|
Author/year | Focus | Geographic area | Study design/sample | Quality score: rigor/relevance (high or low) |
|
Comparisons of care provided |
Gadzinkski et al. (2013) | Differences in utilization and outcome measures between CAHs and non-CAHs for patients undergoing the eight most performed CAH surgical procedures between CAHs and non-CAHs | The U.S. | Retrospective cohort study (2005 to 2009) n = 1282 CAHs and 3612 non-CAHs, representing 6,587,713 surgical admissions | Low/high |
Ibrahim et al. (2016) | Differences in surgical outcomes and costs for patients treated in CAHs vs. non-CAHs | The U.S. | Cross-sectional retrospective review N = 1,631,904 MCR beneficiary admissions; CAHs n = 828, non-CAHs n = 3676 | Low/high |
Ibrahim et al. (2018) | Differences in mortality, serious complications, reoperation, and readmission for patients undergoing emergency colectomies performed at CAHs and non-CAHs between 2009 and 2012? | The U.S. | Cross-sectional retrospective correlational study n = 219,170 patients who underwent urgent or emergent colectomies | Low/high |
Johnston et al. (2019) | Rural, micropolitan, and metropolitan access to primary care, specialists, and hospital beds | The U.S. | Retrospective time-series design, using CMS fee-for-service data n = 66,585,996 weighted patient-years in the final sample. 9.6% rural, 17.6% micropolitan, 72.8% metropolitan. This sample represented 11,581 unique beneficiaries | Low/high |
Joynt et al. (2011) | Comparison of the quality of care at CAHs vs. non-CAHs for three common diagnoses (AMI, CHF, and pneumonia). Identify what contributes to differences in the quality of care between the two types of hospitals | The U.S. | Retrospective observational study using Hospital Quality Alliance data (2008 to 2009) n = 4738 hospitals (1268 CAH; 3470 non-CAH) and 2,351,701 index admissions for AMI, CHF, & pneumonia | Low/high |
Joynt et al. (2013) | Comparison of 30-day mortality for AMI, CHF, and pneumonia in CAHs vs. non-CAHs and rural CAHs vs. rural non-CAHs (from 2002–2010). Identification of the resources those are associated with improvement. | The U.S. | Retrospective observational study using data from MCR FFS patients CAH hospitals (n = 860 in 2002 and 1264 in 2010). Non-CAH (n = 3108 in 2002 and 3255 in 2010). AMI: n = 1,902,586 admissions; CHF: n = 4, 488,269 admissions; pneumonia: n = 3,891,074 | Low/high |
Lichtman et al. (2012) | Differences in 30-day mortality and 30-day readmission rates (both risk-standardized for ischemic stroke patients) at CAHs vs. non-CHAs in 2006 | The U.S. | Retrospective review of MCR data (2006) n = 4546 hospitals (1165 CAHs and 3381 non-CAHs) and 310,381 ischemic stroke discharges (10,267 from CAHs; 300,144 from non-CAHs) | Low/high |
Natafgi et al. (2017) | Surgical patient safety outcomes in CAHs in comparison to comparably sized PPS rural hospitals in a four-state analysis | Colorado, North Carolina, Vermont, and Wisconsin (<8% of CAHs in the U.S.) | Retrospective review of HC-UPSID and AHA annual survey data n = 35,674 discharges (14,296 from 100 CAHs and 21,378 from 36 small PPS hospitals) | Low/high |
Windorski et al. (2019) | Differences in ICU LOS, ventilator requirements and duration, hospital LOS, and mortality for trauma patients treated initially in a CAH ED vs. those initially transferred to a level 1 trauma center | The U.S. | Retrospective correlational study using chart review data (2009–2014) n = 1478 patients; 73.3% transferred from CAH, 26.7% transported directly to level 1 trauma center | Low/high |
|
Analyses of hospital closures and openings |
Anderson et al. (2019) | How the expansion of trauma care centers in N.M. has translated into improved access to trauma care services at the population level and the characteristics of those with access to care | New Mexico | Retrospective cross-sectional analysis using spatial data at the census block level, 2007–2017 n = 2,082,669 general population | High/low |
Burkey et al. (2017) | How hospital closures based upon optimization models differ from actual closures (2001–2015); whether short-term optimization models are stable over time; how recommended closures affect service equity/equality; and whether recommended closures disproportionately impact rural communities | North Carolina, South Carolina, Virginia, and Tennessee | GIS-informed optimization modeling design using AHA and Census Bureau Data N = 4 states | High/high |
Busingye et al. (2011) | The changes in spatial and temporal disparities in access to emergency cardiac and stroke care that occurred in Middle Tennessee from 1999 to 2010 | Middle Tennessee | Retrospective network analysis and two-way comparison design n = 30 counties, 250 census tracts, and 1.2 million people | High/high |
