by hospitals. In addition, the facs that more Michigan clinics
are suburban and that more are hospital-owned are not contradictory since
hospitals open FACCs in suburban areas (for proﬁt) as well as in central city
areas (for service). Actually, suburban dominance is quite logical since both
hospitals and non-hospitals will choose suburban locations for their F ACCs.

Almost half (44%) of the respondens belonged to chains and the average
chain included six clinics. Fifty-ﬁve percent of the clinics opened before 1985
(19% before 1980). Chi-square tess for relationships between clinic age and

 

91
ownership categories showed that hospital-owned clinics were likely to be

newer (opened after 1985) than nonhospital FACCs (Chi-square=6.89, p=.009).
This suppors the hypothesis that hospitals entered the FACC market relatively
late and in response to the competition from physician and corporate-owned
FACCs.

Employment of physicians and other medical personnel can be an
indication of the need for locations near these professionals (as an
agglomeration economy). Employment averages and ranges are given in Table
22. These resuls are similar to those reported by NAFAC (1990), except that
the clinics in this survey employed somewhat fewer nurses, x-ray technicians,
and nonmedical staff members than the 1990 respondens. The Mann-Whitney
U statistic was used to test for differences in employment between ownership
and chain / non-chain categories. Hospital-owned clinics and chain clinics were
found to employ significantly more full time nurses than the other categories
(U=209.5, p=.005 for hospital-owned clinics; U=193, p=.027 for chain clinics),
while chain clinics also employed signiﬁcantly more full time physicians than
non-chain clinics (U=219, p=.002).

The services provided by the clinics are detailed in Table 23. These results
are similar to the 1990 survey except that primary care, orthopedics, drug
dispensing, and dermatology are offered somewhat less by this survey's
respondens, while gynecology, obstetrics, and physical therapy are offered
more often by this survey's respondens. Occupational medicine, spors

medicine, and laboratory services were not analyzed in the 1990 survey. Chi-

92

Table 22 Employment in F ACCs.

 

 

Employee Full-time Min Max Part-time Min Max
average average

Physician 2.7 4.0 0 20

Nurses 1.2 2.9 13

X-ray 1.3 1.6 6

technician

Medical 2.8 0 15 1,8 0 8

assistant

Nonmed. 3.7 0 25 2.0 0 12

staff

 

“:3

Ln .l'l

93
square analyses show that suburban clinics are more likely than central city

clinics to offer gynecological services (Chi-square=5.8, p=.016), while non-
hospital owned clinics are slightly more likely than hospital-owneds to offer
industrial-occupational medicine (Chi-square = 4.3, p=.038). These resuls
support hypotheses which stated that suburban and non-hospital owned
FACCs would be more likely to offer "specialized" services. This suppors the
general hypothesis which states that suburban FACCs are more likely to be
located for proﬁt rather than purely service motives.

The average clinic sees 55.5 patiens per day (compared to 47.5 for the 1990
survey) and a vast majority of clinics accept Medicaid (73%) and Medicare
(91%). Hospital-owned FACCs are more likely to accept Medicaid (though not
Medicare) than non-hospital owned clinics (Chi-square=13.2, p=.000). This
result is not surprising given the traditional role of hospitals in providing care
for the poor. It is also possible that since hospitals are required to accept
Medicaid/ Medicare, their ancillary ofﬁces (such as FACCs) are also required to
accept these methods of payment.

Although only 15 respondens gave information concerning their
revenues, Pearson correlation coefﬁciens were calculated and a step-wise
regression analysis was performed on the answers received. Revenues were
correlated with the number of physicians employed full time (r=.68), the
number of patiens per day (r=.59), the number of medical assistans employed
part time (r=.50), and clinic size (r=.46). There were 15 independent variables

originally entered into the model, including all ratio variables obtained from

 

94

Table 23 Proportion of Clinics Providing Certain Services.

 

 

Service °/o
Urgent Care 94
Industrial Med. 85
Primary Care 66
laboratory 64
Physical Therapy 52
Drug Dispensing 45
Pediatrics 42
Gynecology 41
Orthopedics 23
Spors Med. 22
Diet Control 19
Cancer Screening 16
Obstetrics 13
Dermatology 10
Drug Abuse Treatment

Psychology / Psychiatry

Dentistry

 

95
the survey (Table 24). The model with revenues as the dependent variable

yielded only one significant independent variable, the number of full time
physicians employed by the clinics. The variables were positively related
(t=2.9, p=.012) with an R2 of .395. The number of full time physicians was
highly correlated with all other full time employment groups and was
correlated as well with clinic size (r=.46) and the number of patiens (r=.53).
The number of physicians employed full time appears to represent clinic size
and the number of patiens the clinic attracs. Therefore, the conclusion is
simply that larger clinics generate more revenues, as hypothesized. Although
questions were asked on clinic expenses and proﬁs, even fewer respondens
answered them so they were not analyzed.

Relationships

This section asks questions concerning the third main research objective,
understanding the role of the FACC in the health care system and on factor
three, competition. It is expected that competition, the factor which was not
easily studied at either the regional or metropolitan scales, will show is
importance at the site level. The survey respondens have already
acknowledged the importance of competition. This section will obtain further
information on the strength of competitors and the distance FACCs place
between themselves and the competition.

To better understand how FACCs interact with other providers and to

some extent with other businesses, questions were asked concerning patient

96

Table 24 Variables Included in Regression Analysis

 

physicians employed full-
time

physicians part-time
nurses full time

nurses part-time

x-ray tech. full time
x-ray tech. part-time
med. asst. full time

med. asst. part-time

nonmed. staff full time

nonmed. staff part-time
distance to nearest F ACC

distance to nearest
physicians' ofﬁces

distance to nearest hospital
size of clinic

number of patiens seen
daily

 

 

 

 

97
referral, contracs with HMOs/PPOs, physician employment, competition, and

where the clinics obtain certain business services and supplies. Most
respondens indicated that they either "often" (24%) or "sometimes" (45%)
received patiens referred to them from local physicians. Sixty-one percent
said they "often" refer patiens to nearby physicians, and 58% said they "often"

refer patiens to local hospitals. Over two-thirds of clinics have a contract with

‘3

an HMO and 79% have a contract with a PPO. These resuls indicate that '
FACCs are closely tied to other health care providers and that they act most E
often as a ﬁrst-contact for patiens seeking health care. This is not surprising
given that most FACCs position themselves as offering convenient care for
minor injuries and illnesses. However, these responses may also indicate that
patiens are attempting to use FACCs for care beyond that which the clinics
can provide and are therefore being referred to more comprehensive providers
such as physicians and hospitals. The fact that a majority of FACCs are
contracting with HMOs/PPOs pus them in a good position to be a part of the
new health care reform measures which are apparently hinging on HMOs.
At least one physician employed at the clinic also worked elsewhere in 73%
of the responding clinics. Of this 73%, 45% said at least one physician was
also employed at a hospital, 36% had her/ his own practice, 36% had
physicians employed in another FACC, and 3% (1 physician) were also
employed in an HMO/ PPO. (NOTE: physicians at a clinic could be employed
in more than one of these other places). These resuls suggest the integrated

nature of the FACC, in this case as a work place. It is also interesting to note

98
that over one-third of the clinics had physicians (presumably part-time) who

were also working in other clinics (in some cases at least, working for the
competition).

