摘要：Objectives. We identified factors associated with local health department (LHD) adoption and discontinuation of clinical services. Methods. We used multivariate regression with 1997 and 2008 LHD survey and area resource data to examine factors associated with LHDs maintaining or offering more clinical services (adopter) versus offering fewer services (discontinuer) over time and with the number of clinical services discontinued among discontinuers. Results. Few LHDs (22.2%) were adopters. The LHDs were more likely to be adopters if operating in jurisdictions with local boards of health and not in health professional shortage areas, and if experiencing larger percentage increase in non-White population and Medicaid managed care penetration. Discontinuer LHDs eliminated more clinical services in jurisdictions that decreased core public health activities’ scope over time, increased community partners’ involvement in these activities, had larger increases in Medicaid managed care penetration, and had lower LHD expenditures per capita over time. Conclusions. Most LHDs are discontinuing clinical services over time. Those that cover a wide range of core public health functions are less likely to discontinue services when residents lack care access. Thus, the impact of discontinuation on population health may be mitigated. The role of local health departments (LHDs) in offering clinical services is hotly contested in public health practice. Some LHD leaders believe that offering clinical services is critical to their mission 1 and public image. 2 Others embrace the position of the Institute of Medicine (IOM) on the future of public health, which calls for LHDs to focus on core public health functions of assessment, assurance, and policy development and away from clinical services. 3,4 Most LHDs have decreased clinical service offerings over time (discontinued). 1,3,5,6 Some have done so because their leaders believe that offering clinical services is inconsistent with the LHD’s mission, 6,7 diverts resources from population-based services, 3,8 or distracts from core public health functions. 1,7 Leaders of LHDs may rely on non-LHD public health system partner organizations to provide clinical services for vulnerable populations rather than providing them directly. 1 They may see the private sector as more appropriate than LHDs in delivering clinical care. 1,6,9 For example, in some regions, Medicaid managed care organizations collaborate with LHDs to ensure clinical services. 5,10,11 However, some LHDs have maintained or increased their clinical service offerings over time (adopted) because their leaders view clinical services as part of their mission, 1 derive satisfaction from patient contact, 5 or believe that offering such services is part of the core public health function of ensuring access to care to their patients if care is not available elsewhere. 2,3,8,12 They offer clinical services if no private sector alternatives exist, 6,12 or if the LHD is uniquely qualified in dealing with specific vulnerable populations 2,5 or for certain conditions (e.g., infectious disease control). 1 These LHDs may offer clinical services if no other safety net providers are available or community need is high. For instance, 63.3% of LHD directors in 2000 believed that LHDs should offer clinical services when no other organization was available to do so, compared with 23.6% of directors who believed that LHDs should offer clinical services unequivocally. 7 Other LHDs may offer clinical services to generate revenue to fund other operations. 1,13,14 We posit that 3 main drivers underlie LHD decisions to offer clinical services. First, a conflicting goal driver suggests that LHD leaders may view offering clinical services as conflicting with core public health functions, particularly when they have few resources. 1,7 Performance of these functions by LHDs varies with jurisdiction-level sociodemographic factors, and LHD organizational and public health system attributes. 15–18 Second, an assurance driver suggests that LHDs offer clinical services if leaders believe that residents lack access to care, that the LHD has important expertise in providing clinical services to vulnerable patient populations, and that LHDs should provide these services when these services are limited. Third, an entrepreneurial driver suggests that LHDs leverage revenue-generating clinical services to fund needed public health services. 5,13,19 The 3 drivers are not mutually exclusive. For instance, an LHD may stop offering comprehensive primary care because of the conflicting goal driver, and simultaneously start offering tuberculosis screening because of the assurance driver. The drivers provide a framework for understanding LHD decisions about the provision of clinical services, elucidating how LHDs change their clinical service offerings in response to strategic and environmental changes. We explored how the 3 drivers relate to 2 decisions: (1) whether an LHD departs from the majority and the IOM recommendations and maintains or offers more clinical services over time (an adoption decision) and (2) conditional on the LHD’s decision to discontinue services, how many fewer clinical services to offer over time (a degree of discontinuation decision). Institutional theory suggests that conformity pressures lead organizations to become more similar in behavior over time, 20 but others resist such pressures for strategic reasons. 21 Thus, we expected that LHDs adopting clinical services over time (adopters) would do so for different reasons than LHDs following the norm of discontinuing clinical services over time (discontinuers), and that leaders at discontinuer LHDs deciding how many clinical services to discontinue may do so for yet other reasons. Because we compared LHDs that followed the norm of decreasing the number of clinical service offerings 1,3,5,6 with those that departed from the norm, we defined adoption to include offering the same number of clinical services over time. Our focus on the number of clinical services does not mean that the actual mix of clinical services offered stayed the same across time. Table 1 describes adoption and degree of discontinuation patterns we posited for each of the 3 drivers of LHD decisions—that is, conflicting goal, assurance, and entrepreneurial drivers. The categories, theoretically linked to the drivers and findings from previous literature are public health system attributes, LHD autonomy, LHD resources, community need, specialized expertise in serving vulnerable populations, 2,5 number of community clinical service providers, and Medicaid reimbursement levels. Public health system attributes measure the LHD jurisdiction’s delivery on core public health functions across 3 dimensions: differentiation, integration, and concentration. Differentiation indicates the jurisdiction’s emphasis on