hen Thomas Shaw gets worked up, he twists in his chair and kneads his hand. Or he paces about in his tube socks grumbling, “They’re trying to destroy us,” and “The whole thing is a giant scam.” And Shaw, the founder of a medical device maker called Retractable Technologies, spends a lot of time being agitated.

One of the topics that gets him most riled up these days is bloodstream infections. And with good reason—while most people rarely think about them, these are the most dangerous of the hospital-acquired bugs that afflict one in ten patients in the United States. Their spread has helped to make contact with our health care system the fifth leading cause of death in this country.

A few years ago, Shaw, an engineer by training, decided he wanted to do something to help solve this problem and quickly homed in on the mechanics of needle-less IV catheters. Rather than using needles to inject drugs into IV systems, most hospitals have moved to a new design, which involves screwing the threaded tip of a needle-less syringe into a specially designed port. The problem is that if the tip brushes against a nurse’s scrubs, or a counter, or the railing of a hospital bed, it can pick up bacteria. And the rugged threaded surface makes it difficult to get rid of the germs once they’re there. Often, the bacteria go straight into the patients’ bloodstream—which explains why, according to some studies, the rate of bloodstream infections is three times higher with needle-less systems than with their needle-based counterparts.

After months of trial and error, Shaw hit on the idea of surrounding the tip of the syringe with six petal-like flanges, which could flare open to make way for the catheter port. Unlike some of the solutions floated by big medical device makers, such as coating the ports with silver, Shaw’s innovation added only a few pennies to the cost of production. And it seemed to be remarkably effective: a 2007 clinical study funded by Shaw’s company and conducted by the independent SGS Laboratories found the device prevented germs from being transferred to catheters nearly 100 percent of the time.

Given these facts, you might expect that hospitals would be lining up to buy Shaw’s product. But that is not the case, even though his company is offering to match whatever price medical facilities are paying for their current, infection-prone IV catheter syringes. In fact, since the device hit the market two years ago, Retractable has sold fewer than 20,000 units, mostly to one New York hospital. Often, the company’s sales team can’t even get in the door to show their wares to purchasing agents. “The product does exactly what it is supposed to do,” Shaw says. “But it has one fatal flaw. Right there at the bottom of the handle it says Retractable Technologies.”

This is hardly the first time Shaw has found his path to market blocked. In fact, he has spent the last fifteen years watching his potentially game-changing inventions collect dust on warehouse shelves. And the same is true of countless other small medical suppliers. Their plight is just the most visible outgrowth of the tangled system hospitals use to purchase their supplies—a system built on a seemingly minor provision in Medicare law that few people even know about. It’s a system that has stifled innovation and kept lifesaving medical devices off the market. And while it’s supposed to curb prices, it may actually be driving up the cost of medical supplies, the second largest expenditure for our nation’s hospitals and clinics and a major contributor to the ballooning cost of health care, which consumes nearly a fifth of our gross domestic product.

T homas Shaw is a lanky fifty-nine-year-old man with dark eyes and a shock of gray hair that gives him a bit of a mad scientist air. Growing up, he lived in Mexico and Arizona, where his father worked as a chemist (among other things, the elder Shaw invented the first nitrogen test for plants). Shaw describes his childhood home as a kind of frenetic laboratory where science and math problems were worked out on a chalkboard that hung over the dinner table.

After high school, Shaw went on to study engineering at the University of Arizona, and eventually launched his own engineering firm in a former bicycle-repair shop on a rundown strip in Lewisville, Texas. His core business was small-town building projects, like road repairs and structural inspections, but he also dabbled in medical devices. At one point, a friend’s grandmother underwent gallbladder surgery and came out addled and confused. Believing a medication mix-up was to blame, Shaw invented an automated pill dispenser.

Then, one night in the late 1980s, Shaw saw a news program about a doctor in California who had been infected with HIV after being stuck with a contaminated needle. This got Shaw’s attention. One of his oldest friends had recently been diagnosed with AIDS, and Shaw was all too aware of the ravages of the disease. “I thought, I can’t do anything to save my friends,” he recalls. “But maybe I could do something to save other people.”

