SEARCH

SEARCH BY CITATION

Keywords:

  • acute stroke therapy;
  • acute;
  • ischemic stroke;
  • stroke facilities;
  • stroke units;
  • treatment

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Background

Stroke mortality has been found to be much higher among residents in the stroke belt region than in the rest of United States, but it is not known whether differences exist in the quality of stroke care provided in Department of Veterans Affairs medical centers in states inside and outside this region.

Objective

We compared mortality and inpatient stroke care quality between Veterans Affairs medical centers inside and outside the stroke belt region.

Methods

Study patients were veterans hospitalized for ischemic stroke at 129 Veterans Affairs medical centers. Inpatient stroke care quality was assessed by 14 quality indicators. Multivariable logistic regression models were fit to examine differences in quality between facilities inside and outside the stroke belt, adjusting for patient characteristics and Veterans Affairs medical centers clustering effect.

Results

Among the 3909 patients, 28·1% received inpatient ischemic stroke care in 28 stroke belt Veterans Affairs medical centers, and 71·9% obtained care in 101 non-stroke belt Veterans Affairs medical centers. Patients cared for in stroke belt Veterans Affairs medical centers were more likely to be younger, Black, married, have a higher stroke severity, and less likely to be ambulatory pre-stroke. We found no statistically significant differences in short- and long-term post-admission mortality and inpatient care quality indicators between the patients cared for in stroke belt and non-stroke belt Veterans Affairs medical centers after risk adjustment.

Conclusions

These data suggest that a stroke belt does not exist within the Veterans Affairs health care system in terms of either post-admission mortality or inpatient care quality.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

A large body of research has demonstrated that stroke mortality is much higher among residents in the stroke belt region of the United States [1, 2]. The stroke belt region has traditionally been defined as including 11 states: Mississippi, Louisiana, Kentucky, Georgia, Tennessee, North Carolina, Alabama, South Carolina, Arkansas, Indiana, and Virginia [3]. Although higher death rates from stroke have been consistently observed in the stroke belt region for several decades [1, 2], there is little agreement as to its underlying cause or causes [4]. It is not clear whether stroke belt exists within the Department of Veterans Affairs (VA) health care system, how much of the quality of stroke care in medical facilities vary according to their location within or outside the stroke belt, and whether care quality variation explains some of the variance in mortality [4].

Several researchers have examined the quality of acute stroke care in individual stroke belt states such as North Carolina [5] and Georgia [6, 7]. In a recent study on quality of acute stroke care, Reeves et al [7] reported variations in eight performance indicators for acute stroke care between hospitals in Georgia and other non-stroke belt states (Massachusetts, Michigan, and Ohio). We found no literature reports that systematically compared inpatients stroke care performance between stroke belt and non-stroke belt states. The primary purpose of the present study is to compare post-admission mortality and inpatient stroke care quality between VA medical centers (VAMCs) inside and outside the 11-state stroke belt region.

Methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Material and patients

This is a retrospective, observational study. Research data were obtained from an ischemic stroke care database that was developed as part of the VA Office of Quality and Performance (OQP) Stroke Special Project [8]. The OQP database consists of quality indicators, sociodemographic and clinical information for a national sample of 5000 veterans who were hospitalized for ischemic stroke at any of the 131 VAMCs during fiscal year (FY) 2007 (October 1, 2006–September 30, 2007). These patients were initially identified from the FY 2007 VA Medical sas Patient Treatment File by using high-sensitivity International Classification of Diseases (ICD)-9 codes for ischemic stroke [9]. The sample included 100% of cases at small volume VAMCs (≤55 ischemic stroke admissions in FY 2007), and an random sample of 80% of cases at high-volume VAMCs (>55 ischemic stroke admissions in FY 2007). An extensive chart review was conducted by trained chart abstractors to verify the patients' ischemic stroke diagnosis and collect other clinical and process information during the patients' inpatient stay.

