When the Allegheny Health, Education, and Research Foundation (AHERF) declared bankruptcy on July 21, 1998, it left behind $1.3 billion in debt, 65,000 creditors and enough bile to blanket the East coast.

The bankruptcy meant the dismantling of the largest statewide integrated delivery system in Pennsylvania. It led to thousands of layoffs in Philadelphia’s health care community and the sale of six local hospitals to an out-of-state, investor-owned corporation. And it called into question a strategy – popular among academic medical centers in the mid-1990s – of acquiring, at almost any cost, physicians, researchers and medical facilities in order to corner a market that has turned out to be remarkably elusive.

In an article that appeared last week in Health Affairs titled, “The Fall of the House of AHERF: The Allegheny Bankruptcy,” Wharton professor Robert Burns, Penn School of Medicine fellow John Cacciamani, consultant James Clement and nurse manager Welman Aquino analyze the developments behind Allegheny’s spectacular rise and equally spectacular fall. As the authors note, no one single factor led to AHERF’s demise. Instead, there was what Burns calls “a huge breakdown in accountability at almost every level.” Indeed, much of “the intrigue of the story stems from the fact that so many actors, both inside and outside the company, appear to have played a part.” Among those who failed both AHERF and the communities it served were the organization’s top management, board of directors, accountants, auditors and bond-rating agencies.

Even with this cast of hundreds, however, two individuals stand out as particularly deserving of censure. The authors describe at length the center stage roles of AHERF CEO Sherif Abdelhak and CFO David McConnell, whose flawed assumptions, fiscal irresponsibility, and questionable ethical decisions will be debated and litigated for years to come. For now, the two executives, along with several members of AHERF’s management team and board, are the subject of an unsecured creditors’ lawsuit, two grand jury investigations (in Pittsburgh and Philadelphia), and investigations by the SEC, the Pennsylvania State attorney general and the Pension Benefit Guaranty Corp. Additional lawsuits may also be filed against AHERF’s auditor, Coopers and Lybrand, now part of PricewaterhouseCoopers.

The authors also devote a large section of their article to an analysis of the bond market, in part because “nobody knows much about it,” says Burns. “Our findings should be especially useful to trustees of hospitals who may not realize, for example, that an Aaa rating refers to the rating of the insurer, not the rating of the hospital … Misunderstandings like this all along the line helped mask AHERF’s true financial condition.”

The article points out other lessons as well, including the industry-wide need for oversight mechanisms to protect hospitals’ charitable assets, the vast majority of which are endowments and other restricted accounts. “These mechanisms are available but the Allegheny board didn’t use them” until too late, Burns says. “Other hospitals should.”

The larger question – of what AHERF’s bankruptcy means for other academic medical centers – is not easily answered. “Few other centers operate in exactly the same way Allegheny did,” notes Burns, “although all of them have suffered from the 1997 reimbursement changes [which lowered payments to hospitals for treatment of Medicare and Medicaid patients]. I think there will be a number of other bankruptcies down the road – similar situations now face Detroit Medical Center, the University of Pennsylvania Health System and other academic medical centers – but I doubt that it will be widespread.”

The outlook for Philadelphia-area hospitals in particular won’t be helped by AHERF’s bankruptcy filing. Hospital bonds in this region “have become unpopular, both new issues and the secondary market,” the authors note. As a result, “it will now be more difficult and expensive for hospitals to upgrade their existing plant and capacity as they compete.”

In the meantime, Burns notes, “hospitals should continue their current plan, which is to cut costs and raise short-term revenues in whatever way they can.”

A Flawed Strategy

Abdelhak, named CEO of AHERF in 1986, had a mandate to change AHERF and its affiliate, Pittsburgh-based Allegheny General Hospital (AGH), into a premier medical education and research institution. His strategy, the authors write, can be summarized as follows: To develop Pennsylvania’s first statewide integrated delivery system grounded in academic medicine; to build regional market share to leverage managed care payers; to garner contracts from HMOs that covered all the costs of enrollees’ health care; to achieve synergies among the assets acquired; and to use community/suburban hospitals to refer private-pay patients to AHERF’s teaching hospitals.

For various reasons, the article says, “all five elements of AHERF’s strategy were questionable.” First, Pennsylvania has few statewide payers, other than Medicaid and U.S. Healthcare, or employers, other than banks, that wished to contract with a statewide integrated delivery system; second, few of these delivery systems have enough market share to leverage managed care payers, especially in markets such as Philadelphia which has an excess of hospital beds and physicians; third, hospitals’ enthusiasm for assuming risk, and HMOs’ reluctance to pass it on, have meant huge hospital losses; fourth, synergies and economies of scale through mergers are difficult to realize, especially when the organization expands so rapidly, as AHERF did, that it fails to do the necessary integration; and fifth, academic medical centers in Philadelphia had trouble persuading suburbanites to use older teaching hospitals in the city, even as suburban hospitals were developing revenue-generating services that attracted urban patients.

