Return to Emily Friedman home page
First published in Hospitals & Health Networks OnLine, April 5, 2006
Our society takes much greater notice of hospital openings than of hospital closings, but both are important barometers of how our institutions and communities are doing. Why don’t we have decent data to help inform decisions affecting these events?
In some ways, this column is an extension of my previous one on population diversity ("Tapestry", Feb. 7, 2006). That is because the research I conducted for that article sparked some questions regarding population change and health care for which I could not find ready answers--or any answers at all. It led me to wonder why that was the case.
The issue is the opening and closing of hospitals. Most of us know when a hospital opens locally--there is much chopping of ribbons and taking of photographs and tours of the new facility. Press releases trumpet the beautiful building and the many cutting-edge services offered (no pun intended). Although hospital openings are not always greeted with joy--the current intense debate over the explosion in “niche,” physician-owned, for-profit specialty hospitals comes to mind--these events are usually viewed as positive.
The closing of a hospital, on the other hand, is a much quieter affair. True, there is often bitter controversy over whether it should be allowed, but rare is the hospital scheduled for closing that avoids its fate. The remaining patients are discharged or transferred, the staff take their belongings and go their separate ways, neighboring hospitals try to recruit the nurses, and the building is closed--sometimes to re-emerge as a hospital; more often to be reborn as a condominium development, nursing home or homeless shelter; and sometimes to be left as a brick-and-mortar corpse, standing empty, sometimes for years.
The problem is that we don’t have good, timely national data on either closings or openings. When I first became interested in the subject, years ago, such data were easily available. Not now. The Department of Health & Human Services (HHS) is supposed to issue an annual report on closings, but the department is hopelessly behind; the latest full federal report I was able to find was for 2000. In 2001, Stephanie Poley and Thomas Ricketts at the University of North Carolina analyzed available data and reported that 460 general hospitals closed between 1990 and 2000 (150 rural, 310 urban, by their count), but their work has not been updated.
My colleague Alan Sager of Boston University, whose work I cited in my last column regarding hospitals and the racial makeup of communities, has been studying closings for decades. He has found that in 52 U.S. cities, 140 hospitals closed between 1990 and 2003. A 2005 study by State University of New York researchers found that from 1996 to 2002, 36 public and 31 private suburban hospitals closed. A mental health group reported in 2003 that 40 psychiatric hospitals closed from 1993 through 2002.
There are readily available data for some states. The California Hospital Association reports 82 closings from 1998 through 2004, the Illinois association reports 49 from 1980 through 2005, the Massachusetts association reports 27 closings between 1980 and 2003, and the Minnesota association reports 33 closings from 1987 through 2005. In Pennsylvania, the Health Care Cost Containment Council reports 20 from 2001 through 2005. It’s not like trend data are impossible to get; they are collected by all state governments. But access to them can be difficult for outsiders. And current data are almost impossible to obtain. Determined researchers might be able to find them, but by the time they do so, the data are likely to be seriously outdated.
We know even less about recent hospital openings. Modern Healthcare reports that 31 hospitals opened in 2003, but that may not be a complete list; furthermore, 9 of those hospitals were in Louisiana, so some of them may no longer be functioning in this post-Katrina and -Rita world. The federal HHS in 2003 reported that from 1990 through 2000, 129 hospitals opened. There have been no further HHS reports on the subject.
Furthermore, researchers use different standards, different databases (when they exist) and different definitions; data often are not comparable and may be incomplete or inaccurate. So we have little reliable information about what is happening to our hospital system.
Why, in recent years, has information that used to be readily available suddenly become scarce? There are many possible explanations. In terms of closings, it may just be one more instance of the traditional American aversion to bad news. Also, negative press coverage might give a black eye to the government or health system or company that closes a facility.
After all, there have been some bruising battles over proposed closings, including that of HCA’s San Jose (Calif.) Medical Center, which closed in 1994 despite strong community protest, leaving the city with no downtown hospital; Destin Hospital in Florida, which HCA closed in 1994 on very short notice; and the county-owned King-Drew Medical Center in Los Angeles, which has survived (so far) but has closed its once-busy trauma center. Its fate is still uncertain.
