Nearly two decades ago, the files were closed on a very public scandal over excessive health care profits from inflated Medicare reimbursements to hospitals. The Centers of Medicaid and Medicaid Services (CMS), which administers the Medicare program, had accused health care companies of charging Medicare excessive payments in its “outlier program,” which is designed to cover extraordinarily costly episodes of care. The CMS investigation ended with those companies settling the matter without admitting to any wrongdoing.
But the case left unanswered questions for Atul Gupta, Wharton professor of health care management, and a couple of his research colleagues. They sensed that an analysis of how those hospitals used those excessive payments will bring important insights for CMS and regulators, and also inform contract design for any government program that third-party private firms implement.
“We felt it was like investigative journalism,” said Gupta of their findings, which they detailed in a recent paper titled “Turbocharging Profits? Contract Gaming and Revenue Allocation in Healthcare.” Gupta’s coauthors are Ambar La Forgia, a management professor at the University of California, Berkeley, and Adam Sacarny, a professor of health policy and management at Columbia University.”
How a Medicare Loophole Formed
According to the paper, “flawed implementation” of Medicare’s “outlier” payments program created a loophole that allowed hospitals to game the program and inflate their outlier payments. They achieved that by a method called “turbocharging,” where they rapidly increased their list prices or charges for treating patients that involve extraordinarily high costs. In their defense, hospitals accused of gaming the outlier program described their actions as a response to “flawed public policy, not fraud or illegal activity,” at a U.S. Senate hearing in 2003.
The authors “conservatively” estimated that gaming hospitals received $3 billion in excess Medicare payments by gaming the Medicare loophole between 1998 and 2003, before CMS closed that loophole. They chose to term that practice as “gaming” instead of “fraud” because they were unable to determine whether the hospitals acted with intent to defraud Medicare, they explained. The turbocharging hospitals raised their effective Medicare payment rates by 22% at the peak of the episode, the paper noted.
The study analyzed data on 120 gamer hospitals between 1994 and 2006, comprising 42 for-profit and 78 nonprofit hospitals. Its broader sample included nearly 1,400 non-gamer hospitals to study contrasts in hospital charges between gamers and non-gamers.
“We felt it was like investigative journalism.” – Atul Gupta
Identifying the gamers was not a straightforward proposition, and the authors had to devise an algorithm for that, Gupta said. Roughly 3,000 hospitals tap into the outlier program, and the study’s sample of 120 gamer hospitals is “an understatement,” which was arrived at after applying several filters, he added.
The Money Tracks of Nonprofits and For-profits
Although contract gaming imposes significant costs on government budgets and taxpayers, little is known about which firms engage in this behavior and how they allocate the “engineered” windfall, the paper stated.
The study found a consistent theme between how the hospitals used turbocharging to game the program and their ownership structure. Nonprofit hospitals deployed most funds toward operating costs. For-profits took those monies off their balance sheets and distributed a large share to executives and shareholders.
Nonprofit gamer hospitals deployed most of their excess revenue from turbocharging to increased spending on non-labor operating costs that could enhance care delivery, drugs and medical supplies, and costs of operating rooms and emergency rooms. Significantly, the study detected reductions in mortality only at nonprofits in the study sample. Nonprofit gamers reduced mortality rates by 3% following an 8% increase in Medicare spending during the study period, the paper noted.
But for-profit gamer hospitals transferred all of their excess revenue off their balance sheets. The paper noted that Dallas, TX-headquartered Tenet Healthcare, led that trend; it “dramatically increased” executive compensation and stock buybacks during the gaming period, which it estimated at approximately $1 billion.
Tenet was the largest for-profit company in the study sample, with 60 gamer hospitals. In June 2006, Tenet reached a settlement with the Department of Justice, agreeing to pay more than $900 million to resolve allegations of unlawful billing practices. That included $788 million relating to excessive “outlier” payments; other allegations related to claims of kickbacks to physicians, and irregularities in diagnosis codes to increase reimbursements.
Predictably, for-profit hospitals were the dominant turbochargers — their managers have “stronger incentives to maximize profits which they can distribute to themselves,” the paper stated. Nonprofit hospitals were less likely to indulge in turbocharging, which is consistent with their mission statements and their inability to distribute profits to their executives. Almost no government-owned hospital engaged in turbocharging.
According to the paper, the average gaming hospital increased Medicare and total revenue by around 10% between 1998 and 2003, “implying large spillovers on other payers.” Gupta explained that private insurers typically benchmark their payments to hospitals on the Medicare-approved list prices, called the chargemaster rates. The study found that the total hospital revenue increased by nearly $80 million at gamer hospitals. For 120 gaming hospitals, that works out to nearly $9.6 billion in higher revenues, Gupta noted.
