- Email: [email protected]
Semi-intensive stroke unit versus conventional care in acute ischemic stroke or TIA — A prospective study in Germany
Journal of the Neurological Sciences 287 (2009) 131–137
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s
Semi-intensive stroke unit versus conventional care in acute ischemic stroke or TIA — A prospective study in Germany Andreas Walter a,⁎, Guenter Seidel b,⁎, Andreas Thie c, Heiner Raspe d a
National Association of Statutory Health Insurance Physicians, Berlin, Germany Department of Neurology, University of Schleswig-Holstein, Campus Luebeck, Luebeck, Germany Neurology Service, Klinikum Itzehoe, Itzehoe, Germany d Institute for Social Medicine University of Schleswig-Holstein, Campus Luebeck, Luebeck, Germany b c
a r t i c l e
i n f o
Article history: Received 27 April 2009 Received in revised form 6 August 2009 Accepted 10 August 2009 Available online 6 September 2009 Keywords: Stroke units Acute stroke care Semi-intensive monitoring Outcome research
a b s t r a c t German Society of Neurology's stroke-unit concept includes a specialized stroke-unit team and advanced monitoring facilities in the early phase of stroke. Our aim was to evaluate the effectiveness of this semiintensive stroke-unit (SI-SU) concept as compared with conventional care (CC) for patients with acute ischemic stroke (AIS) or transient ischemic attack (TIA). Over a 20-month period starting in March 2001, 755 patients with AIS or TIA were treated under SI-SU (n = 393) or CC (n = 362) conditions within an observational study. All patients were admitted within 24 h after symptom onset. Patients' outcomes were assessed after 1 year of follow-up, considering concept of care, patients' clinical characteristics, and comorbid conditions. The CC patients were older and had a greater frequency of atrial fibrillation. 1 year after ischemia, 715 patients (94.7%) could be reinvestigated. Mortality and disability (mRS > 2) were lower in the SI-SU group compared with the CC group (15.6% vs. 27.0%, P < .05; 33.9% vs. 39.8%, P = .16, respectively). In multivariate analysis, AIS patients (n = 453) treated under SI-SU had significantly lower 1-year mortality and disability compared with the CC-treated patients (odds ratio [OR]: 0.47, 95% confidence interval [CI]: 0.27–0.83 and OR: 0.44, 95% CI: 0.22–0.87; respectively). For TIA patients, (n = 262) SI-SU care showed no significant effect in any outcome variable. Our prospective study provides evidence that SI-SU with advanced early monitoring and treatment for patients suffering from AIS results in a better outcome 1 year after ischemic stroke if compared with conventional care. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The individual, social, and economic consequences of stroke impose a considerable burden on patients, payers, and society in terms of premature death, long-term disability, restricted social functioning, costs of care, lost productivity, and informal caregiver time [1–7]. Metaanalyses, [8–10] as well as further individual studies, show that organized stroke care with a strong focus on early rehabilitation, and with specially trained and motivated staff, favorably influences the outcomes of stroke patients with regard to survival rate, degree of disability, and the ability to lead an independent life [11–15].
Abbreviations: AIS, acute ischemic stroke; CC, conventional care; DGN, German Society of Neurology; ESO, European Stroke Organization; MDK, Medical advisory board of statutory health insurance of Schleswig-Holstein; mRS, modified Rankin Scale; QuSS, quality assurance program for stroke treatment in Schleswig-Holstein; SI–SU, semiintensive stroke unit; TIA, transient ischemic attack. ⁎ Corresponding authors. Seidel, is to be contacted at Tel.: +49 451 5000; fax: +49 451 500 2489. Walter, Kassenaerztliche Bundesvereinigung, Herbert-Lewin-Platz 2, 10623 Berlin, Germany. Tel.: +49 30 4005 1223; fax: +49 30 4005 27 1223. E-mail addresses: [email protected] (A. Walter), [email protected] (G. Seidel). 0022-510X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2009.08.010
The German Society for Neurology (DGN) has recommended [16–19] that stroke units follow a neurological-intensive-care concept for specialized treatment of stroke patients [20,21]. This concept contains the entire European Stroke Organization (ESO) recommended minimum requirements for centers managing acute stroke patients and the most of the recommended additional facilities [21,22]. Until now, only a preliminary evaluation of the effectiveness of the DGN's stroke-unit concept has been conducted [22,23]. Our observational study was part of a quality assurance program for stroke treatment (QuSS). The aim of the study was to compare medical conventional care (CC) and semi-intensive stroke units (SISUs) regarding mortality rate and the proportion of higher degree disabilities within 1 year of symptom onset in patients with transient ischemic attack (TIA) or acute ischemic stroke (AIS). 2. Subjects and methods 2.1. Study design The federal state of Schleswig-Holstein has approximately 2.8 million inhabitants and is located in the northernmost part of Germany. An
132
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
