Case management was developed to address fragmentation in service delivery and is rooted in the social casework tradition (1,2). The theoretical basis for case management is systems theory (3), as reflected by standards established by the Case Management Society of America (4) and the National Association of Social Workers (5), with case managers intervening at the individual, agency, community, and political or policy levels. Case management is used for clients who have extensive, complex, and ongoing needs, such as frail, elderly people or people with severe mental illness (6,7). In practice, case management usually takes place at the individual level and seldom occurs at the agency, community, or political or policy levels because of the immediate needs of clients and the requirements of the organization or supervisor (5).

Little consensus exists on what constitutes case management in traditional settings, such as mental health or aging services, let alone in new settings, such as cancer screening (8-11). Tasks performed by case managers are not always clearly delineated in services that offer case management (12). Holloway and Carson (2) note that literature on the effectiveness of case management in traditional settings is limited primarily to "anecdotal reports and poor quality research studies." Many of the models in the literature are linked so closely to specific client groups that model elements cannot be easily conceptualized for use in other settings. Choices in how to implement case management determine the structure and, therefore, the breadth and effectiveness of any case management program (13).

In preliminary unpublished work for this study, the author reviewed records from the Centers for Disease Control and Prevention's National Breast and Cervical Cancer Early Detection Program (NBCCEDP) and from the New York State Department of Health (NYSDOH) and spoke with staff members at both agencies to learn how case management policy for breast cancer screening programs was conceptualized at the federal and state levels and communicated to local programs. A summary of these findings follows.

States participating in the NBCCEDP were prohibited by federal legislation from using national funding for treatment but were required to ensure their clients received diagnostic testing and treatment as needed. In 1996, NBCCEDP sponsored a case study to determine how states identified resources and obtained these services for clients; this study found the process of identifying resources for follow-up care to be labor-intensive and only a short-term solution for providing client care, in part because approaches for delivering services were fragmented (14). In 1997, NBCCEDP surveyed states about their use of case management and presented a video conference on case management for states and providers. An internal NBCCEDP committee convened in September 1998 recommended that NBCCEDP pursue case management as a component of the NBCCEDP, review the literature on case management effectiveness, and draft guidelines for local programs. The subsequent literature review included only articles describing a systems-oriented approach (15). Committee meeting transcripts show that some NBCCEDP committee staff members viewed case management as instrumental in addressing psychosocial and socioeconomic needs that impede diagnosis and treatment of breast cancer and as distinct from clinically oriented disease or care management.

In October 1998, the Women's Health Research and Prevention Amendments of 1998 (Pub L No. 105–340) amended the law that established the NBCCEDP to add the words "support services such as case management" (16). With passage of this amendment, NBCCEDP could move forward in identifying key elements of case management policy for its participating states, including assessment, planning, coordination, monitoring, evaluation, and resource development.

An NBCCEDP memorandum to states in 1999 indicated that approximately $5.6 million in supplemental funding for case management was available nationwide. Ultimately, state breast cancer screening programs were asked to implement case management services for women with abnormal breast screening results without additional federal funding.

In New York, publicly funded breast cancer screening is organized locally by 53 partnerships comprising more than 2300 partner agencies that provide more than 40,000 mammograms to underserved women each year. When NBCCEDP began emphasizing case management in 1998, NYSDOH staff reported in interviews that they and staff in other states believed clients were well served by the tracking and follow-up services already in place — without case management — to ensure timeliness of testing, diagnosis, and treatment of breast cancer. Nonetheless, NYSDOH moved forward to require case management, viewing NBCCEDP's 1999 policy report, its conference calls about case management with states in 2000, and its 2000 operational plan for case management as a mandate.

In 2001, NYSDOH mailed the following definition of case management to partnerships: "activities that can increase client adherence to screening, diagnostic and treatment recommendations." NYSDOH described case management as having a goal similar to that of tracking and follow-up, the previous model for ensuring timeliness of diagnostic care, but involving "a more direct and personal level of support for clients at risk for not obtaining recommended diagnostic or treatment services." This definition addressed primarily client-level activities and did not include system-level activities such as resource development or advocacy. In interviews with key informants, NYSDOH staff indicated that partnerships were asked to focus on client activities so the partnerships would not be burdened with additional tasks without corresponding funding. Other NYSDOH documents distributed to partnerships focused on the time frames for completing case management activities, reflecting the idea that case management grows out of tracking women with abnormal test results. After the initial year of case management implementation among the partnerships, state funding for case management was provided in 2002.

In 2004, the author suggested that case management in publicly funded breast cancer screening programs differed from case management in traditional settings, such as in case management of chronically mentally ill individuals (17). Differences include a limited time in which to interact with clients; a lack of opportunity to function as a client advocate or influence the medical care system; and preconceived ideas about clients' problems, such as defining the client's problem as noncompliance (17). The research presented examines case management in local breast cancer screening partnerships within New York for a better understanding of case management in a new and distinct practice setting.

