Study ID | Country | Study design | Condition | Intervention | Sample source | Sample size | Length of study |
---|---|---|---|---|---|---|---|
Avdic 2016 [6] | Other Europe Sweden | Controlled observational Analysis of linked administrative datasets | Acute MI | Hospital ED closure | Administrative registers obtained from the Swedish National Board of Health and Welfare (hospitalisations and deaths) | Approximately 374,000 events | 21 years (1990–2010) |
Combier 2013 [7] | France Burgundy region | Uncontrolled observational Before–after study | Obstetric/neonatal complications | Obstetric unit closure | Hospital discharge summary data for all deliveries from 22 weeks’ gestation in the region’s maternity units | 111,001 deliveries | 10 years (2000–2009) |
El Sayed 2012 [8] | USA | Uncontrolled observational Before–after study | General emergency care | Hospital ED merger | Routinely collected EMS and ED data | 5338 EMS transports; 21,685 ED visits | 3 months (June 1 to August 262,010) |
Hansen 2011 [9] | Other Europe Denmark | Uncontrolled observational Before–after study | General emergency care | Hospital ED closure | Danish National Person Registry including all Danish residents | 21,000 residents of Viborg county (2300 from Morso) | 7 years (1997–2003) |
Hsia 2012 [10] | USA California | Controlled observational Cohort | General emergency care Acute MI, stroke, sepsis and asthma/COPD | Hospital ED closure | California Office of State-wide Health and Planning Development database, combined with information on ED closures by year between 1999 and 2009 | 785,385, of whom 67,577 (8.6%) experienced an increase in distance to ED care as a result of an ED closure | 11 years (1999 to 2009) |
Hsia 2014 [11] | USA | Other cross-sectional comparison of existing datasets, compared at T1 and T2 10 years later. | Major trauma Acute trauma aged 20 or older. | Trauma unit closure | Database of trauma centres open at T1 and 10 years later at T2. Patient discharge database. Household demographic database. | 266,023 had no increased drive time, 5122 had increased drive time. | Compared 1999 to 2009 |
Knowles 2018 [3] | UK | Controlled observational Interrupted time series | General emergency care | Hospital ED closure or downgrade | ONS, HES, ambulance dispatch records | Unable to locate, refers to areas only | Two years pre closure and 2 years post closure. |
Mustonen 2017 [12] | Other Europe Finland (Vantaa, Finland’s third-largest city, with approximately 182,000 inhabitants) | Controlled observational Controlled before–after study | General emergency care | Primary care ED closure | Electronic health records plus monthly mortality statistics by age groups | Unclear (34,000 inhabitants in area with ED closure) | 4 years (February 2004 to December 2007) |
Roberts 2014 [13] | UK England only | Uncontrolled observational National data on distance travelled to emergency care plus three case studies of local reconfiguration | General emergency care | Hospital ED closure or relocation | Hospital Episode Statistics plus data on ED attendances from every major (type 1) ED in England | 13 million ED attendances and 5.4 million emergency admissions (2011/12) | 10 years (2001/2 to 2011/12) |
Shen 2012 [14] | USA | Controlled observational Difference in difference approach | Acute MI | Hospital ED closure or relocation | American hospital annual survey, database for California hospitals, Medicare claims | Unclear | 4 years before change to 4 years after change to ED access |
Shen 2016 [15] | USA | Controlled observational | Acute MI | Hospital ED closure or relocation | Medicare records, cost provider systems | 1.35 million patients | 90-day mortality reported in this paper |
Yaghoubian 2008 [16] | USA California (Los Angeles County) | Uncontrolled observational Interrupted time series | Major trauma | Trauma centre closure | Patient records from prospectively collected database (Trauma and Emergency Medicine Information System) | 14,996 | 9 years 2 months (January 1997 to 1 March 2006) |