estimated stroke incidence of 200 per 100,000 inhabitants would result in about 5600 stroke patients per annum [24]. Of the 13 hospitals involved in QuSS, 6 clinics were chosen as candidates for our study because they represented the pure concept of the semi-intensive stroke unit (n = 2) or the conventional medical standard care (n = 4). Because the catchment areas of the 6 hospitals do not overlap, it is highly likely that the regional stroke incidence is more or less completely drained by the respective departments. Group membership is thus mainly determined by place of residence. The 2 neurological semiintensive stroke units cover areas with about 220,000 and 35,000 residents, respectively; the 4 internal medicine departments cover areas with 10,000, 16,000, 17,000, and 21,000 residents, respectively. The study follows a quasi-experimental design with 2 parallel consecutive cohorts—one with stroke victims from 2 semi-intensive stroke units, the other with patients from 4 internal medicine departments without stroke units. Our general hypothesis states that patients with TIA or AIS experience less mortality and severe disability (modified Rankin Scale [mRS] 3–5) 1 year after stroke under SI-SU treatment than under CC treatment. 2.2. Data collection All patients over the age of 18 years suffering from AIS or TIA were recruited in the 6 participating hospitals during the study period from March 2001 to October 2003. Inclusion criteria were AIS or TIA according to the World Health Organization definition [25] and admission to the hospital within the first 24 h after symptom onset. Patients' main place of residence had to be in Schleswig-Holstein. Exclusion criteria were acute stroke other than ischemic or refusal of study participation. One year (± 4 weeks) after AIS or TIA, the surviving patients were visited and questioned in their current homes by the medical service of the statutory health insurance of Schleswig-Holstein (MDK). This careful follow-up procedure guarantee a high data quality. Data collection and documentation followed a uniform study manual. The following data were recorded (Tables 1 and 2): sociodemographic characteristics (age, gender), vascular risk factors, neurological status (type of paresis, dysphagia, dysphasia, or dysarthria), degree of disability according to the mRS [26,27] before stroke onset and at admission, modalities of admission, length of hospital stay, diagnostic and therapeutic procedures, time interval between admission and diagnostic procedures, treatment complications, and clinical stroke classification according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) protocol [28,29]. The data collected in the context of visiting, observing, and interviewing surviving stroke patients after 12 months again included disability according to the mRS. All data were checked for completeness and consistency by a medical doctor not involved in the treatment of the patients. 2.3. Characterization of treatment concepts 2.3.1. Semi-intensive Stroke Unit (SI-SU) The SI-SUs were located in the department of neurology of a university hospital and a department of neurology in a general hospital. The two units work in accordance with the DGN stroke-unit concept [16,22]. This specifies that a patient with an acute stroke or TIA should be transferred to a hospital as quickly as possible for rapid evaluation of stroke etiology. In the SI-SU, an interdisciplinary cooperation among neurologists, internists, and neuroradiologists is mandatory for an immediate start of a pathophysiology- and evidence-based therapy. Various techniques such as cranial computed tomography [CCT], high standard ultrasound examinations of the brain-supplying arteries, electrocardiogram [ECG], and echocardiography are seen as indispensable to reaching that goal. Continuous monitoring of blood pressure, heart rhythm, blood sugar, oxygen saturation, fluid balance, and neurological status should be carried out during the first 24 to 48 h
Table 1 Baseline characteristics.
Female Age, mean [SD] 18–74 y 75–84 y >85 y TIA* Ischemic stroke Paresis Dysphagia Dysarthria/dysphasia
Stroke units
Conventional care
n
%
n
%
P
162 68.1 265 100 28 127 266 283 88 243
41.2 [12.5] 67.4 25.4 7.1 32.3 67.7 72.6 22.5 62.1
210 77.1 127 137 98 145 217 240 46 219
58.0 [10.2] 35.1 37.8 27.1 40.1 59.9 67.0 12.9 61.5
<.001 <.001 <.001
Mod. Rankin Scale before stroke onset 0–1 329 2–3 49 4–5 15 Mod. Rankin Scale on admission 0–1 2–3 4–5 Barthel-ADL-Index, mean [SD] Independent/low dependent (75–100) Moderate/high dependent (0–70) Comorbidity/risk factors Hypertension Prior stroke Atrial fibrilation Other cardial disease Hyperlipidaemia Current smoker Diabetes mellitus
114 90 189 56.2 166
83.7 12.5 3.8
29.0 22.9 48.1 [37.3] 42.3
283 43 36
115 76 171 55.6 151
.030 .100 .001 .860
78.2 11.9 9.9
.004
31.8 21.0 47.2 [38.2] 41.7
.670
226
57.7
211
58.3
278 77 82 124 111 113 99
70.7 19.6 20.9 31.6 28.2 28.8 25.2
290 79 120 90 103 53 99
80.1 21.8 33.1 24.9 28.5 14.6 27.3
.830 .860
.003 .450 <.001 .040 .950 <.001 .500
*TIA indicates transient ischemic attack.
after stroke onset. If necessary a specialized medical and nursing team should implement physiotherapy immediately and both logopedic therapy and ergotherapy as early as possible. Therefore the SI-SUs contain the entire from the European Stroke Initiative recommended minimum requirements for centers managing acute stroke patients and the most of the recommended additional facilities [21]. 2.3.2. Conventional medical care (CC) The four departments of internal medicine representing the CC treatment were located in general hospitals. After admission to the emergency ward, the stroke patient, depending on stroke severity, is either transferred to a general ward or remains in intensive care for initial stabilization and observation. No specialized teams for stroke patient treatment and care were involved. Outside ICU monitoring focused on heart rate and blood pressure measurement. Magnetic resonance tomography (MRT) was not available, and CCT could not be provided on a 24-hour basis. Systemic or local thrombolysis was not possible. Departments of neurology, vascular surgery, or neurosurgery were not available in the hospitals. Corresponding interdisciplinary advice could be guaranteed only on working days during daytime hours by an employed specialist in the hospital or by external services. Four clinics were requested to agree with a uniform medical and nursing protocol. Based on a questionnaire, their infrastructure and procedures were systematically documented before the start of the data-collection phase. 2.4. Data processing and statistics Data were processed using Microsoft Access and Lotus Approach and analyzed with SPSS for PC, version 11.0. To test our primary hypotheses, r x c tables were constructed for chi-square tests (α ≤5%). Continuous variables such as age, Rankin Scale, and the Barthel index of the activities
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
3. Results
Table 2 Characterization of treatment processes.
