formal hiring. The study also found that beneficiaries had a greater sense of security and choice in hiring a family member. Overall, the results of the cash and counseling demonstration, which allowed enrollees to hire family, indicate that the project did not result in misuse of Medicaid funding.
The committee did not attempt to rank the models described above or to recommend one model of care over another. In fact, little evidence exists that one might use to rate the relative effectiveness of these different approaches. Typically, evaluations focus on whether a single model proved to be successful rather than identifying which of several models produced the strongest results. The committee concluded that no single one of the models described above would be sufficient to meet the needs of all older adults. Instead, a variety of models will need to be employed to meet the targeted needs of older adults. For example, preventive home visits may be too costly to expand to all older persons, the majority of whom may not even require that level of care. Similarly, caregiver-support programs may not be sufficient for older adults with more intensive needs. The health care needs of the older population are diverse, and addressing those needs will require varying models of care. Fortunately, the models described above have generally been successful in enrolling mainly those older adults who would best benefit from the expanded services.
After reviewing the evidence on a number of different models of care, the committee concluded that some of the models with the strongest evidence of success in improving care quality, health-related outcomes, or efficiency have common features which may contribute to their success (Table 3-3).
The model components described in Table 3-3 have shown positive outcomes, at least in some circumstances, but these findings need to be interpreted with caution. First, the list is derived from an examination of only those interventions that have been rigorously evaluated and published in the peer-reviewed literature; many others have not yet been thoroughly evaluated. Furthermore, because the models have proved successful in only certain settings, one cannot be certain that they will experience the same success if they are adopted more widely. Adoption of a model in rural areas or at community hospitals, for instance, may not yield the same results as when the initiatives were undertaken at urban academic medical centers. Similarly, there is limited information on the scalability of the models, that is, on whether they could be successfully applied to a much larger population of patients. Finally, the literature review commissioned by the committee focused on identifying interventions that have produced successful results. In some cases, alternative evaluations of the same model may show