One of the four general factors related to FACC location which has not
been adequately analyzed at either the regional or metropolitan scale is
competition. FACCs will be able to identify the strongest competitors. It is
expected that FACCs will then respond to competition by choosing locations
with a certain distance from competitors (each other, physicians, etc.).
Competition should be most intense in the suburbs, since these FACCs are
competing with each other, physicians, and hospitals.

Respondens were asked to rank the intensity of competition (on a scale of
1 to 5, with 5 being most competitive) that their clinic experienced with these
potential competitors: other FACCs, physicians, hospitals, HMOs, and PPOs.
Competition "scores" were calculated by adding the number of Is, 2s, 3s, 4s,
and 58, each competitor received, multiplying those numbers by 1, 2, 3, 4, and
5 respectively, adding the total "score," and dividing by the number of
respondens. Other FACCs were considered the most competitive by the
clinics in this survey as they scored 2.9. Physicians were next with 2.6,
hospitals scored 2.3, HMOs scored 2.1 and PPOs scored 1.9. The Friedman's
two way analysis of variance test indicates that these scores are signiﬁcantly
different (statistic = 12.14, prob. = .016 given a chi-square distribution with 4
degrees of freedom). These resuls correspond closely to those reported in the

1990 survey (N AFAC, 1990). There were some differences in competition

99
"scores" when broken down by category. The Mann-Whitney U statistic

showed that suburban clinics felt that physicians were more competitive than
central city clinics did (U=61.5, p=.03). Similarly, chain clinics experienced
greater competition with other FACCs than non-chain clinics (U =209, p=.006).
The result concerning the greater competition between suburban clinics and
physicians corresponds with the hypothesis that suburban FACCs will be
found more often with physicians. Physicians frequently own suburban
FACCs. They also compete with FACCs as private practitioners.

The average FACC was found between 4 and 6 miles from the nearest of
all types of competitors except physicians. The average FACC was located just
less than 1 mile from the nearest physicians' ofﬁces, although chain clinics
located farther from physicians than non-chains (U=187, p=.049). The resuls
for other FACCs and for hospitals are supported in the literature (Parsons
1987), although the average distance to physicians is much smaller than
expected. On average then it can be said that, either clinics are locating near
physicians' ofﬁces despite competition, or these locations may be the result of
a private practitioner (who would probably have an ofﬁce located near other
physicians) transforming his/ her ofﬁce into an FACC. These clinics may also
be owned by hospitals which are locating them (the FACCs) near physicians'
ofﬁces in order to compete with them. Yet still another possibility is that the
greater numbers of physicians (compared to FACCs, hospitals, HMOs, etc)
means that FACCs cannot easily avoid locating near a physician's ofﬁce,

whereas they can disperse from each other and hospitals.

 

100
FACCs also have a variety of linkages to other businesses. FACCs which

are part of chains or have a parent organization receive most of their business
services through the headquarters or parent (Appendix D). Chain clinics tend
to obtain their own medical and other supplies, while independens usually
receive medical and other supplies from ﬁrms inside and ouside the
metropolitan area. Subcontracting for tax preparation, accounting, legal
services, and advertising is common for non-chain FACCs, while payroll,
personnel, and business management are often performed within the clinic
(Appendix D).

The resuls of the survey lend support to the four general factors expected
to inﬂuence the location of FACCs. The respondens indicated that demand
(population characteristics) was a relevant location factor at both the
metropolitan and site levels. The importance of economic structure,
particularly in terms of agglomeration economies, was demonstrated by the
resuls concerning the interrelationships between FACCs and other health care
providers and other businesses. FACCs need to be near physicians and
hospitals for patient referrals, contracs with HMO and PPOs, and to employ
physicians. They also employ other members of the health services ﬁeld such
as nurses and x-ray technicians. In addition, most FACCs (especially
independent centers) use business services and thus need to be in areas where
these services are available.

Competition (the third factor) also plays an important role. The

respondens indicated is importance at both metropolitan and site levels.

101
FACCs considered each other to be most "threatening," while physicians and

hospitals were also thought to be competitive. In fact, physicians were
considered to be greater competition in suburbs than central cities. Although
there were no statistically signiﬁcant relationships between ownership patterns
and location, the resuls did support other conclusions about ownership and
FACCs. Hospital-owned FACCs are more likely to have social service
functions than other FACCs (as evidenced by their greater acceptance of
Medicaid). Hospital-owned FACCs are also relatively younger than other
types of FACCs, suggesting that their increased participation in the market is
in part competition-driven and in part the result of slower response times from
hospitals to changing market conditions. These explanations are supported in
the literature which shows that early FACCs were dominated by non-physician
corporations, while physician-owned, and slowly, hospital owned FACCs have
more recently gained control of the FACC market. In general, it is possible to
conclude that the hypotheses that the four factors affect location patterns of

FACCS at all scales (to varying extens) are justiﬁed.

 

CHAPTER V
CONCLUSIONS

Summary 5'1
The results for all scales of analysis reveal that the four factors, demand, I
economic structure, competition, and ownership all inﬂuence FACC location
patterns. At the regional scale, factors two and three, demand (population)
and economic structure (employment growth) were found to be important
determinans of FACC location. The South and West (areas of population and
employment growth) have maintained dominance in numbers of FACCs.
However, there has been a dispersal of the clinics between 1985 and 1993. The
South and West have declined in dominance, while the North Central region
gained. Similarly, population and employment growth in the South and West
has slowed, while population and economic growth in the North central region
has increased in recent years. The Northeast, however, still lags in the number
of FACCs. Many possible reasons for this avoidance of the Northeast exist,
including, of course, the recent economic decline of the region and population
out-rnigration. The presence of higher malpractice insurance coss for
physicians (AMA, 1992), stricter state-level governmental regulations of FACCs
(GAO, 1990), and the strong, well-established presence of other health care

102

103
providers in the Northeast are additional reasons for the relative lack of

FACCs in this region.

In three of the four regions, central city locations were found to be
dominant, rather than the expected suburban locations. However, this result is
likely to be a function of regional differences in city boundaries, rather than a
preference for central city over suburban locations. That is, Northeastern cities
are "underbounded" while cities in the other regions are more likely to be
"overbounded." Thus, a FACC with an address in Atlanta is more likely to be
suburban in character than a FACC with an address in New York City.

At the MSA scale, demand and economic structure are once again revealed
as signiﬁcant factors in FACC location. In central cities, population variables
measuring demand (FACC users) and socioeconomic status were found to be
relevant to FACC location. The resuls indicated positive relationships
between demand and the FACC / Population ratio as expected. The
socioeconomic status variables were somewhat contradictory as FACCs were
negatively related to the unemployment rate, but also to measures of income.
At the county level, the importance of population as a determinant of FACC
location declined while the importance of economic structure rose. In the
suburban counties, FACCs are positively correlated to employment in services,
including business services and FIRE and they are negatively correlated to
manufacturing employment. Of course, population variables, as measures of
demand and socioeconomic status were not irrelevant to FACC location in

suburban counties.

