Volume 78 Number 2, 2000
Community Accountability among
Hospitals Affiliated with Health
Care Systems
JEFFREY A. ALEXANDER,
BRYAN J . WEINER,
and MELISSA SUCCI
The shift from local, community-based organizations to more complex
delivery systems raises critical questions about the community orientation
and accountability of health systems and their affiliated organizations
(Berry and Seavey 1994). Unlike freestanding hospitals, whose
loyalty and accountability to the local community are clearly established
in their bylaws and are traditionally reflected in the composition of their
governing boards, systems are often characterized by multiple layers of
governance and health care operations comprising multiple vertically or
horizontally linked provider organizations. Under these arrangements,
power, resources, and services within the system reside not in a single
organization but across its multiple constituent organizations and in
its administrative authorities. This creates a situation whereby community
members, patients, their families, and advocates find direct access
to leadership more problematic than is the case in freestanding hospitals.
Further, the extralocal managing entities of health care systems
may have no particular ties to the community served by the hospital,
and their priorities or strategic goals may differ from those envisioned
by the local community as well (Relman 1996; Manning 1997; Scott
1997).
A fundamental question, then, is whether affiliation with a health
care system changes the traditional accountability of hospitals to the
local community and, if so, in what ways. The question is particularly
relevant to nonprofit hospitals, whose tax-exempt status rests primarily
on providing benefit to the community as a whole, rather than to
investors or parent corporations.
In this study, we examine how community accountability differs between
system-affiliated hospitals and freestanding hospitals and whether
community accountability differs in degree or form across hospitals affiliated
with different types of systems. We particularly emphasize hospital
governance, which traditionally has been the principal means by which
the community informs the hospital of local health needs and holds the
institution accountable for meeting them. Based on a national sample of
2,079 hospitals, we compare system-affiliated hospitals and freestanding
hospitals according to how they exercise community accountability in
the composition, structure, and activity of their governing boards. Further,
in a subsample of 587 hospitals, we perform similar comparisons
among system-affiliated hospitals on three potentially important system
attributes: size, geographic scope, and ownership.
Conceptual Framework
Accountability implies relationships of monitoring, control, and answerability
to superiors or public constituencies by administrators and
policy makers (Chisolm 1995; Fry 1995; Gamm 1996). Community
accountability, more specifically, applies these relationships to the interests
and needs of all individuals "residing within a reasonably circum-scribed
geographic area, in which there is a sense of interdependence and
belonging" (Griffith 1987; Sigmond 1995, 36). Among health-services
delivery organizations, community accountability has traditionally focused
on certain objectives, such as care for the poor, access to emergency
care, cost containment, and ensuring high-quality care. More recently,
the notion of community accountability has been broadened to include
intentions to improve overall community health, reduce the gap between
the health and well-being of rich and poor (or at least not permit it to
widen), and, in general, support linkages between health and community
development (Gamm 1996).
Historically, hospitals have relied on their governing boards as the
principal mode of exercising community accountability (Weiner and
Alexander 1998). Hospital boards represent an interface between the
hospital and its environment insofar as members are drawn from, and
have a primary affiliation with, organizations and agencies outside the
hospital (Middleton 1987). This dual alliance with the organization and
the outside locality not only permits direct exchanges of information and
resources across boundaries but also serves as a mechanism for holding
hospitals accountable to serving the needs of their communities (Pfeffer
and Salancik 1978; Griffith 1987).
The literature on hospital and health-system governance suggests that
boards can exercise community accountability in a variety of ways. Some
literature has highlighted board structure and composition, stressing
that hospitals have greater potential to exercise community accountability
when their boards comprise a mix of experts and community representatives
(Griffith 1987; Hash 1989; Arkus 1993; Gamm 1996). Other
writers have highlighted board activities, suggesting that hospitals are
more likely to exercise community accountability when their boards are
engaged in monitoring the local environment and gathering information
(Proenca 1998). Finally, a third research perspective stresses the role of
governing boards in creating links with community agencies to enhance
community health (American Hospital Association 1993; Proenca 1998;
Gamm 1998). Drawing from past literature, we have classified hospitals'
exercise of community accountability into three categories:
Each of these dimensions is discussed below.
Board Structure and Composition
The existence and legal form of the hospital board itself may either
advance or deter community accountability. We assume that the existence
of a legally constituted governing board at the hospital advances
community accountability by providing greater local control. A legally
constituted hospital governing board bears sole and ultimate responsibility
for hospital affairs and has nondelegable powers that are recognized
by statute and common law (Pointer and Ewell 1994). Hospitals that do
not possess legally constituted governing boards either have no board at
all or they have an "advisory board" that exists at the discretion of a health
care system or some other higher authority to which the hospital is accountable
(e.g., university, religious order, or unit of government). These
boards often have limited powers or responsibilities and exist primarily
to advise the management or board of the system or higher authority
(Pointer and Ewell 1994). Although such "advisory boards" provide a
structural interface between the hospital and the community and often
offer a channel for communicating local issues to the hospital, their lack
of fiduciary powers renders them less influential in advancing community
accountability than legally constituted governing boards. (Claxton,
Feder, Schachtman, et al. [1997] offer a similar argument.)
