ARTICLE IN PRESS

Patterns of Care and Temporal Trends in Ischemic Stroke

Management: A Brazilian Perspective

Monique Bueno Alves, RN, MsC,*, Gisele Sampaio Silva, MD, MPH, PhD,*,
Renata Carolina Acri Miranda, RN, Rodrigo Meireles Massaud, MD,
Andreia Maria Heins Vaccari, RN, Miguel Cendoroglo-Neto, MD, PhD,*, and
Solange Diccini, RN, PhD*

Background and Purpose: Demonstration of an improvement process of quality

indicators in stroke care is essential to obtain certification as a primary stroke center
(PSC). Our aim was to evaluate factors that influence temporal trends in quality
indicators of ischemic stroke (IS) in a Brazilian hospital. Methods: We evaluated
patients discharged with IS from a tertiary hospital from January 2009 to Decem-
ber 2013. Ten predefined performance measures selected by the Get With the
Guidelines-Stroke program were assessed. We also compared 5 quality indica-
tors available from a secondary community hospital for the first year of the series
to those found in the tertiary hospital. Results: We evaluated 551 patients at the
tertiary stroke center (median age 77.0 years [interquartile range 64.0-84.0]; 58.4%
were men). The quality indicators that improved with time were the use of cholesterol-
lowering therapy (P = .02) and stroke education (P = .04). The median composite
perfect care did not consistently improve throughout the period (P = .13). After a
multivariable adjustment, only thrombolytic treatment (odds ratio [OR] 2.06, P < .01),
dyslipidemia (OR 2.03, P < .01), and discharge in a Joint Commission Internatio-
nals (JCI) visit year (OR 1.8, P < .01) remained as predictors of a perfect care index
of 85% or higher. The quality indicators with worse performance (anticoagula-
tion for atrial fibrillation and cholesterol reduction) were similar in the tertiary
and secondary community hospitals. Conclusions: We found a significant im-
provement in some quality indicators across the years in a PSC located in Latin
America. The overall perfect care measure did not improve and was influenced
by being discharged in a JCI visit year, having dyslipidemia, and having under-
gone thrombolytic treatment. Key Words: Strokequality indicatorsstroke
careischemic strokestroke centeroutcomes.
2017 National Stroke Association. Published by Elsevier Inc. All rights reserved.

From the *Universidade Federal de So Paulo, So Paulo, Brazil;

and Programa Integrado de Neurologia and Instituto de Ensino e
Pesquisa, Hospital Israelita Albert Einstein, So Paulo, Brazil.
Received February 8, 2017; revision received April 16, 2017; accepted
May 5, 2017.
Address correspondence to Monique Bueno Alves, RN, MsC, 50
Ventura Ladalardo Street, Apt21B, So Paulo, So Paulo 05704-140,
Brazil. E-mail: monique_bueno@hotmail.com.
1052-3057/$ - see front matter
2017 National Stroke Association. Published by Elsevier Inc. All
rights reserved.
http://dx.doi.org/10.1016/j.jstrokecerebrovasdis.2017.05.008
Introduction
Several studies have demonstrated that better perfor-
mances are achieved in facilities with implemented quality
programs for stroke as well as for other diseases such as
heart failure and myocardial infarct. The potential con-
sequences of a higher adherence to quality indicators include
lower rates of readmission and mortality based on the
practice of evidence-based medicine.1,2 Quality programs,
such as the Get With the Guidelines (GWTG), monitor
and encourage adherence to quality indicators. To obtain

