From the Institute for Clinical Evaluative Sciences (Dr Tu and Ms Gong), the Division of General Internal Medicine and Clinical Epidemiology, Sunnybrook and Women's College Health Science Centre (Dr Tu), and the Departments of Medicine, Public Health Sciences, and Health Policy Management and Evaluation (Dr Tu), University of Toronto, Toronto, Ontario.
Several interventions have been shown to be of benefit to patients with stroke (hereafter referred to as stroke patients) in clinical trials, but the net effect of these interventions in the general stroke population has not been established. The purpose of this study was to evaluate temporal trends in the characteristics, treatments, and outcomes of acute stroke patients in the province of Ontario.
We conducted a population-based retrospective cohort study using linked administrative databases of all 91 419 patients discharged with a most-responsible diagnosis of acute stroke from acute care hospitals in Ontario from April 1, 1992, to March 31, 1999 (fiscal years 1992-1998).
The average age and proportion of stroke patients with co-existing diseases increased over time. The proportion of elderly patients 65 years and older who received warfarin sodium (Coumadin) and statins increased during the study period (14.6% to 19.6% [P = .001] and 2.7% to 15.0% [P<.001], respectively). Declines in the median length of stay (11 to 8 days [P<.001]) and risk-adjusted in-hospital mortality (21.9% to 18.9% [P<.001]) were significant, but the 30-day mortality rates for acute stroke stayed relatively constant (19.7% to 19.0% [P = .18]). We found a moderate decline in risk-adjusted 1-year mortality (34.1% to 32.0% [P<.001]) and stroke readmission rates (12.1% to 9.9% [P = .001]).
Improvements in the outcomes of stroke patients have occurred in Ontario during the 1990s, despite an increasing proportion of elderly stroke patients with multiple comorbidities. Increasing use of secondary prevention medications may explain this trend.
STROKE IS the leading cause of long-term disability in adults and the third most common cause of death in North America.1,2 The 1980s and 1990s have witnessed the development of several interventions that improve the short- and long-term outcomes of patients with stroke (hereafter referred to as stroke patients). Interventions that improve the outcomes in acute stroke include thrombolytic therapy for patients presenting with an ischemic stroke within 3 hours of stroke onset and aspirin administration within 24 hours of admission.3,4 Admission to a stroke unit has also been associated with better outcomes.5 A number of therapies have been developed that reduce the risk for recurrent strokes. These include aspirin, ticlopidine hydrochloride, clopidogrel bisulfate, statins, and warfarin sodium (Coumadin).6- 10 In addition, carotid endarterectomy has been shown to be effective in preventing strokes in patients with high-grade carotid stenosis.11
Although the efficacy of these interventions has been well established, fewer data establish their net effectiveness in the real world. It is not known to what extent these interventions have diffused into routine clinical practice, and whether they have had a net positive impact on the outcomes of stroke patients in the 1990s. For these reasons, we conducted a population-based study using linked administrative databases in the province of Ontario to evaluate temporal trends in the characteristics, treatment, and outcomes of acute stroke patients from April 1, 1992, to March 31, 1999 (fiscal years 1992-1998).
Three administrative databases were used to create the stroke cohort for this study. Acute stroke patients were identified from the Canadian Institute for Health Information (CIHI) hospital discharge database, which contains demographic, length of stay, comorbidity, in-hospital mortality, and discharge destination information on all patients admitted to Ontario hospitals. The most responsible and up to 15 secondary diagnoses are coded in the CIHI database according to the International Classification of Diseases, Ninth Revision (ICD-9) coding system.12 The CIHI database also codes as many as 10 procedures according to the Canadian Classification of Procedures.13 Prescription claims for aspirin, ticlopidine, warfarin, and statins in stroke survivors 65 years and older were obtained by linkage to the Ontario Drug Benefit database from April 1, 1992, to June 30, 1999. The Ontario Drug Benefit database contains all outpatient prescriptions filled in Ontario by persons 65 years and older for prescription medications from a minimally restrictive formulary. Mortality status of the patients at 30 days and 1 year was obtained from the Registered Persons Database.