“[The Centers of Medicaid and Medicaid Services] has an outsized effect on the economy, because a lot of private firms follow the rules that CMS sets.” – Atul Gupta
The Rise and Fall of Outlier Gaming
The paper traced the practice of turbocharging to 1983 when Medicare implemented a prospective payment system to reimburse hospitals for inpatient stays. In order to incentivize hospitals to admit patients who would be costly to treat, it created the outlier program. Under that program, hospitals could tap Medicare to pay 80% of costs that exceeded the deductible in its standard pricing formula for inpatient stays.
But hospitals could game that program by inflating their costs of treatment. “Obviously, this gives an incentive to a hospital to say its costs are 1,000, and not 100, because CMS will have to give them 80% of 1,000 instead of 80% of 100,” Gupta said. CMS did audits of the cost-to-charge ratios, but these audits took a long time. If the audits took four years, CMS used the old cost-to-charge ratios to pay hospitals, relying on the costs that hospitals reported.
CMS calculated those costs in “a convoluted fashion,” the paper noted. Hospitals reported the list price or “charges” for each patient stay, and CMS deflated this list price using a cost-to-charge ratio to arrive at the expected cost, the paper explained. The $3 billion worth of gaming with the outlier program size was a small fraction of CMS’s overall budget ($382 billion in 2002, and $1.5 trillion in 2024), and so, “it was not on its radar,” Gupta noted.
Hospitals could continue to game the system because they have “wide latitude to set these list prices, untethering them from actual costs,” the paper stated, citing other research. They could also exploit delays in finalizing their cost reports by increasing their cost-to-charge ratios by more than actual costs, it added.
The paper pointed to three other developments in the 1990s which encouraged hospitals to game Medicare’s outlier program. First, Medicare directed more funds to the outlier program by lowering the deductible for high-cost outlier payments, which increased the number of patients triggering the payments. Second, bureaucratic delays led to longer lag times to settle cost reports, and CMS deflated hospital charges with older cost-to-charge ratios for payments to hospitals. Third, a 1997 law called the Balanced Budget Act reduced Medicare payments to hospitals for standard procedures, but it left the outlier program largely untouched. The paper cited a hospital association’s suggestion that gaming occurred because of the pressures its members faced from the Balanced Budget Act.
The gaming scandal was first exposed by October 2002 by a financial analyst who showed that Tenet relied more heavily on outlier payments than previously known. A flurry of media reports followed, raising public concerns. Around that time, a whistleblower lawsuit accused a dozen hospitals in New Jersey and Pennsylvania of manipulating the outlier program, and the Justice Department began investigating Tenet.
The CMS responded quickly by closing the gaming loophole with several policy changes. It asked contractors to use more recent cost reports to compute cost-to-charge ratios, and created a framework to claw back excess payments. It also began expediting the audits and capped the growth in chargemaster rates at a certain level instead of using the charge ratios hospitals submitted. Consequently, the outlier payments showed a sudden drop in 2004. Federal agencies sued dozens of hospitals and health systems for fraudulent billing.
“In prioritizing fraud prevention energy and dollars, [government fraud prevention programs] should probably focus on for-profit firms.” – Atul Gupta
Unfinished Work After Plugging the Loophole
Although the CMS has effectively plugged the loophole, “in theory a hospital could still mess with the outlier program (or game it),” Gupta noted. CMS does not require hospitals to maintain separate accounts for how they use the outlier payments, so it is hard to track them, he added. In fact, the study provided “evidence that gaming was far more widespread than the set of hospitals that were sued,” the paper pointed out.
According to Gupta, the government, CMS and regulators have their work cut out in light of the study’s findings. He identified three takeaways. First, government fraud prevention programs “might get a greater bang for their buck” if they focus on for-profit vendors, he said. Second, CMS and other government agencies must factor in the “spillover effect” in contract design and how they set prices, which affect a greater number of participants than those in their immediate sights.
“CMS has an outsized effect on the economy, because a lot of private firms follow the rules that CMS sets,” he said. In the outlier gaming scandal, much of the excess payments came from employer-sponsored insurance, and “that is ultimately from people like you and me,” he added. “When the government considers investing in fraud prevention and data security to ensure the integrity of its programs, they have to internalize the fact that the effects on the economy are much bigger than on the programs themselves.”
The third takeaway is in the heterogeneity of how gaming hospitals showed divergence in how they allocated their excess revenues, based on their ownership structure: for-profits spent that money on executive compensation and stock buybacks, while nonprofits invested in operating expenses. “This just strengthens the first takeaway … that in prioritizing fraud prevention energy and dollars, you should probably focus on for-profit firms.”
“Contract gaming and fraud contributes to the hundreds of billions of dollars the federal government spends annually on improper payments,” the paper noted, citing prior studies. Design flaws or ambiguities in these contracts provide opportunities for “gaming” in which firms strategically exploit these contract imperfections to increase revenue beyond the intention of policymakers, it explained. Medicare accounts for 12% of federal expenditures but 34% of improper payments, the paper noted, citing government data.
