NEW YORK (Money Magazine) - There are mysteries mortal man was never meant to fathom. One example: why hospital bills are what they are. o For the past few years, inflation has been close to nonexistent and the prices of a lot of things have fallen. Health care, though, exists in its own financial Bizarro Universe where the rules are different.

Total health-care spending in the United States has been rising at double-digit rates, propelled by ever more expensive drugs and medical devices, a tight labor market for nurses and what health-care bureaucrats call increased utilization -- meaning that patients get more procedures and treatments than they used to. All that means bigger hospital bills, and bigger hospital bills are among the main reasons that your health-care insurance premiums, co-payments and deductibles have been going up.

But even as charges have gone up, a lot of hospitals complain that their reimbursements aren't keeping pace with their costs. Some claim they can't even afford to pay their light bills. To which the rational consumer can only reply, "Wha?"

Robert Chaloner, CEO of Cabrini Medical Center, a perennially money-losing 300-bed nonprofit general hospital on Manhattan's East Side, says, "I'm constantly being asked the question, 'Hospitals charge so much, how can they be losing money?' It's a system that doesn't make a lot of sense."

Which is probably lesson No. 1 for any innocent taking a first look at the economics of American hospitals. It took us a while to realize that it's futile to apply the rules you would use to analyze any other kind of business. The way hospitals bill and get paid for their services is contradictory, opaque and labyrinthine almost beyond comprehension. The prices hospitals charge have almost nothing to do with what they actually collect. And what they collect has little to do with their costs. And their costs? Don't ask.

Like the man said, it doesn't make a lot of sense. But we'll try to explain it anyway.

Pay this amount: $0.00

The short answer to the question that plagues Chaloner -- how can hospitals charge a ton of money and still go broke? -- is that while hospitals do charge a lot, they usually collect much less. How much less depends on who's paying.

Generalizing about hospitals is a treacherous business; each one has its own problems and advantages depending on where it's located, who runs it and what kind of population it serves. But as a window into more or less typical hospital finances, let's consider the bill for one recent Cabrini patient.

Cabrini is an old-school nonprofit urban hospital. Of the roughly 5,000 acute-care general hospitals in the U.S., about 80 percent are nonprofits; the rest are run by corporations like Tenet and HCA. Cabrini was founded by an order of Catholic nuns, the Missionary Sisters of the Sacred Heart, more than 100 years ago to serve a part of New York City that was then a slum filled with Italian immigrants. The neighborhood has changed since then, but the hospital still takes care of a lot of people with little money.

This March, a 76-year-old woman was admitted to the hospital early one morning by her doctor, an orthopedic surgeon who had decided she needed to have one of her hips replaced with a plastic-and-metal prosthesis. The procedure, in doc vernacular a "total hip," is common, although Cabrini typically does only two or three a month. The woman had no private insurance, but she was covered by both Medicare, the federal program for people 65 and over, and Medicaid, the combined federal and state program for the poor. Medicare, which covers 40 million Americans, is by far the largest purchaser of medical services in the country.

The woman spent seven days in the hospital, a little on the long side for a total hip. Her recovery was slowed by some respiratory problems.

The master bill that Cabrini compiled for her stay runs to six single-spaced pages containing 178 items, among them "tablet $9.00; CAPSULES $9.00; INJCTNS & TOPICAL $37.00." There are also line items for $854 for time in a pre-op ready room, $669 for anesthesia, $284 for sutures, $6,000 for one day in intensive care and $8,612 for the metal and plastic hardware she is now carrying around in her hip. The grand total for seven days in the hospital: $48,906. That did not include her surgeon or anesthesiologist, who billed her insurers separately.

Forty-nine thousand bucks is a pretty breathtaking number to see on one bill, and it's not hard to imagine how you'd react -- hyperventilation, nausea, stroke -- if you got such a bill in the mail. Nothing like that happened in this case, though. Nobody paid the $48,906, and Cabrini's accounts receivable department would have been stunned if anyone had. The patient herself was sent a one-page summary of the bill, at the bottom of which was the happy command: "PAY THIS AMOUNT: $0.00." Medicaid kicked in a nominal amount, $840. Medicare paid $15,005. Cabrini called it even.

So, on a bill of nearly $49,000, Cabrini was willing to take a loss of roughly $33,000, right? You'd think so. We thought so when we first saw that bill. But we were wrong.