Journal of Stroke and Cerebrovascular Diseases, Vol. , No. (), 2017: pp 1

 ARTICLE IN PRESS
2 M.B. ALVES ET AL.
and maintain the certification as a primary stroke center
(PSC), for example, the demonstration of an improve-
ment process of quality indicators in stroke care is essential.3
Therefore, following temporal trends is an important part
of the certification process by which hospitals can eval-
uate and improve their performances.
Factors influencing adherence to performance mea-
sures for patients with acute stroke or transient ischemic
attack have been previously evaluated in the American pop-
ulation. However, such predictors have been scarcely studied
outside the United States.4,5 Brazil is a country with cul-
tural aspects and health-care practices that could influence
predictors of adherence to stroke protocols. Brazil is also
a country with very limited availability of stroke quality
programs, with only 3 stroke centers certified by the Joint
Commission International (JCI) as PSCs and 1 center cer-
tified by the Canadian Stroke Network.4 Knowledge on
the predictors of adherence to performance measures in
Brazil could facilitate the implementation of stroke pro-
tocols worldwide. Our aim was to evaluate factors that
influence temporal trends in indicators of health-care quality
performance of ischemic stroke (IS) in Brazil using a com-
posite measure based on the GWTG program for IS in a
Brazilian tertiary hospital.
Methods
We evaluated a database of consecutive patients with
IS discharged from a private tertiary hospital in So Paulo
from January 2009 to December 2013. The hospital was
certified by the JCI as a PSC since 2007 and had 625 beds
at the time of the present study. We included patients aged
18 years or older with a confirmed diagnosis of IS admitted
by the emergency department. In-hospital strokes, transient
ischemic attacks, and transferred patients were excluded.
Data collected included demographics, stroke risk factors,
National Institutes of Health Stroke Scale (NIHSS) score
at admission, in-hospital treatment, modified Rankin Scale
score at discharge, neuroimaging characteristics, and throm-
bolysis status using a standardized, structured questionnaire.
Risk factors were considered if noted on the patients chart
or if medications for known risk factors were used before
hospital admission or at discharge.6 As part of our hospital
stroke program, all patients admitted with a diagnosis of
IS were identified and followed up daily by a case manager
nurse who prospectively evaluated medical records. IS was
defined as an episode of neurological dysfunction caused
by focal cerebral infarction confirmed by neuroimaging.7
Ten predefined performance measures selected by the
GWTG stroke program as targets for stroke quality im-
provement were evaluated8-10:
1) Intravenous (IV) recombinant tissue plasmino-
gen activator (rtPA) in patients who arrive 2 hours
or less after symptom onset and who were treated
within 3 hours of symptom onset (IV rtPA 3
hours)
2) Antithrombotic medication (antiplatelet or anti-
coagulant) prescribed within 48 hours of admission
(early antithrombotics)
3) Deep venous thrombosis (DVT) prophylaxis (war-
farin, heparin, lowmolecular-weight heparin, other
anticoagulants, and pneumatic compression
devices) within 48 hours of admission in
nonambulatory patients at risk of DVT (DVT
prophylaxis)
4) Antithrombotic medication (antiplatelet or anti-
coagulant) prescribed at discharge (antithrombotics
at discharge)
5) Anticoagulation prescribed at discharge in pa-
tients with documented atrial fibrillation (AF) or
atrial flutter (anticoagulation for AF)
6) Lipid-lowering medication prescribed at dis-
charge if the low-density lipoprotein (LDL) is
higher than 100 mg/dL, if the patient was treated
with a lipid-lowering agent before admission, or
if the LDL was not documented (cholesterol
reduction)
7) Smoking cessation intervention (counseling or med-
ication) at discharge for current or recent smokers
(smoking cessation)
8) Door-to-computerized tomography (CT) read time
in patients presenting with stroke symptoms within
3 hours or less (door-to-CT time target = 45
minutes)
9) Door-to-needle time in patients who received IV
rtPA (door-to-needle time)
10) Stroke education provided to a patient or caregiver,
including 5 components: explanation of which are
the modifiable stroke risk factors, stroke warning
signs and symptoms, how to activate emergen-
cy medical services, need for follow-up, and
medications prescribed (stroke education)
Eligible patients were those admitted without any
medical contraindications documented as reasons for
nontreatment for each of the applicable measures. Dis-
charge and subacute measures excluded patients who
died.3,4 All quality indicators not documented in the pa-
tients chart were scored as not done; therefore, we had
no missing data. Data for IV rtPA of 3 hours or less, early
antithrombotics, DVT prophylaxis, antithrombotics at dis-