The next day, Shaw set to work trying to invent a safer syringe. He began buying pigs’ feet from the local butcher and using them to simulate injections. He outfitted every room in his engineering firm with chalkboards so he could draw design ideas whenever they popped into his head. To make time for the syringe venture alongside his regular work, he started pulling ninety-five-hour weeks. And even when he was on vacation, he rarely stopped obsessing. “I remember being in South Padre Island with my wife and kids,” Shaw recalls. “Everyone wanted to go out and play. I wouldn’t go anywhere until I figured out what to do with the back corner of the syringe. I told my wife, ‘I have to work on it all the time until I get it or I’m dead.’”

It took four years and more than 150 design permutations, but Shaw finally came up with a crude prototype and found a local physician to test it on him—an event Shaw’s wife documented with a shaky handheld camcorder. In the video, the doctor holds up a saline-filled syringe about the size of a kielbasa sausage. Then he jabs the needle into Shaw’s arm and pauses for a second before pushing in the plunger. First the saline empties, and then the needle snaps back into the barrel with a pop.

Shaw had just invented the first retractable syringe, a fact that drew the attention of public health officials. In 1993, the National Institutes of Health gave him a $600,000 grant to shrink it down to the size of an ordinary hypodermic and produce 50,000 of them for clinical trials. Shaw was now able to bring on a team of engineers and product designers, and turn a cinderblock bay adjoining the old bicycle shop into a clean room. By the mid-1990s, he had the final design in hand.

Around this time, Shaw launched Retractable Technologies and began searching for funds to build a factory in Little Elm, Texas. Eventually, he raised $42 million, much of it from doctors at Presbyterian Hospital in Dallas. “Everyone was eager to invest,” recalls Lawrence Mills, who was then chief of thoracic surgery at the hospital and invested $95,000 in Shaw’s company. “We all thought it was just a matter of time before it became the standard in the industry.”

In 1996, Shaw returned to Presbyterian to conduct a final round of clinical trials. The nurses who took part gave his syringe uniformly high marks (though some complained in the follow-up survey that the packaging was hard to open and that the air bubbles were difficult to get out), and Presbyterian’s top medical brass was clamoring to get it into the supply rooms. Edward Goodman, the hospital’s director of infection control, wrote a letter to the purchasing department, saying Shaw’s product was “essential to the safety and health of our employees, staff and patients.” But Shaw soon learned that the enthusiasm of health care workers was not enough to gain him entrée; the hospital initially promised him a contract, only to back out three months later. Though he didn’t realize it at the time, Shaw had just stumbled into the path of a juggernaut.

B reaking into the medical supply market has always been tough, in part because for decades the business has been dominated by a handful of behemoth suppliers. In the case of syringes, the incumbent heavyweight has long been Becton Dickinson, or BD, a New Jersey–based company that controls 70 percent of the syringe market and has a lengthy history of trampling competitors. As early as 1960, BD was brought up on Justice Department charges for its anticompetitive practices—among them price fixing, buying up patents to kill its rivals’ innovations, and forcing hospitals to buy its syringes to get other essential supplies, some of which were only produced by BD.

Often, these large companies used their clout to squeeze hospitals on prices. To keep costs in check, in the 1970s many medical facilities began banding together to form group purchasing organizations, or GPOs. The underlying idea was simple: because suppliers generally give price breaks to customers who buy large quantities, hospitals could get better deals on, say, gauze or gloves, if a group of them came together and bargained for ten cases, rather than each hospital buying a case on its own.

Originally, these purchasing groups were nonprofit collectives and were managed and funded by the hospitals themselves. But in the mid-1970s, the model began to shift. Some large hospital chains started to spin off for-profit GPO subsidiaries, which other hospitals could join by paying membership dues, much the way members of buying clubs like Costco pay dues to get bulk-buying discounts. By decade’s end, virtually every hospital in America belonged to a GPO.