 

104
As expected, population variables (representing demand) were important

to FACCs in central cities and central counties. These FACCs are providing a
social service, an alternative to the hospital emergency room. Thus, their focus
is on the level of demand and where the service might be most needed. Also
as expected, FACCs in the outlying suburban counties showed a greater
attraction to the economic structure variables, particularly employment in the
services, business services, and FIRE. These FACCs focus on the maximization Fl
of proﬁt, thus the location in economically prosperous areas.

At both the regional and metropolitan scales, the third and fourth factors,
competition and FACC ownership, were not easily measured. However,
available data did show FACCs to be positively correlated with both the
hospital bed/ population ratio (central cities) and the physician/ population
ratio (suburban counties). These resuls were expected since hospital owned
FACCs are more likely to be in central cities (where hospitals are) while
physician-owned F ACCs are more likely to be in the suburbs (where
physicians are). In addition, it seems apparent that the relationships between
FACCs and the ﬁrst two factors, population and economic structure, also relate
back to the ownership of FACCs. F ACCs in central cities (hospital-owned,
service—oriented) are related to population characteristics. FACCs in the
suburbs (physician and corporate owned, proﬁt-oriented) are related to
economic structure. Based on these relationships, it can be said that FACC
location patterns are in a very real sense most closely correlated with

ownership. Ownership and the consequent market strategy ultimately

105
determine location by determining which other characteristics of the area

(population, economic structure) are important. Ownership even determines
the effect of competition (factor three) on FACC location patterns.

FACCs have come to exist for two main reasons: 1) in response to patient
desires for convenience and an alternative to the hospital emergency room for
urgent care (demand) and 2) because corporations, and (later) other health care
providers, saw that FACCs could be proﬁtable. There are, of course, other
beneﬁs that support the existence of FACCs (such as the opportunities for
physician employment), but demand and proﬁt continue to be the primary
reasons for the FACCs' success. It is no surprise then that other health care
providers, physicians, and hospitals, would take advantage of the perceived
lucre of FACCs. By becoming FACC owners, physicians and hospitals could
be seen as responsive to patient needs and they could generate revenues if
successful. The FACC, more directly than traditional health care providers, is
both service and business.

The survey of Michigan FACCs allowed a site level analysis of the four
general factors expected to inﬂuence FACC location patterns. Although the
factors of demand and economic structure were not directly addressed in the
survey, the resuls did lead to conclusions regarding the importance of these
two factors, as well as the importance of competition and ownership. The
survey resuls indicated that population characteristics (demand) and economic
structure (in the form of linkages and agglomeration economies) were

important to FACC location. FACCs show a variety of linkages with other

106
health care providers (as well as with other businesses). FACCs employ

physicians, nurses, and other medical professionals. They also use business
services such as accountans, tax preparation services, and lawyers. These
linkages mean that FACCs are locating in metropolitan areas where there are
other health care professionals and other necessary services.

Factor three, competition, also plays an important role. At the site level,
FACCs do tend to locate away from their competitors, especially other FACCs.
Although no statistically signiﬁcant relationships between ownership and
location were found in the survey, the importance of ownership on FACC
geography can not be overlooked. This study has been able to infer the
importance of ownership in affecting the relevance of other factors such as
demand, economic structure, and competition.

One of the main objectives of this dissertation was to understand the role
of the FACC in the American health care system. The survey provided the
ideal format from which to obtain this objective. The survey resuls indicated
that FACCs are often owned by physicians and hospitals. In addition, FACCs
refer patients to physicians and hospitals; FACC physicians often work in
private practice, have hospital appointmens, and are even employed by other
clinics. Many FACCs have contracs with HMOs and PPOs. Given these
relationships, it can be said that FACCs are not merely competitors of
traditional providers, but are interdependent with physicians and hospitals.

In conclusion, it can be said that the determinans of FACC location,

particularly at the regional and metropolitan scales, are similar to those of

 

107
other services, including health care providers. The factors which affect the

location of service and retail businesses, such as demand, economic growth,
and competition, also inﬂuence the geographical patterns of FACCs. The
literature on FACC location in fact states that these centers locate according to
"feasibility studies" and "market analysis techniques" much as a branch bank or
retail outlet store might do. Although it could not be directly conﬁrmed in
these analyses, it can be inferred that what ses the FACC apart from other
services is the affect of ownership patterns and the market strategy associated
with these patterns. In particular, the involvement of other health care
providers such as physicians and hospitals in the FACC industry dictates
which aspecs of the "feasibility studies" (if any) will be appropriate for the
placement of a physician or hospital-owned FACC.

This study has shown that FACCs are businesses, with the needs of
businesses, as evidenced by their attraction to areas with higher demand,
economic growth and agglomeration economies. However, the FACC is
unique in that it is a health care provider and as such it is a representative of
our changing health care system. Many social, economic, and institutional
transformations have allowed FACCs to exist and ﬂourish. Society's desire for
convenient medical care, the rising coss of private practice for physicians, and
the proﬁs expected from health care (especially during the 19805) all
necessitated a response from traditional health care providers (and others)
interested in keeping up with modern health care delivery. In addition, the

FACC as health care provider has demanded the attention of institutions such

 

108
as the American Hospital and Medical Associations (AHA, AMA), insurance

companies, and of course, state and local governmens. These entities are
concerned with FACC licensure and accreditation and with making sure that
the public is aware of the services that FACCs provide (they are not equal to
hospital emergency rooms). Thus, along with demand and economic factors
which affect all businesses, FACC location studies must also consider less
tangible inﬂuences such as government regulations and the involvement of

other members of the health care system in the FACC industry.

Future Research

The resuls presented here lead to many other research avenues. For
example, a smaller scale analysis (e.g., census tracs) may be necessary to
understand exactly how the population ﬁs into the FACC location decision,
particularly in non-central city areas. Similarly, data on the economic structure
in central cities would allow for better testing of the hypothesis that
population factors are more relevant than the economic structure for FACCs in
central cities. Other site-level characteristics such as commercial zoning,
building type and availability, and ofﬁce rent or overhead coss are also likely
to be important in FACC site selection. In addition, a complete study of FACC
location (at all scales) would need to include information on ownership (the
extent of involvement of physicians and hospitals), and on other institutional

and economic factors such as government regulations, the support or

109
opposition of local physicians and hospitals to FACCs, and the economies of

physicians' practices.

Implications of F ACCs for patiens and physicians, although mentioned,
were not directly addressed in this research. For example, the issue of
accessibility to FACC services would be an appropriate topic for future
research. Although there is no concrete evidence that FACCs shun the poor or
uninsured (they do accept Medicare and Medicaid), it would be necessary to
study the question more from the patient's point of view (perhaps through
another survey) to really understand the problem of access. In addition, the
very concept of the F ACC raises questions, not only of ﬁnancial access, but of
temporal access and of how the use of this service (primarily for episodic care)
affects the long-term health outcomes of patiens. It would also be interesting
to determine the role of physicians in the FACC. Are they "merely"
employees? How are medical decisions made? How are proﬁts divided? The
answers to these and other questions are important to understanding the
future of physicians and other health care providers if physician employment
becomes a more common trend.

