Time Trends in Patient Characteristics Treated on Acute Stroke-Units
Results From the Austrian Stroke Unit Registry 2003–2011
Background and Purpose—Demographic changes, increased awareness of vascular risk factors, better diagnostic, progress in medical care, and increasing primary stroke prevention influence the profile of patients admitted to stroke-units. Changes in patient population and stroke type have important consequences on outcome and management at stroke-units.
Methods—Data from the national database of the Austrian Stroke Unit Registry were analyzed for time-trends in demography, risk factors, cause, and stroke severity.
Results—Data of 48 038 ischemic and 5088 hemorrhagic strokes were analyzed. Between 2003 and 2011, median age increased significantly for ischemic strokes from 68 to 71 years in men and from 76 to 78 years in women, respectively. Ischemic stroke patients showed significantly increased rates of hypertension, hypercholesterolemia, and atrial fibrillation. In hemorrhagic strokes an increase for hypercholesterolemia and cardiac diseases other than atrial fibrillation and myocardial infarction were only found in men. A small but significant decrease in stroke severity was found for ischemic strokes from 4 to 3 points on the National Institutes of Health Stroke Scale in men and from 5 to 4 in women, and for hemorrhagic strokes from 9 to 6 points in men and from 9 to 7 in women. Cardioembolic strokes increased slightly, whereas macroangiopathy decreased.
Conclusions—Significant time trends were seen for characteristics of ischemic and hemorrhagic stroke patients admitted to acute stroke-units in Austria. These include trends for older age and toward milder strokes with more cardioembolic causes. This signals a need for increased resources for managing multimorbidity and enabling early mobilization.
Demographic changes, increases in lifestyle-related risk factors of stroke such as physical inactivity, obesity, and diabetes mellitus, better primary prevention, and advances in medical care in the general population, suggest corresponding changes within the stroke population. Indeed, vascular risk factors, stroke severity, and stroke types of patients admitted for stroke have been found to change over the last 20 to 30 years.1–5 However, the detected changes differ between study populations, possibly as a result of different settings, diagnostic criteria, or referral bias.
Benefits of stroke-unit care reflected by a decreased mortality and dependency after stroke have been repeatedly demonstrated,6,7 and patients are increasingly admitted to stroke-units. Since the introduction of stroke-units, different referral strategies and better awareness of stroke symptoms may have led to changes in the characteristics of patients admitted to stroke-units.
Changes in the profile of stroke patients have consequences for the outcome and may necessitate adaptations of stroke-unit care to new needs. The aim of this study is to analyze data of the national wide Austrian Stroke Unit Registry for time trends between 2003 and 2011 in stroke subtypes, stroke severity, and risk factor profiles of stroke patients.
Since 2003, the Austrian Stroke Unit Registry prospectively collects data on baseline characteristics, management, and outcome of all stroke patients admitted to 33 of 34 Austrian stroke-units. Data entry, administration, and scientific analysis are approved and coordinated by an expert committee from the Austrian Stroke Society. Data collection, ratings and data entry are performed by experienced stroke neurologists at the times of admission to and discharge from the stroke-unit (for more details see Reference 8).9
Vascular risk factors were determined according to medical history, prestroke medication, or newly diagnosed during the stay at the stroke-unit. The risk factors recorded were hypertension, hypercholesterolemia, diabetes mellitus, previous stroke, atrial fibrillation, myocardial infarction, other cardiac diseases, peripheral arterial disease, nicotine, and regular alcohol consumption. Hypertension was defined by a history of known hypertension with a systolic blood pressure ≥140 mm Hg or diastolic blood pressure ≥90 mm Hg (according to the criteria of the Seventh Report of Joint National Committee on Prevention and Detection, Evaluation, and Treatment of High Blood Pressure),10 or antihypertensive treatment. Hypercholesterolemia was defined by a history of total cholesterol level ≥240 mg/dL or in the blood sample taken during hospital stay (according to the National Cholesterol Education Program),11 or if patients were treated with lipid-lowering therapy. Diabetes mellitus was diagnosed according to the World Health Organization12 if the fasting plasma glucose level ≥126 mg/dL or 2h plasma glucose level ≥200 mg/dL have been reported in the medical history or if patients were treated