Compared with their individual care counterparts, family care adults had significantly lower total health care expenditures (b = -.15; 95% CI, -.30 to -.01). After adjustment, additional measures significantly associated with reduced expenditures included being married, living in the southern United States, having better perceived health status and role functioning on the MOS scales and greater medical skepticism scores, while measures associated with increased expenditures included Medicaid, female sex, and current smoking Table 3. Retransforming these results indicated that family care was associated with 14% ($51) lower expenditures in adults. Retransforming the other significant covariates revealed associations with reductions of 21% ($75) for being married and 27% ($99) for living in the South, and increases of 42% ($153) for Medicaid, 67% ($242) for female sex, and 25% ($92) for current smoking. For each 1-point increase on the perceived health status score and for each 1-point increase on the role functioning score, expenditures decreased by 1% ($3). Each increase on the medical skepticism score was associated with a reduction of 4% ($16).
Discussion
Our findings, using data from a large representative sample of US households, show that intergenerational family care, when compared with individual care, was significantly associated with modestly lower adjusted total health care expenditures for adults, with similar, although not significant, findings for children. These findings suggest that emphasizing family care might be an effective means of reducing health care costs.
Strengths
The validity of these analyses is supported by a number of strengths. First, because having health care expenditures and the amount of expenditures were modeled separately, the lower costs associated with family care for adults do not reflect simply a lower likelihood of using care. In fact, the odds of having any expenditures did not differ significantly between the 2 groups. Second, because total health care expenditures were examined, cost shifting between outpatient, inpatient, or other settings does not explain the relationship between family care and lower expenditures. Third, we adjusted for a wide array of potential patient confounders, including sociodemographics and health status. Finally, we examined a nationally representative survey with an excellent response rate, rigorous data collection methods, and validation of expenditure data.
Limitations
Our findings are subject to some limitations. It is possible that control for health status was not adequate. Evidence, however, supports the validity of self-reports of morbidity.42 The MOS health perceptions scale (adults) and the self-reported health status measure (children) were used to adjust for disease severity. Studies have validated this subjective approach, compared with more objective measures.43,44 Because these measures were predictors of mortality in the NMES,1 their validity as health status measures is supported.
Because the NMES is a cross-sectional survey, causality cannot be proved, and unmeasured confounding may explain the observed relationships. Factors associated with the choice of a personal physician may also be related to expenditures. Most important, people choosing family care may exhibit lower need or demand for care. Although we adjusted for measures of both need and demand according to the Andersen-Newman behavioral health model,45 the possibility of confounding remains. For example, attitudes toward health care affect both the choice of a personal physician and health care utilization,46 although in the present study no significant difference in medical skepticism scores between groups was found.
Our findings are also limited by the use of 1987 data. However, the relative cost savings of family care compared with individual care likely remain relevant, suggesting that policies promoting family care within the current managed health care environment merit consideration. If, for example, emphasizing the family results in lower resource utilization as supported by findings from the Direct Observation of Primary Care24,25 studies, then promoting family care should result in cost savings in current managed care settings. However, the extent to which the family can truly be emphasized, given increasing time demands faced by physicians, is not known. Evaluation of relationships between specific aspects of family care and cost and health outcomes needs to be performed using more recent data sources than the NMES.
There are a number of mechanisms by which family care may save money. Savings may occur when unbilled care is provided during visits by other family members, as suggested by findings from the Direct Observation of Primary Care study.24,25 Such unbilled visits could reduce the need for in-person visits and thus reduce costs. Also, some conditions may be treated more cost-effectively with knowledge of family issues, whether through tailoring interventions on the basis of this knowledge or by enlisting family members in treatment plans.
The Institute of Medicine issued a report in 1996 that defined primary care as “the provision of integrated, accessible health care services by clinicians who are accountable for addressing a large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of family and community.”10 Although this definition lends theoretical support for family care, our analyses provide empirical evidence of an association between family care and cost saving. We believe that policies promoting family care may result in appropriately lower health care expenditures. Assessments of relationships between family care and expenditure subcomponents, such as ambulatory visits and diagnostic tests, are needed to identify where expenditure differences are greatest. Further, studies that disentangle aspects of longitudinal continuity from intergenerational family care are warranted. Although we conclude that policies promoting family care may help contain health care costs, research is also needed to explore relationships between family care and health outcomes.