An equally worthwhile avenue of study would be to return to this subject
in ﬁve years to discover what changes have occurred with FACCs and their
locations. Have suburban locations become dominant? Are F ACCs more or
less integrated into the health care system than they are today (particularly in
the light of coming health care reform measures)? How has the use of FACC

services affected patiens and who is most likely to use the clinics? Do FACCs

110

continue to show the same relationships to other service and non-service
industries and to other health care providers?

The future of the FACC may, in fact, depend on is interaction with other
providers, in particular, with HMOs and HMO-type organizations. The
Clinton Administration's proposals for health care reform seem to emphasize
regional health alliances. Apparently, these alliances would function similarly
to an HMO in that every person would be assigned to an insurance plan
(health alliance). As the plan is currently propdsed, physicians would be

 

driven out of the fee-for-service type of private practice since they would not
be allowed to charge fees which were above the fees set by the health
alliances. In essence, if a physician can not charge more to see a patient
ouside his/ her alliance (i.e., HMO), there would be no incentive to see other
patiens and therefore, no private practice in the traditional fee-for-service
sense. Obviously, it is unlikely that the Clinton proposal will be adopted in
this original form. However, for the sake of argument, consider the place of
the FACC under Clinton's health alliance plan. In an HMO-type system, the
FACC would be a suitable place for patients to obtain primary care. For
patiens, the FACC would be convenient (no appointmens or long waiting
times); for physicians, the FACC would provide a practice setting free from the
worries of malpractice insurance and ofﬁce and staff coss (all of which would
presumably be taken care of by the health alliance). The FACC would take the
place of physicians' ofﬁces and would perform the function of "gatekeeper,"

directing patiens to higher levels of care when necessary.

111
The FACC as gatekeeper plan would, of course, have is own problems.

For example, would FACCs really provide an ideal setting for primary care?
Their emphasis on episodic rather than preventative care, on convenient care
rather than meticulous and careful diagnosis, may in the long run be harmful
to patiens. There is currently no evidence that FACC care is of poor quality.
However, today's FACC is used mainly for episodic care - minor injuries and
illnesses. We have yet to see if these clinics could handle all types of primary
care patiens would need in lieu of private physicians. Also, physicians may
ﬁnd employment in FACCs less than ideal. In exchange for a salary and
beneﬁs, they may lose income, autonomy, and professional status. However,
the knowledge and expertise of physicians will likely keep them from
becoming totally subject to "management" whether it be corporations or
government.

The above scenario places FACCS in a pivotal position within the US.
health care system by actually having them replace private physicians in
primary care. In reality, the position of FACCs may be much more precarious.
In some ways, the FACC is a product of the drive for proﬁt and
"expansionism" which affected the health care industry in the 19805. Even
those FACCs owned by hospitals, presumably for social service reasons, were
(and are) expected to make a proﬁt for the hospital. Today, however, the
health care industry is experiencing cut-backs, hospital closings, and pressure
from both business and government to reduce coss. Many FACCs were

opened when capital was readily available and it was considered proﬁtable to

 

112

respond to the consumer's need for convenience in medical care. Now, these
clinics may be one of the ﬁrst services to be sacriﬁced if they are deemed
unproﬁtable. FACCs located in nonmetropolitan areas or inner cities may, in
fact, be at the greatest risk of closure if problems arise.

Whether the FACC becomes a prominent provider of primary care or
quietly disappears into health care history, this study has provided insight into
is location patterns and relationships with health care providers and other
businesses which should be of interest to any one concerned with our

changing health care system.

J

LIST OF REFERENCES

LIST OF REFERENCES

American Hospital Association. 1991. American Hospital Association Guide to
the Health Care Field. Chicago: American Hospital Association.

American Medical Association. 1992. Physician Characteristics and Distribution
in the U.S.. Chicago: American Medical Association.

Berliner, HS. and Burlage, RK. 1987. The walk-in chains: the proprietarization
of ambulatory care. International Journal of Health Services 17(4): 585-94.

Bohland, J. 1984. Neighborhood variations in the use of hospital emergency
rooms for primary care. Social Science and Medicine 19: 1217-26.

Brown, LA. 1981. Diffusion Innovation: a New Perspective. London: Methuen.

Brown, S. 1989. Retail location theory: the legacy of Harold Hotelling. Journal
of Retailing 65 (4): 450-70.

Burns, LA. and Ferber, MS. 1985. Public policy issues. In Freestanding
Emergency Centers, ed. P.M. Friend and J.M. Shiver, Rockville, MD: Aspen
Publications, 17-30.

Chapman, K and Walker, D. F. 1991. Indusz Location: Principles and
Policies. Cambridge, Massachusetts: Basil Blackwell, Inc.

Daniels, PW. 1985. Service Industries: a Geographical Appraisal. London:
Methuen.

Daniels, P. W. 1982. Service Industries: Growth and Location. London:
Cambridge University Press.

Derber, C. 1984. Physicians and their sponsors: the new medical relations of
production. in Issues in the Political Economy of Health Care J.B. McKinlay,
ed. New York: Tavistock Publications.

DeVise, P. 1973. Physician migration from inland to coastal states: antipodal
examples of Illinois and California. Journal of Medical Ed ucatian 48, 141-51.

113

 

114

Dillman, DA. 1978. Mail and Telephone Surveys: the Total Design Method.
New York: John Wiley and Sons.

Dorsey, IL. 1969. Physician distribution in Boston and Brookline, 1940 and
1961. Medical Care 7 (6):429-40.

Elesh, D. and Schollaert, P. T. 1972. Race and urban medicine. factors affecting
the distribution of physicians 1n Chicago. Journal of Health and Social
Behavior 13: 236-50.

Ermann, D. and Gabel, J. 1985. The changing face of American health care.
Medical Care 23 (5): 401-20.

Ernst, KL. and Yett, DE. 1985. Physician Location and Specialty Choice Ann
Arbor, MI: Health Administration Press.

Friedman, E. 1983. Declaration of independence: more and more physicians
ﬁnd salaried status attractive. Hospitals July 1: 73-80.

Friend, PM. and Shiver, J.M. (eds). 1985. Freestanding Emergency Centers: A
guide to Planning, Organization, and Management. Rockville, MD: Aspen
Publications.

Gesler, W.M., and Meade, MS. 1988. Locational and population factors in
health care seeking behavior in Savannah, Georgia. Health Services Research
23: 443-62.

General Accounting Ofﬁce (GAO). 1990. Health care: limited state efforts to
assure quality of care outside hospitals. U.S. GAO report, January.

Gober, P., and R. J. Gordon. 1980. Intraurban physician location: A case study
of Phoenix. Social Science and Medicine 14Dz407-17.

Goe, W.R. 1990. Producer services, trade, and the social division of labor.
Regional Studies 24 (4): 327-42.

Gordon, R. 1., Meister, IS, and Hughes, RJ.. 1992. Accounting for shortages of
rural physicians: push and pull factors. In Health in Rural North America: The
Geography of Health Care Services and Delivery, ed. W. M. Gesler and T. C.
Ricketts, New Brunswick, NJ: Rutgers University Press, 153-78.

Gray, EH. 1986. For-Profit Enterprise in Health Care. Washington DC:
National Academy Press.

 

 

 

115

Grumet, G.W. 1989. Health care rationing through inconvenience: the third
party's secret weapon. New England Journal of Medicine 321 (9): 607-12.

