The dangerous Medicare loophole of observation status

On the morning of Sept. 23, 2014, Caroline Giada woke up in her home in coastal New Jersey. She rose from bed, walked over to the bathroom and fainted, hitting her back on the sink cabinet on the way down. When she came to, she couldn’t move. So she called 911. 

The 76-year-old was taken to the emergency room, where the attending nurses looked her over and booked her into a bed. Over the next six days, she was given every kind of test imaginable: X-rays, electrocardiogram, CT scan, ultrasound, MRI. Nurses came and went. So did doctors. Finally, a week later, the problem was uncovered: a fracture in the L5 segment of Giada’s lower spine. They called in a specialist, but he said he couldn’t operate until the following Thursday. 

Then the hospital kicked her out.

“They told me they couldn’t keep me, because they couldn’t do anything else for me,” says Giada. (At her request, her name has been changed to maintain her privacy.) Her choices were to go home or to a costly private nursing facility nearby. She chose the latter, thinking that given her intense pain and lack of mobility, she’d need the care the nurses could provide. And even though the facility demanded payment upfront, Medicare would reimburse her, she thought. It was a decision she would come to regret. “When you are in a lot of pain, you really don’t think too clearly,” she says. 

After her surgery and rehab, when she was ready to go home, Giada was hit with another surprise: a bill. Medicare had covered the surgery and the rehab, but not the nursing facility stay, and the facility told her she owed $2,360. “I couldn’t understand it,” says Giada. “They said, ‘Well, Medicare isn’t paying, because the hospital put you under observation instead of as an inpatient.’” That was the first she’d heard of it.

Observation is a designation that was meant to be exactly what it sounds like: a short period during which the hospital observes you to assess whether you need to be checked in for longer-term inpatient care or whether you can be quickly treated as an outpatient and sent home. According to the claims-processing manual published by the Centers for Medicare and Medicaid Services, or CMS, observation services are a “well-defined set of specific, clinically appropriate services” and in “the majority of cases” a patient is either admitted as an inpatient or discharged in less than 24 hours; “in only rare and exceptional cases” do they last more than 48 hours. 

But that definition doesn’t square with the data. In 2012, a group of researchers at Brown University undertook the first investigation of observation-status data, reviewing Medicare claims from 2007 to 2009. The results were striking and in complete discordance with the assessment in the manual. More than 10 percent of beneficiaries were placed on observation status for over 48 hours, for example, and the problem seemed to be growing: In those two years, there was an 88 percent increase in the number of patients held in observation status for more than 72 hours. A 2014 report to Congress published by the Medicare Payment Advisory Commission, an independent agency, found that there were 1.8 million observation claims submitted in 2012, an 88 percent increase from six years earlier. 

There are real consequences to the seemingly innocuous coding decision of observation vs. inpatient. Observation stays are considered outpatient services, which means they are paid for by Medicare Part B, whereas inpatient services are paid for by Part A. And while Part A covers the costs of all services after the deductible has been met, Part B generally only covers 80 percent, and it usually doesn’t cover the costs of drugs. Then there’s the even costlier problems observation status creates with skilled nursing facilities, or SNFs, which are care centers where patients usually recuperate postoperation. Medicare is happy to cover the SNF costs — as long as a patient has stayed in the hospital for a minimum of three nights beforehand. Time in observation status does not count toward those three nights. So when Giada was “observed” for seven days at the hospital, she never clocked in her three nights and was charged for her SNF stay. 

When she and her daughter Sarah Giada decided to appeal the charges, the nursing-facility administrators said their hands were tied, because the hospital had already coded the stay as observation. The hospital staff told Giada they were following Medicare rules, she says. The Medicare billing center said the rehab facility never filed the claim for reimbursement for the time Giada spent there. “It was almost like the rehab facility was blaming the hospital, the hospital was blaming Medicare, and Medicare was blaming the rehab center,” says Giada’s daughter. “It’s a constant circle.”