Referral by General practitioner Emergency physician Discharge Inpatient (other department) Geriatrics Rehabilitation Nursing home At home
3.1. Patients included
Stroke units
Conventional care
n
%
n
%
P
103 146
27.2 38.6
143 98
43.1 29.5
<.001
19
7.0
19
5.9
<.001
41 57 15 140
15.1 21.0 5.5 51.5
12 95 27 167
3.8 29.7 8.4 52.2
Time stroke onset to admission <3 h 175 Length of hospital stay, mean [d] (SD)
44.5 142 11.22 (5.6)
39.2 13.32 (7.3)
12 330
3.1 85.1
46 117
12.8 32.5
<.001
Extracranial Doppler/duplex none 7 <3 h 220
1.8 56.3
79 30
22.1 8.4
<.001
Transcranial Doppler/duplex none 6 <3 h 222 TTE done 68 TEE done 72 Swallowing assessment 127 Thrombolysis 14
1.5 56.6 18.6 19.8 33.9 3.6
308 11 243 57 111 1
87.7 3.1 69.4 16.4 32.2 0.3
<.001
Physical therapy none Start <24 h
21.5 71.3
124 125
35.2 35.5
<.001
126 2 29 1 30 53 167
59.2 0.9 13.6 0.5 14.1 24.9 78.4
.051 1.000 <.001 .447 .495 .601 .102
CCT none <3 h
84 278
Secondary prevention 1 year after ASS 132 Ticlopidin 2 Clopidogrel 72 Heparine 0 Oral anticoagulation 43 Lipid lowering drugs 71 Antihypertensive drugs 189
stroke onset 50.2 0.8 27.4 0.0 16.3 27.0 71.9
133
.140 .001
<.001 .232 .630 .001
A total of 1484 patients were screened (Fig. 1). Of these, 612 fulfilled the exclusion criteria. For those that were excluded, 255 patients (41.7%) had a time period between symptom onset and hospital admission that exceeded 24 hours, 105 patients (17.2%) had an initial diagnosis of AIS or TIA that was not supported after the complete diagnostic workup, 100 patients (16.3%) had their residence outside of Schleswig-Holstein, 61 patients (10.0%) rejected participation in the study, and 77 patients (12.6%) had a non-ischemic stroke. Another 43 patients (7.0%) were later excluded because of a late refusal or missing written consent. Of the remaining 872 participants, a further 88 patients could not be included in the evaluation because of documentation error. In total, 755 patients were included in the evaluation of the patients' characteristics and treatment processes. These patients were followed up 12 months (± 4 weeks) after AIS or TIA. One month before the planned follow-up, the MDK announced a home visit by physicians or nursing staff. In case of missing feedback, patients were reminded by a phone call a few days before the visit date. When follow-up was refused, an attempt was made to collect data by telephone. Where this was not possible (ill health, complete denial), only the vital status was documented. If a patient was not at home but apparently still alive, another visit was planned 1 to 2 months later. Otherwise, the patient's current address or the date of death was obtained from the respective resident's registration office. Of the 703 contacted patients, 480 could be interviewed personally. The average interval of follow-up was 381 days (SD 48.6). In the period between the end of the AIS or TIA treatment and the planned home visit, 98 patients died. Follow-up could be conducted in 40 cases because of missing information; 52 patients refused an interview. Despite inquiry at the residents' registration office, 27 patients could not be reached. Meanwhile, 6 patients lived outside of Schleswig-Holstein. 3.2. Patient characteristics
CCT indicates cerebral computed tomography; TTE, transthoracal echocardiography; TEE, transesophageal echocardiography; ASS, acetylsalicylate.
of daily living (ADL) were di- or polychotomized using appropriate cutoffs (groups with same size respectively clinical criteria). Intervalscaled variables were analyzed using the t test/Analysis of Variance (ANOVA) test or the U-test/Kruskal–Wallis test as appropriate. To control for possibly confounding variables, multivariate analyses were performed (logistic regression for dichotomous outcomes, Kaplan– Meier estimates with log-rank tests, or Cox regression for time-dependent data). Subgroup analyses were based on age (3 groups), sex, and AIS/TIA status. Our sample size calculation assumed a 16% 1-year mortality rate among patients treated in SI-SU and 32% for patients under CC treatment. Using α = 5% (two-sided), 123 patients per group are required to identify a difference as statistically significant.
At admission, demographic characteristics and health status differed between the 2 groups (Table 1). Significant differences existed with regard to sex, age, diagnosis (TIA/AIS), dysphagia, degree of disability before the event, and the presence of risk factors such as hypertension, atrial fibrillation, other cardiac diseases, and smoking status. 3.3. Characterization of the treatment processes SI-SU and CC patients were admitted by a general practitioner in 27% and 43% of the cases, respectively, and by an emergency physician in 39% and 30% of cases, respectively (Table 2). The rate of admission within 3 hours after symptom onset was higher in the SI-SU group than in the CC group (45% vs. 39%). A CCT was carried out more often for SI-SU patients (97% vs. 87%) and more frequently within 3 h (85% vs. 33%). A similar situation was found with the ultrasound examination of the brain-supplying arteries and with physiotherapy. Only transthoracic echocardiography was carried out more often in the CC hospitals (69% vs. 19%). This was not the case for transoesophageal echocardiography (16.4 % vs. 19.8%). At the time of the follow-up observation 1 year after stroke, no significant differences between the SI-SU and CC patients regarding relevant groups of secondary prevention strategies (Table 2) were measurable. 3.4. Outcomes
2.5. Data protection, ethics The Ethics Committee of the Medical Faculty of the University of Schleswig-Holstein, Campus Lübeck examined and approved the study protocol (May 15, 2001).
