Incentives and Long-term Care

The search for the perfect provider reimbursement method goes on, as does the one for the ideal way to incent patient behavior.  The National Bureau of Economic Research issues a paper examining the impact of incentives in long-term care.   (NBER Paper)   The authors looked at over one million nursing home stays in four large states from 2000 to 2005 to discern how patient cost-sharing and nursing facility payment method and amount might affect length of stay.  Nursing home stays cost over $350 billion a year and more than half of this is paid for by Medicaid, causing a substantial strain to state budgets.  States have been trying to figure out how to encourage more care at home, which is less costly.  As you might imagine, when they have lower occupancy rates, nursing homes are happy to take even relatively low-paying Medicaid beneficiaries rather than have an empty bed, but when occupancy is near capacity, the facilities might be more likely to discharge Medicaid recipients in order to serve more private-pay or commercially insured patients, who typically have higher reimbursements.  And patients might be more willing to stay longer in a nursing home if they are on Medicaid and basically have no out-of-pocket expense, compared to when they do have an out-of-pocket cost.

And the analysis results suggest that while both patient and provider financial incentives matter, providers react more strongly.  At relatively low occupancy rates, the discharge rates for Medicaid patients is about half that of private pay patients, suggesting that the facilities are hanging on to the patients to keep revenue coming in.  At relatively full capacity, there is an increase in Medicaid discharge rates but not for private patients.  There is a larger difference in discharge rates for facilities that have higher private pay rates and there is a greater difference in discharge rates when a facility is both near capacity and has higher private pay reimbursement.  There were no apparent quality impacts from patients having shorter lengths of stay.  Earlier discharge was not linked to more subsequent hospitalizations or greater mortality.  The results strongly suggest that Medicaid programs are paying for unnecessary time in a nursing home.  Most states are still using per day reimbursements, so the authors make the obvious suggestion of moving to an episode form of payment.  In addition, differential treatment of private pay and Medicaid recipients could be limited by moving to some all-payer reimbursement scheme.