Let's back up a bit and take this in steps. First, how did Medicaid come up with that $840? This one is easy: $840 is the deductible the patient would have had to pay under Medicare. One government agency simply paid another.

Next, where did Medicare get that $15,005 figure? This one is more complicated.

Since 1983, Medicare has paid hospitals a fixed amount for each patient with a given diagnosis. The payment is made in accordance with a list of about 500 DRGs, or Diagnostic Related Groups. When a hospital prepares a bill for Medicare, it takes all of the doctors' and nurses' orders on the patient's chart and turns them into a series of numerical codes that ultimately correspond to a DRG. Medicare pays a fee based on its calculation of what it should cost a given hospital on average to treat the condition indicated by that DRG. (Medicare pays different hospitals different amounts for the same DRG because it makes allowances for things like regional labor costs and the number of uninsured people a hospital treats.)

Because no two patients have identical needs, some who fall into the same DRG category will cost less to treat than the average patient and thus be profitable for a hospital; others will be money losers. Over time, though, the hospital should break even if Medicare has set its DRG payment properly. The idea is to give hospitals an incentive to treat and discharge patients as quickly -- and therefore cheaply -- as possible.

Why do hospitals put up with Medicare's dictates? Because they have to.

Cabrini CEO Chaloner estimates that Medicare represents some 50 percent of the hospital's daily business. Medicare insures so many people that most hospitals have no choice but to take what Medicare will pay, even though hospital executives complain mightily that its rates have lagged far behind rising costs. (Reimbursements from private insurers, who use Medicare as a bottom benchmark, are generally higher.)

So, for our lady with the total hip, Cabrini had no choice but to take about $16,000 in payment for a bill of $49,000. Did the hospital lose money on this patient? Probably -- but nothing like the $33,000 difference between what it billed and what it collected.

It is bizarre but commonplace in the hospital biz that nominal prices have nothing whatsoever to do with actual costs -- and thus have nothing to do with profit or loss. Since the advent of hospital insurance back in the 1920s, insurers have been negotiating discounts with hospitals based on how many people they insure. The more volume they could send a hospital, the steeper the discount. Over time, and especially since the advent of Medicare and Medicaid in the mid-1960s, the number of people covered by insurance grew to the point that nominal prices have become a fiction. Nobody -- or almost nobody -- is paying the full sticker price. In one of the most brutal and oft-noted paradoxes of U.S. health care, the only Americans who get hit with the full sticker price these days are those least able to pay anything -- those with no insurance at all, roughly 41 million people.

Hospital sticker prices are ultimately arbitrary. "There's no science to it," admits Chaloner. "Medicare pays a flat fee, so if you charge $30,000 or $300,000, you still get what the DRG says." He says hospital charges are holdovers from the days before the government, in the form of Medicare and Medicaid, began effectively setting prices. And hospitals have no incentive to lower their charges to reflect what they really collect, for two reasons. One is simply inertia -- what business willingly lowers its prices? The second is that Medicare makes extra payments for unusually expensive cases (so-called "outliers"). The higher your costs are to start with, the sooner a complicated case -- say, a total hip replacement patient who gets pneumonia and stays in the hospital for weeks -- becomes expensive enough to qualify as an outlier.

Okay, okay, you cry, but if Cabrini didn't lose $33,000 on the lady with the hip, what did it lose? The answer to that -- and here's where things get terminally, irretrievably weird -- is that nobody knows exactly.

CEO Chaloner estimates that if Cabrini can get a Medicare total hip patient in and out in four days, the hospital will generally break even. After that, it is probably losing about $900 a day. But Chaloner and CFO Charles Wilczynski say their financial concern is not the profitability of any one patient -- it's simply generating enough cash from operations, philanthropy, state aid, bond sales and whatever else they can think of to keep their emergency room, operating rooms and other departments open.

As Wilczynski explains, "There are a lot of fixed costs." The emergency room, for instance, has to be fully staffed and ready to go 24/7, regardless of how many (or how few) patients there may be on hand. For that reason, Wilczynski says he looks at the hospital's financials on an aggregate basis. Plus, he doesn't have the computer systems or the staff he'd need to track his costs in minute detail -- nor does he have the money to pay for such resources.

That detail can get awfully minute. "It's the most complicated job shop in the universe," says Gregory Poulson, senior vice president of Salt Lake City's Intermountain Health, which runs 21 nonprofit hospitals. "People come in off the street with bodies in different shapes, getting different things done. The procedures are infinitely variable.... Unless you have a sophisticated accounting procedure, it's hard to describe what [treating any patient] costs."