Beyond the existence and legal form of the hospital board, its composition
may also be critical to the exercise of community accountability.
Community representation on a board is believed to enhance its ability
to reflect the interests of the whole community and ensure that hospital
policies, strategies, and operations are consistent with those interests
(Scott 1997). Hence, community accountability is apt to be greater
among hospitals with higher percentages of "outside" board members––
that is, board members who hold neither management positions nor
clinical privileges at the hospital or health system. Similarly, community
accountability is greater among hospitals with higher percentages
of board members residing in the community served by the hospital.
Finally, we consider one structural attribute of governance that is
specific to system-affiliated hospitals: the role of the hospital board in
policy development and strategic decision making. We assume that the
exercise of community accountability is enhanced when the hospital
board has authority independent of the system board to make a wide
range of hospital policy and strategic decisions.
Information Gathering and Scanning
of Local Environments
Another area that we examine is how hospital boards exercise community
accountability through the activities executed by board members.
In particular, we are interested in the extent to which hospital boards
reflect a focus on community accountability through their information
monitoring, and more specifically, their monitoring, evaluating, and reporting
activities related to community health. The information that
a board routinely collects and monitors indicates the scope of its attention
and the priorities it holds for the institution that it governs
(American Hospital Association 1997). Similarly, the information that
a board uses to evaluate hospital performance and to issue reports to the
community reflects its views on the range of activities the hospital is
obliged to perform. Hence, there is greater potential to exercise community
accountability when the hospital board conducts information-monitoring
activities related to routine reviews of community health
measures and immunization rates. Further, community accountability is
enhanced when the hospital board undertakes the following initiatives:
uses community health standards, or benchmarks, to evaluate hospital
performance; evaluates the hospital chief executive officer (CEO) on community
health improvement; and disseminates reports to the community
on hospital quality and costs.
Collaborative Relationships with Other Agencies
to Enhance Community Benefits
Hospital boards can also exercise community accountability by direc-ing
hospital management to identify and respond to unmet community
health needs, particularly community-level risk factors that lead
to extensive health-services utilization, increasing health care costs, and
other adverse outcomes (Sigmond 1995; Schlesinger and Gray 1998).
Although no single organization possesses all the resources and competencies
necessary to improve community health, hospitals could increase
their leverage by working with other local organizations and agencies
to address underlying community risk factors (Weiner and Alexander
1998). Hospitals could take certain initiatives to exercise community
accountability:
System Characteristics and Community
Accountability
Comparisons of system-affiliated and freestanding hospitals offer a useful
starting point for investigating potential shifts in their community
accountability, but a narrow focus on such comparisons may mask critical
disparities in the degree and form of community accountability exhibited
by hospitals that operate within different types of health care systems. For
example, policy makers and community leaders have scrutinized real or
perceived differences in the degree of community accountability shown
by hospitals affiliated with either a not-for profit or an investor-owned
(IO) system. Beyond ownership, other system characteristics, such as size
or geographic scope, might also systematically affect the degree to which
system-affiliated hospitals hold themselves accountable to communities.
Larger systems, for example, may cede greater authority to local hospital
boards over community relations because their size precludes effective
centralized decision-making by the system board. Alternatively, however,
larger systems may not be as attentive to any given local community.
Because of their orientation to a wide array of organizations operating
in various localities, they are often forced to subordinate the interests of
any given community and to devise strategies and policies that benefit
the whole system.
In sum, we expect that system-affiliated hospitals and freestanding
hospitals may differ in the ways they exercise community accountability,
given differences in their key stakeholders (i.e., hierarchically
structured delivery systems or local community agencies) and their respective
goals. Further, we expect that hospitals affiliated with different
types of health care systems may differ in the ways they exercise community
accountability, thereby reflecting the different goals of their
respective key stakeholders.
Method
Data Sources
The American Hospital Association's (AHA) 1997 Hospital Governance
Survey provided most of the data for the study. The survey was mailed to
the CEOs of all U.S. general community hospitals, and it generated 2,079
responses (a 42 percent response rate). Compared with the population of
U.S. general community hospitals, respondents overrepresented secular,
not-for-profit hospitals (63 percent versus 59 percent), underrepresented
IO hospitals (10 percent versus 15 percent), overrepresented urban hospitals
(62 percent versus 56 percent), and underrepresented system-affiliated
hospitals (28 percent versus 35 percent). No significant differences
were observed for hospital bed size or for state and local government
hospitals. The AHA's 1997 Annual Survey of Hospitals provided information
on hospital community collaboration activities and data relevant
to measures of health care-system size, geographic scope, and ownership.