Countouris et al. (2014) | The impact of a community hospital closure on older adults in a suburb of Pittsburgh | Pittsburgh, PA | Qualitative focus group study (2010–2012) using a community-engaged research process n = 37 participants | High/low |
Crandall et al. (2016) | The effect of the closure of L.A.’s MLK trauma center on the distribution of admissions to nearby trauma centers and the impact of the closure on mortality rates in these centers and within L.A. County | Los Angeles, CA | Retrospective observational study involving nonpublic patient-level data from the state of California, 1999–2009 N = 37,131 trauma patients | Low/low |
Germack et al. (2019) | The relationship between rural hospital closures and the supply of physicians following a closure | The U.S. | Retrospective design, assessing rural U.S. hospitals that experienced at least one closure, 1997–2016 n = 1541 rural counties over 20 years; equated to 30,820 county-years | Low/high |
He et al. (2017) | The effect of time to definitive care on patient outcomes after the closure of trauma centers | Uncertain | Chart review; pre- and postevent design, 2008–2009 and 2011–2013 n = 27,843 patients | Low (not enough information–conference abstract)/low |
Hsia & Shen (2011) | The populations most affected by the rising rate of hospital trauma center closures (2001–2007) | The U.S. | Retrospective correlational design n = 31,475 zip codes covering 283 million people | Low/high |
Hsia & Shen (2019) | Patient outcomes (in AMI patients) at bystander hospitals (high and nonhigh occupancy) when a nearby E.D. opens or closes | The U.S. | Retrospective time-series probability and fixed effects design n = 1,143,745 patients across 3720 hospitals (1209 of which were high occupancy) | High/high |
Joynt et al. (2015) | Characteristics of hospitals that closed between 2003 and 2011, the effect of closures on mortality and readmission rates, and whether outcomes varied by the acuity of medical condition or rurality | The U.S. | Retrospective correlational design, using MCR data, 2003–2011 n = 32,485,906 MCR beneficiaries, 195 hospital closures, and 2847 hospital service areas (184 w/closures) | High/high |
Liu et al. (2014) | How California E.D. closures (from 1999-2010) affected inpatient mortality rates at nearby hospitals | California | Retrospective analysis, using C.A. Office of Statewide Health Planning and Development’s Hospital Annual Utilization data from 1999-2010 n = 16,246,892 admissions to CA hospitals via ED | High/high |
McCarthy et al. (2021) | The impact of rural hospital closures (2010–2019) on the proportion of the population that can reach a hospital or E.D. by road within 15, 30, 45, and 60 minutes | The U.S. | Retrospective GIS-informed isochrone design, using U.S. Census Bureau population data (2010) and Provider of Services (hospital location) CMS data (2010–2019) n = 9 Census Bureau divisions, 48 contiguous states, and the populations therein | High/high |
Rhudy et al. (2016) | Effects of delayed interventional cardiology access and social factors associated w/excess ACS, NSTEMI, or STEMI mortality | Maine | Retrospective population-based secondary analysis of census and CMS claims data overlaid on a GIS-derived base map for 2013 n = 3126 ACS cases meeting inclusion criteria (2247 NSTEMI; 879 STEMI) | Low/high |
Romero et al. (2012) | The impact of the 2010 closure of St. Vincent’s Catholic Medical in Lower Manhattan on patients, community, and healthcare providers | New York, NY | Community-based participatory qualitative approach, using key informants and focus groups n = 16 key informants and six focus groups (44 group participants) | High/low |
Romero et al. (2012) | The impact of the 2010 closure of St. Vincent’s Catholic Medical in Lower Manhattan on patients, community, and healthcare providers | New York, NY | Descriptive survey study with both closed and open-ended questions n = 1438 respondents (local service providers and residents) | Low/low |
Shen & Hsia (2012) | Effects of increased drive times to the nearest E.D. on mortality rates, the health profile of the patients, and long-lasting outcomes of patients experiencing AMI | The U.S. | Retrospective observational difference-in-differences design, using American Hospital Association and MedPAR data, 1996–2005 n = 1.49 million patient-year observations | High/high |
Yamashita & Kunkel (2010) | The relationship between age-adjusted heart disease mortality and distance to a hospital, adjusted for social factors such as poverty, education, and insurance coverage | Ohio | Retrospective spatial analysis, using U.S. Census Bureau and Ohio Department of Health data for the year 2000 n = 88 Ohio Counties (representing >11 million residents in 2000) | High/high |
|