Rural versus urban academic hospital mortality following stroke in Canada
Rural versus urban academic hospital mortality following stroke in Canada
Richard Fleet 1 2 3
Sylvain Bussières 1 3
Fatoumata Korika Tounkara 1 3
SteÂ phane Turcotte 0 3
France LeÂ gareÂ 3
Jeff Plant 3
Julien Poitras 1 2 3
Patrick M. Archambault 1 2 3
Gilles Dupuis 3
0 Population Health and Practice-Changing Research Group, CHU de QueÂbec Research Centre , QueÂbec, QC , Canada , 4 Department of Family Medicine and Emergency Medicine and Knowledge Transfer and Health Technology Assessment Group, CHU de QueÂbec Research Centre and Evaluative Research Unit , UniversiteÂ Laval, QueÂbec, QC , Canada , 5 Faculty of Medicine, University of British Columbia and Department of Emergency Medicine, Penticton Regional Hospital , Penticton, BC , Canada , 6 Intensive Care Division, Department of Anesthesiology, UniversiteÂ Laval , Quebec, QC , Canada , 7 Department of Psychology, UniversiteÂ du QueÂbec à MontreÂal , MontreÂal, QC , Canada
1 Research Chair in Emergency Medicine UniversiteÂ Laval-CHAU HoÃtel-Dieu de LeÂvis , LeÂvis, QC , Canada
2 Department of Family Medicine and Emergency Medicine , UniversiteÂ Laval, QueÂbec, QC , Canada
3 Editor: Yu Ru Kou, National Yang-Ming University , TAIWAN
Data Availability Statement: Data on adjusted
30day in-hospital mortality following stroke from
2007 to 2011 were obtained on the CIHI website
interactive tool. This tool has since been modified.
Data can now been obtained at the hospital level for
all acute care hospitals in Canada upon request
from the CIHI, www.cihi.ca. CIHI owns the data.
Data are provided to CIHI from the individual
hospitals or health authorities according to
agreements. Access to data must meet CIHI
requirements and those of data contributing
Stroke is one of the leading causes of death in Canada. While stroke care has improved
dramatically over the last decade, outcomes following stroke among patients treated in rural
hospitals have not yet been reported in Canada.
To describe variation in 30-day post-stroke in-hospital mortality rates between rural and
urban academic hospitals in Canada. We also examined 24/7 in-hospital access to CT
scanners and selected services in rural hospitals.
Materials and methods
We included Canadian Institute for Health Information (CIHI) data on adjusted 30-day
in-hospital mortality following stroke from 2007 to 2011 for all acute care hospitals in Canada excluding
Quebec and the Territories. We categorized rural hospitals as those located in rural small
towns providing 24/7 emergency physician coverage with inpatient beds. Urban hospitals were
academic centres designated as Level 1 or 2 trauma centres. We computed descriptive data
on local access to a CT scanner and other services and compared mean 30-day adjusted
poststroke mortality rates for rural and urban hospitals to the overall Canadian rate.
A total of 286 rural hospitals (3.4 million emergency department (ED) visits/year) and 24
urban hospitals (1.5 million ED visits/year) met inclusion criteria. From 2007 to 2011, 30-day
in-hospital mortality rates following stroke were significantly higher in rural than in urban
Funding: This research project was supported by
the Research Chair Emergency Medicine Laval
University ± CHAU HoÃtel-Dieu de LeÂvis Hospital
and the HoÃtel-Dieu de LeÂvis Hospital Research
Fund and Fonds de Recherche du QueÂbec ± SanteÂ
to RF. The funders had no role in study design,
data collection and analysis, decision to publish, or
preparation of the manuscript.
Competing interests: The authors have declared
that no competing interests exist.
hospitals and higher than the Canadian average for every year except 2008 (rural average
range = 18.26 to 21.04 and urban average range = 14.11 to 16.78). Only 11% of rural
hospitals had a CT-scanner, 1% had MRI, 21% had in-hospital ICU, 94% had laboratory and 92%
had basic x-ray facilities.