Measures
Table 1 describes our measures of hospital community accountability.
In addition, we constructed a measure to assess the degree of authority
of system-affiliated hospitals over operational and strategic activities.
This measure was the mean centralization score across 17 policy decisions:
hospital board-member appointment; executive compensation;
CEO evaluation; CEO appointment; capital budget; operating budget;
control of assets; operational quality improvement; clinical quality assurance;
community and government relations; bylaw changes; strategic
planning; mission revision; service additions and deletions; managed-care
contracting; credentialing of medical staff and delineation of their
privileges; and new company formation. Principal components factor
analysis with varimax (orthogonal) rotation suggested a single-factor
solution and thus a single scale (Æ D
:92).
University of Michigan; University of North Carolina at Chapel Hill;
Chapman University
HOSPITALS: REASSESSING ACCOUNTABILITY
ost-containment efforts by managed-care
firms, government agencies, and employer coalitions have precipitated
widespread consolidation in the health-care delivery
system (Shortell, Gillies, and Anderson 1994; Feldman, Wholey, and
Christianson 1999). Along with increasing numbers of closures and
mergers, the combining of hospitals and other delivery organizations
into horizontally and vertically integrated health care systems has perhaps
been the most visible manifestation of this trend (Ackerman 1992;
Burns, Bazzoli, Dynan, et al. 1997). The emergence of large, and sometimes
geographically dispersed, health systems has rendered the free-standing,
locally based hospital, once the mainstay of the U.S. health
delivery system, a less central player in today's health care environment
(Shortell, Gillies, and Devers 1995; Cerne 1994).
1. the presence and structure of a community-based hospital board
2. the existence of initiatives for monitoring and reporting information
about community health and hospital performance
3. collaboration with other local agencies to enhance community
benefits
System affiliation status was defined according to whether or not a hospital was owned, leased, or sponsored by a health care system. Network affiliations or contract management by a system were not components of our measure. In addition, we constructed three measures to compare hospitals affiliated with different types of systems. Health care systems were classified on the basis of three dimensions: size, geographic scope, and ownership. For system size, we constructed our small, medium, and large categories, using the terciles of the system size distribution for sample hospitals. Geographic scope comprises three categories: (1) single market; (2) multimarket; and (3) rural. Each system was assigned to one of the three categories, according to the number and location of their member hospitals operating inside or outside metropolitan statistical areas (MSAs) (Luke, Ozcan, and Olden 1995). Single-market systems manage owned, leased, or sponsored hospitals that operate within a single MSA. Multimarket systems manage owned, leased, or sponsored hospitals operating in two or more MSAs. Rural systems manage owned, leased, or sponsored hospitals that operate in non-MSA (i.e., rural) areas. System ownership consists of three categories:
- investor owned
- not-for-profit, religious
- other private, not-for-profit, secular
We did not include state and local government hospitals in our analysis because of the small number of hospitals in this category that were identified as components of health care systems.
Results
Table 2 compares freestanding and system-affiliated hospitals, based on
their measures of community accountability. System-affiliated hospitals
rated lower than freestanding hospitals on the first dimension of
community accountability: the legal form and structure of a hospital
board. Specifically, system-affiliated hospitals were less likely than free-standing
hospitals to possess a legally constituted, local governing board
(¬ D 328.01, p < :001). Notably, all of the system-affiliated hospitals
in our sample that did not operate a legally constituted governing board
did possess an advisory board. This finding is consistent with those of
other surveys, indicating that most systems have some sort of multi-tiered governance structure, consisting of a system-level board and a subsidiary-level advisory board or governing boards (American Hospital Association 1997; Governance Institute 1997; 1999). Although advisory boards possess no legally mandated authority or responsibility, they do provide a structural interface between the hospital and the community. However, they may be less influential in advancing community accountability than legally constituted governing boards.
Further, system-affiliated hospitals had lower proportions of outside (i.e., nonhospital) members on their boards than freestanding hospitals (t D 10:10; p < :0001) and higher proportions of members who were not residents of the local community in which the hospital operates (t D 6:05; p < :0001). These findings suggest weaker ties to the community. First, boards dominated by insider members (e.g., management or active physicians) may be more likely to reflect the strategic priorities of the hospital or system without the potentially countervailing influence of community representation. Second, the presence of members who are physically (and perhaps emotionally) removed from the community may reduce the probability that the board will identify and address community concerns. Both attributes may reduce the capacity of boards to exercise community accountability.
