Dive Brief:
- The Trump administration is expanding a Medicare ambulance model nationwide requiring providers to get prior authorization before scheduling a non-emergent ambulance transport for a patient. But the implementation timeline is still up in the air and will depend on the state of the coronavirus pandemic.
- Repetitive, scheduled non-emergency ambulance transport (RSNAT) is covered under Medicare Part B for beneficiaries who need consistent medical appointments, usually for dialysis. The model, called the Medicare Prior Authorization Model for Repetitive, Scheduled Non-Emergency Ambulance Transport, was launched in 2014 in a bid to cut down on overuse of RSNAT.
- The model reduced RSNAT use by 63% and spending by 72% for beneficiaries with end-stage renal disease or severe pressure ulcers in its first four years. In the same time frame, it saved Medicare about $650 million without changing care quality or access, according to a September evaluation report.
Dive Insight:
Ambulance services are a prime example of misaligned incentives and variable pricing in the healthcare industry. A single ride can cost between hundreds and thousands of dollars depending on locations, services and the contractual agreements between providers and payers.
They're also a major contributor to surprise out-of-network bills, and are used much more often than they should be. A CMS program measuring improper Medicare payments consistently pegs ambulance services in the top 20 services in Part B with improper payments, according to the agency. An estimated 22.6% and 18.6% of non-emergent ambulance transport payments were improper in 2017 and 2018, respectively.
Between 2013 and 2016, CMS stopped allowing new ambulance suppliers in Medicare in areas of Texas, Pennsylvania and New Jersey over concerns about overuse.
CMS' innovation agency turned to prior authorization, a favored tool in insurers' arsenals for lowering healthcare costs and improper services where the payer approves a treatment or service in advance, to try to ameliorate the problem in the RSNAT model. Doctors generally argue the step adds administrative burden and delays patient access to care.
CMS in 2014 launched the model in states with high rates of RSNAT use: New Jersey, Pennsylvania and South Carolina. Two years later, North Carolina, Virginia, West Virginia, Maryland, Delaware and D.C. joined the model.
States in the first cohort saw average savings of of $440 per beneficiary, while those in the second reported savings of $121 per beneficiary. CMS Administrator Seema Verma called the model a "resounding success" in a late Tuesday statement.
The model was originally scheduled to end in those states in December, but will now continue without interruption, CMS said. The agency plans to release a timeline for the national expansion and implementation dates in more states "as it becomes available."
It's up in the air because of COVID-19, which CMS will take in account when deciding expansion timeline. Confirmed cases of COVID-19 are almost at the 7 million mark and deaths crossed 200,000 this week, as epidemiologists and public health experts continue to warn the situation could worsen due to the upcoming flu season. As of Wednesday, 22 states were reporting an uptick in cases, according to a tracker maintained by Johns Hopkins University.
CMS found no evidence the RSNAT model hurt care quality. It was associated with a small reduction in ER use and unplanned hospitalizations, but didn't have any impact on patient mortality. It also led to a small dip in dialysis use for ESRD beneficiaries but outcomes remained unchanged.
The model did, however, have a significant impact on ambulance companies.
The number of ambulance suppliers in Medicare dropped by half in the first cohort of states following model implementation, according to the evaluation report run for CMS by research firm Mathematica.
Suppliers that stopped billing Medicare for ambulance transportation before the model went into effect were smaller and less rural, and depended much more heavily on RSNAT services for revenue than suppliers that went on to participate in the model.
Suppliers that dropped out after a year also relied much more heavily on RSNAT than suppliers that stayed in the model long term. They could have expected they could absorb the lowered payments in the model, but ultimately elected to leave Medicare instead, researchers said.
The rate of claims denials rose after the model first began, but fell back toward the baseline over time, CMS found.
The RSNAT model lowered fee-for-service Medicare spending by about 2% in the four-year trial. Savings are likely to continue once expanded nationwide, but the savings will probably be smaller overall, researchers said, as the report draws on findings from model states that had a much higher rate of RSNAT use.
Last year, CMS launched another model, called ET3, to try to pare down ambulance costs. ET3 is more expansive than RSNAT, trying to nudge ambulance suppliers toward lower-acuity facilities for unscheduled emergency rides using value-based payments. The model, originally set to go into effect this year but pushed to 2021 due to COVID-19, is voluntary.