“These five assumptions were serious miscalculations, but it’s important to remember that they were common-sense wisdom at the time,” says Burns. “Everyone believed them. Consultants were propounding them. The trade literature was repeating them. After a while, if enough people repeat these things, you end up believing them. A lot of other medical centers, including Penn’s, went the same route.”

AHERF’s expansion into Philadelphia was conducted at a break-neck pace. In 1987 AHERF acquired Medical College of Pennsylvania and its two affiliated hospitals; in 1991, it bought United Hospitals’ four hospitals and Suburban Medical Associates; in 1993, it purchased Hahnemann Medical College and hospital and in 1996-97, Graduate Health System and its six hospitals.

“People commonly refer to the purchase of the Graduate System as the straw that broke the camel’s back,” Burns says. “It was a huge debt load and the hospitals were marginal performers financially. There was no way to manage it all.”

Debt load, in fact, became a major problem for AHERF. The article points out that during the 1980s, when the newly-established AHERF consisted of Pittsburgh-based Allegheny General Hospital, AGH had just $67 million in debt and an enviable 15% margin. It was one of only 40 hospitals nationwide with an Aa bond rating. By 1998, a hugely-expanded AHERF, which by then had been using AGH and other facilities in the western part of the state to subsidize its eastern operations, had a debt load of $1.2 billion. Margins during the mid-1990s fell to between 0% and 3%, compared to 6% to 12% at competing hospitals.

An Inhospitable Environment

As part of his vertical integration strategy, Abdelhak was also buying up PCP (primary care physician) practices in the mistaken belief that this would help secure inpatient referrals to the system’s hospitals. Unfortunately for Allegheny, other Philadelphia systems – including Penn, Temple and Graduate (before it was acquired) – were following the same strategy. The bidding wars that resulted meant that Allegheny frequently overpaid for these practices.

The researchers cite two purchase agreements filed in Common Pleas Court showing that PCPs received $70,000-$150,000 for their assets, an average annual salary of $220,000-$250,000 for five years, and 60% of the revenues above $470,000-$570,000. In addition, the contracts included no means to monitor practice productivity, even as the average age of the PCP provider was reported to be 56. These were individuals “eager to sell and less eager to continue working hard.”

Along with aggressively acquiring PCP practices, AHERF was also recruiting clinical and research faculty, which it hoped would attract federal funding, an enhanced research reputation, more patients and thus new revenue sources. Again, high payment was the norm. AHERF, the article says, recruited three orthopedists in 1997 who received guaranteed salaries of $3.9 million a year.

Meanwhile, AHERF’s precarious financial condition was significantly aggravated by what the researchers describe as Philadelphia’s “totally unforgiving” health care environment. In 1994 Pittsburgh had only five health plans and 15% HMO penetration. By contrast, Philadelphia had 11 health plans and 30% HMO penetration. “Moreover, two plans – U.S. Healthcare and Keystone Blue Cross – dominated the HMO market with 74% share” which meant they had enormous power over the providers. The fact that Philadelphia – with its five major academic medical centers vying for market share and research funding – was “overbedded, overdoctored, overstaffed and overutilized” only increased the HMOs’ bargaining power, the authors note.

On top of all this, in 1997 commercial insurers and public payers (Medicare and Medicaid) all lowered their reimbursement rates. The sudden decline in revenues hit an already overextended AHERF particularly hard.

Yet Abdelhak, as late as January 1998, was praising AHERF’s growth, productivity improvements, growing market share and physician networks. What he wasn’t disclosing, the article says, were the “financing mechanisms used to fuel the growth, including internal subsidies, hidden internal cash transfers, raids on hospital endowments and the enormous debt piled up from all of the acquisitions.”

Behind the Ratings

In their section on Allegheny’s bond-rating agencies, the researchers explain some of the more complex and little-understood issues related to hospital debt. One example: Moody’s Investor Service covered debt issued by five sets of AHERF hospitals, some of which found their debt publicly downgraded during 1996-98 as AHERF’s financial problems grew. (Graduate Health System’s debt, for instance, was junk-bond status – Ba2 – at the end of 1996 and was repeatedly downgraded throughout 1998.)

However, another AHERF hospital set – the Delaware Valley Obligated Group (DVOG), which consisted of Medical College of Pennsylvania, Hahnemann and United Hospitals – “had bonds separately rated before 1996 as Aaa insured (highest rating possible), Baa (barely above investment grade) and Ba (below investment grade). In June 1996, AHERF called the bonds, refinanced and reissued them under DVOG, and insured them,” the article notes. AHERF used MBIA Insurance, the largest health care bond insurer, as its underwriter.

