Home Health & Hospice Week

What measures will your reimbursement depend on?

You can't just ignore pay for performance--the mechanism could determine your Medicare payment rates sooner than you think.

That was the message of multiple presenters at the National Association for Home Care & Hospice's annual policy conference in Washington, DC the last week of March.

The Centers for Medicare & Medicaid Services is shifting its payment policies for many provider types toward "quality and outcomes," noted Laurence Wilson, director of CMS' Chronic Care Policy Group, in a March 27 CMS panel on regulatory and policy issues. Pay-for-performance models are an important part of that focus, Wilson told attendees.

Be prepared: In particular, CMS plans a dem-onstration project to test a home care P4P model, Wil-son revealed. After thorough research and model testing, CMS would roll out a P4P model to the entire industry.

"This is really an imminent change in the payment system," agreed Sharon Bee Cheng, senior analyst for the Medicare Payment Advisory Commission, in a separate March 27 presentation focusing on P4P. Congress has essentially made up its mind to implement P4P and is merely working out the details now before passing requirements, Cheng told conference attendees.

Here are the major details a home care P4P model will have to nail down:

• Measures. Congress and CMS must decide on which measures agencies will be judged--and paid. MedPAC recommends starting with the current outcome-based quality improvement (OBQI) measures agencies are using, particularly those on Home Health Compare, Cheng noted.

But P4P shouldn't end with OBQI measures, Cheng suggested. Instead, the program should evolve to include OBQM measures, patient "experience" measures and process-based measures, which MedPAC appears poised to endorse in its June report to Congress (see Eli's HCW, Vol. XV, No. 12).

Control issues: Providers actually prefer process measures because they have direct control over them, said Amanda Twiss, president of Seattle, WA-based benchmarking company Outcome Concept Systems, in the same presentation. On the other hand, many factors outside of an agency's control tend to influence patient outcome measures.

P4P measures should be well accepted by the industry, evidence-based, based on data collected with a standardized tool and risk adjusted, Cheng said. And providers should be able to improve on them. "We want as great an impact on as many patients as we can," she noted. Moving a score from 98 to 99 percent "is not a lot of bang for the buck."

• Payment structure. The concept behind P4P is to reward providers for quality care, but how much is enough? P4P programs for private payors range from a 1 percent bonus to as much as 10 percent, Twiss noted.

MedPAC recommends starting with a small portion of payments that grows over time, Cheng said. A good place to start might be shaving 1 percent of payments off the base episode payment rate to form an award pool. Agencies could receive different bonus amounts depending on performance, she added.

Warning: But CMS must be careful not to create a two-tiered system of care, warned Chicago, IL-based regulatory consultant Rebecca Friedman Zuber during the P4P session's question-and-answer portion. Taking money from poor performers to pay top performers encourages two levels of care, Zuber criticized.

• Who's rewarded. A P4P model can reward two categories--providers who achieve top measure scores, or providers who improve their measure scores the most, Twiss noted. While a P4P program should reward top performers, it shouldn't create access problems for difficult patient populations.

Policymakers also must decide whether to set an absolute value for agencies to achieve or a percentage of top performers. For example, a P4P model could reward agencies that have a 25 percent hospitalization rate or those who are in the top 10 percent nationwide for hospitalization rates.

An absolute measure is a specific number to shoot for and is easier for providers to grasp, Twiss said. The bonus of a comparative approach is that the bar raises for everyone as outcomes improve.

Problem: "The risk adjustment has to be adequate to pull (P4P) off," Cheng admitted.

Even for the OBQI measures endorsed by the National Quality Forum and displayed on Home Health Compare, the risk adjustment doesn't completely account for patient differences, Twiss noted with data analysis for 2005. Even including risk adjustment, outcomes differed for patients in different diagnosis groups, long stays versus short stays and different regions of the country, the OCS analysis showed.

Chronic patients generally fared the worst, even under risk adjustment, Twiss noted. A P4P program based on the current OBQI measures could create access problems for these patients.

Timeline: Cheng expects Congress to pass P4P legislation this year or next. The Deficit Reduction Act enacted in February requires MedPAC to make home health P4P recommendations in a June 2007 report.

CMS could have a P4P model ready to roll as early as 2008, after a demo, Wilson predicted.

Note: For tips on preparing for P4P, see a future issue of Eli's Home Care Week.