Volume 77 Number 3, 1999
Responses of HMO Medical Directors to Trust Building in Managed Care
DAVID MECHANIC and MARSHA ROSENTHAL
Rutgers University
anaged care organizations (MCOs) are in the
midst of a firestorm of public outrage over their restrictions
on the delivery of care, a situation that in turn has stimulated
federal and state regulatory responses. MCOs have become the targets of
distrust, a reality that is reflected in numerous surveys, media criticism,
and political attacks (Kassirer 1997; Mechanic 1997; Brodie, Brady, and
Altman 1998; Blendon, Brodie, Benson, et al. 1998). Health insurance
companies and HMOtrade organizations are trying to limit the political
damage and have succeeded in defeating the most onerous regulatory
initiatives, but the industry itself is divided on the extent and nature of
regulation that is acceptable (Kilborn 1998a). With the lack of consensus
over federal bills designed to augment the power of patients vis- `a-vis
health plans, the political battle is likely to continue (Pear 1998).
Perhaps more important than regulation in the long run will be the
MCOsí own efforts to develop structures and practices that build and
maintain trust. In the last three years, state regulators have reported a
50 percent rise in complaints about HMOs by patients and physicians,
particularly regarding service denials and delays (Kilborn 1998b). These
complaints may reflect the publicís increasing distrust of managed care
rather than a true decline in quality of care. Traditionally, patients have
relied on trust in medical professionalism to minimize the stresses and
uncertainty of illness. If patients worry about their physiciansí control
under the strictures of managed care, then the trustworthiness of the
HMO itself becomes a major factor in how they view their physicians
(Gray and Schlesinger 1997). We will report on a survey of medical directors
of MCOs concerning the practices they have introduced both to
enhance the public credibility of their organizations and to encourage
individuals to place more trust in their physicians.
The Concept of Trust
Trust can be defined here as the extent to which organizations and clinical personnel are perceived to be functioning in the interests of patients and the public, acting as their agents and as advocates for their needs and welfare (Mechanic 1996). Patients expect medical institutions and their physicians to place patientsí needs ahead of other considerations and to be willing to sacrifice organizational and personal interests to a degree that is not expected of most institutions in society (Mechanic, Ettel, and Davis 1990). Although trust is multidimensional, it primarily functions in three areas: assurance of competence; advocacy for the patientís welfare; and appropriate control over the health care process (Mechanic 1998a).
Trust is always precarious because it represents a prediction of the health planís or the physicianís response to an individualís needs. A recent survey suggests that patients enrolled in ëëheavy managed careî are particularly uncertain about the response of their health plans to any needs that might arise: 34 percent of respondents in traditional plans and 61 percent in ëëheavy managed careî were either very, or somewhat, worried that their health plan would be more concerned with saving money than with devising the best treatment for them when they are sick (Blendon et al. 1998). This high level of uncertainty makes trust building a rational organizational strategy for MCOs (Sleeper, Wholey, Hamer, et al. 1998). Although trust in oneís physician is largely determined by personal experience over time, trust in large organizations and institutions tends to be shaped by media attention and public discourse. Reputation remains a critical dimension of public confidence in larger institutions, including MCOs (Mechanic 1998a).
Institutional efforts to build public trust can be usefully conceptualized
as technologies that involve various organizational strategies, which
are then implemented through carefully designed steps. Although we
tend to perceive ëëtechnologyî as a label for medical equipment and other
hardware, we must keep in mind that new organizational processes for
enhancing patient satisfaction, quality of care, and trust also can be described
as ëëtechnology.î HMOs and other organizations build their pub-lic
reputations by the way they organize their affairs, by their responsiveness
to community needs and patient concerns, and by demonstrating a
high level of expertise (Mechanic 1998a; 1998b). The successful implementation
of HMO administratorsí strategies for enhancing trust may
be a condition of managed care structures, which range from high centralization
to wide dispersal. In this study, we asked one group of influential
HMO administrators (the medical directors) to respond to questions
about trust building in their organizations, and we report on the
different responses among varying types of HMOs. We would no doubt
have obtained even more diverse views if we had surveyed nonmedical
managers as well.
