NEW YORK (Money Magazine) - Three years ago my dad, Chuck Sherman, hopped on a treadmill and flunked a stress test. He was scheduled for bypass within the week.

When the cardio-thoracic surgeon at the GW University Medical Center in Washington, D.C. emerged from the operating room after the five-hour procedure, he commented on how nice it was to work on someone who clearly took good care of himself. My dad was 65 and looked a decade younger.

After a remark like that, everyone anticipated a smooth recovery. My dad watched his salt, walked his miles and waited to feel better. It never happened. Instead, he slowly started getting worse. Climbing stairs became more difficult. Walking to his job a mile away was no longer an option. At his doctors' office, no one was sure what was wrong. He had test after test. Medications were changed. Nothing helped.

Finally, months later -- and with no lack of prodding from my mother -- my dad's internists sent us elsewhere. The cardiologists at Johns Hopkins Hospital in Baltimore looked at the film of an echocardiogram and saw what was wrong: constrictive pericarditis. Because the sack around my dad's heart, the pericardium, was hardening, his heart could not beat properly.

They recommended removing the sack in an operation called a pericardiectomy.

Everyone in my family was so relieved to know what was wrong that when the doctors suggested scheduling surgery for a month later, we didn't question them.

We didn't ask how many of these procedures they'd done or whether there was a better place for the surgery. We were so grateful, we simply said okay.

I, for one, should have known better.

Nine years ago -- two days after our son Jake was born -- my husband and I learned he had a rare congenital heart defect that required complicated surgery. With three weeks to find the best place to have it done, we set about doing our homework, meeting with doctors in New York, where we lived, calling others, faxing Jake's case file and sending his echocardiogram across town.

Our search led us to a conference room at Mount Sinai Medical Center, the hospital where Jake had been born and diagnosed, which we'd heard had one of the best pediatric cardiology departments in New York City.

A group of doctors sat with us and recommended a two-step approach. Jake would have one surgery immediately to start the repair and give his heart time to grow, and another a year down the road. It was August, so the hospital's top surgeon was on vacation, but the doctors said that the No. 2 was perfectly able to handle this.

I don't recall every detail of this meeting -- I was in tears much of the time -- but I do remember asking the physicians if they would have the surgery done at this hospital, by this doctor, if Jake were their child. They all said yes.

Back at home, at 10 p.m. that night, our phone rang. It was one of the doctors from the meeting. Mount Sinai was not the best place to have the surgery, he confessed: "You're going to Boston Children's."

That Massachusetts hospital, he explained, sees many more cases each year and has better technology. "If the worst happens," he said, as reassuringly as possible, "you need to know you did everything you could."

When we arrived in Boston 10 days later, the advice of this brave and caring pediatric cardiologist was confirmed. The surgeon we met told us that he'd stopped doing the two-step procedure that Mount Sinai was recommending a decade earlier. He could repair Jake's heart in a single operation.

Today Jake is in fourth grade. He plays soccer and Little League, and hopes to grow tall enough to be drafted into the NBA. We were incredibly lucky.

My dad's pericardiectomy, on the other hand, was a disappointment. The surgeon at Johns Hopkins was able to remove only half of the tissue, and my dad's problems persisted. My mother and I started looking for other solutions -- and discovered, too late, that despite Johns Hopkins' great reputation, the hospital handles relatively few pericardiectomies a year.

(Dr. Stephen Achuff, my dad's cardiologist at Johns Hopkins, says that they do about eight to 10 pericardiectomies a year but perhaps only one or two similar to my dad's.)

This was a hard lesson: We should have done more homework the first time around. Now, belatedly, we contacted the Cleveland Clinic and the Mayo Clinic, which topped the U.S. News & World Report list of best hospitals for cardiac care. The doctors in Cleveland didn't return calls from my father's internist, but the Mayo Clinic in Rochester, Minn. took the case.

This time, we didn't hesitate to question the surgeon, Dr. Hartzell Schaff. He told us that he and his colleagues remove more pericardiums than any other hospital in the country, some 30 to 40 a year. "We've probably done 500 or 600 over the time I've been here," he said.