Then, in 1986 Congress passed a bill exempting GPOs from the anti-kickback provisions embedded in Medicare law. This meant that instead of collecting membership dues, GPOs could collect “fees”—in other industries they might be called kickbacks or bribes—from suppliers in the form of a share of sales revenue. (For example, in exchange for signing a contract with a given gauze maker, a GPO might get a percentage of whatever the company made selling gauze to members.) The idea was to help struggling hospitals by shifting the burden of funding GPOs’ operations to vendors. To prevent abuse, “fees” of more than 3 percent of sales were supposed to be reported to member hospitals and (upon request) the secretary of health and human services.

But, as with many well-intended laws, the shift had some ground-shaking unintended consequences. Most importantly, it turned the incentives for GPOs upside down. Instead of being tied to the dues paid by members, GPOs’ revenues were now tied to the profits of the suppliers they were supposed to be pressing for lower prices. This created an incentive to cater to the sellers rather than to the buyers—to big companies like Becton Dickinson rather than to member hospitals. Before long, large suppliers began using “fees”—sometimes very generous ones—along with tiered pricing to secure deals that locked GPO members into buying their products. In many cases, hospitals were obliged to buy virtually all of their bandages or scalpels or heart monitors from one company. GPOs also began offering package deals that bundled products together. To get the best price on stethoscopes, a hospital might have to agree to buy everything from pacemakers to cotton balls from the GPO’s preferred vendors. Hospitals went along because they got price breaks, usually in the form of rebates if they met buying quotas.

This situation only grew thornier in 1996, when the Justice Department and the Federal Trade Commission overhauled antitrust rules and granted the organizations protection from antitrust actions, except under “extraordinary circumstances.” Once again, the idea was to help struggling hospitals, this time by allowing the buying groups to grow big enough to negotiate the best deals for their members. But the decision led to a frenzy of consolidation. Within a few years, five GPOs controlled purchasing for 90 percent of the nation’s hospitals, which only amplified the clout of big suppliers.

As it turns out, Shaw’s retractable syringe hit just as these trends were converging. In fact, the year his product came onto the market, three of the nation’s largest GPOs merged to form a company called Premier, which managed buying for 1,700 hospitals, or about a third of all hospitals in the United States. Shortly thereafter, Premier signed a $1.8 billion, seven-and-a-half-year deal with Becton Dickinson. Under the agreement, member hospitals—among them Dallas-based Presbyterian, where Shaw would hit a brick wall—had to buy 90 percent of their syringes and blood collection tubes from the company. Over the next two years, BD landed similar deals with all but one major GPO. As a result, almost everywhere Shaw turned, he found hospital doors were closed to him.

evertheless, Shaw soldiered on and managed to score a few victories. He landed a number of contracts with government agencies, including the VA, that negotiate directly with vendors for supplies. Or he sold his wares to systems so small and poor that they weren’t on the GPOs’ radar—prisons, nursing homes, Indian reservations, and the like. He also teamed up with the SEIU, the nation’s largest union of health care workers, which was lobbying for legislation to curb the needle sticks that were afflicting more than 600,000 health care workers each year. Shaw ended up helping craft a California bill that required hospitals to keep a log indicating which syringes were causing needle sticks and take regular steps to transition to the safer ones. Twenty-one states later passed laws patterned after California’s, and in 2000 the federal government followed suit. That winter, Shaw traveled from Little Elm for the signing ceremony in the Oval Office, and President Bill Clinton gave him a pen he had used to sign the measure into law.

This bumper crop of legislation should have been a boon to Shaw’s company—after all, there was nothing else like his product on the market. BD had released its own safety syringes some years earlier. But the ECRI Institute, the Consumer Reports of the health care industry, had rated its best-selling model “unacceptable” (it was later upgraded to “not recommended”), whereas Shaw’s product received the top rating. And some medical facilities had found that, rather than drive down needle sticks, BD products caused their numbers to rise. After the federal needle safety law passed, Cook Children’s, a Fort Worth–based chain of pediatric clinics, first moved to BD safety needles. But after dropping initially, the number of needle sticks more than doubled, from nine to nineteen a year. So in 2004 Cooks began transitioning to Retractable syringes, and over the next four years the number of sticks fell to zero.