charge, anticoagulation therapy at discharge in patients
with documented AF or atrial flutter, lipid-lowering med-
ication prescribed at discharge if the LDL was higher than
100 mg/dL, and door-to-needle time in patients who re-
ceived IV rtPA were extracted from an electronic medical
prescription system. Door-to-CT times were extracted di-
rectly from the electronic medical record and the CT
scanner report. When stroke education and smoking ces-
sation interventions were not documented at the patients
chart, the performance indicators were scored as not per-
formed. The percentage of adherence to these measures
was obtained for each individual patient through an
 ARTICLE IN PRESS
PATTERNS OF CARE IN ISCHEMIC STROKE: A BRAZILIAN PERSPECTIVE
adherenceeligibility index calculated by the ratio between
eligible measures and compliance to them in every patient.
Of note, until 2012, the recommended time window for
IV thrombolysis in Brazil was 3 hours, but besides that
fact, the private hospital adopted the 4.5-hour therapeu-
tic window since 2010 based on the updated statement
from the American Stroke Association supported by the
European Cooperative Acute Stroke Study (ECASS III)
study.11
The quality indicators early antithrombotics, antico-
agulation for AF, DVT prophylaxis, cholesterol reduction,
and antithrombotics at discharge from the first year of
the series were compared to the quality indicators col-
lected at a secondary community public hospital located
in So Paulo using the same methodology. The hospi-
tals institutional review board approved the present study.
Statistical Analysis
Means and standard deviations or medians and
interquartile ranges were used to describe the patients
characteristics. The appropriateness of nonparametric testing
was tested after checking the normality diagnostics
(KolmogorovSmirnov tests and histograms). One-way
analysis of variance and KruskalWallis tests were used
for comparison of continuous parametric and nonpara-
metric variables, respectively, among the different time
periods. Categorical variables were compared with the
chi-square and chi-square trend tests. The composite perfect
care was calculated, giving 10 points for each measure
achieved divided by the total of eligible measures mul-
tiplied by 10 (0 was the lowest score and 100 was the
highest score). Multiple logistic regression was used to
investigate the influence of epidemiological and clinical
data upon the perfect care index dichotomized in 85%
of the adherence (the GWTG program recognizes hos-
pitals with awards if they follow treatment guidelines in
key measures greater than 85% during the measure-
ment period). All variables that showed an association
Figure 1. Flowchart of included patients. Abbre-
viations: ICH, intracerebral hemorrhage; SAH,
subarachnoid hemorrhage; TIA, transient isch-
emic attack.
3
with achieving a perfect care index greater than 85% in
the univariate analysis with a P value greater than .10
were included in the logistic regression. A 2-tailed P value
of .05 was considered statistically significant. Statistical
analysis was performed with SPSS 16.0 software (SPSS
Inc., Chicago, IL).
Results
In the study period, 1417 patients were admitted to
our stroke center. A total of 25.6% patients had had a
transient ischemic attack, and 16.0% had hemorrhagic
strokes, including nontraumatic subarachnoid hemor-
rhages, and were therefore excluded from the present study
(Fig 1). We also excluded from the analyses 276 pa-
tients with IS: 66.2% were transferred-in or transferred-
out patients, 32.6% had in-hospital strokes, and 1.0% were
younger than 18 years old. The final analysis sample con-
sisted of 551 consecutive patients with IS admitted to our
hospital. The median age was 77.0 years (interquartile
range 64.0-84.0]), and 58.4% were men. The median time
from symptom onset to hospital admission was 345.0
minutes (104.5-1417.5). Demographics and clinical char-
acteristics presented at admission are shown in Table 1.
Good outcome at discharge, defined as an modified
Rankin Scale score of 3 or lower, was present in 67.6%
of the patients. The median length of stay was 8.0 days
(4.0-14.0). A total of 42.5% of the patients were discharged
in a JCI visit year (2010 and 2013), and 7.9% of the patients
died (Table 1). Table 1 compares the stroke risk factors
according to the year of discharge. Hypertension was the
most common risk factor (60.3%) followed by diabetes
(31.2%) and dyslipidemia (30.1%). The only significant
difference in risk factors among years was having a pre-
vious stroke or TIA, more frequent in patients discharged
in 2010 (39.0% of the patients, P = .02).
A total of 80 patients were treated with recanalization
therapies (17.1%). The most frequently used technique in
57.5% of the patients was IV rtPA. Intra-arterial therapy
 4
Table 1. Clinical and demographic characteristics according to the year of discharge