In this study, we excluded patients who did not have acute ischemic stroke as the primary diagnosis (n = 534), were already admitted for a non-stroke condition when the ischemic stroke event occurred (n = 200), were admitted only for post-stroke rehabilitation (n = 190), were admitted for elective carotid endarterectomy (n = 89), did not have an acute ischemic stroke ICD-9 codes (n = 22), and those patients who left against medical advice (n = 56). As a result, our final study sample consisted of 3909 patients at 129 VAMCs.

Outcome measures

Mortality

Mortality refers to all deaths that occurred within the initial 30 days and during the 12 months from the patients' stroke hospitalization admission date. Patient vital information was obtained from the Beneficiary Identification and Records Locator Subsystem (BIRLS) death file, a commonly used source for VA health care enrollees' vital status by VA investigators. Furthermore, the VA Medical sas inpatient data set was used to verify the findings from the BIRLS death file [10].

Quality indicators

In-hospital stroke care quality was assessed using 14 quality indicators: dysphagia screening, documentation of the National Institute of Health Stroke Scale (NIHSS), thrombolysis (tissue plasminogen activator), antithrombotic therapy by hospital day two and at discharge, deep vein thrombosis (DVT) prophylaxis, early ambulation, fall risk assessment, pressure ulcer risk assessment, rehabilitation needs assessment using the functional independence measure (FIM), atrial fibrillation management, lipid management, smoking cessation counseling, and stroke education. A detailed description of the 14 indicators can be found elsewhere [11]. These indicators were developed by a multidisciplinary panel of VA stroke and performance measure experts based on evidence-based clinical guidelines for inpatient stroke care and existing performance measures developed by The Joint Commission [12-19]. For each of the 14 quality indicators, a passing or compliance rate was calculated (the number of patients with the process of care present divided by the total number of eligible patients for each component of care).

Independent variable and other covariates

Independent variable

stroke belt status was determined by the location of the VAMCs providing the inpatient ischemic stroke care. In other words, a VAMC was designated as a stroke belt facility if it was located within the 11 southeastern states (i.e. Mississippi, Louisiana, Kentucky, Georgia, Tennessee, North Carolina, Alabama, South Carolina, Arkansas, Indiana, and Virginia) [3]. Otherwise, the VAMC was coded as non-stroke belt.

Risk adjustment

The covariates of interest for this study can be presented in two broad categories: patient characteristics and facility characteristics. Patient characteristics included age, gender, race-ethnicity, marital status, Charlson comorbidity index, the retrospective NIHSS, the modified Acute Physiology and Chronic Health Evaluation III (APACHE III) score, oxygen saturation, smoking status, code status, and pre-stroke ambulatory status. Information for these variables was obtained from the VA National Medical sas database as well as chart review from the electronic medical record.

In this study, race-ethnicity was categorized as Caucasian, Black, and all other. Marital status was coded as married, divorced, and all other. Charlson comorbidity index was used to assess the patients' medical comorbidity where the higher the weighted summary score, the more severe the burden of comorbidity [20]. The NIHSS, a stroke severity measurement, was calculated retrospectively from the admission neurological examination in the medical record. Stroke severity was categorized as mild (NIHSS ≤ 2), moderate (NIHSS = 3–9), or severe (NIHSS ≥ 10) [21]. The modified APACHE III score, a measure of overall disease severity and a predictor of an individual's risk of dying, was used to assess the admission clinical status of the patients [22]. Hypoxia was defined as either an oxygen saturation <90% or PaO2 <60 mmHg at any time during the first four-days of the hospital stay. Smoking status was classified as current smoker (any smoking in the year prior to admission) vs. other (no smoking or no documentation). Admission code status was classified as full code vs. other [e.g. do not resuscitate/do not intubate (DNR/DNI)]. The patient's level of independence prior to stroke was classified as either ambulatory or nonambulatory, with ambulatory defined as living at home without assistance, and nonambulatory being at home on bed rest or with assistance.