But Shaw’s company continued to have trouble breaking into hospitals. In mid-1999, Kaiser Permanente of California signed a one-year contract to buy Retractable syringes, which seemed like an enormous coup. But a month later, Becton Dickinson announced a “unique” three-year, $30 million deal with Kaiser nationwide. After that, Shaw struggled to get his syringes into Kaiser supply rooms—often, he says, they sat locked in warehouses or trucks in distributors’ parking lots. Kaiser spokesman Jim Anderson argues that if Shaw’s products didn’t make their way to hospitals it was because of “significant supply issues” on Retractable’s end. He also says they were prone to malfunction and that, in several cases, needles detached and were left “stuck in the arms of patients.” Whatever the reasons, Kaiser broke off the deal early.

Meanwhile, as Shaw was fighting his battles hospital to hospital, Becton Dickinson was working to extend its hold on the nation’s GPOs. According to confidential documents filed as part of a whistleblower lawsuit, in 1999 BD paid $1 million to Novation, the only major GPO with which it hadn’t yet signed a sole-source contract, in return for a three-year sole-source deal for syringes and needles. This payment, which it dubbed a “special marketing fee,” was on top of more than 3 percent of its sales revenue and other perks valued at hundreds of thousands of dollars. Becton Dickinson’s grip on hospitals was now even tighter than it had been before.

By this point, the struggle was starting to take its toll on Shaw. Now when he came home after long days in the office, he would shut himself in a room and not let anyone in except his children. His marriage was unraveling (he later divorced) and his increasingly confrontational style was starting to put off potential allies. When he was invited to speak at a luncheon of the Medical Device Manufacturers Association, an alliance of small medical suppliers, no one would sit near him; he ate alone, surrounded by twelve empty chairs, and was booed when he stepped to the podium. “They were afraid if we took on the GPOs they would be destroyed,” Shaw recalls. Meanwhile, Retractable Technologies’ stock had lost nearly two-thirds of its value, and its operating capital was dwindling rapidly. After weighing his options, in 2001 Shaw finally filed an antitrust suit against Becton Dickinson, Novation, and Premier.

round this time, GPOs started to come under scrutiny. The New York Times ran an investigative series on their business practices in 2002, and Congress followed suit with a string of hearings. One of the first witnesses was California entrepreneur Joe Kiani, who had invented a machine to monitor blood-oxygen levels. Unlike other similar devices, Kiani’s worked even when patients moved around or had little blood flowing to their extremities, a crucial innovation for treating sickly, premature infants, who tend to squirm and need to be monitored constantly for oxygen saturation—too little and they suffocate, too much and they go blind. But most hospitals couldn’t buy Kiani’s product because his larger rival, Nellcor, had cut a deal with the GPOs.

Kiani’s testimony was followed by a flood of revelations about self-dealing and conflicts of interest among GPOs and their executives. Congress was also given a slew of documents showing that GPOs were collecting upfront payments of up to $3 million from suppliers, including drug makers like Astra-Zeneca, in return for awarding them sales contracts, not to mention a large share of revenues. In one case, a vendor was handing Novation not 3 percent of its revenue on a given product line, but a full 94 percent, according to Novation documents.

These revelations stirred a groundswell of outrage, and there was talk of legislation to rein the GPOs in. Spooked by this threat, in 2002, the industry introduced a voluntary code of conduct, which it promised would foster “a thriving, innovative and competitive healthcare marketplace,” and three years later created a body to oversee compliance. For the first time, it seemed as if these powerful middlemen might actually cede some ground.