3.4.1. All patients Univariate analyses (n[mortality] = 715, n[Rankin] = 480) show a 1-year mortality in the SI-SU group of 16.3% (n = 60) and, in the CC group, of 26.0% (n = 90, P = .001). The percentage of surviving patients with a
134
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
Fig. 1. Participation flow chart.
high disability rate (Rankin Scale 3–5) at 12 months was 32.1% for SI-SU patients (n = 85) and 40.5% for CC patients (n = 87; P = .057). As shown in Table 1, a complete model combining all relevant factors was chosen for the multivariate analysis. The factors included were chosen with respect to their clinical relevance, measurability, and distribution. The model was retained for all outcomes and subgroup analyses. The following factors significantly influence mortality after 1 year (Fig. 2): old age, paresis, dysphagia, and a high degree of disability (Rankin 4–5) before admission or during hospital stay. Atrial fibrillation and other cardiac diseases significantly increase the risk of dying within 1 year after stroke. Controlling for these factors, we found that strokeunit care does not significantly reduce the 1-year mortality risk (odds ratio [OR] 0.67, 95% CI 0.41–1.10) in patients with AIS or TIA. The following independent factors were shown to increase the risk of moderate to very severe disability (Rankin 3–5) of the surviving patients 1 year after stroke: very old age (>84 years), a high degree of disability at admission, paresis, dysarthria or dysphasia, and previous stroke. Treatment of AIS or TIA patients in the SI-SU significantly reduced the risk of disability, irrespective of the other factors (OR 0.53, 95% CI 0.31–0.90) (Fig. 2). 3.4.2. Exploratory subgroup analyses Multivariate analysis of subgroups with respect to differential effects of SI-SU care on 1-year mortality shows that patients with ischemic stroke (OR 0.47; 95% CI 0.27–0.83), male patients (OR 0.34; 95% CI 0.16–
0.72), and patients 85 years of age or older (OR 0.16; 95% CI 0.03–0.79) benefit significantly from treatment in SI-SU as compared with CC care. In comparison with conventional care (Fig. 3), patients with TIA who were treated in SI-SU have a nonsignificant higher risk of dying within 1 year after incident (95% CI 0.9–8.24). SI-SU care showed a significant beneficial effect on disability level 1 year after stroke in patients with ischemic strokes (OR 0.44; 95% CI 0.22–0.87) and in female patients (OR 0.42; 95% CI 0.20–0.91). Adjusted (for sex, age, paresis, dysphagia, dysarthria/dysphasia, mRS before stroke onset and on admission, and comorbidity) survival curves show a significantly lower mortality for ischemic stroke patients treated in SI-SU. A difference in mortality rates between the groups of about 10% was reached after approximately 2 to 3 months (Fig. 4).
4. Discussion For the first time in Germany, this study compared the rates of survival and severe disability 1 year after acute cerebral ischemia onset between patients treated under the SI-SU concept of the DGN/ESO or standard conventional care. We were able to show that the SI-SU concept implied a reduction of disability in patients with an ischemic stroke or TIA. With respect to 1-year mortality, our data show a significant advantage of SI-SU care for patients with acute ischemic strokes but not for TIA patients.
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
135
Fig. 2. Adjusted regression analysis of factors influencing death and disability 1 year after stroke onset (adjusted for influence factors shown in Table 1 with the exception of the Barthel–ADL-Index).
Contrary to the traditional rehabilitative-stroke-unit concept, [30,31] the SI-SU concept is characterized by both a multidisciplinary and semi-intensive care emphasis in diagnosis, monitoring, treatment, and early rehabilitation. Process descriptions show that the concept was put into practice. For the CC therapy, our crude rate for 1-year mortality (26.0%) matches those from other studies. These are summarized in a recent Cochrane Review from the Stroke Units Trialists' Collaboration [9] and
resulted in a standardized mortality rate of 25.2% for general medical wards after 1 year. In our study, the mortality rate of the SI-SU patients was substantially lower (16.3%) than the Cochrane Review's rate, with a weighted rate of 22.2%. However, results from individual studies varied between 5.7% and 45.2% for groups of between 29 and 279 cases. In our study, we refrained from using a composite outcome. In several international studies, [14,30,32,33] mortality was combined with either severe disability or level of dependence and institutional
Fig. 3. Adjusted regression analysis of SI-SU care compared to CC treatment* regarding death and disability 1 year after stroke onset (adjusted for influence factors shown in Table 1 with the exception of the Barthel-ADL-Index).
136
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
Fig. 4. Adjusted survival curve* for ischemic stroke patients treated in the CC or SI-SU concept (adjusted for influence factors shown in Table 1 with the exception of the BarthelADL-Index).