2009 2010 2011 2012 2013 All

ARTICLE IN PRESS
Variable (n = 83) (n = 105) (n = 115) (n = 119) (n = 128) (n = 551) P Value

Age 77.0 (64.0-83.0) 77.0 (66.0-84.50) 75.0 (63.0-82.0) 78.0 (63.0-84.0) 80.0 (62.0-87.0) 77.0 (64.0-84.0) .600
Gender (male) (%) 57.8 61.9 57.4 54.6 60.5 58.4 .824
Median time from symptom onset 530.0 (87.0-1500.0) 420.0 (85.0-1301.0) 205.0 (102.0-655.0) 390.0 (110.0-1440.0) 465.0 (115.5-1440.0) 345.0 (104.5-1417.5) .137
to hospital admission (min)
Mean NIHSS score at admission 16.8 6.2 15.5 7.1 13.3 7.7 12.4 7.0 14.3 8.5 14.2 7.5 .242
Hypertension (%) 51.8 65.7 66.1 53.8 62.0 60.3 .105
Dyslipidemia (%) 27.7 27.6 27.8 31.1 34.9 30.1 .686
Diabetes (%) 31.3 33.3 27.8 26.9 36.4 31.2 .477
Smoking (%) 20.5 11.4 16.5 10.9 9.3 13.2 .113
Alcohol abuse (%) 10.8 5.7 1.7 .2 1.6 4.0 .004
CAD (%) 10.8 12.4 9.6 21.8 14.7 14.2 .065
Atrial fibrillation (%) 22.9 13.3 13.0 16.8 15.6 16.0 .350
Previous stroke or TIA (%) 25.3 39.0 21.7 28.6 22.5 27.2 .028
Length of stay (d) 7.0 (3.0-13.0) 9.0 (5.0-15.0) 7.0 (3.0-14.0) 8.0 (5.0-14.0) 7.0 (4.0-15.5) 8.0 (4.0-14.0) .376
In-hospital mortality (%) 6.0 10.4 13.0 3.3 9.3 8.5 <.001

Abbreviations: CAD, coronary artery disease; NIHSS, National Institutes of Health Stroke Scale; TIA, transient ischemic attack.

M.B. ALVES ET AL.

 ARTICLE IN PRESS
PATTERNS OF CARE IN ISCHEMIC STROKE: A BRAZILIAN PERSPECTIVE 5
Table 2. Recanalization therapies according to the year of discharge

2009 2010 2011 2012 2013 All P

Treatment with reperfusion therapy (%) 12.0 16.2 17.4 13.4 13.5 17.1 .806
IV rtPA (%) 70.0 64.7 65.0 43.8 47.1 57.5 .513
IA therapy (%) 10.0 29.4 35.0 50.0 29.4 33.8
Combined therapy (%) 20.0 5.9 .0 6.3 23.5 8.8

Abbreviations: IA, intra-arterial; IV, intravenous; rtPA, recombinant tissue plasminogen activator.
Combined therapy includes an IV therapy followed by an IA procedure.

Table 3. Adherence to Performance Measures according to the year of discharge

Patterns of care 2009 2010 2011 2012 2013 All P

IV rtPA (3 h) (%) 85.7 100.0 77.8 100.0 91.7 91.3 .397

Early antithrombotics (%) 98.7 100.0 99.1 99.1 100.0 99.4 .673
DVT prophylaxis (%) 100.0 100.0 100.0 100.0 100.0 100.0 ns
Antithrombotics at discharge (%) 90.0 100.0 100.0 95.5 99.1 97.4 <.001
Anticoagulation for AF (%) 66.7 84.6 80.0 73.7 62.5 72.6 .678
Cholesterol reduction* (%) 69.2 78.7 69.0 70.4 88.0 75.6 .003
Smoking cessation (%) 71.4 100.0 70.0 92.3 100.0 84.0 .021
Door-to-CT time (45 min) (%) 89.7 70.7 67.3 67.5 66.7 71.1 .210
Mean door-to-CT (min) 33.3 22.9 45.8 37.8 47.8 41.2 41.6 38.2 45.2 23.4 43.6 34.7 .450
Door-to-needle time (60 min) 33.3 75.0 61.5 50.0 66.7 58.5 .226
Mean door-to-needle time (min) 82.5 28.4 49.0 20.0 59.3 26.1 64.6 28.59 65.3 30.8 62.6 27.8 .118
Stroke education* (%) 73.1 85.1 78.0 74.8 90.6 80.8 .006
Perfect care (%) 83.0 18.9 90.7 11.7 85.6 17.4 85.8 17.7 92.6 11.4 87.8 15.8 <.001