All patients discharged from acute care hospitals in Ontario from years 1992 through 1998 (inclusive) with a most responsible diagnosis (the diagnosis that accounts for most of the hospital stay) of acute stroke (ICD-9 code 431, intracerebral hemorrhage; ICD-9 code 434, occlusion of cerebral arteries; or ICD-9 code 436, acute but ill-defined cerebrovascular disease) were included in this study. Ischemic strokes were identified using ICD-9 codes 434 and 436, whereas hemorrhagic strokes were identified using ICD-9 code 431. Previous validation studies have shown a high accuracy rate of coding for stroke in the CIHI database (>90%) when these ICD-9 codes are compared with the opinions of a neurologist.14 A fiscal year was defined as April 1 of that calendar year to March 31 of the subsequent year. We excluded from the study patients who were younger than 20 years or older than 105 years, those who were non-Ontario residents or were transferred from another institution, and those whose strokes occurred as an in-hospital complication. We also excluded those who had a previous admission within the past year due to an acute stroke, so that our cohort consisted of patients with new strokes only.
The frequency of various stroke risk factors and other comorbid diseases was determined with use of the appropriate ICD-9 codes in the 15 secondary diagnosis fields of the CIHI database. The stroke risk factors and comorbidities included ischemic heart disease, congestive heart failure, atrial fibrillation, peripheral vascular disease, chronic pulmonary disease, peptic ulcer disease, liver disease, malignancy, renal disease, hypertensive disease, and diabetes mellitus. The Charlson Comorbidity Index score as adapted by Deyo et al15 was used to quantify overall comorbid disease status.
We estimated in-hospital, 30-day, and 1-year rates of carotid endarterectomy and the percentage of elderly stroke survivors receiving aspirin, ticlopidine, warfarin, and statins within 90 days of discharge. The outcomes evaluated included in-hospital, 30-day, and 1-year mortality, 1-year recurrent stroke readmission rates, length of stay, and discharge destination.
We calculated the crude rates of carotid endarterectomy use, readmission, discharge destination, drug prescription, and mortality for each acute stroke cohort annually. We obtained the annual risk-adjusted mortality rates by multiplying the overall mortality rate by the ratio of the observed annual mortality rate to the expected mortality rate. Logistic regression models were developed to calculate the expected mortality rates, with age, sex, and the presence of comorbid diseases (eg, ischemic heart disease, atrial fibrillation, hypertension, acute myocardial infarction, congestive heart failure, peripheral vascular disease, chronic obstructive pulmonary disease, diabetes, renal disease, and cancer) included in the model. We assessed changes in demographic characteristics using the Mantel-Haenszel χ2 test for trend and changes in the processes and outcomes of care during the study period using the linear regression of means. All statistical analyses and data linkage were conducted by use of the statistical program SAS release 8.0.16
During fiscal years 1992 to 1998, 107 289 patients were admitted to Ontario acute care hospitals with a most responsible diagnosis of acute stroke. Excluded from the study were 364 patients who were younger than 20 years or older than 105 years, 80 patients who were non-Ontario residents, 5637 patients who were transferred from another acute care hospital, 2594 patients whose acute strokes were coded as a hospital complication, and 7195 patients who had a previous stroke admission within the past year. The final cohort contained 91 419 patients.
Table 1 shows the changes in the demographic characteristics of the acute stroke population from 1992 to 1998. Overall, the mean age of the acute stroke patients was 73.9 years (SD, 11.9 years). The percentage of the study population who were 75 years or older increased from 52.5% to 54.9% (P<.001), whereas the sex distribution among the study population remained the same (P = .39). The most common comorbid conditions were hypertension, diabetes, ischemic heart disease, and atrial fibrillation. We found a gradual increase in the preadmission level of chronic illness, as reflected by the rising percentages of patients with hypertension, atrial fibrillation, diabetes, and a Deyo-Charlson Comorbidity Index score of 2 or higher (P<.001).
