Improved Public Databases Let People Compare Practices and Outcomes; The Importance of Looking Past the Numbers

Published in: Wall Street Journal
Written by: Theo Francis

7/10/2007 – Amid a broad push to bring more accountability to the U.S. health-care system, consumers have access to a growing range of data on hospital quality.

Just a few years ago, only a handful of resources offered such data — some of it too broad to apply to individual hospitals. But now, the federal government, state agencies and a number of private entities are stepping up their push for greater “transparency” on hospital practices. They are mining Medicare data and state records and surveying hospitals to come up with user-friendly databases to help consumers comparison-shop for care.

To be sure, many communities still have just one or two hospitals to pick from — and in an emergency, choosiness may not be an option even in major cities. But hospital systems are expanding, and high-profile institutions are increasingly competing for patients across broad swaths of the country.

The push for more public hospital information is also part of a larger effort to improve health-care quality and ultimately lower health-care costs. Such public comparisons sometimes make hospitals and doctors uneasy, but disclosure can also inspire change, says Marc Volavka, executive director of the Pennsylvania Health Care Cost Containment Council, which analyzes data on hospitals in that state. “If in the light of day, problems are shown,” he says, “that spurs improvement more than anything else you can imagine.”

Here is as look at the mass of data being published online:

Hospital Practices

Perhaps the biggest category of data involves so-called best practices. Such data — available from a variety of public and private entities — tracks whether hospitals adhere to recommended guidelines for certain procedures. So you can find out, for instance, how consistently your local hospital gives heart-attack patients a kind of medication called beta-blockers, or what proportion of surgery patients get antibiotics an hour before surgery. But experts caution that sites that primarily delve into hospital processes — rather than outcomes — won’t tell you how well patients actually fare.

The best-known source for hospital data may be Hospital Compare (, set up jointly by the federal Centers for Medicare and Medicaid Services, hospitals and other groups. The site lets consumers search by city, state or other criteria, and look up a variety of statistics comparing more than 5,000 hospitals against one another and to state and federal averages.

The site covers essentially all acute-care hospitals in the country. And although it draws on Medicare data, its findings can be applicable beyond elderly and disabled patients.

Other groups track similar data on best practices, including the Joint Commission, the independent nonprofit group that accredits most of the hospitals in the U.S. On the group’s site,, consumers can search by hospital, location or type of service and get reports on hospital practices that draw on the commission’s inspections of facilities it accredits.

Much of the best-practices data available focuses on three of the most common areas of hospital care: heart-attack, pneumonia and surgery. There are some groups that include a broader range of procedures.

The Leapfrog Group, a not-for-profit consortium of big health-care buyers like General Motors Corp., provides hospital ratings that are available to the public at Like Medicare’s, many ratings focus on process rather than outcomes, but it collects some data of its own and analyzes 30 different practices at about 1,300 hospitals. Measures include whether hospital procedures consistently encourage hand-washing, whether specialized doctors and nurses staff intensive-care units, and whether doctors enter orders electronically in an effort to avoid errors.

The assumption is that hospitals with the best practices will provide the best care, Chief Executive Suzanne Delbanco says. The group also tracks how hospitals handle “never events” — mistakes that should never happen, like a newborn abduction or amputating the wrong leg — but it doesn’t track how often such events occur. “We’re not counting problems,” Ms. Delbanco says.

Overall, research shows that such “best practices” tend to be good for patients, but they ultimately are just a few of dozens of elements that determine good care, hospital experts warn.

“I frankly don’t care whether I get a beta blocker or not — what I want to know is whether I live or die,” says Mr. Volavka of the Pennsylvania Health Care Cost Containment Council.

Tracking Outcomes

Some resources, including some state governments, are increasingly publishing data on how patients actually fare — at least for some conditions. New York’s state health department, for example, combines state and federal data to let consumers compare mortality rates for cardiac surgery at all hospitals in the state. At, consumers can compare facilities against one another and against state averages. The site also provides some data on how often a hospital does a particular procedure, as well as some best-practices statistics such as those on the Hospital Compare site.