Facility characteristics included VAMC facility complexity. VA has classified all of its VAMCs as low, medium, or high complexity based on patient risk, level of intensive care, number of residency slots, amount of research dollars, and number of physician specialists [23].

This study was approved by the Institutional Review Board and the local VA Research and Development Committee both at Indianapolis, Indiana, and Gainesville, Florida, United States.

Statistical analysis

All statistical analyses were performed using sas version 9·13 (SAS Institute, Cary, NC, USA). First, descriptive statistics were obtained on all the variables. Statistical inference (chi-square test on categorical variables and analysis of variance or Kruskal-Wallis test on dimensional variables) was performed to compare stroke belt and non-stroke belt VAMCs. Second, descriptive statistics for each quality indicator were calculated, and chi-square test was applied to compare each quality indicator between the two types of facilities. Finally, a multivariable logistic regression model was fit to examine the difference in 30-day and 12-month mortality as well as each quality indicator between the two types of facilities, adjusting for patient and facility characteristics as well as VAMC clustering factors. Given the large number of comparisons made in our final analyses, we used a Bonferroni correction (dividing the 0·05 significance level by the number of covariates), resulting in a significance level of 0·0042 for each of the quality indicator model.

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

In Table 1, we summarized the comparison of patient and facility characteristics between stroke belt and non-stroke belt VAMCs. Among the 3909 patients at 129 VAMCs, 28·1% received inpatient ischemic stroke care in the 28 (21·7%) stroke belt VAMCs and 71·9% received the inpatient care in the 101 (78·3%) non-stroke belt VAMCs. Patients who were admitted to the stroke belt VAMCs compared with patients at non-stroke belt VAMCs were younger (mean age 67·1 vs. 68·3), more likely to be married (47·4% vs. 42·2%) and Black (32·8% vs. 19·1%), had more severe strokes (mean NIHSS score 4·9 vs. 4·5), and were less likely to be ambulatory pre-stroke (91·3% vs. 94·1%). We did not find a significant difference in 30-day or 12-month post-admission mortality outcome between the patients in the stroke belt and the patients in the non-stroke belt regions (30-day mortality rate: 7·6% for stroke belt vs. 6·1% for non-stroke belt; 12-month mortality rate: 18·7% for stroke belt vs. 18·6% for non-stroke belt).

Table 1. Univariable comparisons of patient and facility characteristics by stroke belt facility
CharacteristicsSampleStroke belt VAMCsNon-stroke belt VAMCsPa
  1. a

    All P-values were results from bivariate comparisons (analysis of variance or Kruskal-Wallis tests for numerical variables and chi-square tests for categorical variables) between stroke belt VAMCs and non-stroke belt VAMCs.

  2. VAMCs, Veterans Affairs Medical Centers; SD, standard deviation; NIHSS, retrospective National Institute of Health Stroke Scale; APACHE, modified Acute Physiology and Chronic Health Evaluation.

Patient leveln = 3909n = 1098n = 2811 
 Age, mean ± SD68·0 ± 1167·1 ± 1268·3 ± 110·0023
 Female, %2·41·72·70·0757
 Marital status, %   0·0118
Married43·747·442·2 
Divorced30·328·031·2 
All other26·024·626·5 
 Race-ethnicity, %   <0·0001
Caucasian63·060·863·8 
Black22·932·819·1 
All other14·16·417·1 
 Charlson index, mean ± SD4·8 ± 24·7 ± 24·8 ± 20·0822
 NIHSS, mean ± SD4·6 ± 64·9 ± 64·5 ± 60·0373
 APACHE III, mean ± SD12·6 ± 712·7 ± 812·6 ± 70·4186
 Hypoxia, %2·42·72·20·3653
 Smoking status, %35·136·234·80·4052
 Comfort measure, %3·74·53·40·1194
 Do not resuscitate/do not intubate, %13·615·013·10·1129
 Pre-stroke ambulatory, %93·391·394·10·0047
 30-day mortality, %6·57·66·10·1100
 12-month mortality, %18·618·718·60·9832
Facility leveln = 129n = 28n = 101 
 Facility complexity, %   0·0984
High74·873·875·3 
Medium18·520·417·8 
Low6·65·86·9 

