Medicare and other payers are continually tinkering with reimbursement methods in the belief, or hope, that the method used can control cost and/or utilization. Doesn’t always work out as planned, as reflected in a new National Bureau of Economics Research paper. (NBER Paper) Long-term care hospitals are one discharge location for patients following an acute admission and are usually used for pretty ill patients. Up until 2003, Medicare reimbursed these facilities on a per diem basis. Then it switched to a per diem payment for a certain number of days up to a threshold, at which point if the patient stays longer the facility gets one large payment for all additional days. Post-acute care has been growing rapidly; is responsible for much of the geographic variation in Medicare spending and has been the subject of many fraud and abuse investigations. So reimbursement to post-acute care providers has gathered a lot of attention. The researchers look at whether the switch in LTCH reimbursement methods changed provider behavior and whether a better system might be constructed. They used data from 2000 to 2002 and 2007 to 2012 to compare effects.
The number of stays in which the patient was discharged right on the day when the payment jumped went from 2% before the reimbursement change to 9% after. And those patients discharged on that threshold day were actually healthier than those discharged before that day and they had lower post-discharge mortality. But while these patients appear healthier and thus it seems questionable that they were held just long enough for the facility to get extra money, their longer LTCH stays do not appear, on an adjusted basis, to be associated with a decrease in mortality. Nor does the fact that they were discharged right on the threshold, and not kept longer, which would maximize profits for the facility, appear associated with an increase in mortality. Uhhh, and even more damning, for-profit LTCHs showed more of this gaming behavior. So it seems fairly clear that in a number of cases these facilities are just holding patients long enough to get the extra payment, and then discharging them before incurring a lot of additional care costs, without a lot of regard to whether the patients could have been discharged earlier or needed to stay longer. The authors further construct a model that suggests that if the jump in payments was eliminated, Medicare could reap substantial savings; over $10,000 per admission. Just further evidence that many providers are as concerned about financial results as patient outcomes.