But the main reason most hospitals don't look at their costs the way Wal-Mart or GM do is that there has never been much of a financial incentive for them to do so. Hospital billing and reimbursement involve a series of disconnects, with ultimate financial responsibility far removed from the practical everyday decision-making of doctors and nurses.

As Chaloner jokes, "The most expensive medical device is a physician's ordering pen." And there's not much he can do about it: "I can't go onto the floor and say, 'This is too expensive, stop doing it.' I'm not a clinician."

So doctors order treatments, and Medicare and private insurers decide what they'll pay for them. The hospital itself -- never mind the patient -- has little to say about how much will be spent or how much will be collected. In that environment, why would hospital administrators drive themselves nuts trying to calculate the institution's profit or loss on every patient who comes through their doors?

Regina Herzlinger, a professor of business administration at Harvard Business School who specializes in health care and accounting and is the author of three books about our health- care system, says, "American industry spends a lot of time on cost accounting, and they do so because it gives them so much more control. But the health system rewards inefficiency. It's not a competitive system where the supplier names the price -- as a doctor, you can't determine the price, Medicare quotes it for you.... When the one saying it is as big a gorilla as Medicare, you say, 'I'll take it.'"

It's not that hospitals don't have to look at their costs at all. They are required to provide annual cost reports to Medicare for use in calculating reimbursement rates, but those reports are huge documents that contain aggregate numbers on departments, not individual patients. You might think that HMOs and private insurers would want detailed information on hospitals' costs, but generally speaking, you'd be wrong. They pay hospitals in a huge variety of ways, determined solely by what they can negotiate -- a discount from hospital charges, per diem rates, flat annual fees or some combination. (Cabrini, Wilczynski says, has contracts with as many as 100 insurers.) Insurers, for the most part, don't care what hospitals' costs are, as long as they are getting a better price than their competitors.

Gregory Brodek, an attorney specializing in health-care law, says, "The insurance companies will compare the charges on a given procedure in a given area [to decide what they're willing to pay]. How do they know the actual cost? They don't know, and it's irrelevant."

If most hospitals are clueless about how much money they are making or losing on any one patient, they do have an idea about which procedures are profitable. That can vary from hospital to hospital, but the conventional industry wisdom is that big-ticket surgery, especially cardiac surgery -- bypasses, angioplasty and so forth -- is more profitable than any other kind of medicine.

Why should the profit margin on a heart bypass be greater than on a tonsillectomy? We asked a lot of people -- C.P.A.s, economists, attorneys and insurance providers who specialize in health care, CEOs from for-profit and nonprofit hospitals, Medicare advisory committee members, professors from schools of public health, doctors-turned-businessmen, Wall Street analysts -- and nobody could give us a better explanation than "It's always been that way."

So surgery is where the money is in medicine. Doctors and hospitals get paid more by Medicare and private insurers for doing something -- especially for cutting patients open or doing some kind of invasive diagnostic procedure -- than for treating them with drugs, or for simply watching and waiting. In terms of reimbursement, cardiac surgery is generally the most lucrative, with orthopedics, neurology, gastroenterology and the other specialties bringing up the rear. Way down at the bottom of the reimbursement ladder is the lowly general practitioner, who practices medicine mainly with his wits, not a knife.

According to Paul Ginsburg, an economist and president of the Center for Studying Health System Change, a Washington, D.C. think tank, for years hospitals didn't care about the different margins because they could use what they made on surgery to pay for money losers like deliveries. That is changing as it's become medically feasible to perform in clinics and doctors' offices a lot of procedures that used to be done solely in hospitals -- and as specialty hospitals that do only specific kinds of surgery have emerged.

Stand-alone clinics can be more profitable than hospitals because their overhead is much lower -- they don't have E.R.s and maternity wards to worry about. And insurers are often willing to pay more for the same procedure in a clinic setting than in a hospital, on the grounds that it will be cheaper for them in the long run. Hospitals are full of sick people, after all, and if you have a minor operation in a hospital, you are more likely to be exposed to infection and end up staying there longer.