Comparisons of hospitals across attributes associated with the second dimension of community accountability, information-monitoring activities, suggested that system-affiliated hospitals came out ahead. Specifically, system-affiliated boards were more likely than freestanding hospital boards to perform certain types of evaluation on a routine basis:
- evaluate the hospital CEO on community health improvement (¬ D 5:16; p < :05), review community health status measures (¬ D 8:61, p < :01)
- routinely review immunization rates (¬ D 6:24; p < :05)
- use performance benchmarks based on community health standards (¬ D 28:60; p < :001)
- disseminate reports on costs and quality to the community (¬ D 11:42; p < :001)
To the extent that the information-monitoring and processing activity of a board reflects its values and priorities, the boards of system-affiliated hospitals appear to be more responsive to community needs and issues than their freestanding counterparts.
Finally, comparisons of attributes of community accountability associated with the third dimension, collaborating with other agencies to enhance community benefits, demonstrated three statistically significant differences between system-affiliated and freestanding hospitals. Specifically, the former were more likely than the latter to work with others to develop written capacity assessments (¬ D 9:45; p < :01), use assessments to identify unmet needs, excess capacity, or duplicative services (¬ D 4:53; p < :05), and work with outside agencies to collect, track, and communicate clinical and health information (¬ D 13:98; p < :001). Although both types of hospitals frequently organized themselves and collaborated with others to address unmet needs, adjust their capacity, and deal with community issues, system-affiliated hospitals displayed a pattern of results that suggested greater community accountability than their freestanding counterparts.
In sum, our findings indicate that system-affiliated hospitals were
less likely to exercise community accountability in terms of governance
structure and composition. However, they were actually more likely
than freestanding hospitals to monitor information and collaborate with
other health care agencies. It is worth noting that we observed the
same pattern of results even after we removed IO hospitals from our
sample (results available from authors). Thus, the differences observed
between system-affiliated and freestanding hospitals do not appear to
result from the disproportionate presence of IO hospitals in health care
systems.
Community Accountability in Different Types of System Hospitals
Because health care systems are not uniform in their structures or strategies,
global comparisons between freestanding and system-affiliated hospitals
may mask important differences among system-affiliated hospitals.
To address these concerns, we compared hospitals affiliated with different
system attributes––specifically, size, geographic scope, and ownership––
in order to discover how, and to what extent, they express community
accountability. Table 3 presents the results for system size. We found several
significant differences in structural attributes of the hospital board
that are associated with higher community accountability. For example,
hospitals in smaller systems were more likely to have a legally constituted
governing board (¬ D 72:56; p < :001), a higher proportion of outside
(versus inside) members on the board (F D 64:54; p < :0001), and a lower proportion of board members drawn from outside the community ( F D 10:16; p < :0001), all structural indicators of greater community accountability. However, the boards of hospitals in smaller systems retained less authority to make independent decisions (¬ D 3:85; p < :05), a factor that may somewhat offset the greater participation of community members in hospital governance.
Only three additional significant differences were obtained for at-tributes of community accountability that inhere in the other two dimensions (information monitoring and reporting; collaborative relation-ships). Boards of hospitals affiliated with smaller systems were more likely to routinely review immunization rates in the community (¬ D 13:66; p < :001) but less likely to establish benchmarks for community health standards (¬ D 6:52; p < :05). Further, although smaller systems were only somewhat more likely than larger ones to work with others to develop a written capacity assessment, hospitals affiliated with smaller systems were significantly more likely to use these capacity assessments to identify unmet needs, excess capacity, and duplicative services (¬ D 8:38; p < :05).
In sum, hospitals affiliated with smaller systems generally displayed higher community accountability than hospitals affiliated with medium or large systems, particularly on board structure and composition measures.
Table 4 presents the results of comparisons across systems on the
basis of their geographic scope. In general, we observed few statistically
significant differences in community accountability among rural,
single-market, and multimarket systems. To some extent, this may
be an artifact of our classification scheme. Eighty percent of the hospitals
in our sample belong to multimarket systems, whereas only 6
percent belong to rural systems. Thus, unequal cell sizes and consequent
lack of statistical power may contribute to the rather low number
of significant differences we observe. Notwithstanding this potential
limitation, we did observe three statistically significant differences in
the structural forms of community accountability. Specifically, hospitals
affiliated with multimarket systems were less likely than hospitals
affiliated with single-market and rural systems to have either a
legally constituted governing board (¬ D 8:35; p < :05) or a large pro-portion
of outside members on the board ( F D 12:16; p < :0001). We
observed only one significant difference in terms of board information
and management processes. Specifically, hospitals affiliated with rural systems were less likely than those affiliated with single-market or multimarket systems to disseminate reports on hospital quality and costs to the community (¬ D 9:14; p < :01). Finally, we observed no significant differences in the collaborative relationships among hospitals affiliated with rural, single-market, and multimarket systems. It is worth noting that the pattern of results for multimarket systems generally coincides with a centralized, corporate approach to system governance. Multi-market systems are less likely to have legally constituted boards at the hospital level, and they have fewer outside (versus inside) directors serving on the hospital board. Multimarket systems are also more likely to receive community input from separate community-advisory boards.