Rural hospitals in Canada had higher 30-day in-hospital mortality rates following stroke than
urban academic hospitals and the Canadian average. Rural hospitals also have very limited
local access to CT scanners and ICUs. These rural/urban discrepancies are cause for
concern in the context of Canada's universal health care system.
While stroke care has improved dramatically over the last decade, little is known about
poststroke outcomes among patients treated in rural hospitals in Canada and whether they differ
from those treated in academic urban hospitals. Roughly 20% of Canadians live rurally and
more than 3 million patients are treated in rural Emergency Departments (ED) in Canada
each year [
]. Rural hospitals in Canada are often distant from tertiary referral centers and
have limited in-house /access to services that are cardinal in the early diagnosis and
management of acute stroke such as a CT scanner [1, 3±5]. Consequently, rural hospitals may not
meet recent standards for acute stroke care [
] potentially resulting in poorer outcomes than
better staffed and equipped academic urban hospitals.
Stroke is one of the three leading causes of death in Canada [
] and represents more than
$3.6 billion a year in physician services, hospital costs, lost wages, and decreased productivity
]. Stroke survival partly reflects the quality of acute care received [
], and stroke
case-fatality rates are now used for hospital benchmarking within and between member
countries of the Organisation for Economic Co-operation and Development (OECD) . As
Canada takes pride in the universal accessibility of its health care system, (Canada Health Act.
2011), equal access to quality stroke care should be provided to citizens wherever they reside
. To the best of our knowledge, no previous study has examined the rural versus urban
mortality gaps on a national level in Canada.
The primary aim of this study was to describe variation in 30-day in- hospital mortality rates following stroke in rural and urban academic hospitals in Canada.
Materials and methods
The rural hospitals selected were those located in rural small towns that provided 24/7
emergency physician coverage and inpatient beds for acute admissions. Statistics Canada defines a
ªrural small townº as a town or municipality outside the commuting zone of larger urban
centres with populations of 10 000 or more . Our full methodology for rural ED selection and
data collection on hospital characteristics is described elsewhere [
1, 4, 14
]. We have piloted
and published preliminary descriptive data on the same hospitals [
1, 4, 5
]. The urban hospitals
were academic centres designated Level 1 or 2 trauma centres (as defined by Hameed et al.,
]). We chose these facilities as by definition they are required to provide 24/7 in-house
access to a wide range of services including CT scanners, ICUs and extensive access to
speciality care and can thus constitute a potential ªGold Standardº of urban care. We identified these
2 / 11
centres using the Canadian Institute for Health Information (CIHI) database. The CIHI, is an
independent, not-for-profit organization created by the federal, provincial and territorial
governments that collects and analyses information on health and health care in Canada and
makes it publicly available. Stroke data has been reported since 2007. However, Quebec did
not contribute to the CIHI database for the study period, due partly to differences in data
collection methods for 30-day in-hospital stroke mortality rates. Data from the territories
(Northwest Territories, the Yukon and Nunavut) were not included either as they did not meet the
hospital selection criteria of our previous studies described above. Furthermore, we did not
include urban community hospitals as the level of services and speciality care is difficult to
ascertain. Moreover, several urban community hospitals are near the urban academic centers
where patients are likely to be transferred to stroke centers.
Stroke mortality data
We used the 30-day stroke in-hospital mortality data from the CIHI Health Indicators .
This risk-adjusted rate includes all-cause in-hospital deaths occurring within 30 days of first
admission to an acute care hospital with a diagnosis of stroke . This indicator is
determined by the ratio of the number of stroke deaths in the hospital in a year to the expected
number of stroke deaths in the same year. The expected number of stroke deaths is estimated
from a logistic regression model that is fitted with age, sex, type of stroke and selected
preadmission comorbid diagnoses. This ratio is then multiplied by the Canadian average rate of
stroke death to express a rate of in-hospital deaths per 100 first stroke admission episodes. As
risk-adjustment modelling cannot entirely eliminate differences in patient characteristics
among all hospitals in that year, the interpretation of each in-hospital rate has to be compared
with the Canadian average for that year . Data from every Canadian hospital was published
on the CIHI website interactive tool at the time of conducting this study, unless they were
considered low-volume rates (unstable stable data). Rates with denominators between 0 and 4 are
considered low-volume. Details about how this is calculated are presented in S1 Table.
