The Dangers of Hospital Closures Ch 3

II. HOSPITAL CLOSINGS HARM ACCESS TO HEALTH CARE Closing hospitals undermines equity of access to urban health care. It does so in two general ways. The first is geographic. 9 Cities vary greatly in the ratio of hospital beds per 1,000 citizens. Hospital closings do not tend to even out these differences. This is important to consider in light of the very substantial differences in beds per 1,000 people among large counties in New York, and also the inter-borough differences in beds per 1,000 in New York City. 9 Hospitals in black/African neighborhoods are much more likely to close. Over several decades, this pattern has strikingly visible consequences. 9 Hospital closings undermine continuity of inpatient care. Many of the patients displaced by closings take long times to re-weave the fabrics of their health care. The second concerns hospitals and the provision of physicians’ services. 9 Hospitals are very important providers of ambulatory care, especially to black/African-American patients. 9 Hospitals and doctors are mainly symbiotic. Each is needed to sustain the other. Because hospitals need doctors to refer, admit, and care for patients, hospitals are motivated to recruit and sustain doctors who practice nearby. Some hospitals are located in areas where physicians are in short supply, and are therefore motivated to recruit physicians into the area, but these tend to be the very hospitals that lack the money to do so. A. HOSPITAL CLOSINGS UNDERMINE ACCESS GEOGRAPHICALLY 1. Inter-county differences in beds per 1,000 residents Large counties in New York State differ substantially in their ratios of hospital beds per 1,000 people. Data from 2004 show that Manhattan had 7.7 beds per 1,000. At the other extreme, Queens had only 2.0 and Saratoga had only 0.9 beds per 1,000. These ratios count licensed beds, many of which are not actually staffed. A look at staffed beds shows even greater disparities among some of the counties. Evidence from our 52-city study of hospital closings nationwide shows that there is no tendency for hospitals to close in the cities with high beds per 1,000. Indeed, the reverse seems to be true. It seems that the forces that make for more beds generally tend to sustain more beds over time. Even an across-the-board 10 percent cut in beds would do nothing to help equalize the ratios of hospital beds per 1,000 people across counties. Exhibit F Licensed Hospital Beds per 1,000 People, 2004 New York Counties with More than 200,000 People 9.0 8.0 7.7 7.0 6.0 5.0 4.0 4.4 4.0 3.8 3.7 3.4 3.4 3.3 3.3 3.0 3.4 3.0 3.0 3.0 2.9 2.8 2.7 2.6 2.5 2.0 2.0 0.9 1.0 0.0 e rag Ave ga ato Sar s een Qu folk Suf a ga r Nia e nro Mo ss che Du yn okl Bro nx Bro nge Or a nd Isla ten Sta a eid On sau Nas e Eri nd kla Roc aga ond On e om Bro ter hes stc We any Alb ttan nha Ma 2. Race and closings Decade after decade, since the 1930s, hospitals located in black/African-American neighborhoods of 52 U.S. cities have been substantially more likely to close than those located in white neighborhoods. This relationship holds up after controlling for many characteristics of hospitals themselves, their competitive environments, and other factors. Appendix 2 presents the results of logistic regressions for various periods from 1980 to 2003. Exhibit G shows the probability that a hospital will close at any time from 1980 to 2003, depending on the racial mix of its neighborhood in 1980 and its bed size in 1980. As shown in Exhibit G, 9 less than three percent of hospitals with 832 beds that were located in areas that were 1.0 percent black in 1980 were predicted to close during the following 23 years; 9 the average hospital with 340 beds had a 23 percent chance of closing in a neighborhood that was 1 percent black/African-American, but a 43 percent chance of closing in a neighborhood that was 90 percent black/African-American; and 9 at the other extreme, almost three-fourths of 66-bed hospitals that were located in areas that were 90.0 percent black in 1980 were predicted to close. Exhibit G The Predicted Chance of Closing between 1980 and 2003 Is Greater for Hospitals with Fewer Beds in 1980 and Greater Area Percent Black Area % black in 1980 5th %ile mean 75th %ile 95th %ile 95th %ile / 5th%ile 1 30 50 90 Hospital beds in 1980 95th %ile 75th %ile mean 832 443 340 25th %ile 162 5th %ile 66 5th %ile / 95th %ile 17.7 15.1 13.3 10.3 2.9% 3.9% 4.8% 7.1% 15.5% 19.9% 23.4% 31.8% 22.8% 28.6% 33.0% 42.9% 40.2% 47.7% 53.0% 63.2% 51.2% 58.8% 63.8% 72.8% 2.4 2.1 1.9 1.6 1.4 The cumulative impact of decades of closings has been to strip away hospitals from large expanses of U.S. cities. Today, these are neighborhoods where large numbers of black/African-American, Latino, and other people live—people who are often vulnerable to denial of needed health care owing to their insurance status, their lack of primary care physicians, and other factors. The loss of hospitals from these neighborhoods heightens that vulnerability. (Even though race is the main demographic predictor of closing across the 52 cities, older residential neighborhoods in older northeastern and midwestern cities that are heavily black/African-American or Latino tend to lack hospital care.) We display this loss through maps of Detroit, St. Louis, and Washington, D.C. in exhibits H, I, and J. Exhibit H Hospital Survival and Closings in Detroit, 1937 - 2003 Exhibit I Hospital Survival and Closings in St. Louis, 1937-2003 Exhibit J Hospital Survival and Closings in Washington, D.C., 1937-2001 