with insulin or oral hypoglycemic agents. Regular alcohol consumption is defined as >1 standard drink per day (>500 mL beer or >250 mL wine). Atrial fibrillation, if not known prestroke, is diagnosed using ECG monitoring on the stroke-unit. Functional status before stroke was assessed by the modified Rankin Scale. Stroke types were classified on the basis of neuroimaging (CT or MRI) findings according to the International Classification of Diseases (ICD)-10 code into ischemic stroke (I63) or primary intracerebral hemorrhage (I61). Stroke severity is classified using the National Institutes of Health Stroke Scale (NIHSS), and stroke cause is classified into 5 subtypes (macroangiopathy, cardioembolism, microangiopathy, other determined etiology, undetermined etiology) according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) criteria.13
Analysis included data from acute stroke patients, aged >17 years, living in Austria and admitted to one of the Austrian stroke-units between 2003 and 2011. From 2003 to 2011, data of a total 56 552 acute stroke patients were entered in the Austrian Stroke Unit Registry. Of these, 53 126 fulfilled the inclusion criteria, of these 48 038 were ischemic and 5088 hemorrhagic strokes and entered the analysis. All time-trends analyses were performed separately for ischemic and hemorrhagic strokes and for both sexes. In all regression analyses, the date of the stroke event was treated as metric variable (days since the 1.1.1970 divided by 365.25). Age was Box-Cox transformed, and the parameter λ of this transformation was optimized to approximate a normal distribution. Time trends for metric target variables (age and stroke severity) were analyzed using linear regressions. Time trends for age were analyzed using linear regressions with age as target variable and date of stroke as explaining variable. Similarly, time trends for stroke severity were analyzed using linear regressions with stroke severity as target variable and date of stroke as explaining variable; in age-adjusted models age was added as explaining variable. Analyses for time trends in stroke severity were performed for each etiologic subtype. Functional status before stroke was analyzed using proportional odds logistic regression with date as explaining variable. Logistic regressions were used to test for time trends in binary variables (risk factor presence, etiology, sex). For each risk factor and each etiology logistic regression analyses, including the presence of a risk factor or cause as target variable and the exact date of stroke as explaining variable, were performed separately. Models were calculated with and without age adjustment. The statistical environment used was R version 184.108.40.206
Mean age of all acute stroke patients was 71 years (± SD), and 47.8% were women. No time trends were found for sex distribution (P>0.970). Age increased significantly over the years for ischemic stroke in men (P<0.001) and in women (P<0.001) and for hemorrhagic stroke in men (P<0.001) and in women (P<0.003; Figure 1). Similar trends were observed for first-ever stroke: median age increased for ischemic stroke from 67.0 to 69.9 years in men (P<0.001) and from 75.1 to 76.6 years in women (P=0.002) and for hemorrhagic strokes from 66.7 to 69.4 years in men (P=0.003), but did not change significantly in women (P=0.083). In 2003, 76.9% of men and 69.3% of women with ischemic stroke and 79.0% of men and 65.2% of women with hemorrhagic stroke had no functional impairment before stroke (modified Rankin Scale score = 0). This rate has decreased in 2011 to 74.6% in men (P=0.087) and 62.5% of women (P=0.011) with ischemic stroke and to 70.8% of men (P<0.001) and 60.1% of women (P=0.002) with hemorrhagic stroke. After adjusting for age time trends were only significant for men with hemorrhagic strokes (P=0.005).
In ischemic stroke patients, rates for hypertension, hypercholesterolemia, atrial fibrillation, and for cardiac diseases other than myocardial infarction increased significantly. A decrease for diabetes mellitus was only observed in women. These time trends remained significant after correcting for age (Table 1). An increase in myocardial infarction was only observed in women (P=0.029) and was no longer significant after age adjustment. Regular alcohol consumption was mainly recorded for men (overall 13.5% in men compared with 2.6% in women) and decreased in men (P<0.001) and women (P=0.042); after age adjustment this trend was no longer significant in women (Table 1).
In hemorrhagic stroke patients, significant increases in rates of hypercholesterolemia and cardiac diseases other than atrial fibrillation or myocardial infarction were observed in men only (Table 2). Before age correction, rates of hypertension in women (P=0.038) and atrial fibrillation in men (P=0.016) and women (P=0.013) were found to increase.