By this point, Shaw appeared to be on the verge of a breakthrough as well. In mid-2003, Novation, Premier, and another company offered to pay him $50 million to settle out of court and agreed to take steps to give him market access, though the specific terms remain under gag order. As the case was getting ready to go to trial the following year, Shaw received a two a.m. phone call from his lawyer saying that Becton Dickinson was prepared to offer a $100 million settlement. Shaw roused his children, and they piled in the car and drove to the local IHOP for blueberry pancakes.

After the settlement, Shaw started offering his retractable syringes for ten cents a piece, about what other companies were charging for their conventional hypodermics. But even this didn’t boost sales—in fact, Retractable’s sales to non-VA hospitals dropped. Shaw has since come to see the settlement as nothing more than a tool for the GPOs to keep the details of their operations under wraps. “The group purchasing organizations that were BD’s agents paid us $50 million to keep their practices from being reviewed in front of a jury,” he told me. “They took the equivalent of $10,000 from every hospital in the U.S. and gave it to a company in Little Elm, Texas. Either they’ve got minimal trust in the average juror or they’ve got something they don’t want the public to know.”

haw is not the only one who kept running into brick walls after the GPOs’ promised reforms took hold. In 2004, Garrett Bolks, a Tulsa native who had spent twenty-four years working in the medical supply business, brought the first X-ray-detectable surgical towel to market. It was a simple invention—nothing more than a strip of blue waffle-weave cloth about the size of a hand towel, with a flexible ribbon of barium sulfate tucked into one corner of the hem. But it promised to eliminate the problem of towels being accidentally left to fester inside the body after surgery, and it garnered attention in high places.

After learning about the product through a friend in the summer of 2004, then Secretary of Health and Human Services Tommy Thompson invited Bolks out to Washington, where they discussed it over steaks at a local restaurant. Thompson liked what he heard. “It made a heck of a lot of sense to me,” Thompson recalled when I spoke to him in January. “I thought, Why hasn’t anybody thought of this before? This should be the standard in the industry.” After leaving the Bush administration the following year, Thompson agreed to sit on the company’s board and began talking up Bolks’s product in speeches. Bolks also landed a contract to sell his towels to the venerable Cleveland Clinic.

Nevertheless, Bolks couldn’t manage to make inroads with the GPOs, even when his X-ray-detectable towel was the only one on the market—and soon enough he had competitors. By 2006, Bolks had sunk more than $1 million of his own money into the venture, and was running out of capital. So when a Dallas-based GPO named Broadlane put out a bid for surgical towels that year, he decided to go all out. Not only did he offer his towels at pennies above cost, he also called in his connections, including Thompson, who personally put in a call to Broadlane. “I brought out as many big guns as I could,” Bolks recalls. “Because I knew this was my last chance.”

But even this was not enough for him to land the deal. Instead, Broadlane chose to go with ordinary, non-X-ray-detectable surgical towels from two established players, Medical Action and Medline. On its face, this choice made little sense. According to internal Broadlane documents, the quality of Bolks’s towels was on par with competitors, and his bid was nearly 20 percent lower than any other company’s X-ray-detectable products. It was also lower than the non-X-ray-detectable towels Broadlane chose. By all appearances, Broadlane went with a more expensive product that offered fewer benefits for patients.

Broadlane’s executive vice president for supply chain services, Michael Berryhill, said via e-mail that he could not comment on the reasons for the decision, though he emphasized that the company and its member hospitals weigh a number of factors beyond price when choosing which bids to accept, including “the reputation and reliability of each potential supplier” and “the transaction costs associated with having more suppliers on-contract compared to a lean supply chain.” But Diana Smith, a former director of surgical services at Broadlane who was privy to the selection process, sees the situation differently. “It should have been a no-brainer,” Smith told me when I met with her in Dallas. “Garrett had a good product, and it was cheaper than everybody else’s. But GPOs make their money by charging vendors fees. And if you get a percentage of sales, going with a lower bid from a little company just loses you money and pisses off the big vendors with multiple contracts.”