In Table 2, we presented the number of eligible patients and the passing rate for each inpatient quality indicator, as well as the bivariate comparison for each indicator between the two types of VAMCs. There were statistically significant unadjusted differences in performance on eight of the 14 quality indicators. Compared to non-stroke belt facilities, stroke belt VAMCs had a significantly higher passing rate in dysphagia screening before oral intake (23·6% vs. 16·2%), documenting the NIHSS (29·1% vs. 24·5%), providing DVT prophylaxis (pharmacologic or mechanical) by the end of hospital day 2 (78·4% vs. 72·2%), providing smoking cessation counseling (97·2% vs. 92·4%), and documenting stroke education (20·6% vs. 14·1%). On the other hand, stroke belt facilities had a significantly lower passing rate than non-stroke belt facilities on three indicators: completing fall risk assessment by the end of hospital day 2 (75·9% vs. 78·9%), pressure ulcer assessment within 24 h before or after hospital admission (90·2% vs. 92·1%), and documenting patient FIM score as means of assessing rehabilitation needs (77·4% vs. 80·2%).

Table 2. Univariable comparisons of inpatient stroke quality indicators by facility type
Quality indicatorsSampleStroke belt VAMCsNon-stroke belt VAMCs 
EligibleCompliant (%)EligibleCompliant (%)EligibleCompliant (%)Pa
  1. a

    All P-values were from chi-square tests comparing stroke belt VAMCs vs. non-stroke belt VAMCs.

  2. VAMC, Veterans Affairs medical center; NIHSS, National Institute of Health Stroke Scale; FIM, functional independence measure; tPA, tissue plasminogen activator.

Dysphagia screening before oral intake3594656 (18·2)987233 (23·6)2607423 (16·2)<0·0001
NIHSS documented3607930 (25·8)1004292 (29·1)2603638 (24·5)0·0049
Thrombolysis (tPA) given30619 (6·2)764 (5·3)23015 (6·5)0·6934
Antithrombotics: by hospital day 234963324 (95·1)966917 (94·9)25302407 (95·1)0·7966
Deep vein thrombosis prophylaxis1043773 (74·1)320251 (78·4)723522 (72·2)0·0339
Early ambulation30092537 (84·3)822681 (82·8)21871856 (84·9)0·1749
Fall risk assessment36382840 (78·1)1016771 (75·9)26222069 (78·9)0·0480
Pressure ulcer assessment37493433 (91·6)1050947 (90·2)26992486 (92·1)0·0577
Rehabilitation needs assessment/FIM35312806 (79·5)975755 (77·4)25562051 (80·2)0·0649
Antithrombotic therapy: discharge35293373 (95·6)979936 (95·6)25502437 (95·6)0·9596
Atrial fibrillation management447307 (68·7)12180 (66·1)326227 (69·6)0·4763
Lipid management30442452 (80·6)859695 (80·9)21851757 (80·4)0·7556
Smoking cessation counseling12721193 (93·8)364354 (97·2)908839 (92·4)0·0012
Stroke education2526403 (15·9)727150 (20·6)1799253 (14·1)<0·0001

As shown in Table 3, there appeared no significant difference in 30-day and 12-month mortality between the VAMCs located inside and outside the stroke belt regions, even after controlling patients' sociodemographics, stroke severity, and facility complexity. Other factors significantly associated with short- and long-term mortality included older age and severity measures such as NIHSS score, APACHE III score, comfort measure, and DNR/DNI. In addition, being Caucasian, diagnosed with hypoxia and low-level hospital complexity were also associated with 30-day mortality; heavier burden of comorbid conditions and pre-stroke admission independence at nonambulatory level were also associated with 12-month mortality, respectively.