Background for the Survey
To identify initiatives for building trust in various health institutions, we made brief inquiries of members of the Institute of Medicine––National Academy of Sciences (specifically those working in the area of health care delivery), medical school deans, and members of the American Academy of Physician and Patient. Despite believing that trust was central to high-quality medical care, many of those we queried had difficulty conceptualizing how this quality might be put into operation at an organizational level (Mechanic 1996). A number of respondents, for example, had difficulty giving us concrete examples of how trust worked, beyond noting that it depended on the delivery of high-quality care or on being a good and dedicated physician. Some were cynical, believing that loss of trust was so pervasive in our commercialized health care system that no initiatives to rebuild it would be likely to succeed. Their perceptions generally were not attuned to the idea that creating social and interpersonal trust could be part of a well-defined technology of structural innovations, positive incentives, teamwork, interpersonal skills, and disease management initiatives (Scott, Aiken, Mechanic, et al. 1995; Suchman, Botelho, and Hinton-Walker 1998; Landon, Wilson, and Cleary 1998).
However, many of these physicians identified trust-related programs and initiatives in various health care institutions and provided the information on which we based a more structured survey schedule for medical directors of managed care organizations (Mechanic 1996; 1998a; 1998b). By giving our respondents specific, detailed choices, this second survey allowed us to solidify the medical directorsí perceptions about organizational trust. The survey inquired about the role of trust in man-agerial and operational policies, asked about ways to elicit public trust and communicate with the public, noted the existence of 26 types of programs that were reputed to contribute to trust (and solicited the respon-dentsí opinions about whether they actually do so), and asked questions about sponsorship of support groups.
HMO medical directors are in a unique position to respond to questions about programs to build patient trust in physicians because of their dual role on behalf of the organization and its physicians. However, the responses of medical directors, although valuable, do not provide a complete picture of how MCOs elicit trust. Organizational policies and strategies regarding referrals, timeliness of response, handling of complaints, scheduling of patient visits, dealing with urgent requests, and many other administrative matters are also likely to have a strong impact on how patients and the public view the organization.
One major aim of the study was to identify the ways that organizational
characteristics might affect the medical directorsí responses. Although
many medical directors have undergone similar professional
training and adhere to comparable ethical standards in their dual role,
we expected that organizational characteristics, such as HMOmodel and
tax status, would influence the type of programs offered by the MCO
and affect the medical directorsí responses. The structures of MCOs vary
widely and, to some extent, shape their attempts to address client satisfaction
and trust. Larger networks, for example, can offer patients more
choice of doctors and other caregivers, a factor that in itself contributes
significantly to trust relationships (Kao, Green, Zaslavsky, et al. 1998;
Kao, Green, Davis, et al. 1998). Group and staff––model HMOs restrict
choice to a smaller number of doctors, a traditional barrier to enroll-ment
growth. Moreover, many studies have indicated over the years that
patients in closed-panel HMOs tend to be less satisfied with their
relationships with physicians than are those enrolled in open-panel or.
fee-for-service structures (Mechanic 1976; Luft 1987; Miller, and Luft
1994). To address these concerns, group and staff models have to manage
their staff relations carefully (Freidson 1975). In contrast, it is
easier for centralized organizations than for network plans, whose physiciansí
offices may be located throughout the region, to offer on-site programs
to their physicians and staff. In addition, physicians in group and
staff––model HMOs are more likely to receive their patients through a
single managed care plan, an arrangement that facilitates the implementation
of group strategies. Network physicians, on the other hand, tend
to draw their patients from a number of plans.
Sample and Research Design
We used the Interstudy National HMO Census 6.2 as a sampling frame, which provided information as of January 1, 1996, on 630 HMOs from the United States, including the District of Columbia and Guam. The Census compiled names and addresses of medical directors of HMOs, as well as organizational data and market characteristics. A mail questionnaire was sent to all medical directors (or medical executives where no medical director was listed), with mail, telephone, and fax follow-ups of nonrespondents. Fourteen HMOs had the same medical director as other HMOs in the sample, and these directors were only surveyed for one organization. Thirteen HMOs were either no longer functioning or could not be located. The resulting sampling frame was 603 HMOs. Two hundred and fifty-two directors responded after three mailings, a postcard follow-up, and phone and fax contacts to obtain cooperation, yielding a response rate of 42 percent.
