Late data for CMS bundled-pay program gives providers little decision time

Hospitals and physician groups that want to participate in the CMS' new Medicare bundled-payment demonstration are sweating because the agency is behind schedule in giving providers the claims data they need to decide which care bundles to select.

The CMS said it would share data in May with all provider groups that applied to participate in the voluntary Bundled Payment for Care Improvement Advanced program, which starts in October and runs through 2023.

The delay in receiving claims data creates a time crunch for applicant providers, who must tell the CMS by Aug. 1 which of the 32 bundled-payment clinical episodes they want to participate in. They need to see their past Medicare claims data to determine how financially viable participating in the different bundles would be. At this point they can still drop out partly or entirely.

In an email sent to program applicants Monday, the CMS said it had transmitted a large portion of applicants' pricing data. It added that over "the next few days" it would finish up sending those files and "will work towards providing raw claims files."

"We think it will take a month to build the analytical platform, which would give our clients only a month to go through the episode selection process," said Jonathan Pearce, a principal at Singletrack Analytics who helps providers with data analytics for bundled-payment programs. "They want to be able to make some decisions, and the CMS just isn't giving them any time."

The CMS did not respond to a request for comment by deadline Monday.

The agency hired a different contractor, Acumen, to administer BPCI Advanced. Mathematica Policy Research administered the original BPCI demonstration.

Under BPCI Advanced, the CMS will pay providers a fixed price for an episode of care, with that price announced in advance. The episode starts with an initial hospital admission or outpatient procedure and includes all care during the next 90 days.

Providers will make money if they keep total costs below a benchmark price, discounted by 3%. That's a lower target price than in the current BPCI program, which includes just a 2% discount. They will be at risk for up to 20% of costs that exceed the target price. Savings payments will be adjusted based on performance on seven quality measures.

Unlike in the original BPCI program, there will be a risk-adjustment formula for payments. It will take into account the illness severity of each provider's patient population, the type of hospital and other factors.

The new program has the potential to pull a larger group of hospitals and physician groups into bundled payments than the 1,000 or so providers now participating in BPCI. That could significantly accelerate the shift from fee-for-service to value-based payment models, which HHS Secretary Alex Azar has said is a high priority for the Trump administration.

Physician groups are attracted by the fact that participating in the new demonstration qualifies as an advanced alternative payment model under Medicare's Quality Payment Program, making doctors eligible for performance-based 5% bonuses.

Many observers hope BPCI Advanced encourages more provider groups to take on complex episodes of care such as congestive heart failure and chronic obstructive pulmonary disease. The goal is to redesign care to achieve cost and quality improvements for these challenging patient groups.

Many providers participating in the original BPCI demonstration have successfully improved quality and reduced costs for patients receiving total knee and hip replacements. That's the bundle type with the widest participation by far in the original BPCI demonstration.

In contrast, providers have struggled with managing congestive heart failure patients, who can be particularly challenging due to the unstable nature of their condition, multiple comorbidities, and the difficulty of changing bad diet and health habits.

Fred Bentley, vice president for the provider practice at Avalere, said there's substantial opportunity for providers to reduce readmissions and post-acute utilization for patients with complex conditions such as congestive heart failure and liver disorders.

A new Avalere analysis found that readmission rates for certain BPCI bundled-payment episodes were up to six times higher than rates for other conditions. For instance, the rate for liver disorders was 43%; congestive heart failure, 36%; acute myocardial infarction, 34%; COPD, 29%; and renal failure, 28%. In contrast, the readmission rate was 8% for lower joint replacement procedures, which is the most popular bundle in the original BPCI program.

Providers can succeed financially with bundled payments for those complex episodes of care by making gradual, sustained improvements, according to Bentley.

"Conditions with the highest readmission rates should be the first things providers should look at because they're a huge driver of costs," he said. "It takes a lot of work to change that. But if providers can get their rate to the market average, they can succeed in the program."

Bentley said some provider groups that have applied for BPCI Advanced are collaborating with physicians who have no previous experience with bundles but are interested because it offers them the potential of receiving a 5% bonus potential under Medicare's Quality Payment Program.

The CMS claims data will be an eye-opening educational experience for these newbie physicians. "It's fascinating because this will be the first time some of these physicians are seeing their performance data compared to the market," he said. "They've never had an incentive to look at this before."

The data also will bring tough choices for hospitals. They will have to look at their past cost performance and pick episodes of care where they think they can achieve enough cost savings to make money on the CMS target price.

Once provider groups select from the 29 inpatient and three outpatient episodes of care by the Aug. 1 deadline, they will be locked into the program from Oct. 1, 2018, through the end of 2019. But they're only going to have a few weeks to make that decision.

"I think this will drive down participation," Pearce said. "If an applicant doesn't have time to do due diligence on an episode before Aug. 1, they'll probably choose not to participate in that episode."