Descriptive statistics such as frequency, percentage, median and interquartile range were used
to characterise emergency services in rural and urban hospitals. To study the variation in
emergency services in Canadian rural hospitals, we considered: 1) access to local ICUs; 2) 24/7
in-hospital access to laboratory services, basic X-rays, CT scanners and magnetic resonance
imaging (MRI); and 3) number of annual visits to the ED. As 30-day in-hospital stroke
mortality rates were not available for each year for every hospital, we compared characteristics of
rural hospitals with CIHI data to those without for each year using chi-square or Fisher's exact
test for the presence of medical equipment and Wilcoxon signed-rank test for the number of
annual emergency visits. Mean and 95% confidence interval (CI) were calculated for the
30-day in-hospital stroke mortality rate for urban and rural hospital separately. Risk-adjusted
rates with confidence intervals that do not contain the Canadian average can be considered
statistically different from the Canadian average. All analyses were performed with SAS 9.4
(SAS Institute, Inc., Cary, North Carolina, USA).
Characteristics of participating hospitals
Our eligibility criteria were met by 286 rural and 24 urban hospitals. Of the 24 urban hospitals, 23 (96%) provided complete data for 30-day in-hospital stroke mortality rates between 2007
3 / 11
Fig 1. Flowchart of hospitals with CIHI available data.
and 2011. Of the 286 rural hospitals, 185 (65%) provided complete or partial information on
stroke mortality rates between 2007 and 2011. Consequently, the number of rural hospitals for
each year varied between 118 to 138, Fig 1. Rural hospitals received a total of 3 451 885 ED
visits and urban hospitals received a total of 1 549 730 visits. In terms of services offered, every
urban hospital had the five services (ICU, laboratory, radiology, CT-scanner and MRI)
available 24/7. For the rural hospitals, 21% had in-hospital ICUs, 94% had laboratory, 92% had
radiology, 11% had CT-scanners and 1% had MRI services.
Because of the potential for reported bias in the rural results, we compared the characteris
tics of rural hospitals for which CIHI data was available (rural reporting) to those without data
(non-reporting). Non-reporting rural hospitals had significantly lower-volume EDs and less
access to CT scans and ICU. They also had less access to basic radiology and laboratory
services (Table 1).
Thirty-day in-hospital mortality following stroke
1 One missing data on characteristic data
2 Two missing data on characteristic data
3 Three missing data on characteristic data
4 Four missing data on characteristic data
5 2007: 138 hospital with CIHI available data; 2008: 131 hospital with CIHI available data; 2009: 134 hospital with CIHI available data; 2010: 118 hospital with CIHI
available data; 2011: 119 hospital with CIHI available data
6 Chi-square or fisher exact test for count data and Wilcoxon test for continuous data
5 / 11
1 Difference between the rural or urban average and Canada average
following stroke were higher in rural than in urban hospitals and higher than the Canadian
average for every year except 2008 (rural average range = 18.26 to 21.04 and urban average
range = 14.11 to 16.78). For all years, the Canadian average was the same as the urban hospital
This is, to the best of our knowledge, the first study to describe variation in Canadian rural and
urban academic in-hospital post-stroke mortality rates. We found that 30-day in-hospital
mortality rates following stroke were higher in rural than in urban hospitals and higher than the
Canadian average for every year since data collection began except for 2008. We also observed
that fewer rural hospitals had in-hospital access to CT scanners, ICUs than their urban
counterparts. Our findings lead us to make the following observations.