beginning of physiotherapy and logopedic therapy, and the use of gastric tubes. Between the CC clinics, there was a great variability in the use of almost all diagnostic measures and physiotherapy, as well as logopedics. The diagnostic and therapeutic profile seemed to depend more on the availability of resources than on a unified concept or existent guidelines [42]. Therefore, caution is required when speaking of the SI-SU and the CC care. Despite some limitations, this study showed that the DGN strokeunit concept, which contain the ESO recommendations for stroke centers [21], offer a beneficial effect with respect to 1-year mortality and disability, as well as length of hospital stay, in patients suffering from ischemic stroke. Our results are in line with a recent metaanalysis comparing models of inpatient stroke care [43] and strength the notion that the beneficial effect of acute stroke unit treatment on outcome is statistically significant. These findings support the notion that stroke care should be focused on larger centres. Further studies are necessary to evaluate the effects of the SI-SU on subgroups of TIA patients. Acknowledgments and Funding
care. If we had combined mortality and disability (Barthel index 0–70), our study would have resulted in a significant benefit for all SI-SU patients (OR 0.55; 95% CI 0.35–0.87). An extension of our regression model to thrombolysis, led to no substantial change of the results. Concerning mortality of ischemic stroke patients the beneficial effect of the SI-SU-treatment did not change (OR 0.48; 95% CI 0.27–0.85) and thrombolysis did not show a significant influence on outcome (OR 0.70; 95% CI 0.14–3.42). The consideration of this item violates our approach of the program evaluation. Nevertheless we conducted this analysis, since the thrombolysis represents a substantial element of the SI-SU-treatment. The effects of the SI-SU cannot obvious be reduced alone to thrombolysis. Others elements of semi intensive care were not evaluated because this study focus on the outcome effects of the whole concept and not on isolated factors. Our study has several limitations. Participating clinics were to report every suspected stroke to the coordinating center. Based on routine administrative data (Section 301 of the German Social Security Code V) from each hospital, we could estimate the proportion of all stroke victims that were successfully recruited. It varied between 28% and 64% of all eligible patients for the CC group. For the 2 SI-SUs, it amounted to 63% and 100%, respectively. This may have led to a certain and, unfortunately, uncontrollable degree of selection bias. Follow-up information was not available from all patients. Mortality and disability data were missing in 5.3% (n = 40) and 11.3% (n = 85) of all patients, respectively. So we performed a best-worst case scenario analysis. In the worst case scenario for AIS patients the beneficial effects of SI-SU was not longer present (outcome 1 year after stroke onset for mortality: OR 0.639, 95% CI 0.39–1.06 and for disability OR 0.61, 95% CI 0.34–1.11). The evidence for TIA patients for both conditions (worse and best) and the beneficial effect for AIS patients in the best condition scenario did not change. Because of organizational reasons, we were not able to conduct a randomized-controlled trial [32,34–40]. Furthermore, knowing the advantages of stroke units as described in existing studies, it appears to be ethically rather questionable whether to admit patients to what is presumably a less effective facility merely on the basis of randomization [41]. Therefore, we used a quasi-experimental approach capitalizing on 2 concurrent treatment concepts associated with different hospitals with non-overlapping catchment areas. The two groups of participating hospitals were requested to coordinate the key points of their therapeutic approach. However, an analysis of the process data revealed significant differences in certain areas within the groups. Between the stroke units, these included, in particular, type of aftercare (geriatrics vs. rehabilitation), ultrasound examinations of the brain-supplying arteries, utilization of MRT, the
The quality assurance program for stroke treatment (QuSS) and this embedded study were financed by the federal government, the statutory health insurance, and the medical advisory board of the statutory health insurance of Schleswig-Holstein (MDK). The following departments participated in the QuSS program: Department of Neurology, University of Schleswig-Holstein, Campus Lübeck (head: Professor Dr. Kömpf, MD); Clinic for Neurology at the Medical Center in Itzehoe (head: Prof. Dr. Thie, MD); Department of Internal Medicine at the Ostholstein Medical Centers in Oldenburg/Holstein (head: Dr. Jakobeit, MD) and Eutin (head: Dr. Gützkow, MD); Department of Internal Medicine at the Husum Clinic (head: Dr. Haacke, MD); and the Department of Internal Medicine at the District Hospital of Plön in Preetz (head: Dr. Holst, MD); Department of Internal Medicine at the Malteser Hospital, St. Franziskus, Flensburg (head: Professor Dr. Saal, MD); Clinic for Neurosurgery (head: Dr. Schmidt, MD) and Internal Medicine (head: Professor Dr. Machraoui, MD), Diakonissen Hospital, Flensburg; Clinic for Internal Medicine at the District Hospital of Rendsburg (head: Professor Dr. Herrlinger, MD); Clinic for Internal Medicine (head: Dr. Keck, MD) and Rehablitation and Geriatric Medicine (head: Dr. Stamm, MD) at the Medical Center in Heide (Westküstenklinikum); Clinic for Internal Medicine (head: Professor Dr. Bethge, MD), Clinic for Geriatric Medicine (head: Dr. Naurath, MD) and Clinic for Neurology and Psychiatry at the Medical Center in Neumünster (head: Professor Dr. Hansen, MD), Friedrich-Ebert Krankenhaus; Department of Neurology, University of SchleswigHolstein, Campus Kiel (head: Professor Dr. Deuschl, MD); Department of Neurology, Clinic for Psychiatry, Neurology und Rehabilitation, Schleswig (head: Dr. Leonhardt, MD). The authors thank the medical advisory board of statutory health insurance of Schleswig-Holstein (MDK) for supporting the QuSS program and especially Dr. Buss, MD and Dr. Erben, PhD, for their assistance in preparing the study files and the statistical workup. References [1] Evers SM, Struijs JN, Ament AJ, van Genugten ML, Jager JH, van den Bos GA. International comparison of stroke cost studies. Stroke 2004;35:1209–15. [2] Feinberg WM. Primary and secondary stroke prevention. Curr Opin Neurol 1996;9: 46–52. [3] Ghatnekar O, Persson U, Glader EL, Terent A. Cost of stroke in Sweden: an incidence estimate. Int J Technol Assess Health Care 2004;20:375–80. [4] Humphrey PR. Management of transient ischaemic attacks and stroke. Postgrad Med J 1995;71:577–84. [5] Taylor TN, Davis PH, Torner JC, Holmes J, Meyer JW, Jacobson MF. Lifetime cost of stroke in the United States. Stroke 1996;27:1459–66. [6] Weimar C, Lüngen M, Wagner M, Kraywinkel K, Evers T, Busse O. Cost of stroke care in Germany. An analysis of the stroke data bank of the German Foundation Stroke-Aid. Aktuelle Neurologie 2002;29:181–90.