Table 3. Mortality results from multivariable logistic regression analyses
Factors30-day post-admission mortality12-month post-admission mortality
Odds ratio95% CIPOdd ratio95% CIP
  1. CI, confidence interval; VAMC, Veterans Affairs medical center; NIHSS, National Institute of Health Stroke Scale; APACHE, modified Acute Physiology and Chronic Health Evaluation; DNR/DNI, do not resuscitate/do not intubate.

VAMC stroke belt: yes vs. no0·90·5–1·50·62610·90·8–1·20·7456
Age: numerical1·11·0–1·1<0·00011·01·0–1·1<0·0001
Married: yes vs. no1·00·8–1·10·64851·31·2–1·40·9202
Caucasian: yes vs. no1·11·0–1·10·04141·11·0–1·10·2918
Charlson index: numerical1·01·0–1·10·89011·01·0–1·1<0·0001
NIHSS: numerical0·90·5–1·5<0·00011·00·8–1·3<0·0001
APACHE III: numerical1·91·0–3·40·00851·10·9–1·4<0·0001
Hypoxia: yes vs. no4·01·5–10·30·00451·70·9–3·50·1137
Smoking status: yes vs. no1·00·5–1·90·96651·20·9–1·50·2507
Comfort measure: yes vs. no34·414·7–80·6<0·000128·48·3–97·3<0·0001
DNR/DNI: yes vs. no2·31·2–4·10·00671·71·3–2·30·0003
Pre-stroke ambulatory: yes vs. no1·10·5–2·40·75530·60·4–0·90·0081
Hospital complexity: low vs. medium and high0·20·1–0·80·01831·10·7–1·80·5398

We presented the results from our multivariable logistic regression analyses in Table 4. For stroke belt facilities, the odds of providing smoking cessation counseling were three times [odds ratio (OR) = 3·3; 95% CI = 1·3–8·5] higher than at the non-stroke belt facilities, even after adjusting for patient and facility characteristics as well as VAMC clustering effect. It should be noted that although our smoking cessation consulting model showed a P-value of 0·0136 and adjusted OR of 3·3 (stroke belt VAMCs vs. non-stroke belt VAMCs), this significance, however, disappeared after Bonferroni correction. The rest of the seven significant bivariate differences were no longer significant after adjusting for the patient and facility level characteristics. The patient characteristics, which were significantly associated with the quality indicators, included patient pre-stroke ambulatory status, race-ethnicity, stroke severity, age, hypoxia, marital status, Charlson comorbidity score, and facility complexity.

Table 4. Inpatient care indicators results from multivariable logistic regression analysesa
IndicatorsStroke belt VAMCs (control = non-stroke belt VAMCs)
Odds ratio (95% CI)Pb
  1. a

    The results for each quality indicator in this table were adjusted for patient age, gender, race-ethnicity, marital status, Charlson index sum score, APACHE III score, retrospective NIHSS, DNR/DNI, pre-stroke ambulatory, hypoxia, comfort measure only, and hospital complexity.

  2. b

    sas PROC GLIMMIX was applied to adjust for the clustering of patients within VAMCs.

  3. VAMC, Veterans Affairs medical Center; CI, confidence interval; FIM, functional independence measure; tPA, tissue plasminogen activator.