Table 2 demonstrates the changes in processes of care during the study period. The percentage of patients receiving carotid endarterectomy procedures after a stroke did not change during the 7-year time frame (P = .84, .69, and .69 for the in-hospital, 30-day, and 1-year periods, respectively) and was very low. Among patients who were 65 years or older and who were discharged home, the percentage who received statin and warfarin prescriptions within 90 days of hospital discharge increased from 1992 to 1998 (2.7% to 15.0% [P<.001] and 14.6% to 19.6% [P = .001], respectively), whereas the percentage of patients receiving aspirin decreased during the same period (50.9% to 36.4% [P = .001]). Part of the decline after 1996 is possibly due to the introduction of a co-payment for medications such as aspirin in the Ontario Drug Benefit program. The percentage of patients receiving physiotherapy, occupational therapy, and speech therapy significantly increased over time (P = .046, P<.001, and P<.001, respectively).
We found several significant outcome changes from 1992 to 1998 (Table 3). There was a steady decline in the mean hospital stay from 1992 to 1998. The median length of the hospital stay declined from 11 days in 1992 to 8 days in 1998 (P<.001). The percentage of patients who survived their index admissions but were transferred to a rehabilitation facility increased from 11.1% to 17.5% (P<.001), and the percentage of patients who were discharged to a home-care program also increased over time (P = .007). The percentage of patients who were discharged home decreased from 49.3% to 41.2% (P<.001). During the same period, the risk-adjusted in-hospital mortality rate dropped from 21.9% to 18.9% (P<.001), but the risk-adjusted 30-day mortality rate was relatively constant (19.7% vs 19.0% [P = .18]). The risk-adjusted mortality rate within 1 year of admission also dropped from 34.1% to 32.0% (P<.001). Finally, the readmission rate for a second acute stroke within 1 year of the index admission declined from 12.1% to 9.9% (P = .001).
This study provided an opportunity to evaluate changes in the characteristics, treatment, and outcomes of acute stroke patients in Ontario during the 1990s. Our study demonstrated a number of important findings. First, the proportion of elderly patients with stroke is increasing, with an associated increase in the prevalence of coexisting comorbidities. Positive increases were found in the rates of prescription of various secondary prevention medications, with an increased proportion of elderly patients receiving prescriptions for statins and warfarin. Although the length of stay and in-hospital mortality rate declined, the 30-day mortality rate for acute stroke patients stayed relatively constant, with a moderate decline in 1-year poststroke mortality and stroke readmissions. The proportion of stroke patients transferred to rehabilitation facilities increased, whereas the proportion sent directly home decreased. These results suggest that although improvements have occurred, significant opportunities for further improving the outcomes of stroke patients in Ontario remain.
The results of our study demonstrated an increase in the average age of stroke patients in Ontario. This may in part reflect the aging of the Canadian population or the effectiveness of the primary prevention strategies (eg, blood pressure control) in younger groups. We observed considerable changes in the available measures of process of care in our study. The increased use of ticlopidine and warfarin over time, relative to aspirin, may reflect the occurence of more primary failures of aspirin therapy or intolerance to aspirin and the effectiveness of warfarin therapy in preventing strokes in patients with atrial fibrillation.8 Clopidogrel was not available in Ontario until 1998, and prescription required that special limited-use forms first be completed. The increased use of statins largely predated the publication of meta-analyses, suggesting benefits of statins in stroke prevention, and may reflect their increased use in the coronary heart disease population. The declining 1-year mortality rates and stroke readmission rates may reflect increases in the use of secondary prevention medications. Nevertheless, the data show that a significant minority of Ontario stroke patients was still discharged without any of these medications, and the rates of carotid endarterectomy in the study population were extremely low.
