How far to the hospital?: The effect of hospital closures on access to care

https://doi.org/10.1016/j.jhealeco.2005.10.006Get rights and content

Abstract

Do urban hospital closures affect health care access or health outcomes? We study closures in Los Angeles County between 1997 and 2003, through their effect on distance to the nearest hospital. We find that increased distance to the closest hospital increases deaths from heart attacks and unintentional injuries. This finding is robust to several sensitivity checks. We also find that, for residents with health insurance, increased distance shifts regular care towards doctor's offices. While most residents are otherwise unaffected, we find some evidence that seniors perceive more difficulty accessing care.

Introduction

Just prior to the November 2002 elections, Los Angeles County announced that without a US$ 350 million bailout it would be forced to close several area hospitals and clinics. High on the list of proposed closures were Harbor-UCLA and Olive View-UCLA Medical Centers, hospitals that serve a disproportionate share of the county's Medi-Cal and uninsured populations. Since Harbor-UCLA is a Trauma I center, its closure would mean the loss of significant trauma and emergency care services in the Los Angeles area. The passage of a ballot initiative (Measure B) that increased tax funding for emergency rooms and trauma centers has reduced pressure on the county's health care system though, even with this additional funding, the system is still projected to face a deficit of between US$ 300 and 600 million over the next 3 years. Thus, the possibility of imminent hospital closures remains real.

The proposed closures are part of an ongoing trend in Southern California. Between 1997 and 2002, Los Angeles County lost roughly 10% of its initial 131 hospitals (see Table 1). Since 2002, nine more general acute care hospitals have closed in Los Angeles County. Although considerable media attention has focused on the potential deleterious effects of these closures on access to care and health outcomes, surprisingly little is known about the impact of urban hospital closures on patients.

The bulk of the literature on urban closures focuses on the supply side of the market: the determinants of closure and the operating efficiency of hospitals remaining in the market (see Lindrooth et al. (2003) for a good summary). This literature finds that nationally, closed hospitals tend to be small (fewer than 100 beds), financially distressed, for-profit facilities, operating with excess capacity. They also tend to offer fewer services, such as neonatal intensive care units, or specialized cardiac or emergency care services.

Scheffler et al. (2001) confirm that poor financial performance was a key predictor of hospital closures in California between 1995 and 2001. As shown in Table 2, the hospitals that closed in the Los Angeles Region between 1997 and 2002 were typical of closures on other dimensions as well. Most were small (the mean number of beds was 88) and all but two were for-profit facilities. Nonetheless, these hospitals did supply services that are critical for certain patients. For example, about two-thirds offered emergency medical or cardiac services, such as by-pass surgery or cardiac catheterization. The impact of closing this type of hospital on the health and health care needs of residents in surrounding areas is ultimately, however, an empirical question.

Research on the impact of closures on access to care and health more generally has focused largely on rural hospitals (Bindman et al., 1990, Mullner et al., 1989, Rosenbach and Dayhoff, 1995, Succi et al., 1997, US GAO, 1991). For obvious reasons, such studies have, at best, limited implications for considering the consequences of hospital closures in urban areas, such as Los Angeles County. A notable exception, Vigdor (1999), examines the effect of changes in the density of hospitals in Los Angeles County between 1984 and 1995 on rates of avoidable hospitalizations and deaths in the hospital from heart attacks and motor vehicle accidents. As pointed out by the author, however, by focusing solely on hospital discharges, Vigdor (1999) cannot assess the effect of closures on the health of people who never make it to the hospital in an emergency or on people who rely on hospital-based outpatient facilities.

In this paper, we address the gap in the literature by assessing the impact of hospital closures in the Los Angeles Region on perceived access to care, health care utilization and health outcomes. We consider closures through their effect on distance from a resident's home to the nearest hospital. Past work has shown that patients typically choose both providers and hospitals, particularly for acute conditions, based on proximity and reduced travel time (Cohen and Lee, 1985, Dranove et al., 1993, Hadley and Cunningham, 2004, Luft et al., 1990, McClellan et al., 1994, McGuirk and Porell, 1984). Thus, increased distance may translate to reduced access to care. While patients affected by a closure in urban areas often have other hospitals nearby,1 the reduction in hospital supply may lead to increased crowding at and reduced access to the facilities remaining in the market. As a result, some may forgo or delay care when obtaining it becomes more difficult.