Guzick, D5. and Jahiel, RI. 1976. Distribution of private practice offices of
physicians with speciﬁed characteristics among urban neighborhoods. Medical
Care 14 (6): 469-88.

Hall, BB. 1983. Neighborhood differences in retail food stores: income versus
race and age of population. Economic Geography 59: 282-95.

Hansen, N. 1990. Do producer services induce regional economic
deveIOpment? Journal of Regional Science 30 (4): 465-76.

Harrington, J.W. and Lombard, JR. 1989. Producer-service ﬁrms in a
declining manufacturing region. Environment and Planning A 21: 65-79.

Hartshorn, TA. and Muller, RC. 1989. Suburban Downtowns and the
transformation of metropolitan Atlanta's business landscape. Urban geography
10 (4), 375-95.

Holloway, S. and Wheeler, JD. 1991. Corporate headquarters relocation and
changes in metropolitan corporate dominance, 1980-1987. Economic Geography
67: 54-74.

Hone, L. 1985. History and evolution of the FEC. In Freestanding Emergency
Centers, ed. P.M. Friend and J.M. Shiver, Rockville, MD: Aspen Publications,
20—36.

Hosking, J.B., Magill, JR, and Maurer, RP. 1985. Site selection and evaluation
criteria. In Freestanding Emergency Centers, ed. P.M. Friend and J.M. Shiver,
Rockville, MD: Aspen Publications, 63-78.

Hoskins, B. 1985. Freestanding emergency centers. Missouri Medicine 81 (5):
266-7.

Illinois Medical Journal. 1983. Freestanding emergency centers in Illinois.
September: 208-12.

Johnston, R]. 1991. Multivariate Statistitical Analysis in Geography. New
York: Longrnan Scientiﬁc and technical.

Jones, JP. and Kodras, J.B. 1990. Restructured regions and families: the
ferninization of poverty in the US. Annals of the Assoc. of American
Geographers 80 : 163-83.

116

Joseph, A.E. and Phillips, DR. 1984. Accessibility and Utilization:
Geographical Perspectives on Health Care Delivery. New York: Harper and
Row.

Kaplan, KS, and Leinhardt, S. 1973. Determinants of physician ofﬁce location.
Medical Care 11: 406-15.

Kellerman, A. 1985. The suburbanization of retail trade: a US. nationwide
view. Geoforum. 16 (1): 15-23.

Kirn, T.J. 1987. Growth and change in the service sector of the U.S.: a spatial
perspective. Annals of the Association of American Geographers. 77 (3): 353-
72.

Knaap, G]. and Blohowiak, D. 1989. Intraurban physician location: new
empirical evidence. Medical Care 27 (12): 1109-16.

Koska, M. T. 1991. Community participation drives rural recruiting campaigns.
Hospitals 65:26-28.

LaTour, S.A., et al. 1985. Strategic marketing: a new methodology. In
Freestanding Emergency Centers, ed. P.M. Friend and J.M. Shiver, Rockville,
MD: Aspen Publications, 37-62.

Lowell-Smith, EC. 1994. Alternative forms of ambulatory care: implications for
patients and physicians. Social Science and Medicine, 38 (2): 275-83.

Lowell-Smith, E.G. 1993a. Regional and intrametropolitan differences in the
location of freestanding ambulatory surgery centers. Professional Geographer,
45: 398-407.

Lowell-Smith, E.G. 1993b. Patterns of dentist location: Lansing, Michigan, 1950-
1991. Social Science and Medicine, 36 (4): 569-73.

Mattingly RF. 1991. The changing location of physician ofﬁces in
Bloomington-Normal Illinois, 1870-1988. Professional Geographer. 43: 465-74.

McKinlay, J.B. and Arches, J. 1985. Towards the proletarianization of
physicians. International Journal of Health Services 15 (2): 161-95.

McKinlay, J.B. and Stoeckle, JD. 1988. Corporatization and the social
transformation of doctoring. International Journal Health Services 18 (2): 191-
205.

117

McLafferty, S. 1986. The geographical restructuring of urban hospitals: spatial
dimensions of corporate strategy. Social Science and Medicine 23 (10): 1079-86.

Miller, K.E., Lairson, D.R., Kapadia, A.S., and Kennedy V.C.. 1985. Patient
characteristics and the demand for care in two freestanding emergency centers.
Inquiry 22: 418-25.

Morrill, R.L., Earickson, R.J., and Rees, P. 1970. Factors inﬂuencing distances
traveled to hospitals. Economic Geography 46: 161-71.

Morrill, RL. 1987. The structure of shopping in a metropolis. Urban
Geography. 8: 97-121.

Muller, PO. 1986. Transportation and urban fornustages in the spatial
evolution of the American metropolis. in The Geography of Urban
Transportation. S. Hanson (ed). New York: Guilford Press, 24-48.

Mullinax, CW. 1980. Urgency or Emergency? Ohio State Medical Journal
October: 622-5.

NAFAC. 1990. ACC Factor 1]]: the ambulatory care industry in 1990. The third
statistical report of the ACC industry, conducted for the National Association
for Ambulatory care by First Source Research.

Nickel, P., and Wertheimer, A.J. 1979. Factors affecting consumers' images and
choices of drugstores. Journal of Retailing 55: 71-78.

Nordstrom, RD, and Steinke, E. 1987. How do patients select a doctor?
Health Marketing Quarterly 4: 37-46.

Noyelle, T. 1983. The implications of industry restructuring for spatial
organization in the United States. in Regional Analysis and the New
International Division of Labor. Applications of a Political Economy
Approach, ed. F. Moulaert and P.W. Salinas. Boston: Kluwer Nijhoff Pub.: 113-
33.

O’Connor, SJ. and Lanning, J.A. 1992. The end of autonomy? Reﬂections on
the postprofessional physician. Health Care Management Review 17 (1), 63-72.

O'hUallachain, B. and Reid, N. 1991. The location and growth of business and
professional services in American metropolitan areas, 1976-1986. Annals of the
Association of American Geographers 81 (2): 254-70.

O'hUallachain, B. 1989. Agglomeration of services in American
metropolitan areas. Growth and Change Summer: 37-49.

118

Olsen, L.M., and Lord, JD. 1979. Market area characteristics and branch bank
performance. Journal of Bank Research 10: 102-10.

O'Mullane, D. M., and Robinson, M. E. 1977. The distribution of dentists and
the uptake of dental treatment by schoolchildren in England. Community
Dental and Oral Epidemiology 5: 156-59.

Parsons, R.J., Willson, W.I(., and Youkstetter, DR. 1987. Applied feasibility
analysis for urgent care centers. Alternative Delivery Systems, ed. N.
Goldﬁeld and SB. Goldsmith, Rockville, MD: Aspen Publications, 19-32.

Phelps, CE. 1992. Health Economics. New York: Harper Collins Publishers.

Powills, S. 1985. Size of FEC market a matter of opinion. Hospitals May 16,
1985: 43-50.

Price, D.G. and Blair, A.M. 1989. The Changing Geography of the Service
Sector. London: Belhaven Press.

Pyle G. 1989. Physician ofﬁce locations within Charlotte, North Carolina.
Southeastern Geographer. 29: 118-35.