In the analyses that follow, we link the responses of the medical directors to five characteristics of their organizations obtained from the Interstudy database (Interstudy 1996a): tax status (nonprofit; for-profit); whether the organization is part of a chain (as a member of a national managed care firm operating in two or more states and receiving management services only from a central entity or as an affiliate of Blue Cross/Blue Shield in any state); model type (group/staff; IPA/network; mixed model); HMO enrollment (as of January 1, 1996); and managed care penetration (HMO enrollment as a proportion of each stateís population according to the Industry Report) (Interstudy 1996b). These are crude categories, self-assigned by HMO respondents, and may involve ambiguities and inconsistencies. Managed care penetration is a particularly crude measure because state percentages do not accurately describe market areas.
In this sample, we had a good mix of organizational types. Group/staff HMOs were mostly nonprofit (58 percent), and network/IPAs and mixedmodel HMOs were mostly for-profit (71 percent and 67 percent). Nevertheless, 69 nonprofit HMOs in the sample were either network/ IPA or mixed models. We performed extensive analyses of nonresponse to assess sample bias and found the sample to be reasonably representative. Medical directors in IPA/networks were less likely to respond relative to their proportion in the population (61 percent versus 66 percent; p <.05). Respondents from for-profit organizations (68 percent versus 72 percent; p <.05) and those affiliated with a national chain (59 percent versus 64 percent; p <.05) were also less likely to respond. There were no statistically significant differences in directorsí response rates relative to the proportion in the population of the type of HMO organizational model they were affiliated with––group/staff or mixed model––or to their organizationís enrollment size or degree of managed care penetration. The distribution of respondents by state is also representative of the overall distribution of HMOs among the states. Respondents characterize a wide range of HMO types and market features.
There are obvious limitations in these data. Although the sample is reasonably representative of the organizational structures, we cannot exclude the possibility that physicians whose plans had trust programs, or who were more interested in the survey, may have been more likely to respond. Similarly, this is a study of medical directors who report having trust-building programs and their judgments about the programs. As such, it does not tell us either how many patients utilize the programs or about the program intensity. Furthermore, the study does not document whether these programs are specifically offered to increase patient trust or whether they are effective in doing so; it only tells us what medical directors believe about the kinds of programs described in our survey.
The analysis reports descriptive results by linking directorsí responses
to organizational characteristics, applying two-tailed tests of chi-square,
t statistics, Pearsonís r, and ANOVA as appropriate, based on a .05 probability
criterion for statistical significance. Responses to questions about
communication and trust-building programs were factor analyzed to validate
our conceptual framework (Mechanic 1996; 1998a; 1998b), and
resultant factors were analyzed in relation to organizational type. When
more than one organizational variable was significantly associated with
a dependent measure, and the independent predictors were themselves
correlated, we used regression analysis to assess the relative importance
of these predictive variables. Logistic regression was used for dichotomous
dependent variables.
Results
Group/staff––model HMOs were most likely to have programs believed
to build trust. Nonprofit organizations and those not affiliated with a
chain also were more supportive of public service and community programs,
the types of programs that go beyond direct provision of care to
enrollees. In addition, we found that for-profit HMOs and HMOs associated
with chains were less likely to sponsor support groups, one of the
most prominent ways that HMOs can provide trust-building services
for the chronically ill and their families.
Programs That Promote Trust
The core survey listed 26 types of initiatives (shown in table 1) that
had emerged from our earlier exploratory efforts, and these were introduced
with the following question: ëëMedical organizations often de-velop
programs that they believe contribute to their effectiveness, the
quality of care they provide, and public trust in their institutions. For
each of the following, does your organization have such a program and
how useful are such programs in building public and client trust?î Respondents
could indicate that the programs were very, somewhat, or not
useful in building trust. Many respondents who reported that their organization
lacked particular programs did not rate the usefulness of such
programs. Nevertheless, the results reveal which organizations offer potentially
trust-building programs and how HMO medical directors perceive
the programs when they are offered by their organization.
Table 1 shows the proportion who reported that their organizations
had such programs, the proportion who reported the program as being
very useful, and the proportion with such programs in their organization
who reported the program as being very useful. An examination of table
1 shows that approximately ten program areas were found in three-fifths
or more of HMOs. Almost all engaged in efforts to improve patient satisfaction,
and more than three-quarters used focus groups and mediation
for resolving disputes. Almost three-quarters reported that they used
patient clinical pathways, and two-thirds reported sponsoring community
development. Some programs that receive much attention in the
literature and the media, such as computer bulletin boards, video aids,
interactive informational technologies, ethics consultation, and primary
hospital nursing, were not commonly used. Also relatively uncommon
were formal programs to improve the interpersonal skills of doctors and
nurses.