Stroke Type, Cause, and Stroke Severity
The percentage of hemorrhagic strokes decrease slightly but significantly over the years (P=0.008). Stroke severity decreased significantly in both stroke types, in men and women, and for all etiologic stroke subtypes (P<0.05; Figure 2). These time trends were independent of age. Ischemic strokes attributable to cardiac causes increased from 2003 to 2011 (P<0.001), whereas macroangiopathy (P<0.001) decreased and microangiopathy did not change (P=0.959). Within the other causes, a decrease of hemodynamic causes was observed (P<0.001).
Over the last 9 years the profile of patients admitted to Austrian stroke-units has changed: Patients became older and had milder strokes, more often associated with cardiac causes. Vascular risk factors were more often hypertension, hypercholesterolemia, and atrial fibrillation.
Decreases of smoking rate, blood pressure, and serum cholesterol and increases in body mass index, rates of diabetes mellitus, atrial fibrillation, and blood pressure lowering medications have been observed in the general population since the 1960s.15–17 These time trends are only partly reflected in shifts of risk factor prevalence in stroke populations.1–5
Hypertension, the strongest risk factor for stroke, is 78% for ischemic stroke higher in the Austrian stroke-unit population than the reported 66% in the Importance of Conventional and Emerging Risk Factors for Stroke in Different Regions of the World and in Different Ethnic Groups (INTERSTROKE) study18 and was found to increase since 2003. The Austrian Stroke Unit Registry does however not distinguish between previous treated and newly detected hypertension. This is comparable with results of 2 hospital-based studies3,5 which found increases in hypertension, and to the population-based Oxford Vascular Study,1 in which decreases of prestroke blood pressure values were observed together with increases in treated hypertension. The hospital-based Lausanne Stroke Registry,2 on the other hand, found a decrease in the proportion of patients with a history of hypertension, and no change was found in the population-based Dijon register.4
Similar to other studies in stroke patients, the number of hypercholesterolemic patients increased over time.1,2,4 Values of blood cholesterol are however not recorded in the register. Results of the Oxford Vascular Study showed a decrease of premorbid mean blood cholesterol, whereas the treatment with lipid-lowering medications is increasing.1
The prevalence of diabetes is increasing worldwide,16,19 however this trend is not reflected within the Austria stroke-unit population. Yet, the real number of diabetic patients is probably underestimated, because oral glucose tolerance tests (OGTT) are not performed in all patients without previous history of diabetes mellitus.20 Results from other stroke populations are conflicting: diabetes mellitus rate was found to increase in the Dijon population4 and the Barcelona register before age adjustment,3 to decrease in the Lausanne registry,2 and no time trends were detected in the Oxford population1 and in the Swedish Register.5
Despite a decreasing rate of smoking in the general and in several stroke populations,1–3,5 no time trends were found in the Austrian Stroke Unit Registry. Rates of current smokers are comparable with those reported for 2002 to 2004 in other studies.1,2 It might be that the decrease in smoking rate has slowed since then, and that longer observation periods are necessary to identify time trends.
The prevalence of atrial fibrillation was found to increase over the years. This is also reflected by an increase of cardioembolic strokes. The observed change is probably not attributable to better diagnostics, as all patients undergo an admission ECG and ECG monitoring. Increased rates of atrial fibrillation were also observed in other studies1,3,5 and may reflect changes in the general population.17 Despite of the admission of more cardioembolic strokes, stroke severity has decreased since 2003. This may be partly explained by a decrease of large-vessel atherosclerotic stroke, but may also reflect better primary prevention or changes in referral. This is in line with other studies.1,5
Case-fatality has decreased in the last decades in high-income countries, likely reflecting advances in stroke care and secondary prevention.21–23 A possible consequence of the increasing number of stroke survivors might be an increase of recurrent strokes and comorbidities in stroke patients. Indeed, in our population, patients were observed to be more often functionally impaired before the stroke. However, this trend can mainly be attributed to the increasing age of the population. Furthermore, during the last decade no changes in the rate of recurrent strokes were observed in the Austrian stroke-unit population. Data on premedication are not recorded in the Austrian Stroke Unit Registry. Thus, we can only speculate that the expected increase in stroke recurrence attributable to stroke survivors is counterbalanced by better secondary prevention. In the Swedish Stroke Register the proportion of recurrent stroke was found to decrease since 1995.5
Hemorrhagic strokes are more severe and occur at younger age. Hypertension is the most common risk factor and stronger for intracerebral hemorrhages than for ischemic stroke. Alcohol intake is associated with an increased risk of hemorrhagic stroke, whereas the influence of other vascular risk factors such as cholesterol and smoking are inconclusive or absent as in the case of diabetes mellitus or heart diseases.18,24 Accordingly, our data showed a higher prevalence of hypertension for hemorrhagic compared with ischemic strokes but lower or similar rates for all other vascular risk factors. The time trends observed for risk factor prevalence in ischemic strokes could not be verified for hemorrhages strokes, except for increasing rates of hypercholesterolemia and cardiac diseases in men. Considering the different characteristics of hemorrhagic and ischemic stroke patients, changes of vascular risk factors in the general population caused by an increasingly unhealthy lifestyle may be reflected stronger in ischemic stroke patients. On the other hand, small changes may be less detectable in hemorrhagic strokes because of their lower frequencies.
