The Balanced Budget Act of 1997 included a number of changes to Medicare managed care. The newly created Medicare+Choice program differs from its predecessor with regard to payment policies, enrollment and disenrollment policies, and the types of plans that can contract to provide care to Medicare beneficiaries. These changes were designed, in part, to expand the number and mix of private health plan choices available to beneficiaries in different markets across the country.
In fact, no new plan types were offered in the first year of Medicare+Choice and some existing plans withdrew from the program or reduced their service areas. It is difficult to know whether these changes were one-time responses to the new program or represent the types of annual changes that are likely to occur, since little is known about the factors that affect plans' decisions to participate in Medicare. The goal of this analysis is to identify those plan and county factors that are associated with the availability (or not) of Medicare+Choice plans in different markets. The analyses address two main outcomes: participation in Medicare+Choice in 1998 (and inclusion (or not) of each county from commercial service area in Medicare service area) and discontinuing participation (or dropping county from Medicare service area) in 1999.
Data are used from three sources. First, the universe of commercial health maintenance organizations (HMOs) is taken from Interstudy'sCompetitive Edge reports, which include self-reported data on commercial service areas, total enrollment, and other key measures. Second, Medicare+Choice plans and service areas are taken from the Health Care Financing Administration's (HCFA(now known as CMS)) contract reports. Some plans listed by HCFA(now known as CMS) as Medicare+Choice plans could not be matched to a plan in the Interstudy data, so are not included as Medicare+Choice plans in the analysis. Finally, unemployment rate, relative size of the elderly population, and other information about each county is taken the Area Resource File (ARF).
Controlling for other plan and market factors, the plan characteristics associated with higher probabilities of participating in Medicare in 1998 include: not being an independent practice association (IPA) model plan, being federally qualified, also participating in the Federal Employee Health Benefits Plan (FEHBP), having relatively large total enrollment, having relatively fewer enrollees per plan physician, having relatively more enrollees per plan hospital, and undergoing service area consolidation. The same model suggests that important county characteristics include: more commercial HMOs, relatively high Medicare risk index, urban location, higher price-adjusted Medicare payment rate, higher share of the population over 65, low hospital admission rates, high physician-to-population ratio, and high unemployment rate.
With regard to dropping out in 1999, some of the factors significantly associated with increased probability of participating are also significantly associated with staying in the program in 1999. For example, FEHBP participation is significantly associated with participating in Medicare and with NOT dropping out. Urban location, price-adjusted payment rates, and elderly population share all had this same pattern. Some factors, however, have the opposite effect on the drop-out decision than they did on the enrollment decision - fewer enrollees per physician, for example, is associated with higher probability of participating AND with higher probability of dropping out. Commercial service area changes, dramatic changes in total enrollment, number of commercial HMOs, and unemployment all have this effect of increasing both the chance that a plan participated AND that it dropped out. Finally, affiliation with a national chain and relatively lower shares of total enrollment in traditional HMO plan are significantly associated with the decision to drop out but are not important in the participation decision.