3. Continuity of care Continuity of care rests strongly on hospital survival. Closing a hospital removes inpatient care, emergency room care, and outpatient clinics. Hospitals are not characters in a video game that can just be blown up or moved around with a controller. For many patients, a stable hospital—open for business in one location— is a visible, solid, and trustworthy guarantor that needed health care will be provided. Disruption of inpatient care may also follow hospital closings. Some who are sanguine about hospital closings, relocations, or mergers may regard hospitals largely as interchangeable parts in a health care machine. But this may be inaccurate. Many patients appear to respond conservatively when their customary hospital closes. Shepard found that 30 percent of the inpatients displaced by a hospital closing do not re-appear at surviving institutions nearby, at least not for some time.42 In one eastern Massachusetts community, half of all residents never or only occasionally leave the immediate area to obtain health care. “Eighty-five percent said that distance was the most important factor in deciding where to get care.”43 Bindman, Keane, and Lurie succeeded in following patients of a closed county hospital. In the year after the closing, they found a doubling in the share of patients lacking a regular provider, a rise of more than one-half in the share reporting a denial of care, a drop in perceived health, and a rise in pain. Regular patients of a surviving county hospital experienced no deterioration.44 B. HOSPITALS AND PROVISION OF PHYSICIANS’ SERVICES Some advocates of hospital closings have tried to justify their proposals by claiming that dollars saved could be used to finance vitally needed ambulatory care services. There are several problems with these assertions. One is that money is not saved when hospitals are closed. A second is that savings, even if they did arise, usually can’t be captured and put to other uses. Third, and at least as important, hospitals and doctors are more symbiotic than adversarial. Both are needed—especially in communities that lack adequate medical resources. We find, therefore, that it is a mistake to pit primary care and other physician services against hospital services. There are two distinct reasons. 1. Hospitals are very important providers of ambulatory care We are concerned that closing some hospitals will remove needed services from communities in which inpatient and ambulatory care are already in short supply. Closing some hospitals will do little to advance the financing of important primary care for patients who are already vulnerable to denial of needed services. New investments in primary care services should be guaranteed and actually made—not promised—before hospitals providing ambulatory care are closed. No such guarantees appear to be in place. As shown in Exhibit A, New Yorkers rely very heavily on hospital-based non-emergency outpatient department (OPD) care. New York’s rate of use of the OPD per 1,000 residents is two-fifths above the national average, and the state ranks 13th- highest in reliance on OPD care. This reliance on hospitals for ambulatory physician services might seem unexpected— since New York ranks fifth in the nation in office-based physicians per 10,000 people— but it is real. Unfortunately, office-based physicians are unevenly distributed geographically, and some physicians are not readily available to Medicaid-financed or uninsured patients. Consequently, hospitals are vital providers and sponsors of OPD services to patients who lack good alternative sources of physician care. This is particularly true for blacks/African-Americans, who rely twice as heavily on OPDs and ERs as do whites. (Data on doctor visits by site of care and by race are available; data by ethnicity are not similarly available.) It is probably true as well for Latino patients and others who are vulnerable to deprivation of needed care. Race and OPD/ER use. As shown in Exhibit K, whites and blacks visit physicians almost exactly as often overall, about 400 visits per 100 people in 2003. But 68 of whites’ visits (17 percent of 399 visits per 100 annually) were made in hospital OPDs or ERs, while 130 of blacks’ visits (33.2 percent of 391 visits per 100 annually) were made in OPDs or ERs. 45 This appears to be tied mainly to race, not to insurance status alone. Consider only people aged 75 and above. Virtually all are covered by Medicare. Whites relied on hospital OPDs and ERs for 12 percent of their physician contacts in 2003, while blacks relied on them for 22 percent of their physician contacts. Exhibit K Visits to Physicians per 100 People by Site and Race, 2003 17% of visits in OPD+ER 33% of visits in OPD+ER 400 38 350 71 30 300 59 Visits by site ER 250 Outpatient department 200 332 150 261 MD office 100 50 0 white black Race These are national data but this pattern can be expected to prevail in individual states also. Therefore, it is reasonable to predict that closing of hospitals will disproportionately harm blacks’/ African-Americans’ access to physician services. Latinos/Hispanics are also likely to experience disproportionate harm. Cutting New Yorkers use of hospital OPDs to national levels might, by itself, save money, but it would result in a loss of 11.0 million doctor visits in OPDs annually—a cut of 26 percent. In 2000, blacks were 15.9 percent of New York State’s population. 