This type of bureaucratic stonewalling is common when patients fall into what some call the “observation-status loophole.” Over the phone, Gary Goodman is livid about what happened to his mother, Dorothy Goodman, early last year. After the 92-year-old was rushed to a nearby hospital in rural Idaho following a fall from her wheelchair, a doctor diagnosed her with a broken pelvis and wrote an order for inpatient services. But then the hospital sent the records off-site to a utilization-review committee — a group of administrators whose job it is, according to the American Health Lawyers Association, to “reduce unnecessary hospital admission and to control the length of stay for inpatients.” The verdict came back within hours: In order to comply with Medicare billing requirements, the hospital would have to change her status to observation — and Dorothy Goodman would have to foot the bill for her SNF stay.

Her son has spent the last year reading hundreds of pages of medical records and calling anyone who would listen (he’s even gotten his Congressman involved). He is now working with advocacy groups try to recoup the $22,000 he was forced — wrongly, he says — to fork over for his mother’s lengthy SNF stay following the surgery she underwent. “We can’t get anybody to answer questions,” says Gary Goodman. “Nobody wants to step up to the plate and do their job.”

Many other patients have similar stories, and physicians working in hospitals confirm that the decision often seems out of the attending health care worker’s control. Whatever the nurse or doctor writes down on your charts has to go past a review committee, which usually uses “clinical-decision support” software. “They are using these very impersonal pieces of software to make these decisions,” says Goodman. 

When the decision does not go in their favor, patients tend to blame the care providers, mostly because from their vantage point, there’s a simple solution: Hey, Doc, just change the code on my paperwork. But health care providers say they’re just as frustrated. 

Ann Sheehy, an internist and associated professor at the University of Wisconsin School of Medicine and Public Health, admits that hospitals do sometimes change a patient’s status retroactively, sometimes days into the hospital stay. However, she says, that’s because it’s not entirely up to the attending physician. A doctor, on seeing the patient, might write down “inpatient” only to be told after the fact by administrators that by law the right code was “observation.” 

“We do change statuses, not because we want to, but because we are trying our darndest not to commit Medicare fraud and to follow Medicare rules,” she says. “Sometimes patients bear the brunt of the consequences.”

The real problem, many argue, can be traced back to a flaw inherent in Medicare’s fee structure, instituted in the 1980s and still in use today, which determines reimbursement payments based on codes provided on the insurance claim. Since it was put into place, says Zhanlian Feng, one of the Brown researchers who worked on the 2012 paper, “if you look at inpatient services, there is only a tiny, tiny increase. But if you look at the volume and spending on outpatient services, it is explosive. It’s something like 160 percent.” In part, that’s because medicine has made tremendous technological advances that allow for previously lengthy treatment and recovery periods to be reduced to mere days or even hours.

“But another factor,” says Feng, “is the government effort to rein in cost growth. Outpatient is cheaper than inpatient” — at least for the government, which spends hundreds of billions each year on Medicare. In 2013, the most recent year for which data exist, total Medicare spending was $492 billion, or 14 percent of total federal spending. 

So it makes sense that CMS, mandated to provide health care to the country’s most at-risk citizens and to keep costs down while doing so, would look for efficient ways to keep spending down. One approach it has taken is to establish the recovery audit contractor, or RAC, program, which utilizes contractors to process claims. The program’s mission is to make sure health care providers aren’t overcharging Medicare for services provided — and to lower the overall incidence of Medicare fraud, which a 2011 report estimated could cost up to $98 billion each year.

Though well-intentioned, Sheehy says, the program doesn’t work as intended because the “business model favors overzealous auditing.” The RACs work on a contingency fee structure: The auditors are only paid for the Medicare dollars they recover for the government on cases audited. (They take a 9 percent to 12.5 percent fee.) “These contingency incentives favor aggressive auditing, without transparency, accountability or repercussions for cases that should never have been audited,” Sheehy testified last year at a meeting of the House Committee on Ways and Means Subcommittee on Health.

In a paper published just last month in the Journal of Hospital Medicine, Sheehy and a team of researchers looked at the audit and appeals data for Medicare Part A claims at three large medical centers. They found that RACs audited 8 percent of the hospitals’ Medicare Part A claims, well above the 0.3 percent baseline projected by CMS. In addition, not a single case audited and denied was a result of unnecessary medical care; all were about inpatient versus outpatient status. 