Abbreviations: AF, atrial fibrillation; CT, computerized tomography; DVT, deep vein thrombosis; IV, intravenous; ns, not significant; rtPA,
recombinant tissue plasminogen activator.
*P < .05 in the chi-square trend test.

was performed in 33.8%, and combined therapy was 2.03, CI 1.36-3.02, P < .01), and discharge in a JCI year of
performed in 8.8% of the patients (Table 2). There was visit (OR 1.8, CI 1.29-2.65, P < .01) remained in the model
no difference among years in the following measures: IV as predictors of higher perfect care index (Table 4).
rtPA (3 hours), DVT prophylaxis, antithrombotics in the The quality indicator DVT prophylaxis was significantly
first 24 hours of admission, anticoagulation therapy for better at the tertiary hospital when compared with the
AF, door-to-CT read time, and door-to-needle time. Table 3
shows the performance measures according to the discharge
year. Table 4. Factors associated with compliance to the adherence
The quality indicators that improved along the years index on a multivariate logistic regression model
were cholesterol-lowering therapy (range 69.2%-88.0%,
P = .02) and stroke education (range 73.1%-90.6%, P = .04) 95 % CI
for exp (B)
(Table 3). The median composite perfect care was different
(range 83.0 18.9 to 92.6 11.4, P < .001) but did not con- Exp Lower Upper
sistently improve along the years (P = .13 in the chi-square Variable Sig. (B) CI CI
trend test). The highest scores were obtained in 2010 and
2013 (Table 3). Age .297 1.006 .995 1.017
In the univariate analysis, being discharged in a JCI Female sex .049 .703 .495 .999
visit year (odds ratio [OR] 1.84, confidence interval [CI] Thrombolytic treatment <0.01 2.06 1.21 3.51
Discharged in a joint <0.01 1.834 1.285 2.617
1.29-2.61, P < .01), female gender (OR .68, CI .48-.96, P = .03),
commissions year
dyslipidemia (OR 1.95, CI 1.33-2.88, P < .01), and undergoing
Dyslipidemia <0.01 2.03 1.36 3.02
thrombolytic treatment (OR 1.81, CI 1.08-3.02, P = .02) were Length of stay .996 1.000 .896 1.004
found to be associated with a perfect care index of 85% or Constant .514 .753
higher. After the multivariable adjustment, only thrombolytic
treatment (OR 2.06, CI 1.21-3.51, P < .01), dyslipidemia (OR Abbreviations: CI, confidence interval; sig., significance.
 ARTICLE IN PRESS
6 M.B. ALVES ET AL.
community hospital (100% versus 84.6%, P < .01). Con-
versely, antithrombotics at discharge were more frequently
prescribed at the community hospital (100% versus 90%,
P < .01). The quality indicators with worse performance (an-
ticoagulation for AF and cholesterol reduction) were similar
in the tertiary and secondary community hospitals (Fig 2).
Discussion
We found a significant improvement across the years
in 2 quality indicators (cholesterol reduction and stroke
education) in a PSC located in Brazil. The overall perfect
care measure did not consistently improve over time and
was influenced by being discharged in a JCI visit year,
having dyslipidemia, and undergoing thrombolytic treat-
ment. The GWTG-Stroke program has improved the quality
of stroke care in U.S. academic and community hospi-
tals, with important implications for the entire country.
The generalizability of GWTG-Stroke across distinct na-
tional and economic realities remains unanswered.8,10
Our hospital was the first institution outside the United
States to be accredited by the JCI in 1999, when the in-
ternational program was launched. In June 2007, Albert
Einstein Hospital became the first institution in Latin
America and one of only three worldwide to have a PSP
certified by JCI.4 Therefore, it is expected that the find-
ings from our study likely reflect better performance in
acute stroke treatment than the overall national perfor-
mance data.12 Actually, data evaluating quality of care
indicators in Brazilian stroke centers are largely unavail-
able. A previous study assessed the quality of hospital
care for acute IS in the Brazilian Unified National Health
System, but none of the quality indicators suggested by
the GWTG stroke program were reported. Another study
evaluating patients at the northeast region of the country
reported a mean door-to-CT time of 75.6 minutes, but
data on other quality indicators were also not available.13,14
In our series, the presence of certain risk factors might
have triggered the clinician to be more attentive to the treat-
ment measures, it could be observed in patients with
dyslipidemia with better perfect care. Likely for similar