These results suggest that there may be an association between the higher stroke mortality
rate in rural hospitals and their lack of stroke- pertinent emergency services. In the 2015
update of the Canadian Stroke Best Practice Recommendations, the hyper-acute stroke care
guidelines recommend early identification and treatment of suspected stroke patients eligible
for acute treatments (tPa or newer endovascular therapies) with ªimmediateº CT imaging
]. The lack of an in-hospital CT scanner in 89% of rural hospitals would hinder physicians
from timely diagnosis of or administration of acute treatments when indicated. As previously
reported, at least 40% of these rural hospitals are more than 300 kms from an academic center
making direct bypass to a stroke center unlikely1 Other variables that we did not measure,
such as the presence or absence of an organized stroke team with a dedicated training
program, inter-facility transfer capability, support from academic centres and even rural patient
characteristics and geographic factors (e.g., distance from ED) could also have contributed to
this mortality difference, as has been reported elsewhere [
While we cannot yet confirm the precise nature of the association between higher rural stroke mortality rates and lack of services in rural EDs, our findings are cause for concern because rural hospitals receive more than 3 million annual ED visits and stroke is a significant medical emergency. While this is the first Canadian study to report stroke data in rural
6 / 11
hospitals, our results are congruent with those of Lichtman et al.,[
] who found rural US
hospitals (Critical Access) had higher 30-day risk-standardised mortality following stroke than
mainly urban (non-Critical Access) hospitals. ªCritical Accessº is a designation given to certain
rural hospitals by the Centres for Medicare and Medicaid Services in the U.S. whose goal is to
maintain essential services in rural communities [
]. A recent study comparing rural to
urban citizens in Ontario, Canada's largest province, found a similar trend of higher 30-day
stroke mortality among rural than urban residents (adjusted hazard ratio 1.14; 95% confidence
interval 0.99±1.32) between 2008 and March 31, 2011 [
]. The study also found that rural
patients were less likely than urban patients to receive stroke unit care, brain imaging within
24 hours, carotid imaging, and consultations from neurologists, physiotherapists, occupational
therapists and speech language pathologists, and were less likely to be transferred to inpatient
rehabilitation facilities. Other population studies in the U.S. and China have shown that stroke
mortality is not only higher in rural areas but that this mortality gap is increasing [
advanced stroke care is improving faster in urban than in rural centres. As universal
accessibility is a cardinal feature of Canada's health care system, the stroke care discrepancies that our
study has uncovered need to be better understood and rectified.
Is the United States' rural hospital system a model to follow?
Our study results suggest that the time has come to reduce inequities in rural health care in
Canada. Looking south of the border may be a logical first step. The U.S. has passed legislation
improving access to medical care and imaging services in small rural hospitals through its
Critical Access Hospital (CAH) designation in 1997 [
]. An acute care general hospital can be
designated a CAH by Medicare if it (1) is located in a rural area with a State Medicare Rural
Hospital Flexibility Program, (2) provides 24-hour emergency care services using on-site or
on-call staff members, (3) has no more than 25 inpatient beds, (4) has an average annual length
of stay of 96 hours or less, and (5) is located more than 35 miles from the nearest hospital (15
miles in mountainous terrain) or is state certified as a necessary provider. Evidence suggests
that CAHs have improved rural access to medical care, including to imaging services [
recent report on critical access hospitals showed that 95% of CAH have a CT scanner and 83%
can perform MRIs [
]. These critical care access hospitals are in reality very similar to
Canada's rural hospitals, and Canada's discrepancy in access to care is thus particularly
reprehensible given its claim to a universal-access health care system. Nevertheless, as mentioned earlier,
rural versus urban stroke mortality gaps are also present in the U.S [
Future interventions for improving stroke care in rural hospitals in Canada
Multiple stakeholders are currently addressing rural versus urban discrepancies in access to
care. In February 2017, the College of Family Physicians of Canada, the Society or Rural
Physicians of Canada in the presence of Canada's Federal Health Minister Jane Philpott presented a
strategic road map to improve access to care for rural citizens and indigenous peoples [
includes the development of specific resources, infrastructure, and networks of care within
local and regional health authorities to address access issues. It also proposes to develop
strategies to guide implementing system-wide, coordinated, distance technology to enhance and
expand local capacity, and improve access and quality health care in rural communities.
In line with this initiative, and for stroke care in particular, several interventions could contrib
ute to reducing the rural-urban stroke mortality discrepancies. Urgently, a closer look at
improving in-hospital access to CT scanners in rural hospitals that are distant from referral centers would
be a first step. Improved access to CT scanners would consequently link distant rural hospitals to
telestroke programs and improve upon current effective integrated stroke networks.