46 But, applying blacks’ higher national rate of use of OPDs, 59 OPD visits per 100 people, we estimate that about 31.5 percent of OPD visits in New York State are made by blacks. This means that blacks would lose 3.5 million of the 11.0 million reduction in OPD visits necessitated to cut use of OPDs in New York State down to national levels. This calculation assumes that the loss of OPD care were evenly distributed throughout the state. Were hospital closings and associated OPD losses disproportionately concentrated in black/African-American neighborhoods, blacks’ share of loss of OPD care would be substantially greater. 2. In areas where vulnerable patients reside, hospitals and doctors are symbiotic, not competitors In some communities, physician groups build free-standing ambulatory surgery or MRI centers to try to compete with doctors for well-insured patients or patients with profitable diagnoses. But in neighborhoods where patients who are vulnerable to denial of needed care reside, hospitals and doctors are mainly symbiotic. Each is needed to sustain the other. Because hospitals need doctors to refer, admit, and care for patients, hospitals are motivated to recruit and sustain doctors who practice nearby. In under-served areas, symbiosis means that without hospitals, many physicians can be overwhelmed or experience greater difficulty in earning a living. As the share of U.S. physicians in primary care shrinks, and as extensive areas of many cities experience loss of physicians, hospitals that depend on physicians in private practice are vulnerable to closing. The neighborhoods with high minority population share and lower incomes, that are likeliest to lose their non-teaching hospitals, tend to start with fewer physicians.47 Other things equal, closing a hospital generates greater need for more non-hospital care— from physicians, free-standing ambulatory surgery centers, community health centers, pharmacies, and the like. But does the amount of such care rise after a nearby hospital closes? Physicians, for example, might move in to care for patients displaced by a hospital closing. And community health centers might expand. Alternatively, though, physicians who had depended on a closed hospital to care for their patients— physicians who had been in and out of a hospital daily or more often— might choose to retire or relocate their practices after their hospital closed.48 49 In a very large expanse of one large U.S. city, large-scale demographic change resulted in a pronounced change in the racial mix. Most of the physicians in private practice gradually relocated their offices to suburban areas, to follow their patients. Some retired. Most physicians see themselves as serving patients, not places. As a result, this sector of the city had far fewer physicians than before the demographic succession. Still, the area retained most of its hospitals, and these wished to remain open and available to address the medical needs of the new residents surrounding them. The great majority of these new residents were insured. Unfortunately, without enough doctors to admit patients and care for them, the hospitals could not remain open. In time, all closed or relocated to suburban areas. Like cars with well-tuned engines and good tires, they could not function—because they lacked transmissions. Doctors play that role in a hospital. Some hospitals are located in areas where physicians are in short supply, and are therefore motivated to recruit physicians into the area, but these tend to be the very hospitals that lack the money to do so. They are willing but not able. In summary, hospital costs in New York State are very high. The state does have an above-average number of hospital beds per 1,000 residents. Therefore, it seems logical to close hospitals to save money. Unfortunately, this will not work, and is likely actually to be counter-productive. It won’t work because excess beds account for only a very small share of the state’s high costs. Also, it is counter-productive. There are two reasons. First, hospitals that close tend to be less costly than survivors. Second, closing some hospitals will boost the market power of surviving hospitals, allowing them to raise their prices. This will force us and our insurers to pay them even more money. At the same time, the pattern of hospital closings has been racially and ethnically imbalanced. The cumulative effect of past closings has left many neighborhoods without needed hospital care. Imbalances in hospital beds per 1,000 people among the boroughs of New York City are also clear. Hospital closings are sometimes sold as a means to save money to finance badly needed primary care. But usually, there are no mechanisms to capture the savings on hospital care—if any actually materialize—and use them to finance more primary care. At the same time, hospital closings deprive many neighborhoods of vitally needed services provided by doctors in outpatient clinics and emergency rooms. Hospital closings look attractive for many reasons—because they seem like a logical solution, and because hospitals can actually be closed. Unfortunately, they don’t save money and they do harm access to care. Their promise greatly exceeds their performance. Closing some hospitals will benefit the surviving hospitals, but that will not lower cost or protect access to care. It will, though, protect the survivors. Until the next round of closings. By boosting cost and degrading access, hospital closings accelerate the drift of health care toward fewer services for fewer people at greater cost. This is medically, socially, economically, and politically unwise.