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137 [7] Kolominsky-Rabas PL, Heuschmann PU, Marschall D, Emmert M, Baltzer N, Neundorfer B, Schoffski O, Krobot KJ. Lifetime cost of ischemic stroke in Germany: results and national projections from a population-based stroke registry: the Erlangen Stroke Project. Stroke 2006;37:1179–83. [8] Collaborative systematic review of the randomised trials of organised inpatient (stroke unit) care after stroke. Stroke Unit Trialists' Collaboration. BMJ 1997;314:1151–59. [9] Organised inpatient (stroke unit) care for stroke. Stroke Unit Trialists' Collaboration. Cochrane Database Syst Rev. 2004. [10] Langhorne P, Williams BO, Gilchrist W, Howie K. Do stroke units save lives? Lancet 1993;342:395–8. [11] Agency for Health Care Policy and Research (AHCPR). “Post Stroke Rehabilitation”. [No. 16]. 1995. AHCPR Publication No. 95-0062. Clinical Practice guideline. [12] Fritze J. Die Evaluation von Stroke Units als medizinische Technologie. Schwartz FW, Köbberling J, Raspe HH, editors. [15]. 1999. Nomos Verlagsgesellschaft BadenBaden. Health Technology Assessment. [13] Schroeder A, Heiderhoff M, Köbberling J. Stroke Units: Update des HTA-Berichts “Die Evaluation von Stroke Units als medizinische Technologie”. Deutsche Agentur für Health Technology Assessment des Deutschen Instituts für Medizinische Dokumentation und Information, editor. [Band 11]. 2004. Health Technology Assessment. [14] Candelise L, Gattinoni M, Bersano A, Micieli G, Sterzi R, Morabito A, PROSIT Study Group. Stroke-unit care for acute stroke patients: an observational follow-up study. Lancet 2007;369:299–305. [15] Scottish Intercollegiate Guidelines Network (SIGN). National Clinical Guideline “Management of Patients with Stroke, Part I-IV” recommend for use in Scotland by the Scottish Intercollegiate Guidelines Network (SIGN). 1998. [16] Berlit P, Busse O, Diener C, Ringelstein EB. Guidelines for the implementation of stroke units from the Committee on Stroke Units of the Deutsche Gesellschaft für Neurologie. Aktuelle Neurologie 1996;23:171–5. [17] Berlit P, Popescu O, Klotzsch C, Diehl RR, Berg-Dammer E. Treatment of acute stroke on the stroke unit. Initial experiences with an acute stroke unit in Germany. Nervenarzt 1997;68:122–8. [18] Ringelstein EB. Recommendation for organization of special stroke departments (stroke units). 1997 implementation. Nervenarzt 1998;69:180–5. [19] Ringelstein EB. Concept of interregional and regional stroke treatment in Germany. Report of the recommendation of Commission 1.06 “Stroke Units and Acute Stroke Therapy” of the German Society for Neurology. Nervenarzt 2000;71:322–4. [20] European Stroke Initiative recommendations for stroke management. European Stroke Council, European Neurological Society and European Federation of Neurological Societies. Cerebrovasc Dis 2000;10:335–51. [21] European Stroke Organisation (ESO) Executive Committee. ESO Writing Committee. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25:457–507. [22] Busse O. Stroke units and stroke services in Germany. Cerebrovasc Dis 2003;15 (Suppl 1):8–10. [23] Sachverständigenrat für die Konzertierte Aktion im Gesundheitswesen. Bedarfsgerechtigkeit und Wirtschaftlichkeit. Band III: Über-, Unter- und Fehlversorgung. 2001. Gutachten 2000/2001.
137
[24] Kolominsky-Rabas PL, Sarti C, Heuschmann PU, Graf C, Siemonsen S, Neundoerfer B, Katalinic A, Lang E, Gassmann KG, von Stockert TR. A prospective communitybased study of stroke in Germany—the Erlangen Stroke Project (ESPro): incidence and case fatality at 1, 3, and 12 months. Stroke 1998;29:2501–6. [25] Stroke–1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989;20:1407-31. [26] Sulter G, Steen C, De Keyser J. Use of the Barthel index and modified Rankin scale in acute stroke trials. Stroke 1999;30:1538–41. [27] van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988;19:604–7. [28] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, Marsh III EE. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. [29] Ringelstein EB, Zeumer H, Schneider R. Contribution of computer tomography of the brain to differential typology and differential therapy of ischemic cerebral infarct. Fortschr Neurol Psychiatr 1985;53:315–36. [30] Kalra L, Evans A, Perez I, Knapp M, Donaldson N, Swift CG. Alternative strategies for stroke care: a prospective randomised controlled trial. Lancet 2000;356:894–9. [31] Ronning OM, Guldvog B. Stroke units versus general medical wards, I: twelve- and eighteen-month survival: a randomized, controlled trial. Stroke 1998;29:58–62. [32] Silva Y, Puigdemont M, Castellanos M, Serena J, Suner RM, Garcia MM, Davalos A. Semi-intensive monitoring in acute stroke and long-term outcome. Cerebrovasc Dis 2005;19:23–30. [33] Cavallini A, Micieli G, Marcheselli S, Quaglini S. Role of monitoring in management of acute ischemic stroke patients. Stroke 2003;34:2599–603. [34] Sulter G, Elting JW, Langedijk M, Maurits NM, De Keyser J. Admitting acute ischemic stroke patients to a stroke care monitoring unit versus a conventional stroke unit: a randomized pilot study. Stroke 2003;34:101–4. [35] Garraway WM, Akhtar AJ, Hockey L, Prescott RJ. Management of acute stroke in the elderly: follow-up of a controlled trial. Br Med J 1980;281:827–9. [36] Kaste M, Palomaki H, Sarna S. Where and how should elderly stroke patients be treated? A randomized trial. Stroke 1995;26:249–53. [37] Sivenius J, Pyorala K, Heinonen OP, Salonen JT, Riekkinen P. The significance of intensity of rehabilitation of stroke–a controlled trial. Stroke 1985;16:928–31. [38] Wood-Dauphinee S, Shapiro S, Bass E, Fletcher C, Georges P, Hensby V, Mendelsohn B. A randomized trial of team care following stroke. Stroke 1984;15:864–72. [39] Lincoln NB, Willis D, Philips SA, Juby LC, Berman P. Comparison of rehabilitation practice on hospital wards for stroke patients. Stroke 1996;27:18–23. [40] Indredavik B, Bakke F, Solberg R, Rokseth R, Haaheim LL, Holme I. Benefit of a stroke unit: a randomized controlled trial. Stroke 1991;22:1026–31. [41] Hacke W. A late step in the right direction of stroke care. Lancet 2000;356:869–70. [42] Dodel RC, Haacke C, Zamzow K, Paweilik S, Spottke A, Rethfeldt M, Siebert U, Oertel WH, Schoffski O, Back T. Resource utilization and costs of stroke unit care in Germany. Value Health 2004;7:144–52. [43] Foley N, Salter K, Teasell R. Specialized stroke services: a meta-analysis comparing three models of care. Cerebrovasc Dis 2007;23:194–202.