The only item rated as very useful for promoting trust among a majority of respondents was ëëspecific efforts to improve patient satisfactionî (67 percent). Three-fifths or more of directors who reported having particular programs rated them as very useful in five instances: patient representatives/ombudsmen programs (70 percent); specific efforts to improve patient satisfaction (68 percent); programs to involve patients formally in treatment decision making (66 percent); formal supportive/educational programs for patients and families (62 percent); and formal staff educational programs to improve patient-centered skills (60 percent).
We factor analyzed these 26 programs and identified five factors that were clearly interpretable, suggesting that these various initiatives cluster in a meaningful way. The first, ethical consultation, comprises three items: ethics consultation for health professionals (loading of .84); ethics consultation services (.82); and informed consent videos for common procedures (.61). The second factor, which we call public and patient education, comprises four items: health information services and resource centers for the public (.69); open houses for the public and potential clients (.59); lecture series for the public (.57); and supportive educational programs for patients and families (.43). The third factor, sensitivity to patients, comprises five items: focus groups (.63); financial incentives to encourage physicians to communicate effectively (.61); sponsorship of community development (.57); diversity training for staff (.50); and programs to increase confidentiality awareness (.50). The fourth factor, new informational technologies, comprises use of computer bulletin boards (.70); interactional, patient-oriented technologies (.67); and videos at the time of diagnosis (.51). The fifth factor, staff interaction skills, is based on two items that represent programs to improve the staff ís patient-centered skills (.77) and the interpersonal skills of doctors and nurses (.74)
Group/staff––model HMOs were significantly higher on ethics consultation than other models ( p <:05), as were organizations that were not part of chains ( p <:001) and nonprofit organizations ( p <:01). When these variables together were regressed on ethics consultation, group/staff HMOs and organizations that were not part of a national chain remained statistically significant in predicting ethics consultation programs.
Higher enrollment was significantly associated with the public and patient education (r D:17; p <:01) and sensitivity to patients factors (r D:14;p <:05). The use of new informational technologies was reported more com-monly in chains ( p <:05) and in organizations with greater enrollment ( p D:051). Finally, programs to improve staff interaction skills were significantly greater in group/staff––model HMOs compared with IPA/ networks ( p <:001) and mixed models ( p <:01), in organizations that were part of a chain ( p <:05), and in organizations with larger enrollments ( r D:17; p <:01). When the significant predictors are regressed together on staff interaction skills, only group/staff HMOs are significantly associated with having such programs.
Table 2 illustrates some of these results by showing just one program
in each of the five program areas that had the highest loading on its
program factor. The table is intended to suggest ways in which organizational
factors may relate to the provision of certain types of programs.
A different selection might yield different results. We also added a sixth
item, ëëformal supportive educational programs for patients and families,î
because we examine ëëpatient support groupsî in more detail later.
As the table illustrates, the single most important variable, resulting in
statistically significant differences in four of the six examples, was the
type of HMO model. Group/staff models were significantly more likely
to have programs in staff education to improve patient-centered skills,
health information and resources for the public, formal supportive educational
programs for patients and families, and ethics consultation for
professionals.
Sponsorship of Patient Support Groups
One major type of educational program for patient and families is the
sponsorship of support groups. We gathered more detailed information
on sponsorship of six types: AIDS, Alzheimerís disease, diabetes,
cancer, mental illness, and survivorsí support groups. Reporting formal
educational programs was associated with sponsorship of at least one
support group (r D:22; p <:01) and number of types of support groups
sponsored (r D:22; p <:01). However, a majority of the medical directors
(56 percent) reported that their organizations did not sponsor any
support groups for patients or families, and 29 percent reported that
their organizations sponsored two or more. Diabetes support groups were
most commonly reported (37 percent), followed by cancer (21 percent)
and mental illness (21 percent). Other types were less common.
