The ICD Impasse: Medically Right But Legally Wrong

A disconnect between best care and what Medicare is willing to reimburse for it has put some electrophysiologists and their hospitals in a bind. By charging Medicare for what they deem to be medically necessary implantable cardioverter-defibrillators (ICDs) for certain patients, they may be at risk of fraud and its penalties. Last October, that hypothetical scenario became a reality for 457 hospitals. 

Stuck in time

From the medical community’s perspective, the crux of the problem lies in Medicare’s national coverage determination (NCD) for primary prevention ICD therapy. The NCD, revised most recently in 2005, outlines what the Centers for Medicare & Medicaid Services (CMS) deemed reasonable and necessary treatment for its beneficiaries based on evidence from key clinical trials. Like most trials, they excluded some patients for ethical or practical reasons, making the results applicable to only a subset of the patients cardiologists treat in everyday practice. Over the decade, science progressed but the NCD remained unchanged, further widening the gap between regulations and clinical care.

“The problem with ICDs is that our national coverage determination is out of date,” says John D. Day, MD, director of Heart Rhythm Services at Intermountain Medical Center in Salt Lake City and

president of the Heart Rhythm Society (HRS). “Also, 10 to 20 percent of our patients do not fit nicely into these very tightly defined clinical scenarios that the NCD covers. When you have a very legalistic interpretation, it leaves 10 to 20 percent of patients who we know will benefit with this life-saving therapy out of luck because the NCD is either out of date or too tightly defined.”

The Patient Protection and Affordable Care Act included a vow to crack down on fraud and the many billions of dollars it costs the federal government. Using the False Claims Act as its enforcement tool, the Department of Justice (DOJ) expanded its targets from what Isaac D. Buck, JD, describes as conventional fraud to overtreatment, or care that Medicare sees as medically unnecessary. ICDs came under the DOJ’s scrutiny after a study concluded that about one in five implantations fell outside practice guidelines (JAMA 2011;305[1]:43-49). The NCD requires, among other things, that cardiologists wait at least 40 days after a myocardial infarction (MI) and 90 days after revascularization before implanting an ICD.

“This is one of the most perfect examples of the era we are in in healthcare fraud enforcement, where no longer is it defined by intentional fraudulent schemes; it is defined by excessive medical care,” says Buck, an assistant professor at Mercer University’s law school in Macon, Ga., who studies the tension between clinical and legal standards. “The battle becomes who says what is excessive.”

Not so cut & dry

On Oct. 30, 2015, the DOJ announced that it reached 70 settlements totaling $250 million with healthcare systems and their 457 hospitals that the DOJ claimed violated the False Claims Act by implanting ICDs in Medicare patients contrary to the NCD. During the expansive investigation, the agency released medical review guidelines and a resolution model for hospitals that outlined categories, or “buckets,” where it would not enforce penalties or seek repayment, plus the documentation hospitals were required to provide.

The buckets included scenarios such as coding “MI” in the prohibited timeframe that later was determined to not be an MI, and patients who met all of the criteria for a permanent pacemaker and an ICD except for the waiting-period limit. Coders sometimes misguidedly interpret an elevated troponin level as an MI when it actually is due to cardiac arrest, Day explains. Waiting three months after graft surgery to replace a pacemaker with an ICD may prove counter to quality care, adds Andrea M. Russo, MD, co-chair of 2013 ICD appropriate use criteria (AUC)

designed to inform treatment decisions for scenarios not addressed in the guidelines.

“Not only is that not in the best interest of that patient but it can be dangerous for that patient because we know revisions have a higher risk of infections or problems occurring after that,” says Russo, director of the electrophysiology and arrhythmia service at Cooper University Hospital in Camden, N.J. “On top of that, it would cost the healthcare system a lot more money than putting the ICD in upfront.”

The DOJ reviewed “clinical realities on a case-by-case basis,” DOJ spokesperson Nicole Navas wrote in an email response to questions. “Such determinations were informed by a panel of leading cardiologists retained by DOJ specifically for that purpose. The panel included Drs. Douglas Zipes, Alfred Buxton, Andrew Epstein, Robert Myerburg, John DiMarco, Mariell Jessup, Sharon Hunt, Joseph Alpert, Rita Redberg and Harlan Krumholz.”  

Russo, who as co-chair of the AUC document was asked to meet with DOJ members but was not a consultant, commends them for making the effort to understand the clinical nuances facing physicians. “They really did due diligence,” she says.

Buck also notes that the DOJ seems to be “giving some leeway in an enforcement mechanism, which I think is a good thing. But then it also indicates how out of step the reimbursement standard is.”

The DOJ emphasized that its resolution model did not replace or update the NCD. In a viewpoint article published after the AUC came out, Russo and colleagues also cautioned that the resolution model was applied retrospectively and that it did not give cardiologists a green light to disregard the NCD (J Am Coll Cardiol 2014;63[1]12-14). “We needed to do some kind of statement to help raise this awareness and help clarify that the AUC are not what is being reimbursed,” she says. “The resolution of the DOJ investigation doesn’t mean now you will be covered for these things, even if they decided not to [find] damages in that particular area.”

Policy pitfalls

The authors offered options for cardiologists and electrophysiologists who decide an ICD that falls outside the NCD is medically appropriate: consider an external defibrillator; don’t charge Medicare for the device or proactively contact a Medicare intermediary about the situation; or ask the patient to sign an Advance Beneficiary Notice acknowledging that he or she may have to pick up the tab. They advised documenting the case in the medical record with a detailed account of the reasoning, a strategy HRS advocates as well.

“This has created a lot of challenges,” Day says. “We are recommending that our members extensively document their decision-making process,” including citations to findings that support use of the therapy. At Intermountain, “if we have a patient who we know, based on medical science, will benefit but won’t fit a very tightly defined, outdated NCD, our hospital is donating the device rather than risk retribution. That is not right, either, but that is what is happening in the marketplace.”

The fear of a legal challenge for hospitals and physicians who want to treat outside the NCD may have other negative consequences as well, Buck points out. It may stymie the advancement of care in the field. “In the last 10 years, there have been presumably breakthroughs in the area where the old standard has become somewhat outdated,” Buck proposes. “If we have a government agency or enforcement mechanism applying standards from 10 years ago, then that will basically completely arrest the standard-of-care development. That is a big concern.”

Ultimately, Medicare beneficiaries may lose the most from this disconnect, according to electrophysiologists. Hospitals can negotiate with private payers in cases that don’t fit within the scope of the guidelines to ensure their patients get appropriate care, but that option is not feasible with Medicare, Russo says. “It doesn’t give you room for more individualized patient care. …You want to do the right thing for your patients, and you don’t want the Medicare patients to get inferior care to private patients,” she says. “You don’t want to treat patients differently.”

“What has been lost in this [situation] is the patient perspective and access to care,” Day agrees. “This gap is really putting patients and physicians at risk.” 

HRS and other cardiology societies are working with CMS to update the NCD, “but it is a slow process,” Day says.

Their experience with pacemakers may offer a tiny glimmer of hope. The 1984 NCD for pacemakers was updated in 2004 and then revised in 2013. With pacemakers, they assembled a consensus document that included “a huge body of evidence,” according to Russo. The data for ICDs in the three-month period after revascularization, on the other hand, is limited. “We clearly need more research in this area.”