The analysis indicates that the likelihood that Medicare beneficiaries transition to MME status is associated with the types of services they use, particularly nursing home care (Table 2). For example, among younger Medicare-only beneficiaries with no service utilization--"no ER visit leading to an inpatient admission, no SNF stay, and no non-SNF stay in the past six months"--only 4.9 per thousand (or 0.49 percent) transitioned to MME status. The transition rate increases slightly to 7.1 per thousand when beneficiaries have an "ER visit resulting in an inpatient stay." Thus, an "ER visit resulting in an inpatient stay" is associated with a 2.2 per thousand (7.1 minus 4.9) increase in the chance of transitioning to MME status relative to the base transition rate associated with no service utilization.
|TABLE 2. Predicted Rates of Transition from Medicare-Only to Medicare-Medicaid Enrollment for Various Pathways, by Age Group|
|Service Utilization in Past Six Months||Age Under 65||Age 65 and Above|
|No ER Visit Resulting in Inpatient Stay, SNF Stay, or Other Nursing Home Stay||4.9||---||0.8||---|
|ER Visit Resulting in Inpatient Stay Only||7.1||2.2*||1.6||0.8*|
|SNF Stay Only||12.3||7.5*||3.9||3.1*|
|ER Visit Resulting in Inpatient Stay, and SNF Stay||16.5||11.6*||5.4||4.7*|
|ER Visit Resulting in Inpatient Stay, and Other Nursing Home Stay||16.5||11.7*||7.6||6.8*|
|Other Nursing Home Stay Only||16.7||11.8*||7.5||6.7*|
|SNF and Other Nursing Home Stay||26.1||21.2*||9.3||8.5*|
|ER Visit Resulting in Inpatient Stay, SNF Stay, and Other Nursing Home Stay||27.2||22.3*||10.8||10.0*|
|SOURCE: Mathematica analysis of 2008 and 2009 MBSF, CCW Timeline, and Medicare Claims Files.
NOTES: 2,480,450 person-month observations for the under 65 model. 14,327,183 person-month observations for the 65 and above model. Observations are weighted to be representative of the population.
* Statistically different from base rate at the 1% level.
Regardless of the age group, a nursing home stay of any kind is associated with increases in the rate of transition to MME status whether or not that stay is preceded by the use of acute care services. Among beneficiaries under 65, a nursing home stay is associated with increases in transition rates ranging from 7.5 per thousand among those who only have a SNF stay to 22.3 per thousand among those who have an "ER visit resulting in an inpatient stay, a SNF stay, and a non-SNF nursing home stay." Similarly, among older beneficiaries, a SNF stay is associated with a 3.1 per thousand increase while an "ER visit resulting in an inpatient stay, a SNF stay, and a non-SNF nursing home stay" is associated with a 10 per thousand increase.
The service combination that is associated with the highest probability of transitioning to MME status is when a beneficiary has an "ER visit resulting in an inpatient stay, a SNF stay, and a non-SNF nursing home stay" all within the past six months: 27.2 per thousand for under 65 and 10.8 per thousand for 65 and above. This combination likely captures the pathway where a beneficiary first suffers an acute medical episode for which they enter the ER and then end up being admitted for an inpatient stay; the beneficiary is then discharged to a Medicare-financed SNF stay for subacute care or rehabilitation, followed by ongoing nursing home care after the Medicare benefit runs out. The data do not tell us if the beneficiary spent down to meet Medicaid resource criteria because of this adverse sequence of events, or would have had sufficiently low resources irrespective of the events, but presumably some do.
Notably, the magnitude of the transition rates, both in levels as well as differences relative to the base rate of no service utilization, are substantially higher among the under 65 than the 65 and above, which is consistent with what Borck et al. (2013) find in their descriptive study of the transitions to MME status.9 For example, the pathway "ER visit resulting in an inpatient stay, a SNF stay, and a non-SNF nursing home stay" described earlier is associated with a transition rate of 27.2 per thousand among the under 65, which is more than twice the transition rate of 10.8 per thousand for the same pathway among the 65 and above. Moreover, the respective differences relative to the base rate are 22.3 and 10 per thousand, respectively--again a more substantial transition rate among the under 65.10
We also find that the other predictor variables are associated with an increased probability of transitioning to MME status (estimates are found in the Appendix, Table A1). We find that a planned inpatient stay is associated with a 4 percent and 2 percent higher odds of entering MME status for the under 65 and the 65 and above, respectively. An outpatient ER visit is associated with a 7 percent and 9 percent higher odds, respectively, of transitioning to MME status, perhaps reflective of a more severe medical episode. A week of home health use is associated with less than a 1 percent increase, though this is not trivial for a long duration of home health use. The model we estimated also included chronic condition indicators, which we will address in a later section.