The specialty-hospital phenomenon shows (among other things) the economic danger inherent in hospitals' financial dependence on their medical staffs. If a hospital is ultimately a big store where medical services are sold, the doctor is the salesman. As far as the hospital is concerned, the crucial question is whether to admit a patient, and that decision is made by the doctor. As a result, a hospital administrator is under a lot of pressure to get the doctors associated with his institution to send him their patients rather than referring them to another hospital where they have admitting privileges. As Chaloner says, "My life and death is determined by my relationship with the medical staff."

Specialty hospitals aren't on the scene in New York State (local regulation makes it a tough market for them to enter and a nearly impossible one for them to flourish in), but Cabrini is nonetheless trying to compete with a range of ambulatory clinics and doctors' offices that do minor surgery. Chaloner is thinking about opening a gastro-intestinal center, a separate corporate entity connected to Cabrini that could share some of its profits with doctors. Hospital administrators all over the country are experimenting with similar ideas.

Keeping the doctors on board, Chaloner says, is critical to the hospital's long-term survival: "If I lost my G.I. practice, it could eventually kill us."

Between competition from for-profit institutions of various types and the eternal grinding struggle to get cash out of the government and private insurers, it sounds pretty grim for traditional hospitals. But they are not helpless.

During the late 1990s, there were a large number of hospital mergers, acquisitions and other kinds of affiliation arrangements -- 1,076 deals from 1996 through 2002, according to Modern Healthcare, an industry trade publication -- although that pace has slackened recently. It's not hard to understand the appeal: Put two 300-bed hospitals together, and you get a 600-bed entity that will have twice as much juice when negotiating contracts with everybody from vendors to insurance companies. In some cities, insurers have lately found themselves having to deal with a single medical center that controls most of the hospital beds in the area and therefore has enormous leverage.

The prevailing dynamic in the health-care business doesn't have much to do with keeping costs down or increasing efficiency. Mainly it's a question of allocation -- who pays. The whole thing is a bit like an overstuffed suitcase. Push down on one side and something pops out on the other. During the late 1980s and early '90s, the popularity of managed care made it possible for HMOs and insurance companies to push the financial burden of rising costs onto hospitals in the form of lower reimbursement. With the demise of managed care as a dominant force in the private insurance marketplace (ultimately as a result of consumers screaming about limited choices), hospitals have learned how to push some of the burden back to insurers.

Insurers have responded by upping premiums and co-pays (as well as by creating various kinds of tiered plans that offer greater degrees of choice for higher prices). Employers have responded by cutting coverage or asking their employees to kick in more. Which is where we came in.

What's likely to happen? If the past is any guide, the buck on higher costs will keep getting passed around, with the various players in this game of hot potato screaming until they can figure out a way of making somebody else pay. The complexity of the way we pay for hospital care in America may seem pointless, but it does do one thing: It keeps all these questions of who is paying for what as obscure as possible. The person ultimately stuck with the bill -- in the form of higher premiums, co-pays or higher taxes -- is always you.

Hospitals run as moneymaking businesses tend to claim margins in the vicinity of 14 percent; for nonprofits, that number is 4 percent. Why?

Most hospitals in the U.S. are charitable institutions, but about 20 percent are for-profit businesses. The for-profit world is dominated by two large chains, HCA and Tenet, the latter of which has been getting a lot of bad press recently. (In August, Tenet agreed to pay the federal government a $54 million fine over allegations that two of its doctors in California performed unnecessary cardiac surgery.)

For-profit hospitals are mainly in the South and West, where there is less state regulation and labor unions are weak. o To the patient, for-profit and nonprofit hospitals seem much the same. Financially, though, they play by different rules. For one thing, as charities, nonprofits don't pay taxes. A financially healthy nonprofit generates a profit margin of about 4 percent, which goes back into operations rather than being paid out to shareholders. In the for-profit world, margins tend to be in the low teens. Why the difference? It depends on who you ask.

Some say it's accounting. For-profits generally prefer to report earnings on the basis of EBITDA (earnings before interest, taxes, depreciation and amortization), while nonprofits generally deduct all those costs (except taxes, of course) from their bottom line. EBITDA can be much higher than earnings under GAAP (generally accepted accounting principles), especially if a hospital has a lot of debt or capital expenditures. Stephen O'Neill, a health-care industry analyst with the investment firm Hilliard Lyons, cautions that it's hard to generalize about profitability differences between the two kinds of hospitals. He says, "If you were the dominant hospital in a rural area, for instance, you might be more profitable because of the lack of pricing pressure."