7 / 11
The CIHI data was available for only 65% of the hospitals we considered to be rural and for
96% of the urban academic hospitals. CIHI does not publicly report data on hospitals with
fewer than five stroke cases (hospitals are identified). Further, because rural hospitals treat low
volumes of stroke patients, more than 80% of reported rural rates were considered in the
potentially "unstable" data range (5< = denominator<50, or expected episode<1 with
numerator>0) according to CIHI's criteria. Results must thus be interpreted with caution.
A legitimate question is whether hospitals with missing data could have had lower mortality
rates than those whose data are reported, thus minimizing the rural-urban stroke mortality
discrepancies reported here. Further analysis suggests this is unlikely. Rural hospitals with
available stroke mortality data received significantly more ED visits and were more likely to
have a CT scanner and ICU than hospitals without available stroke data. Limited clinical
exposure to cases in smaller centres is generally linked to poorer rather than better outcomes.
Another limitation is that data on stroke mortality was not available for the province of
Quebec for the study period. According to CIHI, data collection methods and certain access
restrictions to Quebec government data preclude the CIHI from publishing data on its website.
This is unfortunate as Quebec is Canada`s second most populated province (8 million). Que
bec rural hospitals also have better access to CT scanners than other Canadian provinces (74%
vs. less than 10%) [
]. Further comparison of stroke in Quebec rural hospitals with the
rural hospitals in the rest of Canada will be of interest considering their differential access to
in-hospital services. CIHI data on stroke mortality in Quebec will be available in the future.
We also did not collect data on EDs in the Northwest Territories, Nunavut and Yukon as
they did not uniformly meet inclusion in our initial rural ED studies. CIHI 30-day in hospital
mortality following stroke data was also not uniformly available for our study period. The
territories have differently managed health care systems, are more distant from tertiary centers
and have a higher proportion of Indigenous peoples. Thus, potential variation in stroke
mortality rates is expected. Future studies on access to rural and remote emergency services should
include the territories.
Other potential limitations include the fact patient level characteristics, known to be predic
tive of stroke mortality, are not available in this database. CIHI data-base reports hospital-level
data versus patient±level-data for this indicator. Further multivariate analyses with
adjustments for other factors (not adjusted for by CIHI) such as presence of local
CT/MRI/ICU/consultants would require patient-level data and will be the focus of future studies. Moreover,
rural versus urban times from symptom onset and stroke treatment were not available in the
CIHI database as well. Future, prospective, patient±level studies that include a measure of time from estimated symptom onset and treatment are required to further elucidate the concerning rural versus mortality gaps observed in this study.
This study has several strengths. We used an established national health outcomes database
(CIHI) and presented stroke data for every year since they have been available. Rural
definitions and hospital selection were achieved using rigorous criteria and levels of services were
obtained directly from the hospitals themselves using brief structured interviews with hospital
managers (methodology described elsewhere) [
To the best of our knowledge, this is the first report on disparities in 30-day stroke mortality in rural vs urban Canadian academic hospitals. While the rural/urban gap in access to emergency care resources such as specialists and CT scanners is increasingly acknowledged, this is the first report on disparities in patient outcomes. This information is cause for concern in the
8 / 11
context of Canada's universal health care system. Future studies should address the underlying
causes of such disparities.
S1 Table. This table describe the risk-adjusted rate of all-cause in-hospital death occurring
within 30 days of first discharge from an acute care hospital with a diagnosis of stroke. It is
available at: http://indicatorlibrary.cihi.ca/display/HSPIL/30-Day+Stroke+In-Hospital
We wish to thank the CIHI staff for their help to have access to the data. We also wish to thank
Luc Lapointe and Nicholas Bergeron for work in verifying references and reviewing the final version of the manuscript. Finally, we also thank Ms. Louisa Blair for English editing assistance.
Conceptualization: Richard Fleet.
Data curation: Sylvain Bussières, Gilles Dupuis.
Formal analysis: Fatoumata Korika Tounkara, SteÂphane Turcotte.
Investigation: Richard Fleet.
Methodology: Richard Fleet, Fatoumata Korika Tounkara, SteÂphane Turcotte, Gilles Dupuis.