Dysphagia screening1·5 (0·9–2·6)0·1094
NIHSS completed1·3 (0·3–7·1)0·7230
Thrombolysis (tPA) given0·8 (0·2–2·6)0·6959
Antithrombotic therapy day 21·0 (0·6–1·5)0·8802
Deep vein thrombosis prophylaxis1·3 (0·8–2·1)0·3430
Early ambulation0·7 (0·3–1·3)0·2509
Fall risk assessment0·9 (0·2–3·3)0·8838
Pressure ulcer risk assessment0·9 (0·5–1·7)0·7989
Rehabilitation consultation/FIM0·7 (0·5–1·2)0·1827
Antithrombotic therapy at discharge1·1 (0·7–1·9)0·6087
Atrial fibrillation management0·9 (0·5–1·5)0·6645
Lipid management1·0 (0·8–1·6)0·4493
Smoking cessation counseling3·3 (1·3–8·5)0·0136
Stroke education1·0 (0·3–3·8)0·9606

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This is the first examination of whether a stroke belt exists within the VA health care system. We did not find evidence to support the concept of a stroke belt within the VA health care system either in terms of post-admission mortality or stroke quality of care.

Our finding that post-admission mortality was similar for patients cared for at stroke belt and non-stroke belt VAMCs differs with most non-VA research reports, which have found higher mortality rates in the stroke belt region (vs. non-stroke belt region) [1, 2, 24]. The difference between the VA and non-VA findings could be due to the unique characteristics of the VA health care system. The Veterans Health Administration under the Department of Veterans Affairs is an integrated health care system, with coordination of inpatient and outpatient care and administrative focus on consistent quality of care across the entire national system. The distinctive structure, administration, and clinical practice of the VA health care system may explain why geographic variation in mortality among the ischemic stroke patients was not observed across VAMCs.

In the unadjusted analyses, stroke belt VAMCs provided higher quality of care for five quality indicators (dysphagia screening, NIHSS, DVT prophylaxis, smoking cessation counseling, and smoking education) and lower quality of care for three quality indicators (fall risk assessment, pressure ulcer risk assessment, and rehabilitation consultation). After risk adjustment, however, stroke belt VAMCs were three times as likely to provide smoking cessation counseling and no other differences in quality of care were observed. Because we are unaware of any reports comparing stroke belt vs. non-stroke belt inpatient stroke care, we are unable to compare our findings to those outside the VA. Given that the proportion of patients who smoked was similar in stroke belt and non-stroke belt VAMCs (36·2% vs. 34·8%, P = 0·41), it is unlikely that a difference in the prevalence of smoking can account for the observed difference in the smoking cessation counseling process of care.

It should be noted that our chart review data were collected through remote review of patients' electronic medical records. Electronic records may sometimes not include information that is included in the paper chart. To ensure the quality of our data, the extracted data were sent to all VAMCs (which had access to the paper charts) for review, confirmation, and correction if needed.

In summary, we did not find a significant difference in either adjusted post-stroke mortality or inpatient quality of care between stroke belt and non-stroke belt VAMCs.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