Relman, AS. 1991. Shattuck lecture - the health care industry: where is it
taking us. New England Journal of Medicine 325 (12), 854-60

Riddel, J.B. 1970. On structuring a migration model. Geographical Analysis 2:
403-09.

Rizos, J. P. Anglin, Grava-Gubins, I. and Lazar, C.. 1990. Walk-in clinics:
implications for family practice. Canadian Medical Association Journal 143:
740-43.

Rosenberg, M. W. 1984. Physician location behavior in metropolitan Toronto.
Canadian Geographer 28 (2): 156-70.

Rylko-Bauer, B. 1988. The development and use of freestanding emergency
centers: a review of the literature. Medical Care Review 45 (1): 129-63.

Shannon, G.W., Cromley, E.I(., and Pink, J.L. 1985. Pharmacy patronage
among the elderly: selected racial and geographical patterns. Social Science
and Medicine 20: 85-93.

Shumsky N .L., Bohland J. and Knox P. 1986. Separating doctors' homes
and ofﬁces: San Francisco, 1881-1941. Social Science and Medicine 23: 1051-57.

 

 

119

US. Bureau of the Census. 1990. Census of Population and Housing.
Washington DC: US. Government Printing Ofﬁce.

US. Department of Commerce. 1990. County Business Patterns. Washington
DC: US. Government Printing Ofﬁce.

US. Department of Commerce. 1992. Statistical Abstract of the United States.
Washington DC: US. Government Printing Ofﬁce.

Watson, C]. 1980. The relationship between physician practice location and
medical school area: an empirical model. Social Science and Medicine 14D: 63-
69.

Weiss, K. 1982. Corporate medicine: what's the bottom line for physicians and
patients? The New Physician 9: 19-25.

Wheeler, J0. and Muller, PD. 1986. Economic Geography (second edition).
New York: John Wiley and Sons.

Wheeler, JD. 1990. The new corporate landscape: America's fastest growing
private companies. Professional Geographer 42 (4): 433-44.

Williams, M.J. and Stukenberg, RH. 1985. Feasibility studies for freestanding
emergency centers. in Freestanding Emergency Centers, ed. P.M. Friend and
J.M. Shiver, Rockville, MD: Aspen Publications, 91-106.

Yunker, R., Levine, M., and Sajid, A.. 1986. Physicians practicing in Texas'
freestanding emergency centers. Texas Medicine 82: 22-5.

Yunker, R., Levine, M., and Sajid, A.. 1985. Free-standing emergency centers
and the patient population of family physicians. Journal of Family Practice 21
(1): 63-9.

 

 

 

APPENDICES

Appendix A

MSAs Used in Analysis

Birmingham AL

Mobile AL

Anchorage AK

Little Rock-North Little Rock AR
Phoenix AZ

Tucson AZ

Anaheim-Santa Ana CA
Bakersﬁeld CA

Fresno CA

Los Angeles-Long Beach CA
Oakland CA
Oxnard-Ventura CA
Riverside-San Bernadino CA
Sacramento CA

San Diego CA

San Francisco CA

San Jose CA

Stockton CA
Vallejo-Fairﬁeld—Napa CA
Denver CO
Bridgeport-Milford CT
Hartford CT

New Haven-Meriden CT
Washington DC

Wilmington DE

Ft Lauderdale-Hollywood-Pompano Beach FL
Jacksonville FL
Miami-Hialeah FL

Orlando FL

Tampa-St Petersburg FL
West Palm Beach-Boca Raton-Delray Beach FL

120

 

 

 

121

Atlanta GA

Honolulu HI

Boise City ID

Chicago IL

Lake County IL
Indianapolis IN
Gary-Hammond IN

Des Moines IA

Wichita KS

Louisville KY-IN

Baton Rouge LA

New Orleans LA
Portland ME

Baltimore MD

Boston MA

Springﬁeld MA

Detroit MI

Grand Rapids MI
Minneapolis-St Paul MN
Jackson MS

Kansas City MO-KS

St Louis MO-IL

Billings MT

Omaha NE-IA

Las Vegas NV

Nashua NH
Bergen-Passaic NJ

Jersey City NI
Middlesex-Somerset—Hunterdon NJ
Monmouth-Ocean NJ
Newark NJ

Albuquerque NM
Albany-Schnectady-Troy NY
Buffalo NY
Nassau-Suffolk NY

New York NY

Rochester NY

Syracuse NY
Charlotte-Gastonia-Rock Hill NC-SC
Greensboro-Winston—Salem-High Point NC
Raleigh-Durham NC

F argo-Moorhead ND-MN
Akron OH

Cinicinnati OH
Cleveland OH

 

122

Columbus OH
Dayton-Springﬁeld OH
Toledo OH
Youngstown-Warren OH
Oklahoma City OK

Tulsa OK

Portland OR

Allentown PA
Harrisburg-Lebanon-Carlisle PA
Philadelphia PA-NJ
Pittsburgh PA
Scranton-Wilkes Barre PA
Providence RI

Charleston SC

Columbia SC
Greenville-Spartanburg SC
Sioux Falls SD

Knoxville TN

Memphis TN-AR-MS
Nashville TN

Austin TX

Dallas TX

Ft Worth TX

El Paso TX

Houston TX

San Antonio TX

Salt Lake-Ogden UT
BurlingtonVT

N orfolk-Virginia Beach-NeWport News VA
Richmond VA

Seattle WA

Tacoma WA

Charleston WV
Milwaukee WI
Cheyenne WY

 

Appendix B

Reconstituted Regression Coefﬁcients

Steps in determining reconstituted partial regression coefﬁcients:

1. Identify independent and dependent variables.

 

2. Standardize dependent variable.

3. Submit ALL independent variables to a Principal Components Analysis
(rotated using varimax procedure) and extract COMPONENT SCORES and
COMPONENT SCORE COEFFICIENT MATRD(.

4. Regress COMPONENT SCORES on standardized dependent variable
and extract STANDARDIZED REGRESSION COEFFICIENTS for each

component.

5. Using matrix multiplication, multiply the FACTOR SCORE COEFFICIENTS
MATRD( by the vector of STANDARDIZED REGRESSION COEFFICIENTS:

The results of this multiplication are the RECONSTITUTED REGRESSION
COEFFICIENTS.

 

123

Appendix C

Sample Survey of Michigan FACCs

These ﬁve questions ask about your clinic's location.
1. Please indicate the number which bat describes your location: (circle number)
1. Inside city limits
2. In a suburban area
3. In a rural or nonmetropolitan area
2. In what type of structure is your clinic located? (circle number)
1. Freestanding building 2. Non-hospital medical building 3. Strip mall

4. Enclosed mall 5. Adjacent] attached to hospital 6. Multi-use building

3. Please describe parking at your clinic: (circle one answer in each column)

1.Free 1.Yourownparkinglot

2. Validated parking 2. A shared lot

3. Pay parking 3. Ina nearby parking lot
4.8treetparking

4. Please describe the type of street your clinic is located on: (circle number)
1. 6/8-lane highway
2. 4-lane highway

3. 2-lane road

5. Is your ofﬁce located at a major intersection? (circle number)
1. Yes

2. No

124

 

 

125

The next two questions ask about the reasons tor your clinic's present location at both the METROPOLITAN and SITE levels.

