Hospital 'deserts' a growing problem in major urban centers

Guest post by Alan Sager

In 1960, 42 acute care hospitals with 8,000 beds served Detroit's 1.7 million residents. But only four hospitals with 2,700 beds survived to serve 700,000 residents in 2010.

All of the survivors are costly major teaching hospitals. Especially in winter, ambulances are sometimes challenged to provide rapid response to the people of a 139-square-mile city whose residential side streets rarely see snow plows. 

Detroit is not alone. Hospital "deserts" are conspicuous--and growing--in broad expanses of St. Louis, Cleveland, Washington, Atlanta, several Texas cities, New York City's boroughs, and elsewhere. 

These findings stem from a study of changes in some 1,200 acute care hospitals in 52 cities over the past 75 years. Of the 774 hospitals open in these 52 cities in 1960, 484 (over three-fifths) had closed by 2010. 

What are the characteristics of the hospitals that survive? Decade after decade, hospitals are more likely to remain open if they are larger teaching hospitals, have more accumulated wealth relative to their size, and are located in neighborhoods that are heavily caucasian. Unexpectedly, hospital efficiency--either alone or after controlling for other factors--has had no value in predicting which hospitals survive.

But aren't these changes just one part of a steady, desirable, and money-saving shift of healthcare from inpatient to outpatient services? Why should we worry or do anything about them? 

First, as a nation, the United States has about 30 percent fewer hospital beds per 1,000 people than we did in 1940. And we have only about three-fifths as many hospital beds per 1,000 people as the average of the world's 31 rich democracies, ranking 4th-lowest, above only Sweden, New Zealand, and the U.K. Even so, we suffer inferior health outcomes coupled with much higher spending. 

For decades, experts have predicted alternating perceived shortages and surpluses of both hospitals and physicians. It would be unfortunate to compound this record of failure by closing hospitals that will later have to be rebuilt at $1-3 million per bed. 

Second, the actual pattern of hospital closings in cities has clear real-world consequences for access to all types of hospital care. Loss of hospitals subtracts inpatient care, emergency services, and other ambulatory services. Hospitals in predominantly African-American neighborhoods much more likely to close, raising travel times disproportionately. 

Indeed, the effects of closings on urban African-American patients are compounded, as they obtain fully one-third of their physician care at hospital ERs and other outpatient sites, compared with one-sixth for caucasians. Also, when hospitals close, many remaining doctors in private practice are more likely to retire or relocate. The balanced provision of healthcare rests on a mix of hospital and physician services. In cities--and in rural areas as well--hospitals and doctors are symbiotic, not substitutes for one another.  

Third, costs rise. Fewer hospitals means less competition, and that makes for higher prices. Also, the share of urban beds in teaching hospitals has risen from 44 percent in 1950 to 77 percent in 2010. Even routine care can be costly in teaching hospitals.   

Fourth, effects on appropriateness and technical quality of care are apparently mixed. Clearly, some hospitals that closed provided care of low quality. Separate was sometimes markedly unequal.  But patients displaced by closings don't always obtain care elsewhere. And, when they do, the larger teaching institutions that serve them have not always provide fully equitable levels of care. 

Fifth, unfortunately, no well-functioning competitive free market exists to fairly reward needed and efficient hospitals. The many requirements for such a market are not met in hospital care. (If it were, efficiency would actually confer survival value.) At the same time, competent government action is also generally lacking. Only one state has either compiled a list of the hospitals and ERs that are considered essential to protecting the health of the public, or worked to assure needed and efficiently run hospitals with sufficient revenue. 

Among the steps that could be taken include enacting state hospital receivership laws that empower public officials and patient groups to petition courts for short-term legal protection of needed institutions. 

These should be backed by short-term financial relief from state or federal stabilization trust funds, financed by modest assessments on the revenues of all hospitals. Long-term protection should take the form of assured budgets for each needed hospital. These should be set with one eye on protecting access to care and the other on containing costs of care--just in case ACOs and other currently fashionable ideas fall flat. 

Editor's note: Alan Sager is a professor of health policy and management at the Boston University School of Public Health.

Read more on