Total number of support groups, and having a support group of each
of the six types, was significantly associated with nonprofit, compared
with for-profit organizations (p <:005), and with group/staff HMOs,
compared with others ( p <:05). For example, 61 percent of nonprofit organizations
sponsored support groups, but only 35 percent of for-profit
organizations did so. Organizations that were not part of chains also were
significantly more likely to have sponsored such groups (with the exception
of AIDS and diabetes support groups, which did not reach statistical
significance). Number of support groups was also associated with organizational
size (r D:18; p <:001). Managed care penetration was significantly
associated with the presence of AIDS, Alzheimerís disease, and
cancer support groups, but not with others.
We carried out a variety of regression analyses to identify the factors
that remained statistically significant predictors of support-group
sponsorship when other predictors were also taken into account. Non-profit
status ( p <:005) and having no connection to a national chain
( p <:05) had significant independent effects on sponsorship of any support
group and on the number of types of support groups sponsored
( p <:05). Other factors independently associated with number of types
of support groups included being a group/staff––model HMO ( p <:05)
and enrollment size ( p <:05). In logistic regression analyses of each type
of group, no organizational variable was statistically significant in the regressions
for all six types of the groups we asked about. In four of the six
instances, being outside a national chain was predictive (with the exception
of AIDS and diabetes groups). In two instances, nonprofit status was
predictive (AIDS and diabetes groups), and enrollment was predictive in
three (AIDS, cancer, and survivor groups). In the case of cancer and mental
illness groups, group/staff––model HMOs were predictive. Managed
care penetration was not statistically significant in any instance.
We should note a discrepancy between these results and the item
concerning ëëformal supportive/educational programs for patients and.
familiesî in table 2. In response to the more general item, differences by
tax status and chain membership were not statistically significant. We
attribute this to the broadness of the category, which allows the respondent
to consider many types of activities for patients and families.
Communicating the Message of Trustworthiness
Several of the survey questions were intended to provide an overall sense of how the HMOs operationalize trust in dealings with their staff and the public. In particular, we asked respondents to rate the methods they used to create trust and to communicate with the public as a way of indicating how the HMOdirectors regarded trust building beyond their own enrolled population.
The responses to the general questions suggest that the medical directors rated trust as a prominent factor in the administrative strategies of HMOs. Sixty-eight percent of respondents reported that public trust often is explicitly considered in developing or modifying managerial or operational policies in their organization, and only 6 percent said this was rare. Sixty-two percent also reported that they emphasized issues of trust when communicating to professionals and other personnel in their organizations.
In soliciting answers about efforts to create trust outside the organization,
respondents were asked to consider which three of eight listed items
were most important for enlisting public trust (table 3). An inspection
of the table shows that the overall reputations of the organization, and
of its hospitals and doctors, are most commonly endorsed, followed by
the quality of clinician––patient relationships and ease of access to care.
The item concerning reputation in table 3 may seem to describe out-comes
rather than processes. However, during our exploratory work, we
gained the impression that respondents thought of reputation as a means
of achieving trust, and that reputation could be established in different
ways.
Most organizational characteristics were not associated with such trust-building approaches. However, directors of nonprofit organizations were marginally more likely to select visible involvement in the community (23 percent compared with 14 percent; p D :063), and direc-tors of organizations that were not part of a chain were more likely to cite the overall reputation of the organizationís hospitals and doctors in their list of the three most important ways of enlisting trust (67 percent compared with 47 percent; p <:005). Medical directors who were not part of chains were also more likely to select the reputation of hospitals and doctors as the single most important factor (24 percent versus 7 percent; p <:001). Medical directors of group/staff HMOs were significantly more likely to select quality of clinician––patient relationships (63 percent) than were IPA/network directors (33 percent) and mixed-model directors (40 percent) ( p <:05). Enrollment and managed care penetration were not significantly associated with any of these selections.
We asked medical directors to indicate which of ten possible methods were used by their organizations to communicate with the general public and potential clients. Almost all used marketing brochures (91 percent), health fairs (85 percent), newspaper/magazine/billboard advertisements (83 percent), radio/television ads (81 percent), public service activities and sponsorship of community activities (81 percent), and presentations to community groups (75 percent). Fewer used a home page on the Internet (59 percent), public service brochures or pamphlets (45 percent), open houses for the public or potential clients (45 percent), and lectures and demonstrations open to the public (43 percent). Advertisements and marketing brochures were predominantly considered to be most important. However, a significant minority of medical directors (25 percent) endorsed public service activities and sponsorship of. community activities and programs as belonging to the two most important communication categories.