Contents lists available at ScienceDirect
Journal of the Neurological Sciences j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / j n s
Semi-intensive stroke unit versus conventional care in acute ischemic stroke or TIA — A prospective study in Germany Andreas Walter a,⁎, Guenter Seidel b,⁎, Andreas Thie c, Heiner Raspe d a
National Association of Statutory Health Insurance Physicians, Berlin, Germany Department of Neurology, University of Schleswig-Holstein, Campus Luebeck, Luebeck, Germany Neurology Service, Klinikum Itzehoe, Itzehoe, Germany d Institute for Social Medicine University of Schleswig-Holstein, Campus Luebeck, Luebeck, Germany b c
a r t i c l e
i n f o
Article history: Received 27 April 2009 Received in revised form 6 August 2009 Accepted 10 August 2009 Available online 6 September 2009 Keywords: Stroke units Acute stroke care Semi-intensive monitoring Outcome research
a b s t r a c t German Society of Neurology's stroke-unit concept includes a specialized stroke-unit team and advanced monitoring facilities in the early phase of stroke. Our aim was to evaluate the effectiveness of this semiintensive stroke-unit (SI-SU) concept as compared with conventional care (CC) for patients with acute ischemic stroke (AIS) or transient ischemic attack (TIA). Over a 20-month period starting in March 2001, 755 patients with AIS or TIA were treated under SI-SU (n = 393) or CC (n = 362) conditions within an observational study. All patients were admitted within 24 h after symptom onset. Patients' outcomes were assessed after 1 year of follow-up, considering concept of care, patients' clinical characteristics, and comorbid conditions. The CC patients were older and had a greater frequency of atrial fibrillation. 1 year after ischemia, 715 patients (94.7%) could be reinvestigated. Mortality and disability (mRS > 2) were lower in the SI-SU group compared with the CC group (15.6% vs. 27.0%, P < .05; 33.9% vs. 39.8%, P = .16, respectively). In multivariate analysis, AIS patients (n = 453) treated under SI-SU had significantly lower 1-year mortality and disability compared with the CC-treated patients (odds ratio [OR]: 0.47, 95% confidence interval [CI]: 0.27–0.83 and OR: 0.44, 95% CI: 0.22–0.87; respectively). For TIA patients, (n = 262) SI-SU care showed no significant effect in any outcome variable. Our prospective study provides evidence that SI-SU with advanced early monitoring and treatment for patients suffering from AIS results in a better outcome 1 year after ischemic stroke if compared with conventional care. © 2009 Elsevier B.V. All rights reserved.
1. Introduction The individual, social, and economic consequences of stroke impose a considerable burden on patients, payers, and society in terms of premature death, long-term disability, restricted social functioning, costs of care, lost productivity, and informal caregiver time [1–7]. Metaanalyses, [8–10] as well as further individual studies, show that organized stroke care with a strong focus on early rehabilitation, and with specially trained and motivated staff, favorably influences the outcomes of stroke patients with regard to survival rate, degree of disability, and the ability to lead an independent life [11–15].
Abbreviations: AIS, acute ischemic stroke; CC, conventional care; DGN, German Society of Neurology; ESO, European Stroke Organization; MDK, Medical advisory board of statutory health insurance of Schleswig-Holstein; mRS, modified Rankin Scale; QuSS, quality assurance program for stroke treatment in Schleswig-Holstein; SI–SU, semiintensive stroke unit; TIA, transient ischemic attack. ⁎ Corresponding authors. Seidel, is to be contacted at Tel.: +49 451 5000; fax: +49 451 500 2489. Walter, Kassenaerztliche Bundesvereinigung, Herbert-Lewin-Platz 2, 10623 Berlin, Germany. Tel.: +49 30 4005 1223; fax: +49 30 4005 27 1223. E-mail addresses: [email protected] (A. Walter), [email protected] (G. Seidel). 0022-510X/$ – see front matter © 2009 Elsevier B.V. All rights reserved. doi:10.1016/j.jns.2009.08.010
The German Society for Neurology (DGN) has recommended [16–19] that stroke units follow a neurological-intensive-care concept for specialized treatment of stroke patients [20,21]. This concept contains the entire European Stroke Organization (ESO) recommended minimum requirements for centers managing acute stroke patients and the most of the recommended additional facilities [21,22]. Until now, only a preliminary evaluation of the effectiveness of the DGN's stroke-unit concept has been conducted [22,23]. Our observational study was part of a quality assurance program for stroke treatment (QuSS). The aim of the study was to compare medical conventional care (CC) and semi-intensive stroke units (SISUs) regarding mortality rate and the proportion of higher degree disabilities within 1 year of symptom onset in patients with transient ischemic attack (TIA) or acute ischemic stroke (AIS). 2. Subjects and methods 2.1. Study design The federal state of Schleswig-Holstein has approximately 2.8 million inhabitants and is located in the northernmost part of Germany. An
132
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