“Some interpreted this as a signal that the insurance companies were beginning to approve AHERF’s business strategy,” the researchers say. “By insuring the debt, AHERF garnered an Aaa rating for DVOG’s debt.” Yet in reality this rating had nothing to do with “any improvement in the underlying financial health of the system issuing the bonds. Rather the rating reflected the underlying health of the insurance company (MBIA’s rating by Moody’s) to insure that debt. The financial troubles at DVOG thus remained hidden from the public.”

Meanwhile, the researchers point out, although Moody’s own internal assessment noted weak operations overall at AHERF, “the common practice at the time was not to publish the underlying ratings unless requested by the issuers.” AHERF understandably didn’t make that request.

It wasn’t until 1998 that Moody’s was pressured by the investment community to release its underlying rating of DVOG. On July 8, 13 days before AHERF declared bankruptcy, debt which had been rated Aaa “suddenly had a publicized underlying rating of B3.” By July 21 it had fallen to Caa1.

“This history suggests that the companies which rate health system bonds are constrained in how much financial oversight they provide the public,” the authors write. “Insurance masks the underlying credit quality of the bonds issued by health systems.”

Other aspects of the AHERF bankruptcy saga reported by the researchers include:

  • Abdelhak’s use of hidden cash transfers to cope with the system’s growing financial problems was what finally led to his ouster by the board in June 1998. This situation came about because AHERF’s corporate bylaws allowed it to move cash from one operating unit (donor) to another, without the consent of the donor or the knowledge of the board. The board discovered this in 1997 and set up a committee to review and approve any such transfers. After Abdelhak repeated the offense in April 1998, the board fired him.
  • The COO of AHERF’s physician network “allegedly received a $15,000 commission” for each PCP practice he acquired – “a strong incentive to cut deals rapidly and to overpay for practices.”
  • AHERF’s external auditors, Coopers and Lybrand, gave AHERF “a clean bill of health in its last audit in June 1997. Included in this audit was a large, improperly recorded loan and financial statements that were later retracted, precipitating an investigation by the SEC.”
  • According to a forensic accountant hired by creditors to go through AHERF’s finances, financial management of AHERF “was deliberately ’placed in boxes’ [by the CFO and CEO] so that each person or entity within AHERF could see only one piece of the overall financial position … Moreover, the revenues and endowment funds from AHERF’s scattered operations were commingled in a ’Byzantine structure’ that reportedly permitted the top two executives to transfer funds between units as needed and manipulate final results to make units look as favorable as possible.”
  • AHERF’s “weak governance structure” included an “enormous parent board” of between 25 and 35 members rather than the 13 to 17 recommended by consultants, and a network of 10 different boards responsible for AHERF’s various operations. There was little overlap in membership.
  • “AHERF also suffered from weak board composition” and “several inherent conflicts of interest.” In April 1998, for example, Abdelhak ordered the repayment of an $89 million loan to a bank consortium including Mellon Bank without board discussion or approval. Five board members were current/former directors or executives with Mellon, including its former chairman, who was AHERF’s chairman.
  • AHERF hospitals in the Philadelphia area were initially valued between $500 million and $550 million in 1998, based on various bids received from Vanguard, a potential buyer. The final sale of AHERF’s Philadelphia area facilities to Tenet for $345 million suggests “a potential welfare loss of $200 million.” Furthermore, the deputy attorney general for Pennsylvania stated that “the eight area hospitals had $206 million of charitable assets, the vast majority of which were endowments and other restricted accounts, that now appear to have been raided.”
  • The sale of AHERF’s Philadelphia hospitals required termination of its pension plans. However, AHERF was $40 million short in funds needed to terminate plans in its Philadelphia operations, a shortfall that is now the responsibility of the Pension Benefit Guaranty Corp.
  • AHERF’s bankruptcy spelled the demise of its western hospitals in Pittsburgh, including its one-time star performer and flagship, Allegheny General Hospital.

“A lot of people stepped in to make sure these hospitals didn’t go under,” says Burns. “In terms of saving jobs, that’s good, but in terms of propping up excess capacity – capacity that is expensive and not needed – it’s bad.” As the researchers conclude: “If AHERF’s troubles were simply the product of managerial decisions that initially succeeded but then failed in the face of new market forces, its bankruptcy is not necessarily an undesirable outcome … If, however, AHERF’s troubles were more the product of unethical behavior, a lack of due diligence and the presence of rigid institutional forces, then bankruptcy may not be desirable. The poor performance of its integration strategies was cloaked by inaccurate, misleading financial results and certain institutional structures that limited the scrutiny and efficient response of the tax-exempt financial markets.”