Two weeks ago, Medicare announced a proposed settlement in a long-running billing dispute with hospitals. The size of the settlement is staggering: if accepted, it could end up costing the federal government hundreds of millions of dollars. The settlement also signals deep problems in one of Medicare’s flagship programs for eliminating wasteful spending.

At its core, the dispute turns on the distinction between “inpatient” and “observation” status. When patients show up at the emergency room, it’s not always clear whether they’re sick enough to require an extended hospital stay. For such patients, hospitals have a choice: they can either admit them or place them on “observation status” and keep them in the hospital for monitoring, sometimes for several days, before admitting or discharging them.

There’s no easy way to tell which patients should be admitted and which should go on observation status. But the decision can have big financial consequences for hospitals. (It also matters for beneficiaries, as I explained in an earlier post.) Hospitals are reimbursed under Part A for inpatient care, but for observation-status patients, they’re reimbursed under Part B, which is typically much less generous. All else being equal, hospitals would prefer to get paid under Part A.

Until recently, Medicare didn’t do much to discourage hospitals from admitting short-stay patients who might have been kept on observation status. That changed over the past decade, when Congress turned to Recovery Audit Contractors (RACs) to ferret out inappropriate billing practices. Spurred by the promise of a bounty—about 10% of whatever money they return to Medicare—RACs began to audit hospitals’ admissions practices. Among other things, the RACs concluded that hospitals were admitting too many patients and putting too few on observation status.

As a result of RAC audits, hospitals were ordered to pay back roughly $2.6 billion dollars over four years. They howled bloody murder, arguing that the RACs were insufficiently sensitive to physicians’ and hospitals’ medical judgments about when inpatient admission was warranted. To challenge RAC decisions, hospitals filed hundreds of thousands of administrative claims, collectively worth $1.5 billion. Of the appeals that have been heard to date, hospitals have won about two-thirds of them.

So many hospitals have filed appeals, however, that it has overwhelmed Medicare. A growing backlog of hundreds of thousands of cases means that newly filed claims will now take years to adjudicate, even though the Medicare statute ostensibly requires a decision within 90 days. By this past December, Medicare was under such pressure that it took the drastic step of suspending administrative appeals for a two-year period.

In other words, the hospitals crashed Medicare’s appeal process, much as hackers crash a website by inundating it with communication requests. Concerned, Congress has put a moratorium on RAC review of inpatient hospital admissions until April 2015. For its part, Medicare issued a rule last year to clarify when patients ought to be put on observation status. Hospitals are hardly mollified—they’ve filed suit to challenge the regulation. In the meantime, hundreds of thousands of RAC audits are languishing on appeal. (Matt Lawrence, a fellow at Harvard Law School, has a terrific draft paper offering some novel procedural options for helping to clear the backlog.)

To get out of this mess, Medicare’s proposed settlement offers hospitals roughly two-thirds of what they claim they’re owed—which, not coincidentally, is about how often hospitals succeed on appeal. Medicare’s hope is that a quick payout will persuade hospitals to drop their appeals. And really, what other choice does Medicare have? The hospitals have it over a barrel.

The settlement offer stands as a tacit acknowledgment of serious flaws in the RAC program. Although RACs tout how much money they’ve restored to Medicare, their high reversal rate suggests that the prospect of bounties has made them overzealous. It doesn’t help to recover money for Medicare if the program must pay most of it back. Nor do capricious after-the-fact reviews give hospitals the right incentives to avoid improper claims in the first place.

Worse, RACs’ auditing practices have provoked such ire from hospitals that Medicare has been brought to its knees. Alienating hospitals also makes it harder for Medicare to get their cooperation on other important programmatic changes. And for what? Medicare spends almost $600 billion per year. The total amount that RACs have saved is barely a rounding error.

We’ve gone down this road before. In 1972, shortly after a congressional report documented egregious waste and fraud in the young program, Medicare enlisted Physician Standards Review Organizations (later renamed Peer Review Organizations) to prevent unnecessary hospital admissions. PSROs did what RACs do now: they reviewed claims that hospitals submitted and rejected those that they deemed inappropriate. But, as Tim Jost has observed, “PSROs never succeeded in meeting the expectations of their supporters or overcoming the criticisms of their increasingly vocal detractors.” They saved no money, outraged physicians and hospitals, and embroiled Medicare in protracted litigation.

Medicare couldn’t audit its way out of wasteful spending back then. It can’t audit its way out of it now. RACs are an ill-fitting patch for a system that needs a much more comprehensive fix.