Advocates have saved the public Rancho Los Amigos, Los Angeles County’s renowned public rehabilitation hospital, but only temporarily. Although it’s a done deal, there was much complaint about the proposed closure of the Army’s fabled Walter Reed Hospital in Washington, D.C. And a battle royal, to put it mildly, has broken out in New York State over Governor Pataki’s proposal to order the closing of hospitals for efficiency and cost containment.
Other analysts have suggested that the lack of data on closures has its roots in ideology, in that the proponents of competition as the solution for all of health care’s problems would be embarrassed by evidence of the failure of this approach. And, obviously, many community development efforts would be stymied if it were known that the nearest hospital is now 50 miles away. All of these factors, and more, make it convenient to sweep hospital closings under the rug whenever possible.
It’s harder to explain the skimpy data on openings. Yes, the bitter feud over niche hospitals and quasi-hospitals could lead a prudent niche facility to open quietly so that it doesn’t add fuel to the fire, but construction of these hospitals has been prohibited for over a year. And, given that openings are usually viewed as good news, it’s a puzzlement as to why we don’t have accurate and comprehensive data on a timely basis. But we don’t.
That brings us to the obvious question: Do we know so little about which hospitals are born and die because it doesn’t make any difference? Isn’t it just that the hospitals are going where the people are, or leaving where the people aren’t? In some cases, probably; in other cases, maybe not.
Let’s take the example of Gilbert, Ariz., an affluent community just east of Phoenix, near Mesa. Arizona is enjoying a population boom, and no place is enjoying it more than Gilbert. In 1980, its population was 5,717; in 2004, it had 164,685 residents, 86 percent of them white. Median household income is $68,032.
To put it mildly, Gilbert has been underbedded--if you can call a community with no hospital “underbedded.” (Perhaps I should say “unbedded.”) As one news report put it, “Gilbert may be family-friendly, but there are still no places to be born or buried.”
Well, once Gilbert’s growth spurt was obvious, the stampede began. Catholic Healthcare West (CHW) announced that it would build Gilbert Mercy Medical Center. Banner Health announced that it would also build a Gilbert Medical Center--across the street from the Mercy facility. Iasis Healthcare announced that it wanted one, too, and acquired land for that purpose. So did Vanguard Health Systems. A freestanding emergency hospital opened earlier this year. Banner and CHW went toe to toe, and Banner withdrew its proposal, only to announce that it would build a hospital elsewhere in Gilbert, due to open in 2007. Gilbert Mercy has already announced expansion plans. And so it has gone.
Of course, Arizona is extremely short of both physicians and nurses, but one presumes they will be found somewhere.
Then there’s Gary, Ind., an old steel town that is heading in the other direction. In 1930, its population was 100,426; the booming steel industry led to population growth, and in 1960, Gary had 178,415 residents. Then the long decline of steel manufacturing began. By 1990, the population was down to 116,646; in 2000, it was 102,746. It is around 99,000 today.
Also, 84 percent of Gary’s population is African-American, with a median household income of $27,195--well below the national average.
Gary has one hospital left: Methodist Northlake, whose administrators announced in March 2006 that it would likely close within two years. It has been unable to meet some codes and standards, and 70 percent of its 23,000 patients are on Medicare or Medicaid or are uninsured. Its parent system lost $25 million last year. The state owes the hospital $67.5 million from various funds, but it won’t release the money unless the hospital can prove it is a sustainable operation. This, of course, makes no sense, because Northlake might well be sustainable if it had the state money. Well, never mind that. The hospital is eliminating elective surgeries and cutting beds and staff in order to survive. Its prospects, needless to say, are not good. If it closes, 100,000 people will have to leave town for hospital care.
The Methodist system has another hospital in Merrillville, Ind., about 25 miles away. Merrillville is an upscale community of 31,000 people, 70 percent of them white, with a median income of $49,545, nearly twice that of Gary. That hospital is doing fine. But one wonders what the good people of Merrillville will think if all of Methodist Northlake’s former patients start coming to their hospital for care. As a hospital spokesman in a neighboring town, which happens to be 92 percent white, said, “The closure of Methodist Hospital would really have a profound effect, would ripple throughout northwest Indiana at our hospitals. If that were to occur, you would expect longer waits in the emergency room.” That, and other things.