Mr. Volavka’s agency in Pennsylvania collects a broad variety of data from hospitals in the state, and publishes quarterly reports on mortality rates, readmissions and complications for some conditions, and average lengths of stay adjusted for how sick patients are, among other details. The agency’s Web site — — also publishes reports comparing costs and outcomes for specific procedures, like open-heart surgery, and examining the frequency of patients acquiring new infections while at different facilities.

Many states gather data from hospitals for public-health and other purposes, but only about 20 provide public quality reports for consumers, says Denise Love, executive director of the National Association of Health Data Organizations. Among the others: Florida, California and Texas. To see the association’s map of reporting states, see

Last month, Medicare started providing some mortality information for heart-attack and heart-failure cases through Hospital Compare — specifically, measuring how likely such patients are to die of any cause within 30 days of admission at all of the hospitals it tracks. But critics say the information is too vague to be of much use: The site indicates only whether a given hospital is better, no different or worse than national averages — and virtually all hospitals fall into the “no different” category. (Seventeen U.S. hospitals are “better” than the U.S. rate for heart-attack deaths, and seven are “worse,” for example; the rest — 4,453 — are labeled “no different.”)

A private company, Health Grades Inc. of Golden, Colo., also rates doctors and hospitals and provides information on a broader range of outcomes. The company’s site — — rates hospitals on 32 conditions and procedures, from appendectomies to heart-valve-replacement surgery. The site, which offers data on more than 5,000 hospitals, draws on Medicare billing data to calculate the rate of complications or deaths in the wake of each procedure, giving a star rating and specific percentages.

Much of the information is available free on the site, which is searchable by state, procedure and other criteria. Or, for $18, the group will provide more elaborate reports, including average-length-of-stay data and price comparisons.

However, you won’t get much on smaller hospitals: If a facility doesn’t report handling a condition or procedure at least 30 times in three years, Health Grades says its data aren’t statistically significant.

Some insurers and large employers provide members and employees with access to reports from Health Grades or other companies that provide similar information to clients, including Subimo, a seven-year-old Portland, Ore., company, and HealthShare Technology Inc., recently acquired by WebMD Corp. of Elmwood Park, N.J. Don’t stick with just one tool, says Samantha Collier, Health Grades’ chief medical officer. “Go to all the sites you can. See if you’re getting similar data.”

End-of-Life Care

For the intrepid consumer willing to wade through volumes of statistics, the Dartmouth Medical School’s Dartmouth Atlas of Health Care at is a gold mine of information comparing hospital practices across regions and states — suggesting, for example, that some procedures may be overused in some areas.

Aimed primarily at researchers and policy makers, much of the information is too general for evaluating individual hospitals, but it does provide copious hospital-specific data on end-of-life care.

For each hospital, users can explore a variety of data, including how often patients in the past six months of life were seen by specialists and what proportion were admitted to hospice programs, as well as what proportion of patients died in the ICU. The site — which tracks all but specialty hospitals and very small facilities, or about 4,346 facilities — also lets users compare hospitals to one another.

Beyond end-of-life treatment, the data can also help consumers gauge how aggressive a particular hospital is in its overall care, says Kristen Bronner, the atlas’s editor.

“We’re not going to tell you go to this hospital,” Ms. Bronner says. “We’re going to tell you the profile for this hospital is very aggressive, specialist-oriented, high-tech probably — they’re going to try everything and you’re going to see everybody.”

Beyond the Numbers

Although online tools make hospital comparisons more thorough than ever, experts say other factors matter too, including such abstractions as reputation.

“What most consumers have to rely on is not hard data,” says John Conolly, president and chief executive of Castle Connolly Medical Ltd., which publishes lists of top doctors. “Reputation becomes probably the most important thing consumers rely on.”

Major brand-name institutions like the Mayo Clinic, of course, are well-known even far from their home turf. But for other facilities, reputation often boils down to word of mouth. Pay particular attention to recommendations, or warnings, from people who work at hospitals you’re considering, experts say.

Patients should also probe beyond a hospital’s general reputation — even a good hospital may not be good at everything. “You don’t necessarily want to shop your hospital in general,” Dr. Collier says. “You want to shop your specific procedure or condition.”

And consider calling the hospitals directly: Quality departments or medical staff offices can be a good place to start. Most facilities will take pains to answer questions about how often they (and individual doctors) perform procedures, for example. They know it’s bad business to refuse, Dr. Collier says.

Write to Theo Francis at