The Medicare Home health benefit – where eligible patients needing skilled nursing or therapy services are cared for at home – is a growing sector of the U.S. health care system.
Given that payment is partially based upon use of therapy services, there may be financial incentive to over-provide therapy. A new study from Abt Associates suggests that home health payments could change substantially if reimbursements were based on patients’ anticipated needs rather than actual services received.
The study, “Using Predicted Therapy Visits in the Medicare Home Health Prospective Payment System,” was published in Home Health Care Management and Practice in November. Abt researchers Betty Fout, Ph.D., Michael Plotzke, Ph.D. and Thomas Christian, Ph.D., examined the feasibility of using a payment model, based on patient characteristics, that predicts the use of therapy services rather than using the actual amounts received.
The Current Model
Home health is a growing segment of Medicare expenses in the U. S.: Medicare spent approximately $18 billion on fee-for-service home health payments in 2014, up 111 percent since 2000. The number of home health users increased 39 percent from 2.5 million to 3.4 million during that time. These numbers could continue to grow as more baby boomers enroll in Medicare.
Today, Medicare reimburses home health agencies for services provided to beneficiaries within a 60-day window. These payments, in part, consider patient diagnoses and functional characteristics and the use of therapy services. These services are categorized into thresholds or tiers based on the number of therapy visits provided. While this seems to be a reasonable method of paying for care, therapy thresholds have created incentives for home health agencies to perhaps provide unneeded services to reach a certain threshold.
Recognizing the conflict, prior CMS reports and the Medicare Payment Advisory Commission (MedPAC) recommended changes to the home health payment system. MedPAC, in particular, has stated that the existing thresholds creates an incentive for agencies to provide more therapy services regardless of clinical need. Yet, changing a payment system presents challenges.
Testing a New Model without Therapy Thresholds
As part of Abt’s larger work with the Centers for Medicare & Medicaid Services (CMS), Fout and colleagues created a model to predict therapy use based on patient characteristics. The model pays for patients’ anticipated therapy services, with a goal of ensuring that CMS pays only for needed services.
Fout says the team compared the model’s results with actual therapy use and payments from 2013. On average, the model’s predicted visits were similar to actual therapy visits.
“Where the differences show up is among the extremes, the high- and low-use levels,” said Fout, “and that’s where you see how difficult changing health care payment can be.”
The Inherent Conflict of Incentives
The authors explain that the model predicted less than one percent of patients would need high levels of therapy, meaning 20 or more therapy visits during a 60-day episode. In the data, almost six percent of home health episodes in 2013 received high-level therapy.
The other caveat is that the model predicted therapy use for a quarter of episodes when none were used.
“On one hand, we don’t want to incentivize agencies to provide costly therapy services when fewer are needed. But it is problematic to pay for services when none are used,” Fout observed.
Balancing Incentives with Patient Needs
The goal now is to find ways to remove the current home health payment incentive without hurting patients who need high levels of services, Fout said.The predicted therapy model is a good step towards minimizing therapy utilization as a rate determinant, but it does not eliminate it completely.
Abt is working with CMS on a variety of models looking at this issue, including the Home Health Grouping Model (HHGM). Additional details about the HHGM are available online.
“We want to ensure the home health payment system promotes efficient care that aligns payment with high quality services,” Fout says. “This needs to be supported by data analysis and careful consideration of unintended consequences.”
Click to learn more about our home health work presented at AcademyHealth 2016.