Hospitals open for obvious reasons, and they close for equally obvious reasons. But other reasons are not so obvious. Nationally, there is an unsettling pattern of closures in poor African-American communities that is not seen in other low-income areas.
This is not to say that hospitals are racist; indeed, hospitals, as the last line of refuge for people who have been shut out of the rest of the health care system, have done a splendid job of taking care of people of all races, colors and creeds--if the hospitals happen to be in those communities. The problem is that hospitals are less and less likely to be there.
Professor Sager’s research has shown that over time, hospitals in African-American communities are far more likely to close than those in communities with other racial makeups. That, he says, makes for compromised access, because African-Americans are more likely to use the emergency department as their regular source of care. Other research has found that older minority patients of all races are less likely to travel farther if their usual source of care becomes unavailable. Another study found that when hospitals closed in Los Angeles County, low-income residents had less access to care, and seniors were less likely to get flu shots (just what we need). The researchers also found evidence that traveling longer distances for hospital care increases the risk of infant mortality and deaths from injuries and heart attacks.
But this picture is much too complex to allow for simple assumptions. The links among race, poverty and insurance status are so strong that some researchers use race as a proxy for the economic status of people and communities. And rare is the hospital located in a poverty-stricken neighborhood, no matter what its demographics, that can survive for long without help. Sometimes a philanthropic tradition will see it through; sometimes government helps; sometimes a health system can provide cross-subsidization from more solvent members.
But the fact is that sooner or later, in a competitive, dog-eat-dog health care environment, hospitals that care for large numbers of uninsured and underinsured poor patients, and that don’t have outside financial assistance, will go under. As Professor Sager says, “With hospitals, it’s survival of the fattest, not the fittest.” He reports that large tracts in Atlanta, Baltimore, Chicago, Cleveland, Denver, Detroit, Los Angeles County, New York, St. Louis and Washington have already lost enough hospitals that access to appropriate, high-quality care--and sometimes any hospital care at all--is probably or definitely compromised.
And don’t forget Gary.
Maybe there is intentional racism in the policy decisions that are or are not being made about closings and openings; maybe it’s just what happens in a competition-driven, laissez-faire environment. The problem, of course, is that there are not just dead hospitals at the end of this road; there are also dead patients who did not, or could not, make the 50 miles to the next hospital in time.
So my plea for better data is not just a researcher’s yen for more timely and accurate information. It is also a desire for ammunition to better inform policy decisions, or to force policy decisions on the part of those who would rather not make them and who therefore benefit from a lack of knowledge.
Information makes for change. I was amused by a recent news report concerning trans-fats in foods. Once it became a requirement that the amount of trans-fats be reported on food packages along with other nutrition information, a funny thing happened: Manufacturers started removing trans-fats from their products.
If we knew, on an annual basis, where hospitals were closing and opening, and why, and what was happening to their communities as a result, we might be able to prevent calamities in the making and encourage positive developments. As it is, what policy there is seems to be driven by crisis, when it is announced that a critically needed facility is going under. All of a sudden, there is a flurry of activity and pronouncements and calls to the state capital and to Washington for aid, and disaster is averted. For a while. Until next time. As Sager puts it, “It’s policy by spasm, which is almost as intelligent as the twitch of the leg of a dead frog.”
“Closing hospitals,” says Sager, “is not policy or a substitute for policy.” Ignorance about what is happening is not policy, either. Refusing to collect or release data will not make a bad situation go away. Regardless of whether one favors competition or regulation in health care, or both, or neither, making important decisions while flying blind is unwise at best and destructive at worst.
The least we could do is keep track of what is being born and what has died so that we have a snapshot of what is happening to our hospitals--and to the communities they serve.
First published in Hospitals & Health Networks OnLine, April 5, 2006
© Emily Friedman 2006