I. On a scale nnp'ng from EXTREMELY IMPORTANT to UNIMPORTANT, how would you evaluate the following items in terms of their inﬂuence
on your dinic's location at the METROPOLITAN (city) level (for example, why did you locate your clinic in Detroit or in the Detroit areal)? (circle
number in nch row)

Extremely Very Somewhat Slightly
Important Important Important Important Unimportant

I. Personal reman- (e.g. climate of the area. autertitid, city is home) I 2 3 4 5
ZPopdoﬁmdunctaistiohgincomageme. etc) I 2 3 4 5
3. The availability of labor (cg. physidans. nurses, clerical) I 2 3 4 5
4. Competition (eg. otha' clinic. physidam. hospitals) 1 2 3 4 5
Slaplandcosthctorangaxaregulatimondinia) l 2 3 4 5
SDI-musta'parent'fadlityk.‘ headquanamhapital) I 2 3 4 5

 

2.. On a scale ranging from EXTREMELY IMPORTANT to UNIMI’ORTANI, how would you evaluate the following items in terms of their inﬂuence
on your clinic’s location at the SITE level (the exact location)? (circle number in each row)

Extremely Very Somewhat Slightly

Important Important Important Important Unimportanl
I. To be near a target population I 2 3 4 5
2. Near to other health care pmtasiomls (including physicians) I 2 3 4 5
3. Site umerverl by competition I 2 3 4 5
4. Legal and cost fedora (zoning regulations, taxes) 1 2 3 4 5
5.Pltysialaoc-sibilityolyour specilicsite l 2 3 4 5
6. Type of swromttling buinases, retail stores I 2 3 4 5

7. Otha site dtaracteristia (parking. visibility, size, etc) I 2 3 4 5

126

These qwtions ask about your clinic's duracteristics.

1. Select the amwer which best identiﬁes the ownership of your clinic: (circle)
1. Hospital owned 2. Non-physician owned 3. Single-physician owned
4. Mum-physician owned
2. Isyourclinicpartofachainofclinics?
1. Yes (If yes, how many clinics are in the chain, including your own?___)
2. No

3. When did your clinic open (year)?

 

4.Whatareyourhoursofoperation?

Weekdays: from __ (AM/PM) to ______(AM/PM)
Saturday: from . (AM/ PM) to (AM/PM)
Sunday: from (AM/PM) to (AM/ PM)

 

 

5. Does your clinic normallyrequire appointments?

6. What is the approximate size of your clinic (in square feet)?

 

 

 

 

 

 

 

7. How many total physicians do you employ full time?_______ Part-time?

8. How many total nurses do you employ full time? Part-time?

9. How many total x-ray techniciam do you employ full time? Part-time?
10. How many total medical assistants do you employ full time?____ Part-time?
11. How many other (non medical) staff do you employ full time?______ Part-time?

12.Pleasehtdicatehowmanyofeachofdtesestaffmembersamemployedonatypicalshift
Physicians _____Nurses ___X-raytechnicians _____Medicalassistants

Non-medical staff

127

13. Please circle all services that you offer at present:
1. Urgent care 2. Primary care 3. Physical therapy
4. Gynecology 5. Obstetrics 6. Cancer saeening
7. Psychology/ psychiatry 8. Diet control 9. Orthopedics
10. Pediatrics 11. Dentistry 12. Drug dispensing
13. Drug/alcohol abuse treatment 14. Dermatology 15. Laboratory services

 

16. Industrial/occupational medicine 17. Sports medicine 18. Other

14. How are your physicians compensated? (circle all that apply)
1. Salary only
2. Salary and other compensation
3. Other compensation only

4. Physicians are paid directly by patients/ insurance

15. Please indicate your average annual revenues, expenses, and proﬁts:

1. Revenue

 

2. Expenses

 

3. Proﬁts
16. What is your clinic's bad debt as a percentage of total revenues?

17. Please indicate the approximate percentage of your revenue that comes from the following

 

 

 

sourcm
1. Patient self-pay _____% 5. Private insuan 94.
2. Industrial accounts _____% 6. [PO/FPO fees %
3. Worker's comp. ___% 7. Medicare
4. BlueCross/ Shield ____% 8. Medicaid %

 

5. Private insurance _____%

 

 

128

18. Do you accept Medicaid or Medicare payments? (circle number)

MEDICAID: MEDICARE:
1. Yes 1. Yes
2. No 2. No

19. How many patients does your clinic see (on average) per day?

This section asks questions about your clinic's relationships, competition, and linkages.

1. Do any of the physicians employed by you also work tor/at: (drcle all that apply)

1. A hospital

2. An HMO or FPO

3. His/ her own private practice

4. Another ACC

2. Please scla't the answer which bot docrrbes the location where the lollow'mg services/supplies are provitlul lor your clinic. (circle munber in each row)

I'rovidcil in Provided by Provided by l’rovidnl by Provided by Do not use

clinic Iliaulqiatrtcrs lirm ha‘ntnl m linn lin'ntul in liun lowlisl outside service
central city metropolitan arm metro'mlitan arcs:

Medical supplia l 2 3 4 5 6
Other supplies I 2 3 4 5 6
Payroll l 2 3 4 5 6
Tax preparation I 2 3 4 5 6
Astounding/atrium I 2 3 4 5 6
sonic. I 2 3 4 5 6
Iinqnloyment per-sound I 2 3 4 5 6
Advertising I 2 3 4 5 6
“Initials management I 2 3 4 5 6

T

 

129

3. Does your clinic receive patients referred from nearby physicians? (circle number)

1. Yes, Often 2. Yes, Sometimes 3. Yes, Rarely 4. No, Never

4. Does your clinic refer patients to local physicians? (circle number)

1. Yes, Often 2. Yes, Sometimes 3. Yes, Rarely 4. No, Never
5. Does your clinic refer patients to local hospitals? (circle number)

1. Yes, Often 2. Yes, Sometimes 3. Yes, Rarely 4. No, Never
6. Does your clinic have a contract with a local HMO or PPO? (circle number)

HMO. PPO.
1.. Yes 2. No 1. Yes 2. No

7. Please estimate the size (radius in miles) of your market area:

8. Pleasemteﬂtefoﬂowmgcompeﬁtonintemuofﬂtemmpeﬁﬁmfacedbyymudhuc Ratethecompetitorsfroml
t05,withlbeingleastcompetitivearu15beingmostcompetitive. (circlenumberineachrow)

1. Other ACCs 1 2 3 4 5
2. Hospitals 1 2 3 4 5
3. Physicians' ofﬁce 1 2 3 4 5
4. HMOs 1 2 3 4 5
5. PPO l 2 3 4 5

9. Pleueestimatethedistanca(mnuIa)betweenyourdhucandtheNEARESTofeachofthesecompeﬁtas. Ifyou
donotcompetewiﬂtoneofthesehealﬂtcarepmvidasJeaveﬂteﬁneblank:

 

 

 

 

1. Other ACC 4. HMO
2. Physician's ofﬁce 5. PPO
3. Hospital

 

Thank you for taking the time to complete this survey. Any additional comments you may have will be greatly
appreciated.