This research was supported through the Department of Veterans Affairs Health Services Research and Development (VA HSR&D Grant RRP 09–184). The views and opinions expressed in this manuscript reflect those of the authors and do not necessarily reflect those of the Department of Veterans Affairs.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • 1
    Borhani NO. Changes and geographic distribution of mortality from cerebrovascular disease. Am J Public Health Nations Health 1965; 55:673681.
  • 2
    Lanska DJ, Kuller LH. The geography of stroke mortality in the United States and the concept of a stroke belt. Stroke 1995; 26:11451149.
  • 3
    National Heart, Lung and Blood Institute. 1996 Stroke belt initiative. http://www.Nhlbi.Nih.Gov/health/prof/heart/other/sb_spec.Pdf (accessed December 3, 2010).
  • 4
    Howard G. Why do we have a stroke belt in the southeastern United States? A review of unlikely and uninvestigated potential causes. Am J Med Sci 1999; 317:160167.
  • 5
    Goldstein LB, Hey LA, Laney R. North Carolina stroke prevention and treatment facilities survey: RTPA therapy for acute stroke. Stroke 1998; 29:20692072.
  • 6
    Reeves MJ, Arora S, Broderick JP et al. Acute stroke care in the US: results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke Registry. Stroke 2005; 36:12321240.
  • 7
    Reeves MJ, Gargano J, Maier KS et al. Patient-level and hospital-level determinants of the quality of acute stroke care: a multilevel modeling approach. Stroke 2010; 41:29242931.
  • 8
    Bravata D, Ordin D, Vogel WB, Williams L. The quality of VA inpatient ischemic stroke care, FY2007: final national and medical center results of the VHA Office of Quality and Performance (OQP) special study, 2009.
  • 9
    Reker DM, Hamilton BB, Duncan PW, Yeh SC, Rosen A. Stroke: who's counting what? J Rehabil Res Dev 2001; 38:281289.
  • 10
    Jia H, Zheng Y, Reker DM et al. Multiple system utilization and mortality for veterans with stroke. Stroke 2007; 38:355360.
  • 11
    Chumbler NR, Jia H, Phipps MS et al. Does inpatient quality of care differ by age among US veterans with ischemic stroke? J Stroke Cerebrovasc Dis June 6, 2011. [Epub ahead of print].
  • 12
    The Joint Commission. 2004 Specifications manual for national hospital inpatient quality measures. http://www.Jointcommission.Org/stroke/ (accessed December 6, 2010).
  • 13
    Bates B, Choi JY, Duncan PW et al. Veterans affairs/department of defense clinical practice guideline for the management of adult stroke rehabilitation care: executive summary. Stroke 2005; 36:20492056.
  • 14
    Biller J, Feinberg WM, Castaldo JE et al. Guidelines for carotid endarterectomy: a statement for healthcare professionals from a special writing group of the stroke council, American Heart Association. Stroke 1998; 29:554562.
  • 15
    Coull BM, Williams LS, Goldstein LB et al. Anticoagulants and antiplatelet agents in acute ischemic stroke: report of the joint stroke guideline development committee of the American Academy of Neurology and the American Stroke Association (a division of the American Heart Association). Neurology 2002; 59:1322.
  • 16
    Stroke Council, American Heart Association; American Stroke Association. Statins after ischemic stroke and transient ischemic attack: an advisory statement from the Stroke Council, American Heart Association and American Stroke Association. Stroke 2004; 35:1023.
  • 17
    Duncan PW, Zorowitz R, Bates B et al. Management of adult stroke rehabilitation care: a clinical practice guideline. Stroke 2005; 36:e100143.
  • 18
    Sacco RL, Adams R, Albers G et al. Guidelines for prevention of stroke in patients with ischemic stroke or transient ischemic attack: a statement for healthcare professionals from the American Heart Association/American Stroke Association council on stroke: co-sponsored by the council on cardiovascular radiology and intervention: the American Academy of Neurology affirms the value of this guideline. Circulation 2006; 113:e409449.
  • 19
    Adams HP, Jr, Adams RJ, Brott T et al. Guidelines for the early management of patients with ischemic stroke: a scientific statement from the stroke council of the American Stroke Association. Stroke 2003; 34:10561083.
  • 20
    Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chronic Dis 1987; 40:373383.
  • 21
    Williams LS, Yilmaz EY, Lopez-Yunez AM. Retrospective assessment of initial stroke severity with the NIH Stroke Scale. Stroke 2000; 31:858862.
  • 22
    Zimmerman JE, Wagner DP, Draper EA, Wright L, Alzola C, Knaus WA. Evaluation of Acute Physiology and Chronic Health Evaluation III predictions of hospital mortality in an independent database. Crit Care Med 1998; 26:13171326.
  • 23
    Perlin JB, Kolodner RM, Roswell RH. The Veterans Health Administration: quality, value, accountability, and information as transforming strategies for patient-centered care. Healthc Pap 2005; 5:1024.
  • 24
    Howard G, Anderson R, Johnson NJ, Sorlie P, Russell G, Howard VJ. Evaluation of social status as a contributing factor to the stroke belt region of the United States. Stroke 1997; 28:936940.