Varying types of organizations emphasized different ways of communicating with the public. Nonprofit organizations viewed public service activities and sponsorship of community programs as more important ways of communicating with the public (18 percent versus 8 percent listed them in the most important communication category; p <:05), whereas for-profit respondents saw newspaper/magazine/billboard ads as marginally more important (21 percent versus 11 percent; p D :052). Group/staff HMOs marginally gave public service brochures and pamphlets greater importance than did other types of organizations (13 per-cent versus 3 percent for IPA/networks and 4 percent for mixed groups, but the relationship does not attain statistical significance; p D :067).
These eight communication items fall into two distinct factors, based
on a principal component factor analysis. The first, a marketing orientation,
includes the two advertising items and the use of marketing
brochures (all with loadings above .70). The second factor, which is independent
of the first, we call a consumer-information orientation, and it
encompasses the other seven items, with highest loading for public lectures
(.70), community presentations (.61), open houses (.59), and public
services activities (.59). All seven items had loadings above .50. Respondent
organizations were given scores on each factor, but no organizational
characteristic significantly predicted a consumer information orientation.
In contrast, organizations with larger enrollments, nonprofit organizations,
and organizations located in higher HMO penetration areas
were significantly higher on a marketing orientation. We regressed these
three variables as a group on marketing orientation and found enrollment
size to be the only statistically significant variable ( p <:001). In
analyzing the five factors underlying the trust-building programs, as discussed
above, there was a significant association with a high marketing
orientation. When regressed all together on the marketing orientation
factor, each remains highly statistically significant.
Discussion
Managed care organizations substantially differ in the extent to which they have programs that our respondents believed maintain public and patient trust. The survey is useful because it shows how HMO medical directors think about trust building. This is, of course, only one side. of the trust-building equation; these reports do not permit us to assess either the degree of program development or accessibility or their record of success in increasing trust, nor do they tell us about any other strategies that HMO managers might have developed. It is clear, however, that many health plans have been slow to develop the types of programs that might lead to better patient partnerships (Mechanic 1998b), even though they face a public that is increasingly hostile to managed care.
Although many programs are initiated because organizations are competing both to maintain and to increase their market share, it appears that trust-building programs are not viewed strategically as a tool. It is perhaps curious that although respondents recognize the importance of clinician––patient relationships for trust, formal programs to improve the interpersonal skills of doctors and nurses and financial incentives to encourage physicians to communicate effectively with patients were seldom initiated. Respondents from organizations that had such programs, however, did not view them as useful in building trust (45 percent for formal programs and 32 percent for financial incentives). It is likely that HMOs are unwilling to face the resistance and hostility of physicians, who tend to perceive such initiatives as intrusive (Blandin-Clark 1998).
The literature on HMOs reflects the prominent debate about whether the investment in community health and patient welfare by the for-profit and national chains matches those of the nonprofit and more locally situated organizations (Lawrence, Mattingly, and Ludden 1997; Kuttner 1998; Kleinke 1998; Gray 1997). It is alleged that the primary loyalty of the national, profit-making firms is to their stockhold-ers rather than to their communities and enrollees. Because nonprofit organizations retain tax-exempt status, the argument commonly goes, they are more able, and have more of an obligation, to contribute generally to the community. Schlesinger and Gray (1998) are conducting research to measure community benefits provided by for-profit and non-profit managed care plans. There is, of course, considerable variability among all types of organizational entities. However, it is important to note that public service activities and sponsorship of community activities and programs were valued less highly by respondents from for-profit organizations and organizations linked with chains than by respondents from nonprofit and organizations that were not part of a chain. They also were less likely to sponsor support groups, and when they did, they sponsored fewer types.
The conditions for sponsoring support groups and many other initiatives depend on the character of the health organization. It is much more difficult to organize and monitor special programs when networks and clinicians are dispersed and physicians function in small practices. More-over, sponsorship of many activities, like support groups for cancer and mental illness, depends on a sufficiently large and concentrated enrollee population. As we found, enrollment was predictive of support-group sponsorship. Larger organizations also may have more capital to invest in various developmental activities.