That trend, Sheehy says, “is likely because the inpatient claims are the most lucrative for the RACs. They know what to go after for the most money.” And they’re pretty good at what they do. According to CMS’ 2013 RAC report, the auditors recovered $3 billion for Medicare’s strongbox that year alone. But critics say that the presumed financial win is smoke and mirrors, and neglects to take into account the costs hospitals bear in fighting inappropriate payment denials — and the fact that the expenditure supposedly saved might be shifted back onto the very patients whom Medicare is meant to help. 

Hospitals have become increasingly likely to preemptively place a patient in observation status to avoid the lengthy and costly appeals process and to make sure they recoup some of the costs of administering care to these patients that they would otherwise lose if their Medicare claims were denied. 

When asked about the efficacy of the RAC fee structure, a CMS spokesperson responded in an email, “Section 1893 of the Social Security Act declares that Recovery Auditor payment shall be made on a contingent basis for collecting overpayments … The Recovery Auditor fee structure changes are sanctioned by Congress and negotiated through the procurement process.” In other words, it has have no control over the matter; that’s Congress’s job. Sheehy points out plenty of things CMS could do, though. One example would be adding a provision in the contracts that would make an RAC pay back the hospital’s fee to fight a denial that is overturned in the hospital’s favor. 

While there are no bills in the pipeline that would change the way RACs function, there are a number of legislative efforts in the works to close the observation-status loopholes. The one gaining the most traction is the Improving Access to Medicare Coverage Act, which would amend the rules so that time spent in observation status counts toward satisfying the three-day requirement. Though it has died in two previous congresses, the act was reintroduced in the Senate on Tuesday with both Democratic and Republican support. In a press release, one of the sponsoring Senators, Sherrod Brown (D-OH) stated that "this bipartisan bill would make sure seniors receive the care they need after hospitalization without additional costs so they can focus on their health and recovery instead of how they’ll afford their care.” 

In the meantime, many hospitals have already put the two-midnight rule into practice — and are seeing deterioration in the quality of care as a result. Last year, Marna Borgstrom, the president of the Yale New Haven Health System, told the Senate’s Special Committee on Aging that a process to determine status based on time sounds good on paper, but in practice ends up being arbitrary and counterproductive. Instead of making a patient status decision based on diagnoses, doctors are now forced to count hours — what if, for example, a patient comes in at 1 a.m. on a Monday and is discharged at 9 a.m. on Wednesday? She would be automatically considered outpatient, even though she stayed two nights. Missing an arbitrary cutoff hour, hospitalists say, should not matter. 

“CMS’ insistence that a patient’s designation hinge on time rather than clinical judgment means that the outcomes are often out of our hands,” Borgstrom said at the Senate meeting. “Our inability to reliably tell patients something as basic as whether they’re an inpatient or not undermines the trust between a doctor that is fundamental to so many aspects of the care relationship.” The Society of Hospital Medicine, a membership organization for hospitalists, recently published the results of a poll of their members about the two-midnight rule. Only 17.4 percent of respondents felt that the new rule was an improvement; 47 percent said it has had a negative impact on patient care. 

“I think it’s a hastily pushed out policy fix. It sounds quite arbitrary,” says Feng. “I assume CMS has done a lot of homework to get a good fix, but what they ended up with nobody likes. The hospitals don’t like it, and providers are confused.” 

And what’s worse is that patients suffer the most. Most Medicare beneficiaries harmed by the observation-status loophole end up with bigger debts than Giada’s $2,360, but less than Goodman’s $22,000. According to a 2013 report sent to the CMS by the Department of Health’s Office of Inspector General, “beneficiaries were liable for average SNF charges of $10,503.” For most people, especially seniors on a fixed income, that’s not pocket change. 

“Momma lives in an assisted-living place where she receives good care, but it averages $3,800 to $4,200 a month,” says Gary Goodman. “She doesn’t have a lot of money. She would be destitute if she hadn’t sold her home. The $22,000 could have paid for six months of her care. My dad and I worked years paying into this system [and] they figure out ways to deny you access. I think its thievery.”

Editor’s note: This story was updated to reflect the news that the Improving Access to Medicare Coverage Act was introduced today.