Table 5 presents results of comparisons among system-affiliated hospitals on the basis of their ownership. In general, hospitals affiliated with religious systems were more likely than their counterparts in secular, not-for-profit, and IO systems to display board structure and composition attributes associated with community accountability. Specifically, hospitals affiliated with religious systems were more likely to have a legally constituted governing board (¬ D 113:23; p < :001) and more locally based decision-making authority ( F D 198:61; p < :0001). Hospitals affiliated with religious systems were also more likely (along with hospitals affiliated with secular, not-for-profit systems) to have a higher proportion of outside members on the board, relative to hospitals in IO systems ( F D 193:99; p < :0001).
We observed a similar pattern for information-monitoring activities. Boards of hospitals affiliated with religious systems were more likely to evaluate the hospital CEO on criteria related to community health improvement (¬ D 12:34; p < :01) and to review community health indicators (¬ D 8:38; p < :05) and immunization rates (¬ D 22:95; p < :001). Hospitals affiliated with IO systems, however, were more likely than their counterparts in religious and secular not-for-profit systems to establish benchmarks for hospital performance on community health standards (¬ D 5:93; p < :05).
Finally, hospitals affiliated with a religious system were significantly
more likely to collaborate with other agencies to enhance community
benefits, especially compared with hospitals in IO systems. Specifically,
religious-system hospitals were more likely to work with others to
conduct community health-status assessments (¬ D 11:73; p < :01), to
work with others to develop a written capacity assessment (¬ D 18:51;
p < :001), and to use these assessments for identifying unmet needs, excess capacity, and duplicative services (¬ D 10:20; p < :01). Hospitals affiliated with IO systems ranked behind, or tied with, religious and secular not-for-profit system hospitals on each of the four indicators of community collaborative relationships.
It is important to note that these findings do not take into account
the potential association among our predictors and thus do not control
for the possibility of confounding effects. For example, system size
may be associated with geographic scope; hence, the results reported for
size may be attributable to scope. Although we were not primarily concerned
with testing causal arguments regarding the effects of our system
characteristics, we did perform exploratory multivariate logistic and ordinary
least squares regressions to assess the net effects of each system
attribute on community accountability mechanisms (results available on
request from the authors). In each regression model, we used the categorical
specifications of system size, geographic scope, and ownership
(minus a reference category for each) to predict the dependent variables.
Results indicate that differences across size categories for information
monitoring and reporting were no longer significant when other system
characteristics were controlled. Similarly, differences across geographic-scope
categories for board structure and composition were not significant
when controlling for system size and ownership. All other results were
consistent with those obtained in the categorical comparisons reported
previously.
Discussion
Consolidation in the hospital sector has raised many concerns about the accountability of hospitals that are responsible to authorities other than the community. In view of these concerns, we decided to assess how community accountability is exercised by hospitals affiliated with health care systems. We focused on the structure, composition, and activity of hospital governing boards as the primary accountable entities linking hospitals with their communities. Our approach evaluated how community accountability in system-affiliated hospitals differed from freestanding hospitals and among different types of systems. Our results indicate that hospitals demonstrate community accountability in a variety of ways.
Several key findings emerge from comparisons between system-affiliated and freestanding hospitals on measures of community account-ability. The first is that system-affiliated and freestanding hospitals differed significantly and systematically in how they expressed community accountability. In general, freestanding hospitals demonstrated community accountability through hospital governance structure and composition, whereas system-affiliated hospitals elected to have their boards monitor and gather relevant information and collaborate with other agencies in order to identify and respond to community health needs. Thus, system-affiliated hospitals' relative lack of traditional structural and compositional mechanisms for assuring community accountability (e.g., formally incorporated boards that include community representatives) does not seem to preclude them from engaging in activities that reinforce ties to community (cf. Relman 1996; Scott 1997). This finding suggests that policy makers, regulators, and others concerned about the effects of industry consolidation on community accountability need to recognize that hospitals and health systems can meet in multiple ways their legal, social, and institutional obligations to serve the broad interests of the community in which they operate.