The issue is accountability for Harvard Business School's Regina Herzlinger, author of Market Driven Health Care: Who Wins, Who Loses in the Transformation of America's Largest Service Industry: "The not-for-profits are not accountable to the equity markets.... It's what makes not-for-profits maybe less efficient."

Another take comes from Victor Breed, a senior vice president with Soluscient, a health-care data firm. "For-profits are more aggressive in their pricing, their staffing," he says. "They're not as burdened by mission issues." -- P.C.

What a hospital actually collects is determined in large part by how well its coders translate mountains of patient files into diagnoses that are accurate -- and that pay well

When Medicare introduced a new billing system in 1983, its intentions were honorable: to replace reimbursements based on hospitals' per diem costs with payments based on diagnoses. Patients would go home quicker; everybody would save. Bills would be pared down to their essence, with a single code indicating the diagnosis and the amount the hospital would be paid. But in the way of all megasystems designed to simplify the infinitely unruly, coding has its problems.

Let's start with the numbers. There are currently 527 DRGs (Medicare Diagnostic Related Groups), each backed up by one or more of the 15,000-plus ICD-9s (International Classification of Diseases, 9th Revision). The ICD-9-CM Expert, the ne plus ultra of reference books, is as thick as the Manhattan Yellow Pages and revised annually. Hospitals depend on professional coders and special software to find the ICD-9s that reflect a patient's condition and to report them in a way that is both accurate and likely to produce the highest possible reimbursement. "It's an art, not a science," says Paul Gaudio, vice president of health information and case management at Cabrini Medical Center in New York.

At Cabrini, the coders work in what was formerly the board room. The elegant wood and glass bookcases are mostly empty now, save for an occasional plastic binder with a title like DRG Changes Sent to Patient Accounts. Piled on every horizontal surface are patient files, the raw material from which the codes are built. The doctors, nurses, social workers and therapists who have contact with the patient have to write down everything they see and hear, poke and prod. If the patient complains about a tummy ache to every professional who visits on Wednesday, Thursday and Friday, each and every one of them should record it. "If you write down every day why the patient is in that bed," says Laura Bakaleynik, Cabrini's coding manager, "you get paid for it."

Tiresome? You bet. And mostly indecipherable. As Bakaleynik reads through one file, she mumbles, "59-year-old female who something something complains of low sodium something something..." After 10 years in the business, however, Bakaleynik, 44, knows how to pick out the key words that are essential for diagnoses. A native of Ukraine, she came to this country in 1992 with a master's in Russian literature and training as a nurse. "I didn't think there was much need for a Russian literature major here," she says, laughing, "so I got certified as a coder." She can complete a simple chart in as little as five minutes; a more complex one can take 30.

Complexity, in the world of the coder, has less to do with the patient's condition than with the clarity of the notes. Take the case of the 91-year-old female who came into the hospital complaining of shortness of breath. As Bakaleynik flips back and forth in the file, she records that the patient receives a diuretic twice a day, and that she had a myocardial infarction five years ago and a heart valve replacement.

Several pages have handwritten entries from five or more of the legions of professionals who looked in on the patient daily. On the third day of her stay, an argument emerges. One doctor writes that she has CHF (congestive heart failure); another diagnoses COPD (chronic obstructive pulmonary disease). Bakaleynik checks back to the beginning of the file -- the E.R. diagnosed COPD and pneumonia. She types in the key words and lets the software turn the entries into ICD-9s.

The result: eight diagnoses, one primary and seven secondary. Which diagnosis is primary is critical: If it's COPD, the DRG is 088 and Cabrini will receive $5,535.04. If it's pneumonia, DRG 089, the hospital will collect $6,388.37.

Here is coding's most vulnerable spot. Coders depend on doctors to record precisely what is wrong with a patient. "It isn't our job to diagnose," Bakaleynik points out. But M.D.s have neither the time nor the inclination to write notes according to insurance protocols. They have other patients to see and, besides, it's not their money. Let the billing department take care of it.

Combing the records for details she may have missed, Bakaleynik notices that the word pneumonia disappears on the third day, but "antibiotics" and "infiltration" appear on the fourth; both are indicators of pneumonia. When she gets to the fifth day and reads that the X-ray is "same as before," she makes a decision. "Finis," she pronounces. On the screen -- Principal diagnosis: DRG 089. Reimbursement: $6,388.37.