reasons, thrombolysis treatment was also an independent
Figure 2. Comparison of 5 quality indicators
between a tertiary private hospital and a public sec-
ondary community hospital for the year of 2009.
The dashed line indicates the 85% goal suggested
by the Get With the Guidelines-Stroke program. Ab-
breviations: AF, atrial fibrillation; DVT, deep venous
thrombosis.
predictor of a higher adherenceeligibility index. Physi-
cians might be more concerned with patients who could
achieve more favorable functional outcomes and there-
fore could be more compliant with treatment measures.
Finally, the finding of a higher perfect care index in the
years of JCI visits points out to the importance of a more
continuous evaluation of the quality indicators by the ac-
creditation process using methods like unannounced surveys
or periodic data reports. Audit and feedback, as used in
the JCI evaluation process, is a common approach to
promote the implementation of evidence-based practices.
Overall, one quarter of the audit and feedback interven-
tions have a large positive effect on quality of care. The
identification of factors that distinguish more and less suc-
cessful interventions, including the frequency of visits and
the adaptation of the quality evaluation process to each
country reality, constitutes a real challenge.15,16
Our higher adherence mean rates were for IV rtPA use,
early antithrombotics, DVT prophylaxis, and antithrombotics
at discharge across the years. Smoking cessation counsel-
ing, anticoagulation at discharge for patients with AF, and
cholesterol reduction had lower rates above the target of
85%. Interestingly, in a GWTG cohort, these 3 measures
also had the lowest compliance and major improvement
across the period of 4 years evaluated.8
In the United States, being treated in a PSC increases
the chance of undergoing IV thrombolysis.17,18 The rates
of thrombolysis treatment in Brazilian studies vary from
1.1% to 11.2%.4,14,19,20 A study in Massachusetts concluded
that the PSC designation program improved the throm-
bolysis rates in early admissions.21 Our overall rate of use
of reperfusion therapies was 17.1%.
Early admissions were not related to higher perfect care
scores, and their frequencies did not change across the study
period. At least 25% of our patients arrived in a therapeutic
time window for IV rtPA. The finding of a stable propor-
tion of patients arriving in a therapeutic time window was
also reported in a previous international study from the
GWTG Stroke Program.22 Albert Einstein Hospital is a
private institution in So Paulo, a city with 11.3 million
inhabitants. Around 1.5 million people have potential access
to our hospital. It is possible that the high socioeconomic
 ARTICLE IN PRESS
PATTERNS OF CARE IN ISCHEMIC STROKE: A BRAZILIAN PERSPECTIVE
and educational levels of the patients who have access to
Albert Einstein explain at least in part the high frequen-
cy of early arrivals observed in our series.
We found an interesting difference in in-hospital mor-
tality between the 5 years (ranging from 3.3% in 2012
to 13% in 2011). The overall mortality rate for the 5 years
was 8.5%, which was similar to other national and in-
ternational stroke series. We could not find a good reason
to explain such differences, except for the slightly low
NIHSS scores in 2012. It is possible that differences in
stroke mortality could be attributable to some unmea-
sured factors such as previous modified stroke scale
differences and clinical complications such as pneumonia.
Brazil is a country of great social inequalities; there-
fore, it is not simple to apply international data to our
reality. Fortunately in the last decade, Brazil had a great
evolution in stroke care. From 2008 to 2012, the number
of stroke centers increased from 35 to 82. In 2012, the
Brazilian national stroke policy was published.23 The policy
includes definitions for stroke centers by level, reim-
bursement values, telemedicine, rehabilitation and training
funding, population awareness, and establishment of a
stroke care program integrating available resources from
other health programs. A National Stroke Registry was
initiated linked to the Ministry of Health to improve quality,
and hospitals were invited to participate. Data from this
registry are not yet published.24 In 2015, the Brazilian
Academy of Neurology launched an important survey
evaluating the medical perception of stroke treatment in
the country. In a structured questionnaire, several dif-