We scheduled another surgery, and Dr. Schaff removed all the damaged tissue. Afterward, when I asked him -- somewhat reluctantly -- if he would have been able to do such a complete job the first time around, he hesitated but answered yes. "It doesn't mean that someone else did it wrong," he said. "That person just might not have had the experience to know how much to get."

My father is now slowly recuperating. Three open-heart procedures in as many years have taken a toll on his emotions as well as his physical abilities. I can't help but wonder what shape he would be in if he'd had only two.

That's why I decided to do this story. Choosing the best hospital and doctor to treat a serious health problem turns out to be much more complicated than any other consumer question I've ever researched -- more so than buying a car or an appliance, or even real estate.

That's because when you're trying to pick out, say, an SUV, you can compare features, look up safety records and read reviews online or in magazines before you make a decision. When you're looking for a doctor or hospital, you'll find very little solid comparative information -- just the advice of doctors and medical administrators who likely want your business.

"It's unbelievable how poorly informed we are, how little information we have to go on," says Debra Ness of the Disclosure Project, a coalition of employers and consumers that is pushing health-care providers to measure and report their performance.

So how do you find topnotch care? First, you should understand the variety of techniques used to measure health-care quality, how to use them and what's being done to improve disclosure. Next, we suggest specific questions to ask your doctor (see the box at right). Finally, on page 29, you'll find our list of the top hospitals, by region, for 20 common medical conditions, which we put together with the help of HealthGrades, a Denver health-care rating and information company.

Keep in mind that seeking out the best care requires work -- insistent or repeated phone calls to find a doctor who'll admit you or to get a referral from a physician. And it may entail higher out-of-pocket costs if the hospital and doctors are outside your insurer's network. My father was fortunate enough to have traditional indemnity coverage, which allowed him to go anywhere, but the majority of Americans are in more restrictive HMOs and PPOs.

Agreeing on a single way to measure quality health care isn't easy. One approach is to count the number of nurses or MRI scanners. Another is to ask whether patients are getting the tests and treatments that are known to be effective -- drugs to lower blood pressure, for example, or cancer screening -- in effect, to track the process of delivering care.

You can also define quality by looking at the outcome: A top hospital, by this definition, is one with lower complication and mortality rates. Finally, you can throw in patient satisfaction.

Hospital quality data on the Web • leapfroggroup.org The Leapfrog Group, a coalition of major employers, is compiling a database of hospitals' compliance with the group's recommended safety standards. • healthscope.org California residents can find hospital ratings for services like high-risk pregnancy care as well as volume figures for procedures like carotid endarterectomies at this site, run by the Pacific Business Group on Health, a non-profit coalition of major employers. • mhqp.org The Massachusetts Health Quality Partners, another coalition of health-care providers and purchasers, has posted results of a 1998 survey of residents' experiences at local hospitals. • myhealthfinder.com New Yorkers can look up mortality rates, procedure volume and patient survey reports for hospitals in the state at this site run by the Alliance for Quality Health Care and Niagara Health Quality Coalition. • phc4.org If you're searching for medical care in Pennsylvania, you can see how the state's hospitals rate on 71 procedures and conditions based on outcome, volume and even cost. • ncqa.org You can look up HMO ratings from the nonprofit National Committee for Quality Assurance at this site.

And whatever your definition of quality, you have to measure it in a fair way. Take mortality rates. Some hospitals and doctors routinely treat older and sicker patients than others do, notes Brian Schilling, spokesman for the National Center for Quality Assurance (NCQA) in Washington, D.C., an organization that evaluates the care delivered by HMOs. So you cannot simply compare mortality statistics and conclude which doctors and hospitals are the best.

As a patient, however, quality comes down to one thing: getting better. Health-care experts believe that many approaches to quality, from having sufficient resources to getting the right tests on time, contribute to that. But to date, there's been little research on what quality measures best predict a better outcome.

When it comes to surgery, however, one piece of data has been found to point to a good result: volume, or how often the surgery is performed. It's important to distinguish between two types of volume: how often a hospital does an operation and how many cases a doctor handles.

