Conceptual Framework for Evaluating the Consequences of Uninsurance for Families
The conceptual framework used in this report, and the variations on this framework as described in Chapter 5, are closely related to the conceptual framework introduced in the Committee’s first report, Coverage Matters (IOM, 2001). In the paragraphs that follow, the Committee’s initial framework is described and its modification for use in this report is clarified.
Figure A.1 depicts the conceptual framework used in Coverage Matters. The framework is based on Andersen’s model of access to health services, which incorporates ideas from the behavioral sciences to understand the process of health services delivery and health-related outcomes for individuals (Andersen and Davidson, 2001). In addition, it draws on an economic model of insurance status and the impact of out-of-pocket costs on health care demand.
To describe the roles of factors at the individual, family, and community levels (e.g., an individual’s health insurance status) that influence both the process of services delivery and the consequences of health care experiences, the framework uses Andersen’s grouping of variables into three categories: (1) resources that foster or enable the process of obtaining health care; (2) personal or community characteristics that favor or predispose action related to obtaining health care; and (3) needs for health care, as articulated by those in need, determined by health care providers, or identified by researchers and decision makers. Arrows and spatial relationships among the boxes indicate hypothesized causal and temporal relationships. For example, the young child of a lower-income, uninsured parent may be eligible for Medicaid, but if the parent does not enroll him or her, the family may not be able to afford health care for the child, and the child’s health and psychosocial development may suffer as a result. This case can be followed through the model in Figure A.1.
To depict the economic consequences of uninsurance, the Committee creates links within Andersen’s framework among determinants of health insurance and such factors as family economic well-being, the institutional viability of health services, and community-level socioeconomic conditions. This expansion allows the Committee to assess hypothesized interactions between economics and health and, in particular, the growing literature on the psychosocial dimensions of family well-being and childhood development. The Committee recognizes that insurance is one of many factors that can influence the health of an individual and that of the family unit.
CONCEPTUAL FRAMEWORK INTRODUCED IN COVERAGE MATTERS
The left panel of Figure A.1 addresses the main economic forces affecting the insured or uninsured status of individuals and families. Individual- and family-level characteristics include financial resources, categorical eligibility for public health insurance, labor market determinants of employment-based insurance, and the requisite skills to enroll and maintain coverage. In the example mentioned above, the differing eligibility of child and parent for public insurance would together function as a determinant of coverage, the top box in Panel 1. Community-level factors include public program eligibility standards, labor market characteristics that determine the availability of employment-based health insurance, and the commercial market for individual health insurance.
The center panel of Figure A.1 is based almost directly on Andersen’s model of access to health care (Andersen and Davidson, 2001). The boxes labeled “individual and family level” and “community level” each contains individual- and aggregate-level variables, respectively, believed to influence how people obtain access to health care. Community-level variables are ecologic or aggregate measures to describe the context or environment within which individuals and their families seek and use health care. For example, the community’s morbidity rate for whooping cough might indicate the need for an immunization campaign. Because health care services are provided and consumed locally, the term community refers to a residential or geographic grouping. For example, the family of mixed insurance status described earlier may be more likely to seek health care for its child for immunization, if it lives in an area with low-cost services available through a community health clinic or public health department program, factors that would be included in the lower left box of Panel 2.
Implicit in the categories of resources, characteristics, and needs are judgments about how much a particular variable may be susceptible to change. Variables labeled “resources” are considered, at least theoretically, to be more open to change. Those termed “characteristics” are considered less flexible or manipulable, and those called “needs” comprise a mixed or heterogeneous grouping, with some needs being more changeable than others.
As a whole, community-level and individual- and family-level variables
describe many potential scenarios for accessing health care. The variables within the box labeled “health care” describe how these potentials may be realized, with particular attention to the role of health insurance coverage. The process of health care delivery is characterized in terms of three types of variables; (1) personal health practices (e.g., dietary habits, physical exercise); (2) use of health services (e.g., number and kind of physician visits within a year); and (3) processes of care (e.g., adherence to clinical practice guidelines). In the example given, the child’s lack of immunization could be described as underutilization of health services, a variable related to processes of care. The Committee focuses most of its attention on the literature concerning the processes of services delivery and the utilization of health services while recognizing that personal health practices may be influenced by insurance coverage and access to care.