public health needs, with high differentiation indicating that the jurisdiction offers many core programs or services. Integration indicates how different organizations interact in providing these services, with high integration indicating that many partnering organizations offer these services. Concentration measures the LHD’s role, with high LHD concentration indicating that the LHD bears primary responsibility for offering these services. Mays et al. 22 described these 3 dimensions in further detail. TABLE 1— Conceptual Model Describing the Conflicting Goal, Assurance, and Entrepreneurial Drivers and the Factor Categories Associated With Adoption and Degree of Discontinuation of Clinical Services by Local Health Departments Factor Category Conflicting Goal Assurance Entrepreneurial Public health system attributesa Differentiated systems − + Integrated systems + − LHD concentration − + + LHD autonomy − + LHD resources + − Community need + − Specialized expertise in serving vulnerable populations + Community clinical service providers + − − Medicaid reimbursement levels − − Types of services Increase services consistent with core public health functions; decrease most services, except for those consistent with core public health functions Increase services such as maternity and immunizations, where LHDs have expertise; decrease services most likely offered by other clinical service providers, such as those Medicaid reimbursable Increase services that are Medicaid reimbursable; decrease services consistent with core public health functions, because these LHDs view clinical services as instrumental to offering other functions Open in a separate window Note . LHD = local health department. The “+” indicates that we expect a positive relationship between the category and an LHD adopting services or discontinuing fewer services, based on the specified driver. For instance, the “+” for community need under the assurance driver indicates that the LHD may be more likely to adopt services in jurisdictions with higher need. The “–“ indicates that we expect a negative relationship between the category and an LHD adopting services or discontinuing fewer services, based on the specified driver. For instance, the “–“ for community clinical service providers under the assurance driver indicates that the LHD may be more likely to adopt services in jurisdictions with fewer community clinical service providers. The degree of discontinuation model shows discontinuer LHDs discontinuing more services, so for this model, the signs indicated in this table are reversed. aDelivery system attributes describe the public health system’s orientation on core public health activities. Differentiation measures the number of core programs or services delivered in the jurisdiction, with high differentiation indicating that the system offers many core activities. Integration measures the extent to which these services were offered by different organizations, with high integration indicating that there are many partnering organizations. LHD concentration measures the extent to which an LHD is primarily responsible for those services, with high LHD concentration indicating that the LHD bears primary responsibility. For more detail, please refer to Mays et al. 22 Under the conflicting goal driver, LHDs are more likely to adopt clinical services over time and discontinue fewer services over time if they operate in jurisdictions with low LHD concentration, because LHDs experience less conflict between performance of core public health functions and clinical services if they bear less responsibility for the former in the jurisdiction. In addition, the conflicting goal driver may lead to adoption by LHDs with less autonomy because LHDs may be required to offer certain services by a centralized state agency. 5 Moreover, the conflicting goal driver of LHD adoption and discontinuation of clinical service is likely to dominate LHD decision-making when LHDs have more community clinical service providers available in their jurisdictions or when they operate in public health delivery systems with high integration of system partners (because LHDs contract or partner with these organizations to offer clinical services), 1 and they have more LHD resources per capita. By contrast, under the assurance driver, LHDs are more likely to adopt clinical services over time or discontinue fewer services if they are in local public health delivery systems with low differentiation, low integration, and high LHD concentration because few other organizations ensure core public health services. In addition, the assurance driver may lead to adoption when LHDs have autonomy in decision-making related to the provision of clinical services and when they operate in jurisdictions with higher need by the community and vulnerable populations, but few community clinical service providers and lower Medicaid reimbursement levels. Finally, under the entrepreneurial driver, LHDs are more likely to adopt clinical services over time and to discontinue fewer services over time if they operate in public health delivery systems with high differentiation and high LHD concentration because these LHDs have more need for revenue than LHDs offering fewer core public health functions. These LHDs have lower per capita LHD resources because LHD leaders may find generating revenue by providing Medicaid-reimbursable services more attractive than LHDs that are well resourced. 5,13,19 Furthermore, their jurisdictions have lower community need, fewer community clinical service providers, and lower Medicaid reimbursement levels because there are more competitors for Medicaid revenue than in jurisdictions with higher need, more community clinical service providers, and higher Medicaid reimbursement levels. 3,5,11 Table 1 highlights the types of clinical services LHDs are likely to adopt or discontinue depending on the dominant driver of LHD decision-making. Because our conceptual model suggests that all 3 drivers may be present in an LHD making an adoption or degree of discontinuation decision, and the patterns of factors are complex, examining the individual clinical services being adopted or discontinued helps clarify which driver predominates LHD decision-making. We used LHD organizational survey data and area resource data to describe factors associated with adoption and degree of discontinuation decisions to update previous research, and to explore the usefulness of the 3 drivers in explaining changes in LHD clinical services provision over time. First, we examined factors associated with whether an LHD was a discontinuer or adopter from 1997 to 2008. Second, among discontinuer LHDs, we examined factors associated with how many clinical services were discontinued over time. Finally, we examined the types of clinical services that adopter and discontinuer LHDs gained and lost by 2008.