Project administration: Sylvain Bussières.
Software: Sylvain Bussières.
Supervision: Richard Fleet.
Validation: Richard Fleet.
Visualization: France LeÂgareÂ, Jeff Plant, Julien Poitras, Patrick M. Archambault, Gilles
Writing ± original draft: Richard Fleet.
Writing ± review & editing: Fatoumata Korika Tounkara.
9 / 11
Minister of Justice. Canada Health Act. In. Ottawa: Government of Canada; 1985.
Canada S. Definitions of rural. Rural and Small Town Canada Analysis Bulletin. 2001; 3(3):1±17. 2011.
16. Information IfH. 30-Day Stroke In-Hospital Mortality.
10 / 11
1. Fleet R , Poitras J , Maltais-Giguere J , Villa J , Archambault P. A descriptive study of access to services in a random sample of Canadian rural emergency departments . BMJ Open . 2013 ; 3 ( 11 ): e003876 . Epub 2013 /11/29. https://doi.org/10.1136/bmjopen-2013- 003876 [pii]. PMID: 24285633; PubMed Central PMCID : PMC3845037 .
2. Statistics Canada . Population, urban and rural, by province and territory 2011 [updated 2011- 02 -04; cited 2013 September 6th] . Available from: http://www.statcan.gc.ca/tables-tableaux/sum-som/l01/ cst01/demo62a-eng.htm.
3. Blackwell T. If you're in a rural area and in need of emergency care, you're better off in Quebec than Ontario or B .C.: study. Available at: http://news.nationalpost.com/news/canada/if -youre-in-a-rural-areaand-in-need-of-emergency-care-youd-better-off-in-quebec-than-ontario-or-b-c-study ( Assecced September 16 , 2015 ). National Post . 2015 .
4. Fleet R , Audette LD , Marcoux J , Villa J , Archambault P , Poitras J . Comparison of access to services in rural emergency departments in Quebec and British Columbia . CJEM . 2014 ; 16 ( 6 ): 437 ± 48 . PMID: 25358275 .
5. Fleet R , Pelletier C , Marcoux J , Maltais-Giguere J , Archambault P , Audette LD , et al. Differences in access to services in rural emergency departments of Quebec and Ontario . PLoS One . 2015 ; 10 ( 4 ): e0123746. https://doi.org/10.1371/journal.pone.0123746 PMID: 25874948; PubMed Central PMCID : PMCPMC4398492 .
6. Bayley M , Lindsay P , Hellings C , Woodbury E , Phillips S , Canadian Stroke S. Balancing evidence and opinion in stroke care: the 2008 best practice recommendations . CMAJ . 2008 ; 179 ( 12 ): 1247 ±9. https:// doi.org/10.1503/cmaj.081536 PMID: 19047599; PubMed Central PMCID : PMCPMC2585121 .
7. Coutts SB , Wein TH , Lindsay MP , Buck B , Cote R , Ellis P , et al. Canadian Stroke Best Practice Recommendations: secondary prevention of stroke guidelines , update 2014. Int J Stroke . 2015 ; 10 ( 3 ): 282 ± 91 . https://doi.org/10.1111/ijs.12439 PMID: 25535808 .
8. Krueger H , Lindsay P , Cote R , Kapral MK , Kaczorowski J , Hill MD . Cost avoidance associated with optimal stroke care in Canada . Stroke. 2012 ; 43 ( 8 ): 2198 ± 206 . https://doi.org/10.1161/STROKEAHA.111. 646091 PMID: 22627985 .
9. Paul CL , Ryan A , Rose S , Attia JR , Kerr E , Koller C , et al. How can we improve stroke thrombolysis rates? A review of health system factors and approaches associated with thrombolysis administration rates in acute stroke care . Implement Sci . 2016 ; 11 : 51 . https://doi.org/10.1186/s13012-016-0414-6 PMID: 27059183; PubMed Central PMCID : PMCPMC4825073 .
10. Rodrigues FB , Neves JB , Caldeira D , Ferro JM , Ferreira JJ , Costa J . Endovascular treatment versus medical care alone for ischaemic stroke: systematic review and meta-analysis . BMJ . 2016 ; 353 :i1754. https://doi.org/10.1136/bmj.i1754 PMID: 27091337; PubMed Central PMCID : PMCPMC4834754 .
