WASHINGTON — Medicare's pay-for-performance criterion for hemoglobin A1c level is 9%, a value so outside the therapeutic target that endocrinologist Carlos Hamilton considers the measure to be useless.
“If we're going to ask doctors to fill out these forms, and ask them to do it for nothing, [the exercise] really needs to be meaningful,” Dr. Hamilton said at a meeting of the Practicing Physicians Advisory Council, which advises the Centers for Medicare and Medicaid Services. “If you ask them meaningless questions and then don't pay them to answer those questions, it just irritates them.”
Medicare's new Physician Voluntary Reporting Program, which began last January, rewards participating physicians with confidential feedback reports about how they are doing, compared with their peers. One of the program's 16 performance measures relates to whether patients with diabetes have hemoglobin A1c levels at or below 9%. A hemoglobin A1c level of 7% or less is considered to be the treatment goal by most endocrinology groups.
“You really have created so much angst on behalf of the endocrinology community,” Dr. Hamilton, who is executive vice president for external affairs at the University of Texas, Houston, told CMS representatives at the meeting. “I can't count the number of e-mails I've personally received about what this level really ought to be.
“Everyone knows that 9% or 8.6% is better than 13% or 14%, but it isn't enough better to really be important. What it ought to be is around 7%, or 6.5%, or 7.5%.”
Furthermore, just setting a single, arbitrary hemoglobin A1c value fails to acknowledge therapeutic progress among patients with difficult-to-control diabetes, Dr. Hamilton said. “What is much more important is whether that level has come down,” he added. For example, the reporting program could ask the treating physician whether the patient's hemoglobin A1c level is above or below 7.5%.
“If it's above 7.5%, your next question ought to be: Has this value decreased over the previous year, and has it decreased by 2 percentage points or more? If it has, and [the] value is, say, 8% or 8.5% or even 9.5%—and you've decreased it from, say, 14% to 9%—you've done a very good job. That's a much more meaningful statistic to have,” Dr. Hamilton said.
Dr. Michael Rapp, director of the quality measurement and health assessment group for the CMS's Office of Clinical Standards and Quality, told Dr. Hamilton that the CMS “is not in the business of setting up the standards. We're taking the standards given to us” by outside organizations. But he also noted that the National Committee for Quality Assurance, one of the measurement-development organizations that the CMS uses, recently proposed a quality measure for diabetes patients of less than 7% for hemoglobin A1c. “So I think you'll see there will [soon] be a measure available for us to use which will be hemoglobin A1c less than 7%.”
But taking this step alone only sets the goal and does not differentiate between physician quality and outcome measures in a meaningful way, Dr. Hamilton said.
“With hemoglobin A1c, it's not a matter of, did you order the right treatment for diabetes, or did [the patients] actually get the medicine, or did they refill the medicine after a month?” he said. “It's a matter of, did they take the medication, did they take it the way you prescribed it, and did they take it consistently, not for 30 days or 90 days but for several years?
“What you're testing with hemoglobin A1c is not just the quality of the doctor that wrote the prescription. You're testing the quality of the whole health care system in which patient compliance is by far the major component.”