The absence of a significant decline in 30-day mortality rates was perhaps the most unexpected finding from this study. However, this finding may not be entirely unexpected, since important advances in acute stroke management, such as the use of thrombolytics, offer benefits primarily in terms of functional status rather than mortality,3 and solid evidence of the role of aspirin in acute stroke was not published until 1997.4 Although recent trials and meta-analyses have suggested an important role for acute stroke units in stroke management, a survey of Ontario hospitals showed that only 4% of Ontario hospitals had a stroke unit in 1998.17 The decline in length of stay is encouraging and likely reflects the financial and budgetary pressures on Ontario hospitals to transfer patients out of the acute care sector and greater efficiency on the part of health care providers. However, this finding was associated in part with a greater proportion of transfers to rehabilitation facilities, which may result in less cost savings to the health care system as a whole.
Our study has certain limitations. First, we did not have any data on trends for in-hospital medication use that may have occurred during the 1990s. Second, we were limited in our ability to characterize the severity of strokes using these administrative databases. Third, we did not have information on secondary prevention rates in nonelderly stroke patients in Ontario, although they represent only one fifth of our cohort. Despite these limitations, we believe our data are relatively unique and are unaware of similar studies published in other jurisdictions during our study period.
Little change has occurred in the 30-day mortality rate of acute stroke patients in Ontario during most of the 1990s, but the 1-year mortality rate has declined modestly. This improvement may reflect increasing use of effective secondary prevention interventions in stroke patients in Ontario. The continuing high mortality rates for these stroke patients suggest that although important advances in stroke management have occurred during the past 2 decades, their impact at a population-based level has been modest. Further intensive efforts will be required to improve the outcomes of stroke patients.
Corrresponding author: Jack V. Tu, MD, PhD, FRCPC, Institute for Clinical Evaluative Sciences, G106-2075 Bayview Ave, Toronto, Ontario, Canada M4N 3M5 (e-mail: firstname.lastname@example.org).
Accepted for publication June 13, 2002.
This study was supported by an operating grant from the Canadian Stroke Network, Ottawa, Ontario. The Institute for Clinical Evaluative Sciences, Toronto, Ontario, is supported by an operating grant from the Ontario Ministry of Health and Long-Term Care. Dr Tu is supported by a Canada Research Chair in Health Services Research.
The results and conclusions are those of the authors, and should not be attributed to any of the sponsoring agencies.
Thank you for submitting a comment on this article. It will be reviewed by JAMA Internal Medicine editors. You will be notified when your comment has been published. Comments should not exceed 500 words of text and 10 references.
Do not submit personal medical questions or information that could identify a specific patient, questions about a particular case, or general inquiries to an author. Only content that has not been published, posted, or submitted elsewhere should be submitted. By submitting this Comment, you and any coauthors transfer copyright to the journal if your Comment is posted.
* = Required Field
Disclosure of Any Conflicts of Interest*
Indicate all relevant conflicts of interest of each author below, including all relevant financial interests, activities, and relationships within the past 3 years including, but not limited to, employment, affiliation, grants or funding, consultancies, honoraria or payment, speakers’ bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued. If all authors have none, check "No potential conflicts or relevant financial interests" in the box below. Please also indicate any funding received in support of this work. The information will be posted with your response.
Some tools below are only available to our subscribers or users with an online account.
Download citation file:
Web of Science® Times Cited: 18
Customize your page view by dragging & repositioning the boxes below.
Users' Guides to the Medical Literature
All results at
Enter your username and email address. We'll send you a link to reset your password.
Enter your username and email address. We'll send instructions on how to reset your password to the email address we have on record.
Athens and Shibboleth are access management services that provide single sign-on to protected resources. They replace the multiple user names and passwords necessary to access subscription-based content with a single user name and password that can be entered once per session. It operates independently of a user's location or IP address. If your institution uses Athens or Shibboleth authentication, please contact your site administrator to receive your user name and password.