Smith, who provided the information on which bids were chosen, adds that the tricky part for GPO executives is getting member hospitals to sign off on higher-priced contracts, something she says Broadlane did by presenting the statistics in ways that, though technically accurate, were often misleading. In the case of the towel bid, hospital administrators were shown a PowerPoint presentation (a copy of which she gave to me) indicating that going with the Medline and Medical Action bids would save them between 6 and 29 percent. But this was relative to the same companies’ bids the previous year, not the bids offered by other vendors. “Our job was to bamboozle hospital CFOs and purchasing managers,” Smith explained. “My boss used to call it getting them to drink the Broadlane Kool-Aid.”

The Broadlane decision turned out to be the death knell for Bolks’s towel company. But he continues to come up with new devices. Last April, I visited him at his office in Tulsa, which was stuffed with crumpled cardboard boxes full of medical supplies, and he showed off his newest invention—a black handheld wand and a diode about the size of a fleck of pepper with a tiny antenna poking out from one side. He explained that the idea was to embed the diode, which gives off a special frequency, into all kinds of surgical supplies. That way, if objects are left inside patients, the wand can be used to detect them before the incision is even sown back up. “The towel was nice—at the time it was innovative,” Bolks added. “But this was the product I felt could make a major contribution.” However, his savings are too depleted to put it into production, and he has been unable to drum up outside funding. “Investors know how the system works,” he explained. “Without a GPO contract, it doesn’t matter how good your product is. Even if I could wave this wand over your body and cure you from cancer, chances are I couldn’t sell it to hospitals.”

tories like these abound among small suppliers, a number of whom have filed suit against GPOs. But most are wary of speaking out. Several talked to me off the record. At least a half dozen more agreed to speak, only to back out at the last minute or retract their statements after we had spoken. “Most people who know this world wouldn’t speak to you under threat of subpoena,” one former GPO executive told me. “They are terrified.”

As for the GPOs and their advocates, they argue that if small companies have trouble breaking in, it has to do with the quality of their wares. “Why do small manufacturers fail?” Curtis Rooney, president of the Health Industry Group Purchasing Association, the trade organization for GPOs, asked when I met him at his Washington office. “The answer is that they don’t have a product.” He added that GPOs pick vendors through competitive bidding, which puts small companies on equal footing with their larger rivals.

Rooney also stressed that most GPOs adhere to the code of conduct, which he argued assures openness and competition. But while the code sets firm guidelines regarding conflicts of interest—GPO employees are barred from holding stock in companies whose contracts they are in a position to influence, for instance—when it comes to core business practices, it is vague. Rather than setting caps on kickbacks, for example, it merely directs GPOs to take steps to ensure that any financial perks don’t “encroach upon the best interests” of hospitals and clinics. Obviously, this leaves room for maneuvering. And, while the industry generally keeps its business practices under wraps, critics charge that the tactics that raised red flags in the past continue. In fact, there is evidence to this effect. Some GPOs admit in their limited public disclosures to collecting “fees” of 25 percent or more of vendors’ sales. Others continue to pursue aggressive bundling programs—the GPO MedAssets now bundles together everything from sutures and bedpans to blood-oxygen monitors and cafeteria services (although hospitals have a certain number of opt-outs).

In some cases, GPOs have backed away from their old practices only to revive them in modified form. After the last congressional probe, Premier introduced its own stringent code of conduct and began signing contracts with multiple suppliers for most products nationwide. But it has since begun working with regional hospital groups to forge deals that drive sales to a few preferred vendors. Through a recently launched program called ASCEND—a program the company’s president, Mike Alkire, has called “the future model of Premier”—it has also begun locking individual hospitals into sole-source agreements for a wide variety of products. What’s more, Premier’s code explicitly bars it from pursuing sole-source deals and bundling for what are known as “physician preference items,” meaning those that are seen by doctors as affecting the quality of patient care. But during Premier’s official quarterly conference call for suppliers last February, ASCEND’s director, Andy Brailo, suggested that, while hospitals are not required to sign restrictive deals for physician-preference products, the company is taking steps to persuade them to do so. He added that Premier is “investigating things even down to profit sharing with the physicians.” (Premier maintains that either Brailo misspoke or his words were taken out of context, and that the company “does not include physician preference items in the commitment associated to the ASCEND program” or “engage in profit sharing programs of any type with physicians.”)