ferences from the public and private services were
highlighted, with a focus on triage, infrastructure, and
human resources. The survey is still ongoing and its results
will be of utmost importance for understanding stroke
gaps in the country.25
In the United States, a study comparing PSC with non-
PSCs found that all quality measures in stroke care were
significantly higher in certified centers.12 In our series, in
the first year reported (2009), we had already estab-
lished the improvement plan needed for the certification;
therefore, we do not have a real-life baseline performance
to compare to evaluate the real progress of our program.
Although there are several cultural differences between
Brazil and the United States, having a PSC certification
seems to be important in quality indicators in stroke care
improvement. However, certification or award recognition
programs alone may be insufficient to improve quality. They
should be accompanied by critical analysis, accountabili-
ty, and frequent reports of the measures.10,12 The evolution
of the quality measures across the years in our study further
reinforces the importance of a quality program to deliver
evidence-based and guideline-directed stroke care. The JCI
program is effective to evaluate protocols, structure, process,
and quality improvement but does not have a target for
the measures for the certification. Studies found that any
quality improvement program leads to higher adherence
7
to guidelines-based care and should result in a signifi-
cant improvement in patient outcomes.10,12,26 Quality
programs facilitate the implementation of improvement
efforts, certainly influencing the time-dependent process
such as neuroimaging, laboratory analysis, and in-
hospital transportation, and could possibly decrease time-
dependent measures.12
The present study has several limitations. First, we report
data from a single center, with predominantly white pa-
tients and no concurrent controls; therefore, our findings
might not be generalizable to all Brazilian stroke centers.
Despite this limitation, our comparison with the quality
indicators available for the first year of the study from a
secondary public community hospital suggests that there
are similarities between the public and the private systems,
and that indeed the poor performances on some specific
quality indicators might reflect cultural aspects of adher-
ence to protocols that might indeed be generalizable for
the whole country. Second, data were prospectively col-
lected in a period when the certification was already
achieved; therefore, trends in improvement could not be
observed in all quality measures as in 2009 several actions
have already been implemented. Moreover, our final sample
size was relatively small and probably not sufficiently
powered to detect what could still be considered a clin-
ically meaningful change over time, particularly for indicators
with low eligibility numbers. However, at least for the perfect
care score, our main outcome measure, we had a power
of 98% to detect an annual change of 5%. Finally, data from
patients with TIA were excluded from the present anal-
ysis mostly because benchmarks from the literature regarding
quality indicators in the care for such patients are largely
lacking. This finding could be explained by the fact that
TIA patients have a shorter length of stay at hospitals or
can be treated in outpatient clinics, such as TIA clinics.
In conclusion, the predictors of adherence to stroke care
protocols in a Brazilian hospital appear to be similar to
those described in international series, and some quality
indicators tend to improve over the years when moni-
tored through a quality control program. If the existence
of such protocols has proven important, their implemen-
tation across the country would be of utmost importance.
Whether our results could be further reproduced by the
design of a national stroke quality improvement program
addressing the specific national challenges for Brazil
remains to be evaluated.
References
1. LaBresh KA, Fonarow GC, Smith SC, et al. Improved
treatment of hospitalized coronary artery disease patients
with the Get With the Guidelines program. Crit Pathw
Cardiol 2007;6:98-105.
2. DeVore AD, Cox M, Heidenreich PA, et al. Cluster-
randomized trial of personalized site performance
feedback in Get With the Guidelines-heart failure. Circ
Cardiovasc Qual Outcomes 2015;8:421-427.
 ARTICLE IN PRESS
8 M.B. ALVES ET AL.
3. Alberts MJ, Latchaw RE, Jagoda A, et al. Revised and