Group/staff––model HMOs have a distinct advantage in developing preventive and supportive health programs and other approaches that were not examined in this survey, such as disease management strategies and practice guidelines (Luft and Greenlick 1996; Thompson 1996). It was not surprising that group/staff––model HMOs were most likely to have ethics consultation programs, programs to improve staff interaction and patient-centered skills, and formal supportive educational programs for patients and families. Their organizational settings provide the means to introduce organized programs readily, and these types of programs may be particularly useful in managing chronic disease (Wagner, Austin, and Von Korff 1996). These prepaid plans are also typically governed by physician groups and sometimes by their patient members, who may consider such support programs to be an integral part of the mission and culture of the organization. Group/staff HMOs also may have a greater incentive to invest in these areas because studies have consistently associated plans of this type with greater enrollee dissatisfaction, compared with fee-for-service plans, and have noted more complaints by enrollees about their clinicians being less caring and more inflexible (Mechanic 1976; Luft 1987; Miller and Luft 1994).
Whatever the advantages of the group/staff organization in facilitating patient care initiatives, the reality is that it represents a shrinking part of the HMOsector, which has come to be dominated increasingly by IPA networks, point-of-service plans, and other virtual-type health plans (Freeborn and Pope 1994; Zelman 1996). More creative thought must be given to how these looser, more dispersed organizational structures can mobilize to address community concerns, public health, prevention and health promotion, and the improvement of practice patterns (Mechanic 1998c). It remains unclear how they can maintain the public trust in an increasingly skeptical and critical climate.
Perhaps the most important message to emerge from these data is that
administrators and policy makers need to think systematically about.
a wide range of patient care initiatives, including those that enhance
caring and trust (Scott et al. 1995; Suchman et al. 1998; Landon,
Wilson, and Cleary 1998), and to design trust initiatives as part of an
organizational technology rather than simply depending on the talents,
warmth, and decency of clinicians and staff. Patients are increasingly
knowledgeable and well informed and are demanding new forms of relationships
and responsiveness in their interactions with the health care
system. Carefully constructed programs of the types we reviewed, as well
as others, offer significant potential not only for retaining the high esteem
of the population and of patients, but also for improving processes
of care and the quality of patient outcomes.
References
Blendon, R., M. Brodie, J.M. Benson, et al. 1998. Understanding the Managed Care Backlash. Health Affairs 17(5):80––94.
Blandin-Clark, M. 1998. Addressing Patient Complaints: A Consul-tation Service for HMO Physicians. In Partnerships in Healthcare: Transforming Relational Process, eds. A.L. Suchman, R.J. Botelho, and P. Hinton-Walker. Rochester, N.Y.: University of Rochester Press.
Brodie, M., L.A. Brady, and D.E. Altman. 1998. Media Coverage of Managed Care: Is There a Negative Bias? Health Affairs 17(1):9––25.
Freeborn, D.K., and C.R. Pope. 1994. Promise and Performance in Man-aged Care: The Prepaid Group Practice Model. Baltimore, Md.: Johns Hopkins Press.
Freidson, E. 1975. Doctoring Together: A Study of Professional Social Control. New York: Elsevier.
Gray, B.H. 1997. Trust and Trustworthy Care in the Managed Care Era. Health Affairs 16(1):34––49.
Gray, B.H., and M. Schlesinger. 1997. The Profit Transformation of the Hospital and HMO Fields. (Unpublished paper.)
Interstudy. 1996a. Competitive Edge 6.2 HMO Directory. Minneapolis, Minn.
Interstudy. 1996b. Competitive Edge 6.2 Industry Report. Minneapolis, Minn.
Kao, A.C., D.C. Green, N.A. Davis, J.P. Koplan, and P.D. Cleary. 1998a. Patientsí Trust in their Physicians: Effects of Choice, Continuity and Payment Method. Journal of General Internal Medicine 13:681––6.
Kao, A.C., D.C. Green, A.M. Zaslavsky, J.P. Koplan, and P.D. Cleary. 1998b. The Relationship between Method of Physician Payment and Patient Trust. Journal of the American Medical Association 280: 1708––14..
Kassirer, J.P. 1997. Managing Managed Careís Tarnished Image. New England Journal of Medicine 337:338––9.
Kilborn, P.T. 1998a. Debate on Rights and HMOs Engender Unlikeliest of Alliances. New York Times ( July 21.)