Prakash Sethi, president of the International Center for Corporate Accountability at Baruch College and author of a recently published book on GPOs, argues that if the industry hasn’t transformed itself, that’s because the pay-for-play system remains intact. “It’s a gravy train,” he explains. “Why should they get off it? We can’t even begin to talk seriously about GPO reform until we realign the financial incentives so that hospitals, not vendors, are their main clients.”

he multibillion-dollar question is what this incentive system means for health care costs. GPOs maintain that by pooling hospitals’ buying power and getting big medical suppliers to submit to competitive bidding, they are able to negotiate better deals and save hospitals billions of dollars. If this weren’t the case, Blair Childs, Premier’s senior vice president for public affairs, argues there would be no reason for hospitals to join. “They wouldn’t use our contracts if they weren’t competitive,” he told me. “Many of these hospitals have tiny margins. They’ve got to get better products, better prices, better value.”

Industry-funded studies support these cost-saving claims. In fact, one recent study found that GPOs save hospitals as much as $36 billion a year. The problem is that, rather than hard numerical data, this figure is based on surveys of hospital administrators. And while survey takers weren’t asked what yardstick they used to measure savings, the study’s author, Arizona State University professor Eugene Schneller, says that hospitals generally base their figures on the discounts they get off GPO list prices, often in return for agreeing to buy from select suppliers. Obviously, this is a far less meaningful benchmark than what they would pay for the same supplies if they negotiated prices on their own. But, then, most hospitals don’t appear to have that information. An earlier survey of hospital purchasing managers by supply chain expert Lynn James Everard found that most of the managers who claim to know what they are saving through their GPOs know only what their GPOs report to them.

The idea of hospitals outsourcing oversight of their supply budgets may seem hard to fathom. But the price of medical supplies is not always transparent. Makers of the costliest devices and equipment tend to be secretive about pricing and generally require buyers to sign gag clauses promising not to disclose what they’ve paid, which makes it difficult for hospitals to comparison shop. (In fact, this is one reason GPOs maintain their services are necessary.) Also, many larger hospitals hold stakes in GPOs, and even smaller ones have less incentive than outsiders might think to pour over cost reports, since insurance companies and government programs, like Medicare and Medicaid, are picking up the tab for much of their supplies and equipment.

As for independent assessment of GPOs’ effect on costs, they are hard to come by. But the little information that is available suggests that they may actually drive up the price of supplies. A 2002 pilot study by the Government Accountability Office found, for instance, that hospitals that went through GPOs paid more for safety needles and most models of pacemakers than those that negotiated prices on their own—for some pacemakers the median gap was as wide as 39 percent.

Even more unsettling are the findings of MEMdata, a Texas-based company that helps hospitals process their bids for new equipment and captures the quotes in a database, so that administrators can compare the prices they are offered to what others have paid. Shortly after the company opened for business, founder Bob Yancy says he discovered that bids hospitals got through their GPO contracts were substantially higher than the ones he or medical centers that weren’t locked into GPO pricing could get by negotiating directly with vendors for the same equipment. Yancy later had his staff add a field to their database to track just how GPO bids stacked up. Over the last seven years, his company, which serves more than 500 medical facilities, has collected tens of thousands of bids. On average, Yancy says, the GPOs’ prices are 22 percent higher than the ones that hospitals can get on their own. “The bottom line is that hospitals are being systematically overcharged,” he told me, when I met him at a Washington, D.C., restaurant. “GPOs are inflating the pricing.”