Let's take hospitals first.

"The number of procedures a hospital has done clearly has a significant impact on patients' risk of dying," says Dr. John Birkmeyer of Dartmouth College, whose study of hospital surgery volume and mortality was published last year in the New England Journal of Medicine. And the higher-risk the procedure, the more meaningful hospital volume seems to be.

What about doctors? Birkmeyer has also looked at surgeon volume -- that research will be published later this fall -- and here his findings are mixed. "There are a number of procedures for which surgeon volume is the key driver of outcome," he says. In general, the more technical the operation -- cardiovascular ones, for example -- the more the surgeon's experience matters.

But with less complicated procedures, such as treatments for cancer, particularly lung cancer, the doctor's experience is not as important. "In those cases, the attributes of the hospital, particularly the quality of the ICU and nursing staff, matter more," says Birkmeyer. "Of course, patients can optimize their chances by choosing both a high-volume surgeon and a high-volume hospital for most high-risk operations."

The difficulties of defining and measuring quality mean data often aren't collected, says Suzanne DelBanco, director of the Leapfrog Group, an advocacy group made up of some of the country's largest employers. (For more on the few states and organizations that do collect data, see the box at right.) And that's too bad, because anecdotal evidence suggests that when you measure and report on health-care quality, patient health improves.

Step 1: Select a Region Great Lakes Great Plains Northeast Sun Belt West Coast Step 2: ... Step 3: ...

In the late 1980s, for example, New York State began requiring hospitals to report mortality data for coronary bypass surgery. Three years later, mortality had fallen by 41 percent. In another effort, the NCQA, together with the American Diabetes Association, asked doctors to voluntarily report on their diabetes patients' course of treatment and health. After five years, patients had better blood-sugar, lipid and blood-pressure results.

The good news is that quality standards may be on the way. Four years ago, the National Quality Forum (NQF), a coalition of health-care providers, consumers, major employers and other health-care purchasers, was established to create health-care performance measures.

So far, the group has come up with 39 care standards for hospitals that cover eight areas, from heart attacks to prenatal care. There's nothing ground-breaking about these guidelines; they are steps such as quickly getting aspirin to a person during a heart attack and prescribing beta blockers afterward. In fact, they're so commonsensical that you would think they need not be mandated. Yet a recent study in the New England Journal of Medicine reported that half the time hospitals do not take these steps.

Setting standards, however, is where the NQF's reach ends -- the group has no enforcement power. But, says the group's CEO, Dr. Ken Kizer, the major purchasers of health care -- Medicare, say, or General Motors -- will likely require any hospital that wants to do business with them to track and report their performance on NQF's standards.

Though imperfect, measures like volume and mortality are valuable tools for health-care consumers. That's the thinking behind the following list of the best hospitals for 20 common medical conditions, compiled with the help of HealthGrades, a firm that uses Medicare mortality and complication data to rate hospitals.

On the assumption that you may be willing to travel for excellent medical care but prefer staying somewhat close to home, we asked Sarah Loughran and Dr. Samantha Collier of HealthGrades to provide the names of top hospitals in each of five regions. For each area, HealthGrades started with the five-star hospitals in its database for each procedure.

Then, because the research by Birkmeyer and others has found volume to be the best proxy for quality, HealthGrades narrowed the list to the five with the highest volume. You can also find a searchable version of these tables at our website, money.com.

Once HealthGrades identified the hospitals, our reporters asked each facility for the names of the three doctors who treat the most patients with that condition or, if they refused, the name of the department head or other supervising physician. That way, if you or a family member is facing one of these diagnoses, you'll have a person to call.

To date, HealthGrades doesn't have ratings for some conditions, including cancer. And because the firm collects its data from the Medicare database, the medical outcomes are for patients 65 or older, which excludes pediatric procedures, cesarean sections and other treatments common among women of childbearing age. But despite those flaws, we believe it's the best, most comprehensive set of data currently available. We hope you find it helpful when you need it most.