Kilborn, P.T. 1998b. Complaints about HMOs Rise as Awareness Grows. New York Times (October 11):22.
Kleinke, J.D. 1998. Deconstructing the Columbia/HCA Investigation. Health Affairs 17(2):7––26.
Kuttner, R. 1998. Must Good HMOs Go Bad?: The Commercialization of Prepaid Group Health Care. New England Journal of Medicine 338:1558––63.
Landon, B.E., I.B. Wilson, and P.D. Cleary. 1998. A Conceptual Model of the Effects of Health Care Organizations on the Quality of Medical Care. Journal of the American Medical Association 279:1377––82.
Lawrence, D.M., P.H. Mattingly, and J.M. Ludden. 1997. Trusting in the Future: The Distinct Advantage of Non-profit HMOs. Milbank Quarterly 75:5––10.
Luft, H.S. 1987. Health Maintenance Organizations: Dimensions of Performance. New Brunswick, N.J.: Transaction.
Luft, H.S., and M.R. Greenlick. 1996. The Contribution of Group-and Staff-model HMOs to American Medicine. Milbank Quarterly 74:445––67.
Mechanic, D. 1976. The Growth of Bureaucratic Medicine: An Inquiry into the Dynamics of Patient Behavior and the Organization of Care. New York: Wiley-Interscience.
Mechanic, D. 1996. Changing Medical Organization and the Erosion of Trust. Milbank Quarterly 74:171––89.
Mechanic, D. 1997. Managed Care as a Target of Distrust. Journal of the American Medical Association 277:1810––11.
Mechanic, D. 1998a. The Functions and Limitations of Trust in the Provision of Medical Care. Journal of Health Politics, Policy and Law 23:661––86.
Mechanic, D. 1998b. Public Trust and Initiatives for New Health Care Partnerships. Milbank Quarterly 76:281––302.
Mechanic, D. 1998c. Managed Care and Public Health Opportunities. American Journal of Public Health 88:874––5.
Mechanic, D., T. Ettel, and D. Davis. 1990. Choosing Among Health Insurance Options: A Study of New Employees. Inquiry 27:14––23.
Miller, R.H., and H.S. Luft. 1994. Managed Care Plan Performance Since 1980: A Literature Analysis. Journal of the American Medical Association 271:1512––19.
Pear, R. 1998. Senators Reject Bill to Regulate Care by HMOs. New York Times (October 10): A1, A10.
Schlesinger, M., and B. Gray. 1998. A Broader Vision for Managed Care, Part 2: A Typology of Community Benefits. Health Affairs 17(5):26––49.
Scott, R.A., L.H. Aiken, D. Mechanic, and J. Moravcsik. 1995. Organi-zational Aspects of Caring. Milbank Quarterly 73:77––95.
Sleeper, S., D.R. Wholey, R. Hamer, S. Schwartz, V. Inoferio. 1998. Trust Me: Technical and Institutional Determinants of Health Maintenance Organizations Shifting Risk to Physicians. Journal of Health and Social Behavior 39:189––200.
Suchman, A.L., R.J. Botelho, P. Hinton-Walker. Eds. 1998. Partnerships in Healthcare: Transforming Relational Process. Rochester, N.Y.: Uni-versity of Rochester Press.
Thompson, R.S. 1996. What Have HMOs Learned About Clinical Prevention Services? An Examination of the Experience at Group Health Cooperative of Puget Sound. Milbank Quarterly 74:469––509.
Wagner, E.H., B.T. Austin, and M. Von Korff. 1996. Organizing Care for Patients with Chronic Illness.Milbank Quarterly 74:511––44.
Zelman, W.A. 1996. The Changing Health Care Marketplace: Private Ven-tures, Public Interests . San Francisco: Jossey-Bass.
Acknowledgment: This work was supported in part by a Robert Wood Johnson
Investigator Award to Dr. Mechanic. We wish to thank Dr. Elaine Leventhal
for her assistance in recruiting medical directors for participation in the study.
Address correspondence to: David Mechanic, PhD, Institute for Health, Health Care
Policy and Aging Research, Rutgers University, 30 College Avenue,
New Brunswick, NJ 08901-1293 (e-mail: mechanic@rci.rutgers.edu; mrosenth
@rci.rutgers.edu).
© 1998 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.