The communications model of intergovernmental policy implementation, a systems model, was used as a theoretical framework for this analysis (18). The original model was modified to separate implementation at the state level from implementation at the local level because breast cancer screening in New York is provided through local-level partnerships (Figure). Decisional outcomes are viewed as the desire or willingness to implement policy and as separate from the capacity to do so. This study focuses on the relationship between willingness (local decisional outcome) and capacity (local capacity) to implement case management and actual local implementation of case management.

The study addressed the following questions:

In June 2003, a mail survey was administered to the director of each of New York's 53 partnerships. Directors, rather than case managers, were selected as survey respondents because the directors were responsible for developing new case management programs for local breast screening programs. Thirty-nine (74%) directors responded to the survey. The survey comprised two components: 1) a program philosophy scale adapted by the author for cancer screening case management programs and 2) a set of open-ended questions designed to measure local willingness and local capacity.

The program philosophy scale was previously used to assess the operating philosophy of mental health case management programs (19) and was adapted with permission for this study. The original scale included the subscales program clarity, cohesion of work group, and supervisory support; these components were eliminated from this study because the unit of analysis for this study was the partnership rather than the individual case manager. Thirteen subscales were used for the program philosophy scale (Box). Each subscale consisted of four statements; two were worded positively, and two were worded negatively. Respondents were asked to indicate whether they strongly disagree, disagree, are neutral, agree, or strongly agree to each of the 56 statements in the survey (Appendix A).

Local willingness to implement case management was assessed with open-ended questions on partnership goals for case management, perceived client needs, tasks to meet client needs, and partnership preferences for case management. Capacity was assessed with questions on perceived barriers to implementation, staffing level, staffing disciplines, case management caseloads, perceived autonomy in designing the case management program, and awareness of recognized case management standards (Appendix B).

Factor analysis was conducted on subscales of the program philosophy scale to identify underlying common factors that would represent differences in implementation of case management. The consistency within each subscale was assessed using Cronbach's α; subscales lower than 0.60 (interagency orientation and referral advocacy) were eliminated from further analyses. Another subscale (longitudinality of services) had a low communality and also was dropped.

The open-ended questions on willingness and capacity to implement were categorized into themes, coded as dichotomous variables (present or absent), and correlated with the common factors identified in factor analysis. Relationships between some variables that might have been related may not have been detected because of the small number of respondents. P values less than .10 were considered significant. SAS-PC software (SAS Institute Inc, Cary, North Carolina) was used in the quantitative analyses.

The institutional review boards of the New York State Department of Health and the University at Albany, State University of New York, approved this study.

Of the 13 subscales of the program philosophy scale, eight were included in the factor analysis: housing assistance, emergency access, links to entitlements, family orientation, outreach orientation, involvement in job, interest in serving marginalized clients, and innovation. Two common factors — task focus and self-identity focus — were identified. Because these two common factors were somewhat correlated, an oblique rotation was used to aid in data interpretation. Table 1 shows the relationship between the program philosophy subscales and the two common factors.

The first factor — task focus — was so named because it grouped subscales pertaining to the specific tasks of a comprehensive case management program: housing assistance, emergency access, linking to entitlements, family orientation, and outreach orientation. The second factor — self-identity focus — was so named because it grouped subscales pertaining to the partnership's self-perception: involvement in job, interest in serving marginalized clients, and innovation orientation. Self-identity focus reflected the idea of emotional investment in fulfilling a responsibility with creativity and a sense of service to underserved women. One partnership indicated it was "honored" by the women served, who were "poor and have a multitude of problems, yet they are hopeful, strong." Task focus, therefore, reflects what partnerships do, whereas self-identity focus reflects self-awareness of how the partnership does what it does.

The identification of task focus and self-identity focus as common factors showed that partnerships differed in their approaches to case management. One-third of the partnerships surveyed scored high on both task focus and self-identity focus. Another third scored low on both dimensions. Two-thirds of the partnerships scoring high on self-identity focus also scored high on task focus. The remaining one-third of partnerships were almost evenly divided between high self-identity focus and low task focus and low self-identity focus and high task focus.

Willingness and capacity to implement were examined as explanations for differences in implementation (Table 2). Task focus was associated with a broader number of goals and tasks, including client support, client empowerment, and health education. Provision of health education correlated with both task focus and self-identity focus. Health education went beyond breast health to include ensuring "that clients have the ability to maximize their use of preventive health services."

Self-identity focus was negatively associated with perceived barriers to implementation, such as time and difficulties with outside agencies or providers. Working systemically to address change on their own was apparent in partnerships that described "networking for pro bono care for both cancer . . . and non-cancer related conditions," undertaking "legislative advocacy for funding," or forming "a networking and continuing education group for professionals." In contrast, task focus was associated with a preference for additional resources, with lack of funding identified as a barrier to implementation.

Having a nurse as the case manager and awareness of established standards of case management both were related to task and self-identity focus. Having a health educator as the case manager was negatively associated with task focus, although health education as a task was strongly related to task focus.

Among nearly all of the partnerships that identified tracking and documentation as a program task, few noted any other tasks or program goals. Partnerships limited to tracking and documentation scored low on both task focus and self-identity focus and were more likely to frame the task as "follow-up with noncompliant clients" or "maintain tracking database." In contrast, self-identity focus was negatively associated (but not significantly so) with identification of tracking and documentation as a program task.