By: Lisa Remington, President, Remington Health Strategy Group, Publisher, The Remington Report
To increase the equity of payments within each setting, the Commission recommends that the Congress direct the Secretary to begin blending the relative weights of the setting-specific payment systems and the unified PAC PPS in 2019 (i.e., before the implementation of the unified PAC PPS). The recommendation would redistribute payments across patients’ conditions within each setting, but would not affect the level of spending in each PAC setting.
The March 2018 MedPAC Report:
Per the Commission’s recommendation, the unifying of payments across settings would not begin until 2021, with the start of a three-year transition to a PAC PPS. At that point, payments begin to be redistributed across settings using a blend of the setting-specific base payment and the unified PAC PPS base payment.
Before the unified PAC PPS is fully implemented, CMS could use a blend of the unified PAC PPS relative weights and setting-specific relative weights to calculate payments, while keeping total payments to each setting at the recommended level. Over time, the blend would shift from having the setting-specific relative weights “count more” than the unified PAC PPS relative weights in 2019 to having the relative weights of the unified PAC PPS count more in 2020. But within each setting, aggregate payments would remain at the recommended level.
CMS would apply a budget neutrality factor to keep payments within a setting at the recommended level, which would prevent payments from shifting between the settings before the PAC PPS is implemented. Starting in 2021, the relative weights would be based entirely on the unified PAC PPS weights and the three-year transition to the unified PAC PPS would begin using the new system’s base rates.
During the transition to a fully implemented PAC PPS (2021 to 2023), the base payment would be a blend of the setting-specific base rate and the unified PAC PPS base rate (and using the unified PAC PPS relative weights). In the early years, each setting’s base rate would count more and the unified PAC PPS base rates would count less. In the later years, the unified PAC PPS base rates would count more until they are used exclusively to pay PAC providers. For example, in the first year of the transition, the payment for a stay treated in an IRF would be a blend of the IRF base rate times the unified PAC PPS relative weight and the PAC PPS base rate times the unified PAC PPS relative weight. Using the PAC PPS base rate to establish payments would result in the redistributions across settings, with larger shifts occurring as the “weight” of the PAC PPS base rate increases until it is used exclusively to establish payments in each PAC setting.
We estimated the effects of blended relative weights for the years before the implementation of the unified PAC PPS (2019 and 2020) using a 67:33 blend of current setting-specific relative weights and unified PAC PPS relative weights in 2019 and a 33:67 blend in 2020, while keeping payments at the current level of spending. We did not model any provider responses to the proposed changes.
Payments Across Patient Condition
Within each setting, aggregate payments remain the same, but payments would be redistributed considerably across patient conditions. The broad effects on different conditions would be similar across the four settings and illustrate the findings previously reported (Medicare Payment Advisory Commission 2017, Medicare Payment Advisory Commission 2016). Payments within each setting would increase for patients who are medically complex, including those who are chronically critically ill; patients with the highest level of severity; patients with comorbidities that involve multiple body systems; and patients who require severe wound care or ventilator care. For conditions that typically involve the provision of therapy services unrelated to a patient’s condition, payments would decrease for the majority of stays. The redistribution in payments is likely to make providers less reluctant to admit medically complex patients, thereby increasing those beneficiaries’ access to PAC.
For each PAC setting, the magnitude of the effects by condition would vary because the new system’s relative weights are based on the average cost of stays across the four settings, and these differ from each setting’s relative weights. In addition, the volume of a condition can be low for an individual PAC setting, so the effects for a particular condition will be driven by the costs of the stays in the other PAC settings. Further, the incentives of the current PPS and the provider behavior they have encouraged differ by setting. Thus, for example, the effects for conditions that may involve the overprovision of therapy services are likely to vary by setting. In contrast, the effects are far more uniform for medically complex conditions.
"Within each setting, aggregate payments remain the same, but payments would be redistributed considerably across patient conditions."
The effects of redistributed payments on providers within a setting are relevant to the update discussion. Across providers, average payments would be redistributed based on the mix of patients a provider treats, how a provider’s costs compare with the average, and whether the provider typically furnishes rehabilitation therapy that is unrelated to their patients’ conditions (and not based on the provider’s characteristics, per se). Across each setting’s providers, the effects would be consistent by ownership (for profit vs. nonprofit) and type (hospital based vs. freestanding). Average payments would increase for nonprofit providers and hospital-based providers and decrease for for-profit facilities and freestanding providers. To be clear, these changes in payments reflect the mix of patients treated by these providers and their therapy practices, not the provider characteristics themselves. The redistributions would have the effect of raising payments to low-margin providers and lowering payments to high-margin providers.
Three Benefits of Blending the Unified PAC PPS
Blending the unified PAC PPS and setting-specific relative weights has three benefits. First, it would start to correct the inequities of the current PPS, which create financial incentives for providers to favor treating certain conditions over others because the relative profitability of different conditions would narrow. Second, it would give providers even more time to adjust their practices to payments based on patient characteristics rather than the amount of rehabilitation services furnished or coding practices. Providers would have a financial incentive to change their therapy practices and align their costs with the blended payment even sooner than the full implementation of the unified PAC PPS because the changes encouraged by blended payments would be consistent with those that will be required to be successful under the new payment system. During the blending period, providers and CMS could learn important lessons applicable to the unified PAC PPS’s implementation. Last, because payments would be redistributed across conditions and the providers that treat them, policymakers would be less constrained in reducing payments to a level more closely aligned with the costs of care.