The second central finding is that observed differences in how system-affiliated and freestanding hospitals approach community accountability are not necessarily attributable to the relative concentration of IO hospitals in health care systems. That is, we obtained the same pattern of results, even when we took into account the unequal distribution of ownership categories across system-affiliated and freestanding hospitals. This suggests that differences between system-affiliated and freestanding hospitals can be attributed to the fact that all system-affiliated hospitals (regardless of ownership type) are embedded in superordinate organizational structures with multiple agency relationships. These findings imply that community leaders and policy makers concerned about the sale of freestanding, not-for-profit hospitals to IO systems must be careful not to confound "ownership conversion effects" with "system affiliation effects." All system-affiliated hospitals, both IO and not-for-profit, are subject to multiple agency relationships that affect their legal, social, and institutional obligations to serve the broad interests of the community in which they operate. Our results suggest that system affiliation (that is, the act of embedding a hospital in a superordinate organizational structure) may itself change the way in which a formerly freestanding hospital expresses community accountability, regardless of ownership of the system with which the hospital has affiliated (cf. Relman 1996; Scott 1997).
Perhaps a more fundamental question would be: Why do system-affiliated and freestanding hospitals adopt different approaches to com-munity accountability. One possible explanation is that health care systems consider traditional forms of governance too cumbersome and expensive to maintain under conditions of structural complexity and broad strategic and operational goals (Johnson 1995). Coordination costs and response time increase when organizations must contend with multitiered governance arrangements that involve large boards, composed of members with potentially diverse interests (Pointer, Alexander, and Zuckerman 1995). However, systems may still view links to local markets and communities as essential to the ability of affiliated hospitals to compete in their local markets and, increasingly, to assume risk under managed care (Alexander, Zuckerman, and Pointer 1995). This may explain why system-affiliated hospitals emphasize the collection and reporting of information about community health and hospital performance and why they collaborate with local agencies to address community health needs. An alternative explanation may be that multi-institutional, multilayered health systems emphasize data systems and information exchange because they view these mechanisms as a means of coordinating and controlling the disparate elements of the system. Extending this form of internal accountability to the community may simply be a natural extension of a core capability to which considerable resources and value have been assigned. Freestanding hospitals, by contrast, may not face the same coordination demands and do not invest the same level of resources in information and data collection. These hospitals may instead achieve accountability by emphasizing the resources to which they do have access, namely, local citizenry who can serve as board members.
It may also be the case that freestanding hospitals emphasize board structural and compositional mechanisms for assuring community accountability (e.g., formally incorporated boards that include community representatives) because they are subject to strong, deeply embedded values favoring local control and volunteerism (Alexander and Scott 1984; Starkweather 1988; Alexander and D'Aunno 1990). These institutionalized values prescribe traditional governance structures and practices that, in turn, legitimize local hospitals as entities whose primary mission is to serve the needs of the community. System-affiliated hospitals can perhaps afford to circumvent these traditional, prescribed structures and practices because the normative pressures that confront them to adopt or maintain these structures and practices are weakened by their multiple accountabilities: to the community, to system headquarters, and to other system-based organizations.
This discussion raises the interesting question of whether these traditional, prescribed governance structures and practices actually promote community accountability among system-affiliated or freestanding hospitals. As manifestations of institutionalized norms and values, the traditional hospital board may symbolically reflect community accountability through its composition and structure, yet have little real effect in terms of reflecting the interests of the entire community and ensuring that hospital policies, strategies, and operations are consistent with those interests. This raises the question of which attributes of accountability community leaders should emphasize when they are negotiatives with systems regarding the sale or affiliation of their freestanding hospital. Should they advocate for structural and compositional aspects of community accountability, or should they instead seek enforceable agreements on community monitoring, reporting, and collaboration?
Notwithstanding the differences observed between system-affiliated and freestanding hospitals, our results clearly indicate that the community accountability practices of hospitals affiliated with different types of systems differ substantially. For example, hospitals affiliated with religious systems are more actively engaged in community accountability practices than hospitals that are part of secular, not-for-profit, and, particularly, IOsystems. This orientation to community is deeply embedded in the doctrines and values of many religious systems and the orders with which they are associated. Indeed, our results indicate that hospitals in these systems practice what they preach. In addition, we discovered that hospitals affiliated with IO systems had instituted the fewest practices and structures for promoting community accountability. Although the concentration of IO hospitals in health care systems does not account for the differences that we observed between system-affiliated and free-standing hospitals, it does seem that hospitals affiliated with IO systems do not promote community accountability as actively as those affiliated with religious or secular not-for-profit systems. This may indicate that communities whose hospitals are considering affiliation with IO systems may better preserve their community benefit by looking instead to alternative affiliations with religious or secular not-for-profit systems.
We found few statistically significant differences in the forms of community accountability among hospitals affiliated with systems of differing geographic scope. As noted earlier, differences in sample size across geographic-scope categories may have limited our ability to detect significant patterns. Indeed, evidence from other sources, although limited to community collaboration, suggests that this may be the case. Among the system-affiliated hospitals that responded to the 1997 AHA Annual Survey, those belonging to rural systems were significantly less likely than hospitals affiliated with single- or multimarket systems to seek and monitor information as followings:
- work with others to conduct a health status assessment
- work with others to develop a written capacity assessment
- use that capacity assessment to identify unmet needs, excess capacity, or duplicative services
- work with others to collect, track, and communicate clinical and health information
Further research involving larger, more equally distributed sample sizes across geographic-scope categories would help to determine whether systematic, significant differences exist for the other two dimensions of community accountability examined in this study.