To back up these claims, Yancy sent me more than three dozen paired bids, including two quotes for a suite of endoscopy equipment from the same vendor that were issued on the same day. The specs were identical, from the cameras down to the fiber-optic cables. But one had “aggressive pricing” scrawled across the top and came out to $83,000, while the other had the name of a large GPO above the header (Yancy asked that the name and other sensitive details be withheld to protect his business contacts), and was priced at $131,000—or nearly $50,000 more for the same equipment. In other cases, the picture was less clear; there were modest variations between the specs of the two bids, for instance. But the overall pattern was unmistakable.

The experience of hospitals and clinics that have struck out on their own seems to confirm Yancy’s findings. When Iowa Health System, a chain of ten Midwest hospitals, cut ties with Premier some years ago, it immediately shaved $7 million a year off its supply costs, a savings of more than 12 percent, according to the New York Times. Similarly, in 2005 a chain of community clinics affiliated with the University of California, Los Angeles, began going outside its Novation contract to buy chemotherapy drugs and managed to save $800,000 a year.

And yet, despite all the talk about “bending the cost curve down” in the runup to health care reform, GPOs barely entered the conversation. Critics of the system find this baffling, especially since most believe that if GPOs are driving up prices, the problem could be fixed by simply getting rid of the anti-kickback protections. Nevertheless, lawmakers appear to have limited appetite for taking the issue on. Last August, Congress launched an investigation into GPO contracting practices, and the Government Accountability Office followed suit. But Senate staffers now say that hearings on the subject are unlikely to be held this year, and may not be held at all.

Part of the reason interest has waned seems to be that those who know enough about the system to care aren’t eager to change it. “Hospitals are a big constituency in every district,” explains one senior Senate staffer involved in the ongoing GPO investigation. “And hospitals support GPOs. Reform, on the other hand, doesn’t really have a lobby, which can make it difficult to take action.”

or Shaw this unsolvable riddle has become a kind of obsession. He turns it over and over in his head like an engineering problem, as if the fix might come to him if he just looks at it from enough different angles. Perhaps the part he finds most perplexing is that it was largely government grants that paid for him to develop his retractable syringe. “I’ve spent twenty years fighting to return my obligation to the American taxpayer and to a government that turns its head from its responsibility to protect the free market,” he says. “The taxpayers got screwed out of the technology they paid for.”

Even today, Shaw continues to develop new products. In fact, he has brought five of them onto the market in the last two years, including the IV catheter syringe. But his efforts remain consumed largely by the struggle for access. Among other things, he has hired a lobbyist to agitate for the repeal of the anti-kickback exemption and filed a stack of lawsuits, including a second antitrust suit against Becton Dickinson. All this struggle has brought a few scattered victories—most recently last November, when a jury found that BD had used Shaw’s patented technology for its own retractable syringe and ordered the company to pay Retractable another $5 million. (The case is on appeal.) But Shaw still isn’t any closer to breaking into the hospital market, and in the meantime the life on his patents is dwindling. In just four years, the first of them will expire and the game will be over.

This isn’t just bad news for Shaw. Because his company is in the red, he has been unable to pull together the financing he needs to expand his factory in Little Elm. So he has partnered with Chinese companies, which put up money to build assembly lines in China in return for permission to produce his syringes for the Chinese market. When his patents do run out, the Chinese manufacturers will be the ones poised to bring his technology to the world market, meaning all the jobs and economic benefits that could have gone to the local residents will instead go to the people of Gansu Province.

The senselessness of this quandary has driven Shaw to distraction. “We are devoting our entire lives to something we know is going to fail,” he told me during my final visit to his office. “If we expected anything else, it would be devastating. If somebody’s holding you under water and they let you up and you think you’re going to escape, you’re going to go insane.” He was in one of those moods where he paces about, his mind flitting from outrage to outrage so quickly that it can be hard to follow the flow, much less stop it. As I got up to leave, he trailed me down the stairs and out to the parking lot, where he stood amid the gravel and grit in his socks. Even as I backed my car out of the lot, he was still talking. The question is whether anyone out there is listening.