Our survey is limited by the fact that the register encompasses only cases of acute strokes admitted to stroke-units. In Austria, two thirds of all strokes admitted to hospitals are treated at stroke-units.25 However, only data of acute strokes (admitted within 24 hours) are entered into the stroke-unit registry, representing ≈25% of all hospitalized strokes, including chronic cases. Data of the Austrian Stroke Registry can thus be seen as representative for acute stroke patients in Austria. The strength of this prospective register is that over 9 years data of all stroke-units in Austria were entered in a predefined way. Changes in stroke types or risk factors are thus not caused by changes in diagnostic criteria but reflect real changes in patients’ characteristics. However, we cannot distinguish whether these changes are consequences of changes in admission, of better primary prevention associated with increased screening for risk factors or of risk factor changes in the population. Analyses are based on hospital data and thus do not allow conclusions on risk factor changes in the general population or on their relative importance for stroke risk. However, this analysis of nationwide data allows assessing major time trends in stroke patients’ characteristics and inferences for the future management of acute stroke patients. Changes in patients’ characteristics suggest that acute management will have to focus increasingly on older and multimorbid patients suffering milder strokes. So far only one other analysis of hospital data based on more than a single center has been published.1
Significant time trends were seen for characteristics of ischemic and hemorrhagic stroke patients admitted to acute stroke-units in Austria. These include trends for older age and toward milder strokes. This signals a need for increased resources for managing multimorbidity and enabling early mobilization.
- Received September 6, 2012.
- Accepted December 28, 2012.
- © 2013 American Heart Association, Inc.
- Seenan P,
- Long M,
- Langhorne P
- Reiner-Deitemyer V,
- Teuschl Y,
- Matz K,
- Reiter M,
- Eckhardt R,
- Seyfang L,
- et al
- Chobanian AV,
- Bakris GL,
- Black HR,
- Cushman WC,
- Green LA,
- Izzo JL Jr.,
- et al
- 11.↵National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III). Third Report of the National Cholesterol Education Program (NCEP) Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults (Adult Treatment Panel III) final report. Circulation. 2002;106:3143–421.
- 12.↵WHO Consultation. Definition, diagnosis and classification of diabetes mellitus and its complications. Part 1: diagnosis and classification of diabetes mellitus. Geneva: World Health Organisation; 1999. Report No.99.2.
- Adams HP Jr.,
- Bendixen BH,
- Kappelle LJ,
- Biller J,
- Love BB,
- Gordon DL,
- et al
- 14.↵R Development Core Team. A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing: 2010. http://www.r-project.org/. Accessed September 3, 2012.
- Tolonen H,
- Mähönen M,
- Asplund K,
- Rastenyte D,
- Kuulasmaa K,
- Vanuzzo D,
- et al
- Harmsen P,
- Wilhelmsen L,
- Jacobsson A
- Wild S,
- Roglic G,
- Green A,
- Sicree R,
- King H
- Matz K,
- Keresztes K,
- Tatschl C,
- Nowotny M,
- Dachenhausen A,
- Dachenhausenm A,
- et al
- Kunst AE,
- Amiri M,
- Janssen F
- Sarti C,
- Stegmayr B,
- Tolonen H,
- Mähönen M,
- Tuomilehto J,
- Asplund K
- Ariesen MJ,
- Claus SP,
- Rinkel GJ,
- Algra A