updated recommendations for the establishment of
primary stroke centers: a summary statement from the
brain attack coalition. Stroke 2011;42:2651-2665.
4. de Carvalho FA, Schwamm LH, Kuster GW, et al. Get
With the Guidelines Stroke performance indicators in a
Brazilian tertiary hospital. Cerebrovasc Dis Extra
2012;2:26-35.
5. Hsieh FI, Lien LM, Chen ST, et al. Get With the
Guidelines-Stroke performance indicators: surveillance of
stroke care in the Taiwan stroke registry: Get With the
Guidelines-Stroke in Taiwan. Circulation 2010;122:1116-
1123.
6. van Swieten JC, Koudstaal PJ, Visser MC, et al.
Interobserver agreement for the assessment of handicap
in stroke patients. Stroke 1988;19:604-607.
7. Sacco RL, Kasner SE, Broderick JP, et al. An updated
definition of stroke for the 21st century: a statement for
healthcare professionals from the American Heart
Association/American Stroke Association. Stroke 2013;
44:2064-2089.
8. Schwamm LH, Fonarow GC, Reeves MJ, et al. Get With
the Guidelines-Stroke is associated with sustained
improvement in care for patients hospitalized with acute
stroke or transient ischemic attack. Circulation 2009;
119:107-115.
9. LaBresh KA, Reeves MJ, Frankel MR, et al. Hospital
treatment of patients with ischemic stroke or transient
ischemic attack using the Get With the Guidelines
program. Arch Intern Med 2008;168:411-417.
10. Fonarow GC, Reeves MJ, Smith EE, et al. Characteristics,
performance measures, and in-hospital outcomes of the
first one million stroke and transient ischemic attack
admissions in Get With the Guidelines-Stroke. Circ
Cardiovasc Qual Outcomes 2010;3:291-302.
11. Martins SC, Freitas GR, Pontes-Neto OM, et al. Guidelines
for acute ischemic stroke treatment: part II: stroke
treatment. Arq Neuropsiquiatr 2012;70:885-893.
12. Fonarow GC, Liang L, Smith EE, et al. Comparison of
performance achievement award recognition with primary
stroke center certification for acute ischemic stroke care.
J Am Heart Assoc 2013;2:e000451.
13. Rolim CL, Martins M. Quality of care for ischemic stroke
in the Brazilian Unified National Health System. Cad
Saude Publica 2011;27:2106-2116.
14. de Carvalho JJ, Alves MB, Viana GA, et al. Stroke
epidemiology, patterns of management, and outcomes
in Fortaleza, Brazil: a hospital-based multicenter
prospective study. Stroke 2011;42:3341-3346.
15. Ivers N, Jamtvedt G, Flottorp S, et al. Audit and feedback:
effects on professional practice and healthcare outcomes.
Cochrane Database Syst Rev 2012;(6):CD000259.
16. Ivers NM, Sales A, Colquhoun H, et al. No more
business as usual with audit and feedback interventions:
towards an agenda for a reinvigorated intervention.
Implement Sci 2014;9:14.
17. George BP, Asemota AO, Dorsey ER, et al. United States
trends in thrombolysis for older adults with acute ischemic
stroke. Clin Neurol Neurosurg 2015;139:16-23.
18. Schwamm LH. When in Rome, do like the Romans:
certifying stroke centers with the rod of aesculapius or
the medical caduceus of hermes? J Am Heart Assoc 2013;
2:e000120.
19. Cabral NL, Cougo-Pinto PT, Magalhaes PS, et al. Trends
of stroke incidence from 1995 to 2013 in Joinville, Brazil.
Neuroepidemiology 2016;46:273-281.
20. Moro CH, Goncalves AR, Longo AL, et al. Trends of the
incidence of ischemic stroke thrombolysis over seven years
and one-year outcome: a population-based study in
Joinville, Brazil. Cerebrovasc Dis Extra 2013;3:156-166.
21. Rost NS, Smith EE, Pervez MA, et al. Predictors of
increased intravenous tissue plasminogen activator use
among hospitals participating in the Massachusetts
Primary Stroke Service Program. Circ Cardiovasc Qual
Outcomes 2012;5:314-320.
22. Allen NB, Kaltenbach L, Goldstein LB, et al. Regional
variation in recommended treatments for ischemic stroke
and TIA: Get With the Guidelines-Stroke 2003-2010. Stroke
2012;43:1858-1864.
23. SAUDE MD. Protocolos clnicos e diretrizes teraputicas;
2012. Available at: http://bvsms.saude.gov.br/bvs/
saudelegis/gm/2012/PRT0664_12_04_2012.html. Accessed
April 5th, 2017.
24. Martins SC, Pontes-Neto OM, Alves CV, et al. Past,
present, and future of stroke in middle-income countries:
the Brazilian experience. Int J Stroke 2013;8(Suppl
A100):106-111.
25. Medicina CFd. Percepo Mdica sobre as Condies para
Atendimento do AVC. 2016.
26. Reeves MJ, Grau-Sepulveda MV, Fonarow GC, et al. Are
quality improvements in the Get With the Guidelines:
stroke program related to better care or better data
documentation? Circ Cardiovasc Qual Outcomes 2011;
4:503-511.