Finally, we found that system size is related to rather mixed patterns of community accountability. Hospitals within smaller systems were more likely to express community accountability through the structural and compositional attributes of their boards. However, with one notable exception, we saw few consistent or significant effects of system size on community monitoring and reporting or collaboration with community agencies. Interestingly, hospitals affiliated with large systems were significantly more likely to use community health standards or benchmarks to evaluate hospital performance, despite the fact that their boards were no more likely to engage in routine monitoring of community health measures. Perhaps the boards of hospitals affiliated with small systems monitor community health status but do not evaluate organizational performance in terms of community health improvement. Alter-natively, they may evaluate organizational performance in terms of community health improvement, but they adopt a more localized frame of reference than the boards of hospitals affiliated with large systems. Alocal frame of reference might reflect a more provincial orientation, a lack of information about community health benchmarks or standards, or insufficient resources for tracking community health status based on such benchmarks or standards. More focused research is needed to explore these tentative findings.
It is also important to note that our perspective on hospital account-ability is deliberately circumscribed by our interest in how the hospital and its board exercise community accountability, rather than what they are accountable for or the specific benefits they provide to the community. Accountability for CEO performance, protecting the hospital's charitable assets (if nonprofit), and maintaining the quality of the medical staff are clearly responsibilities of the board. Similarly, providing un-compensated care, shortfalls in treating Medicare and Medicaid patients, providing teaching services, and sponsoring research are examples of hospital activities that may specifically benefit the community. Although we have not examined these expressions of accountability elements directly, our exploration of the structures and processes for exercising community accountability potentially subsumes many, if not all, of these elements.
In sum, our study provides evidence that the expression of community accountability differs in both degree and form between system-affiliated and freestanding hospitals and among hospitals affiliated with different types of systems. These findings give rise to an obvious question: Are differences in how (or how much) hospitals express community accountability associated with differences in how (or how much) hospitals provide community benefit? For example, do hospitals whose legally constituted boards comprise local citizens provide more uncompensated care than those that have only advisory boards or whose boards are composed primarily of members residing outside the community? Do community health status indicators (e.g., immunization rates, incidence of teen pregnancy) improve when hospitals routinely monitor such indicators and use them for evaluating executive and organizational performance? Are community health monitoring, evaluation, and reporting enough, or does community health improvement occur only when hospitals also collaborate with other community organizations to identify health care needs and coordinate care? As consolidation continues in the health care sector, such questions represent the next generation of inquiry in this emerging area of health policy and management research.
Our study provides policy makers and community leaders with insight
into how community accountability is, and is not, exercised
differently by freestanding and system-affiliated hospitals and across different
types of systems. In particular, our results challenge the claim that
system-affiliated hospitals demonstrate less community accountability
than freestanding hospitals. The former express community account-ability
differently, but perhaps no less strongly, than the latter. Further,
hospitals affiliated with different types of systems vary in the type and
degree accountability they demonstrate. These findings suggest the need
for tailored policy making to encourage greater community accountability.
Further, they highlight the need for careful consideration of system
characteristics when communities are making decisions about potential
affiliations between local hospitals and health care systems.
References
Ackerman, F.K. III. 1992. The Movement toward Vertically Integrated Regional Health Systems. Health Care Management Review 17(3):81--8.
Alexander, J.A., and T.A. D'Aunno. 1990. Transformations of Institutional Environments: Perspectives on the Corporatization of U.S. Health Care. In Innovations in Health Care Delivery: New Insights into Organization Theory, ed. S. Mick, 53--85. San Francisco: Jossey-Bass.
Alexander, J.A., and W.R. Scott. 1984. The Impact of Regulation on the Administrative Structure of Hospitals. Hospital and Health Services Administration 29(3):72--85.
Alexander, J.A., H.S. Zuckerman, and D.D. Pointer. 1995. The Challenges of Governing Integrated Health Care Systems. Health Care Management Review 20(4):69--81.
American Hospital Association. 1993. Transforming Health Care Delivery: Toward Community Care Networks. Chicago.
American Hospital Association. 1997. Shining Light on Your Board's Passage to the Future. Chicago: Center for Health Care Leadership and Ernst & Young, LLP.
Arkus, C. 1993. Enhancing Community Relations in a Health Reform Environment. Trustee 46(7):14--15.