On the other hand, closures may have beneficial effects for nearby residents. Since closed hospitals are typically low-volume, poor-performers, health care outcomes might improve as residents are forced to choose among the remaining higher volume hospitals. Similarly, closures may shift some patients’ usual source of care from a hospital to physician offices or community clinics, which are generally viewed as more appropriate sources of primary care.

To the extent that closures affect access and utilization, the effects are likely to vary with patient characteristics. We expect the effect of closures to be greatest on seniors, who travel shorter distances to the hospital (Vigdor, 1999) and low-income patients, who are both less likely to travel far and more likely to use the hospital as their “regular” source of care (Weissman and Epstein, 1994).2 Indeed, in a study of hospital choice for maternal delivery in the San Francisco Bay Area, Phibbs et al. (1993) find that Medi-Cal women rely more heavily on public transportation than privately insured women and are therefore more sensitive to distance. Given the higher likelihood among Medi-Cal women of delivering at hospitals lacking specialized neonatal care and with worse perinatal outcomes, the authors interpret distance as a barrier to effective care for the poor. Similarly, in a study using national data, Currie and Reagan (2003) find that central-city black children living further from a hospital are less likely to have had a check-up, regardless of their insurance status. Both studies suggest that to the extent that closures force nearby residents to travel further for care, poor women and children may be particularly adversely affected.3

There may also be important differences with respect to health conditions. Even if the closure of weaker, poorer performing hospitals improves the average quality of hospitals, closures may have negative consequences for certain types of patients. In particular, outcomes for patients experiencing health events requiring fast attention, such as injuries sustained in an accident or a heart attack (AMI) may be affected by small changes in travel distance (Herlitz et al., 1993). In contrast, we would not expect urban hospital closures to affect mortality from conditions like chronic ischemic heart disease, where immediate emergency care is less relevant.

Our analysis is based on two distinct sources of health data: household surveys conducted by the Los Angeles County Department of Health Services (LACDHS) between 1997 and 2002, the period when most of the recent closures were occurring, and annual administrative zip code level mortality data from the California Department of Health Services. With the survey data, which provide information on residential location, we can assess the impact of changes in hospital proximity on perceived health care access and reported health care utilization. The administrative data give us an independent source of information on health outcomes that is not subject to self-reporting bias.

We find little effect of increased distance to the nearest hospital on outpatient utilization and the effects we do find are mixed. On a positive note, we find that increased distance is associated with an increase in the probability that respondents report a regular source of care as well as an increase in the likelihood that this care is sought at a doctor's office. Distance has little effect on perceived access to care in the population generally, though it is negatively related to perceived access for seniors, who may rely more on hospitals. Some models suggest that hospital closures may be associated with reductions in the probability that uninsured residents receive hospital-intensive diagnostic care, such as colon cancer screenings. Not surprisingly, we find no effect of increased distance on the receipt of other types of preventive care, such as HIV tests, pap smears, mammograms and flu shots that are commonly provided in non-hospital settings.

While the survey data point to some beneficial effects of hospital closures, the mortality data tell a different story. We find that increased distance to the nearest hospital is associated with an increase in deaths from acute myocardial infarction and unintentional injuries suffered at home, but not from other causes, such as cancer or chronic heart disease, for which timely care is less important. Thus, these results suggest that even the closure of small, private hospitals in large urban areas presents challenges to the provision of timely emergency care.

Section snippets

Data sources

Our area of study, Los Angeles County, has roughly 10 million residents spread over about 4000 square miles. The county is comprised of 88 cities, the largest of which, the city of Los Angeles, is home to roughly 40% of the county's population but covers only about 10% of its land. Another 10% of the population lives in unincorporated towns/areas.