 

Appendix D

Summary of Survey Results

"‘ Results are based on 33 returned surveys except where otherwise indicated.
(15% response rate)

These questions are concerned with clinic location and reasons for the location.
1. Clinic Location:
39% inside city limits 52% in suburbs 9% in nonmetropolitan areas
2. Clinic is located in --- Structure:
64% in freestanding building 12% in nonhospital medical building 6% in strip mall
9% adjacent/ attached to hospital 9% in multi-use bldg 0% in enclosed mall
3. Parking:

100% : free parking 70% own parking lot
30% shared parking lot

4. Street location:
12% 6-8 lane Highway 58% 4-lane Hwy 30% 2-lane road
5. Major intersection?

32% yes 68% no

6. Reasons for locating clinic at METROPOLITAN level:

1. Competition: 35% said this was extremely important

2. Nearness to parent facility: 26% said this was extremely important (19% also said it
was least important)

3. Population characteristics: 56% said this was very important

4. Competition: 36% said this was very important

130

 

 

131

Scores for metropolitan location: (lower scores = more important reason)

1. competition 1.8
2. Population chars. 1.9
3. Nearness to parent 2.7
4. personal reasons 2.7
5. legal/ cost factors 2.8
6. labor availability 2.8

7. Reasons for locating clinic at SITE level:

1.Nearness to target population: 48% said this was extremely important
2. Site unserved by competition: 52% said this was extremely important
3. Accessibility of site: 52% said this was very important
4. Site characteristics (e.g. visibility, parking): 52% said very important
5. Legal/ Cost factors: 23% said this was only slightly important

‘ no other answers had large numbers of responses

Scores for site level location: (lower scores = more important reason)

1. near to target population 1.7
2. site unserved by competition 1.8
3. accessibility of site 1.8
4. site characteristics 2.1
5. business/ retail structure 2.4
6. legal/cost factors 2.8
7. nearness to health care providers 2.8

These questions are concerned with clinic operations and services

1. Ownership of clinics:

52% hospital owned 3% non-physician owned 24% single-physician owned
21% multi-physician owned

2. 44% of respondent clinics belonged to CHAINS. The average chain included 6 clinics.
3. Age of clinics:

19% opened before 1980 55% opened before 1985 45% opened after 1985
4. Hours:

100% are open more than 8 hours/ day on weekdays

100% are open on Saturdays

70% are open on Sundays

5. 100% do NOT require appointments.

132

6. Average size of clinic: 5,114 sq. ft.

7. Employment. Answers are average number employed full time and part-time. (Numbers
in parentheses are the highest and lowest number employed in a clinic).

FU LL PART
physicians 2.7 (7,0) 4.0 (20,0)
nurses 1.2 (8, 0) 2.9 (13, 0)
x-ray tech. 1.3 (5, 0) 1.6 (6, 0)
med. asst. 2.8 (15, 0) 1.8 (8, 0)
non-med. staff 3.7 (25, 0) 2.0 (12, 0)

8. Employment per shift: average and highest, lowest employment

physicians 1 (6, 1)

nurses 1 (7, 0)
x-ray tech. 1 (2, 0)
med. asst. 2 (14, 0)

non-med. staff 2.8 (10, 0)

9. Services provided by clinics:

94% urgent care 66% primary care 52% physical therapy 41% gynecology
13% obstetrics 16% cancer screening 3% psychology/ psychiatry

19% diet control 23% orthopedics 42% pediatrics 0% dentistry

45% drug dispensing 3% drug/ alcohol abuse treatment 10% dermatology
64% laboratory services 85% indust./occup. med. 2% sports med.

10. Physicians' compensation:

45% salary only 43% salary and other compensation 3% other comp. only
10% paid directly by patients or insurance

11. 'm The results for questions 27, 28 are based on 15 responses.
Average revenues: $1,062,612 ave. expenses: $925,323 ave. proﬁts: $244,588

12. Average clinic's bad debt as a percentage of total revenues: 8.3%

13. “ This question is based on 21 responses. Average percentage of revenue from each
source (numbers in parentheses are the highest and lowest percentages reported by clinics).

Patient self-pay: 20% (60%, 5%)
Worker's Comp/ industrial accounts: 21.3% (5%, 100%)
Blue Cross/ Shield: 17.5% (44%, 0)

Private insurance: 20% (41%, 2%)
lPO/PPO/HMO fees: 12°/o (7 We, 39’s)
Medicare: 6.7% (10%, 2%)

Medicaid: 14.2% (60%, 0)

 

 

133
14. 73% of clinics will accept MEDICAID, 91% accept MEDICARE

15. The average clinic sees 55.5 patients per day.

These questions concern clinic relationships and linkages

1. Percent of Physicians employed elsewhere: (73% of total clinics have physicians with other
jobs)

45% hospital 3% HMO/FPO 36% own practice 36% other FACC

2. Receive patients from nearby physicians?

24% often 45% sometimes 19% seldom 12% never
3. Refer patients to local physicians?

61% often 36% sometimes 3% seldom 0% never
4. Refer patients to local hospital?

58% often 42% sometimes 0% seldom 0% never
5. Contract with a local HMO/FPO?

HMO PPO

69% YES 79% YES
6. Average radius of market area: 12.1 miles
7. Competition:

’ respondents were more likely to say they faced little competition from a competitor than

to say they were highly competitive with the competitor.

Most Competitive Least Competitive

other FACCs 16% 16%
hospitals 16% 30%
physicians 6% 6%

HMOS 10% 35%)
PPOs 6% 39%

Scores on competition: (higher scores = more competitive)

1. FACCs 2.9
2. Physicians 2.6
3. Hospitals 2.3
4. HMOs 2.1
5. PPOs 1.9

134

8. Average distance to nearest competitors:

9. The question on where services/ supplies are provided for your clinic is divided into two
categories: clinics which are part of CHAINS or have a PARENT organization and those which

other FACCs: 5.1 miles

HMO: 4.4 miles

PPO: 5.6 miles

are INDEPENDENT (non-chain).

CHAIN / PARENT

1
2
3
4
5.
6
7
8
9

. medical supplies

. other supplies

. payroll

. tax prep.

accounting / auditing

. legal services

. employment personnel
. advertising

. bus. mgmt.

Independents:

\OQVOUIQOJNH

. medical supplies

. other supplies

. payroll

. tax prep.

. accounting/ auditing

. legal services

. employment personnel
. advertising

. bus. mgmt.

38% (headquarters / parent)
43% (headquarters/ parent)
76% (headquarters / parent)
76% (headquarters/ parent)
81% (headquarters/ parent)
67% (headquarters/ parent)
67% (headquarters/ parent)
57% (headquarters/ parent)
57% (headquarters/ parent)

50% (firm in metro area)
58% (ﬁrm in metro area)
50% (firm in metro area)
58% (firm in metro area)
67% (firm in metro area)
75% (firm in metro area)
50% (own clinic)

42% (ﬁrm in central city)
75% (own clinic)

physician's office: .96 miles hospital: 4.7 miles

3% (ﬁrm outside metro area)
38% (ﬁrm in metro area)

43% (own clinic)

33% (ﬁrm outside metro area)
25% (firm outside metro area)
25% (own clinic)

25% (ﬁrm outside metro area)

25% (ﬁrm in metro area)
33% (ﬁrm in metro area)