11. OECD. Health at a Glance 2015 : OECD Indicators . Paris: 2015 .
14. Fleet R , Archambault P , Legare F , Chauny JM , Levesque JF , Ouimet M , et al. Portrait of rural emergency departments in Quebec and utilisation of the Quebec Emergency Department Management Guide: a study protocol . BMJ Open . 2013 ; 3 ( 4 ). Epub 2013 /05/02. https://doi.org/10.1136/bmjopen2013-002961 PMID: 23633423; PubMed Central PMCID : PMC3641429 .
15. Hameed SM , Schuurman N , Razek T , Boone D , Van Heest R , Taulu T , et al. Access to trauma systems in Canada . J Trauma . 2010 ; 69 ( 6 ): 1350 ±61; discussion 61. https://doi.org/10.1097/TA. 0b013e3181e751f7 PMID: 20838258 .
17. Casaubon LK , Boulanger JM , Blacquiere D , Boucher S , Brown K , Goddard T , et al. Canadian Stroke Best Practice Recommendations: Hyperacute Stroke Care Guidelines, Update 2015. Int J Stroke . 2015 ; 10 ( 6 ): 924 ± 40 . https://doi.org/10.1111/ijs.12551 PMID: 26148019 .
18. Koifman J , Hall R , Li S , Stamplecoski M , Fang J , Saltman AP , et al. The association between rural residence and stroke care and outcomes . J Neurol Sci . 2016 ; 363 : 16 ± 20 . https://doi.org/10.1016/j.jns. 2016 . 02 .019 PMID: 27000213 .
19. Lichtman JH , Leifheit-Limson EC , Jones SB , Wang Y , Goldstein LB . 30 - Day risk -standardized mortality and readmission rates after ischemic stroke in critical access hospitals . Stroke . 2012 ; 43 ( 10 ): 2741 ± 7 . Epub 2012/09/01. https://doi.org/10.1161/STROKEAHA.112.665646 PMID: 22935397; PubMed Central PMCID : PMC3547601 .
20. Joynt KE , Harris Y , Orav EJ , Jha AK . Quality of care and patient outcomes in critical access rural hospitals . JAMA . 2011 ; 306 ( 1 ): 45 ± 52 . https://doi.org/10.1001/jama. 2011 .902 PMID: 21730240; PubMed Central PMCID : PMC3337777 .
21. Howard G. Ancel Keys Lecture: Adventures (and misadventures) in understanding (and reducing) disparities in stroke mortality . Stroke . 2013 ; 44 ( 11 ): 3254 ±9. https://doi.org/10.1161/STROKEAHA.113. 002113 PMID: 24029634; PubMed Central PMCID : PMCPMC3878050 .
22. Zhang XH , Guan T , Mao J , Liu L. Disparity and its time trends in stroke mortality between urban and rural populations in China 1987 to 2001: changing patterns and their implications for public health policy . Stroke . 2007 ; 38 ( 12 ): 3139 ± 44 . https://doi.org/10.1161/STROKEAHA.107.494336 PMID: 17947594 .
23. Khaliq AA , Deyo D , Duszak R Jr. The Impact of Hospital Characteristics on the Availability of Radiology Services at Critical Access Hospitals. J Am Coll Radiol . 2015 ; 12 (12 Pt B): 1351 ±6. https://doi.org/10. 1016/j.jacr. 2015 . 09 .008 PMID: 26614879 .
24. Ginde AA , Foianini A , Renner DM , Valley M , Camargo CA Jr. Availability and quality of computed tomography and magnetic resonance imaging equipment in U.S. emergency departments . Acad Emerg Med . 2008 ; 15 ( 8 ): 780 ±3. https://doi.org/10.1111/j.1553- 2712 . 2008 . 00192 . x PMID : 18783491 .
25. Canada TCoFPo. The Rural Road Map for Action: Directions. Advancing Rural Family Medicine: The Canadian Collaborative Taskforce. Mississauga: The College of Family Physicians of Canada , 2017 .