Berry, D.E., and J.W. Seavey. 1994. Assuring Access to Rural Health Services: The Case for Revitalizing Small Rural Hospitals. Health Care Management Review 19(2):32--42.
Burns, L.R., G.J. Bazzoli, L. Dynan, and D.R. Wholey. 1997. Managed Care, Market Stages and Integrated Delivery Systems: Is there a Relationship? Health Affairs 16(6):204--18.
Cerne, F. 1994. The Fading Stand-Alone Hospital. Hospitals and Health Networks ( June 20):20--33.
Chisolm, L.B. 1995. Accountability in Nonprofit Organizations and Those Who Control Them: The Legal Framework. Nonprofit Management & Leadership 6(2):141--56.
Claxton, G., J. Feder, D. Schactman, and S. Altman. 1997. Public Policy Issues in Nonprofit Conversions: An Overview. Health Affairs 16(2):9--28.
Feldman, P., D. Wholey, and J. Christianson. 1999. HMO Consolidation: How National Mergers Affect Local Markets. Health Affairs 18(4):96--104.
Fry, R.E. 1995. Accountability in Organizational Life: Problem or Opportunity for Nonprofits? Nonprofit Management & Leadership 6(2): 181--95.
Gamm, L. 1996. Dimensions of Accountability for Not-For-Profit Hospitals and Health Systems. Health Care Management Review 21(2): 79--91.
Gamm, L. 1998. Advancing Community Health through Community Health Partnerships. Journal of Health Care Management 43(1):51--66.
Governance Institute. 1997. Governance Trends and Practices in Health Systems. San Diego, Calif.
Governance Institute. 1999. Governance Trends and Practices in Leading Health Systems across North America. San Diego, Calif.
Griffith, J.R. 1987. The Well-Managed Community Hospital. Ann Arbor, Mich.: Health Administration Press.
Hash, M.H. 1989. Board Composition and the Community. Trustee 42(10):21.
Johnson, R.L. 1995. Commentary on "The Challenges of Governing Integrated Health Care Systems." Health Care Management Review 20(4):82--7.
Luke, R.D., Y.A. Ozcan, and P.C. Olden. 1995. Local Markets and Systems: Hospital Consolidation in Metropolitan Areas. Health Services Research 30:4(October):555--76.
Manning, D.L. 1997. Investing in the Twenty-First Century Hospital. Health Affairs 16(2):108--11.
Middleton, M. 1987. Nonprofit Boards of Directors: Beyond the Governance Function. In The Nonprofit Sector: A Research Handbook, ed. W.W. Powell, 141--53. New Haven: Yale University Press.
Pfeffer, J., and G. Salancik. 1978. The External Control of Organizations. New York: Harper & Row.
Pointer, D., J.A. Alexander, and H.S. Zuckerman. 1995. The Governance Challenge: Preserving Community Mission with Integrated Health Care Systems. Frontiers of Health Services Management 11(3):3--37.
Pointer, D.D., and C.M. Ewell. 1994. Really Governing: How Health System and Hospital Boards Can Make More of a Difference. Albany, N.Y.: Delmar.
Proenca, E.J. 1998. Community Orientation in Health Services Organizations: The Concept and Its Implementation. Health Care Management Review 23(2):28--38.
Relman, A.S. 1996. WhyWe Need Not-For-Profit Health Care. Frontiers of Health Services Management 12(3):47--52.
Schlesinger, M., and M. Gray. 1998. A Broader Vision For Managed Care, Part 1. Measuring the Benefits to Communities. Health Affairs 17(3):52--168.
Scott, L. 1997. Communities Strike Back. Modern Health Care (April 28): 26--32.
Shortell, S.M., R.R. Gillies, and D.A. Anderson. 1994. The New World of Managed Care: Creating Organized Delivery Systems. Health Affairs 13(5):46--64.
Shortell, S.M., R.R. Gillies, and K.J. Devers. 1995. Reinventing the American Hospital. Milbank Quarterly 73:131--60.
Sigmond, R.M. 1995. Back to the Future: Partnerships and Coordination of Community Health. Frontiers of Health Services Management 11(4):5--36.
Starkweather, D. 1988. Hospital Board Power. Health Services Management Research 1(2):74--86.
Weiner, B.J., and J.A. Alexander. 1998. The Challenges of Governing
Public-Private Community Health Partnerships. Health Care Management
Review 23(2):39--55.
Address correspondence to: Jeffrey A. Alexander, PhD, Department of Health Services Management and Policy, University of Michigan, 1420 Washington Heights, Ann Arbor, MI 48109 (e-mail: jalexand@umich.edu).
© 2000 Milbank Memorial Fund. This file may be redistributed electronically as long as it remains wholly intact, including this notice and copyright. This file must not be redistributed in hard-copy form. The Fund will freely distribute this document in its original published form on request.