The right panel of Figure A.1 describes the ways in which the Committee anticipates that health insurance status may affect the health, economic, and social characteristics of individuals, families, and communities by means of access to and the process of health care delivery. These effects of realized access to health care cascade from the smallest unit of analysis, the individual, to increasingly larger units, first the family and then the community. The consequences linked to health insurance influence community-level and individual- and family-level variables that describe the process of obtaining access to health care and also of gaining or losing health insurance coverage. This last panel should make it clear that the process is dynamic with multiple feedbacks. For example, employment status and income affect family insurance status, which affects current and future health status. Health status, in turn, can influence future employment status, bringing us full circle.
CONCEPTUAL FRAMEWORK FOR THIS REPORT
Figure A.2 depicts a version of Figure A.1 modified to reflect the focus of the Committee in this report. The modified version draws on the same theories and conceptual approaches to health insurance as does the initial version in Figure A.1. Figure A.2 also shares the same overall structure of three panels, with a first panel depicting determinants of coverage (left side of diagram), a second panel depicting the process of obtaining access to health care (center of diagram), and a third panel depicting selected consequences of uninsurance (right side of diagram). Where both frameworks are similar, the text description is shortened in Figure A.2, for example, in the individual- and family-level and community-level boxes where the same variables are used in Figures A.1 and A.2.
Key differences between the two frameworks are as follows:
Rather than treat the resources, characteristics, and needs of individuals and their families as a whole, Figure A.2 acknowledges the differences between the situations of children and those of their parents or, alternatively, between those
of a breadwinner and those of family members who depend economically on the breadwinner. In the example described above, the differing coverage status of child and parent may translate into different experiences with the health care system, and different consequences of being uninsured for families with at least one uninsured member, than if both child and parent were covered by public insurance. Likewise, parents’ knowledge of both their children’s and their own eligibility for insurance and their beliefs about the value of coverage may affect whether they enroll themselves and their children. Their knowledge of and beliefs about health care and the availability of services may influence whether they seek care for themselves and their children. The availability and costs of coverage at the community-wide level (lower left box) may influence aggregate enrollment and use of health care in the community.
Within the box labeled “health care,” Figure A.2 merges the three categories listed in Figure A.1 (i.e., personal health practices, utilization of health services, processes of services delivery) into two categories that highlight the influence of a parent or family breadwinner’s decision making on the process of obtaining access to health care. The quality, quantity and continuity of the services obtained can affect the box labeled health outcomes for families in the right panel on consequences as well as the community level in the box effects on communities.
In Figure A.2, the box labeled economic and health outcomes for family members is roughly equivalent to the box in Figure A.1 labeled effects on families. These consequences include effects on family resources (e.g., current and anticipated future income, economic status, long-term asset accumulation), some of the family’s characteristics (e.g., decisions about marital status and childbearing), and the health of all family members (e.g., health status, family stress and well-being, childhood development). These consequences may have different effects on the family as a unit from those on individual members.
The “effects on communities box” in Figure A.2 reflects the influence on communities of the aggregate efforts of families to obtain access to health care. A community’s uninsured rate may influence the configuration of services, providers, and institutions and their costs and revenues. The uninsured rate in the community also may influence the quality of care that providers and institutions deliver, the health care needs of its residents, and their use of care. These community-level effects will be examined in detail in the next report of the Committee.
To tailor the conceptual framework for use in assessing the literature on health outcomes for pregnant women and newborns (in Chapter 6), Figure 6.1 reflects the specific processes and outcomes studied in the literature reviewed. Figure 6.1 emphasizes the role of health insurance for pregnant women in obtaining access to health services and enumerates specific health outcomes for both women and their newborn infants that may be consequences of the mother having been uninsured. It makes linkages between the process of health services delivery and the care received to decisions of providers of care that reflect health insurance status.