We use several independent sources of data to study this area. The first is household level data from the Los Angeles County Health Surveys (LACHS),

Descriptive statistics: LACHS

Table 3 presents summary statistics for LACHS respondents overall, for those living in zip codes unaffected by closures, those living in affected zip codes before a change in distance and those living in affected zip codes after a change in distance. In addition to being directly relevant to our analysis that uses the LACHS, these figures also provide useful context for interpreting the zip code level mortality analysis.

For the full sample, the average driving distance to the nearest hospital

Conclusions

While urban communities across the country have experienced a string of small hospital closures over the past decade, Los Angeles County is unique in just how many closures have occurred. Should excess hospital capacity continue to grow in other urban areas, however, the Los Angeles experience may become more common. The present study will provide useful lessons for these communities.

Like past work showing that urban hospital closures improve the efficiency of the health care system by shifting

Acknowledgements

We thank Tom Y. Chang, Sendhil Mullainathan, Heather Royer, Antoinette Schoar, Gary Solon, Robert Town and seminar participants at the NBER Summer Institute Health Care meetings, the University of Michigan, the University of Illinois, Urbana-Champaign, the University of California, Irvine's Health Policy Research Center, the Association for Public Policy Analysis and Management's 2003 Fall Conference and the International Health Economics Association's Fifth World Congress for helpful comments.

References (37)

  • R.C. Lindrooth et al.

    The effect of urban hospital closures on markets

    Journal of Health Economics

    (2003)
  • C. Ruhm

    Healthy living in hard times

    Journal of Health Economics

    (2005)
  • American Heart Association

    Heart and Stroke Facts

    (2003)
  • L.C. Baker et al.

    Excess cost of emergency department visits for non-urgent care

    Health Affairs

    (1994)
  • D.J. Ballard et al.

    Referral selection bias in the Medicare hospital mortality prediction model: are centers of referral for Medicare beneficiaries necessarily centers of excellence?

    Health Services Research

    (1994)
  • M. Bertrand et al.

    How much should we trust differences-in-differences estimates?

    The Quarterly Journal of Economics

    (2004)
  • A.B. Bindman et al.

    A public hospital closes: impact on patient's access to care and health status

    Journal of the American Medical Association

    (1990)
  • B. Bloom

    Access to health care Part 1: children, National Center for Health Statistics

    Vital Health Statistics

    (1997)
  • B. Bloom

    Access to health care Part 2: working-age adults, National Center for Health Statistics

    Vital Health Statistics

    (1997)
  • T. Buchmueller et al.

    How Far to the Hospital: The Effect of Hospital Closures on Access to Care

    (2004)
  • A.C. Cameron et al.

    Regression Analysis of Count Data

    (1998)
  • M. Cohen et al.

    The determinants of spatial distribution of hospital utilization in a region

    Medical Care

    (1985)
  • J. Currie et al.

    Distance to hospital and children's use of preventive care: is being closer better and for whom?

    Economic Inquiry

    (2003)
  • D. Dranove et al.

    Segmentation in local hospital markets

    Medical Care

    (1993)
  • B. Efron et al.

    An Introduction to the Bootstrap, in Applied Statistics and Probability, no. 57

    (1994)
  • D.C. Goodman et al.

    The distance to community medical care and the likelihood of hospitalizations: is closer always better?

    American Journal of Public Health

    (1997)
  • J. Hadley et al.

    Availability of safety net providers and access to care of uninsured persons

    Health Services Research

    (2004)
  • W. Härdle et al.

    Bootstrap methods for time series

    International Statistical Review

    (2003)
  • Cited by (193)

    • Economic consequences of hospital closures

      2023, Journal of Public Economics
    • National clinical guidelines and treatment centralization do not guarantee consistency in healthcare delivery. A mixed-methods study of wet age-related macular degeneration treatment in Denmark

      2022, Health Policy
      Citation Excerpt :

      Although studies suggest that centralization may contribute to reaching consistency in care [11,12], the empirical relationship between the number of units and the degree of consistency in care across units is largely undocumented. A few studies highlight that centralization of care implies more inconsistency in the delivery of care as patients living in rural areas may be exposed to increased travel distance to acute care necessary for survival [21,22]. For non-acute care, the empirical relation between centralization and consistency remains an open question.

    View all citing articles on Scopus
    View full text