estimated stroke incidence of 200 per 100,000 inhabitants would result in about 5600 stroke patients per annum [24]. Of the 13 hospitals involved in QuSS, 6 clinics were chosen as candidates for our study because they represented the pure concept of the semi-intensive stroke unit (n = 2) or the conventional medical standard care (n = 4). Because the catchment areas of the 6 hospitals do not overlap, it is highly likely that the regional stroke incidence is more or less completely drained by the respective departments. Group membership is thus mainly determined by place of residence. The 2 neurological semiintensive stroke units cover areas with about 220,000 and 35,000 residents, respectively; the 4 internal medicine departments cover areas with 10,000, 16,000, 17,000, and 21,000 residents, respectively. The study follows a quasi-experimental design with 2 parallel consecutive cohorts—one with stroke victims from 2 semi-intensive stroke units, the other with patients from 4 internal medicine departments without stroke units. Our general hypothesis states that patients with TIA or AIS experience less mortality and severe disability (modified Rankin Scale [mRS] 3–5) 1 year after stroke under SI-SU treatment than under CC treatment. 2.2. Data collection All patients over the age of 18 years suffering from AIS or TIA were recruited in the 6 participating hospitals during the study period from March 2001 to October 2003. Inclusion criteria were AIS or TIA according to the World Health Organization definition [25] and admission to the hospital within the first 24 h after symptom onset. Patients' main place of residence had to be in Schleswig-Holstein. Exclusion criteria were acute stroke other than ischemic or refusal of study participation. One year (± 4 weeks) after AIS or TIA, the surviving patients were visited and questioned in their current homes by the medical service of the statutory health insurance of Schleswig-Holstein (MDK). This careful follow-up procedure guarantee a high data quality. Data collection and documentation followed a uniform study manual. The following data were recorded (Tables 1 and 2): sociodemographic characteristics (age, gender), vascular risk factors, neurological status (type of paresis, dysphagia, dysphasia, or dysarthria), degree of disability according to the mRS [26,27] before stroke onset and at admission, modalities of admission, length of hospital stay, diagnostic and therapeutic procedures, time interval between admission and diagnostic procedures, treatment complications, and clinical stroke classification according to the Trial of Org 10172 in Acute Stroke Treatment (TOAST) protocol [28,29]. The data collected in the context of visiting, observing, and interviewing surviving stroke patients after 12 months again included disability according to the mRS. All data were checked for completeness and consistency by a medical doctor not involved in the treatment of the patients. 2.3. Characterization of treatment concepts 2.3.1. Semi-intensive Stroke Unit (SI-SU) The SI-SUs were located in the department of neurology of a university hospital and a department of neurology in a general hospital. The two units work in accordance with the DGN stroke-unit concept [16,22]. This specifies that a patient with an acute stroke or TIA should be transferred to a hospital as quickly as possible for rapid evaluation of stroke etiology. In the SI-SU, an interdisciplinary cooperation among neurologists, internists, and neuroradiologists is mandatory for an immediate start of a pathophysiology- and evidence-based therapy. Various techniques such as cranial computed tomography [CCT], high standard ultrasound examinations of the brain-supplying arteries, electrocardiogram [ECG], and echocardiography are seen as indispensable to reaching that goal. Continuous monitoring of blood pressure, heart rhythm, blood sugar, oxygen saturation, fluid balance, and neurological status should be carried out during the first 24 to 48 h
Table 1 Baseline characteristics.
Female Age, mean [SD] 18–74 y 75–84 y >85 y TIA* Ischemic stroke Paresis Dysphagia Dysarthria/dysphasia
Stroke units
Conventional care
n
%
n
%
P
162 68.1 265 100 28 127 266 283 88 243
41.2 [12.5] 67.4 25.4 7.1 32.3 67.7 72.6 22.5 62.1
210 77.1 127 137 98 145 217 240 46 219
58.0 [10.2] 35.1 37.8 27.1 40.1 59.9 67.0 12.9 61.5
<.001 <.001 <.001
Mod. Rankin Scale before stroke onset 0–1 329 2–3 49 4–5 15 Mod. Rankin Scale on admission 0–1 2–3 4–5 Barthel-ADL-Index, mean [SD] Independent/low dependent (75–100) Moderate/high dependent (0–70) Comorbidity/risk factors Hypertension Prior stroke Atrial fibrilation Other cardial disease Hyperlipidaemia Current smoker Diabetes mellitus
114 90 189 56.2 166
83.7 12.5 3.8
29.0 22.9 48.1 [37.3] 42.3
283 43 36
115 76 171 55.6 151
.030 .100 .001 .860
78.2 11.9 9.9
.004
31.8 21.0 47.2 [38.2] 41.7
.670
226
57.7
211
58.3
278 77 82 124 111 113 99
70.7 19.6 20.9 31.6 28.2 28.8 25.2
290 79 120 90 103 53 99
80.1 21.8 33.1 24.9 28.5 14.6 27.3
.830 .860
.003 .450 <.001 .040 .950 <.001 .500
*TIA indicates transient ischemic attack.
after stroke onset. If necessary a specialized medical and nursing team should implement physiotherapy immediately and both logopedic therapy and ergotherapy as early as possible. Therefore the SI-SUs contain the entire from the European Stroke Initiative recommended minimum requirements for centers managing acute stroke patients and the most of the recommended additional facilities [21]. 2.3.2. Conventional medical care (CC) The four departments of internal medicine representing the CC treatment were located in general hospitals. After admission to the emergency ward, the stroke patient, depending on stroke severity, is either transferred to a general ward or remains in intensive care for initial stabilization and observation. No specialized teams for stroke patient treatment and care were involved. Outside ICU monitoring focused on heart rate and blood pressure measurement. Magnetic resonance tomography (MRT) was not available, and CCT could not be provided on a 24-hour basis. Systemic or local thrombolysis was not possible. Departments of neurology, vascular surgery, or neurosurgery were not available in the hospitals. Corresponding interdisciplinary advice could be guaranteed only on working days during daytime hours by an employed specialist in the hospital or by external services. Four clinics were requested to agree with a uniform medical and nursing protocol. Based on a questionnaire, their infrastructure and procedures were systematically documented before the start of the data-collection phase. 2.4. Data processing and statistics Data were processed using Microsoft Access and Lotus Approach and analyzed with SPSS for PC, version 11.0. To test our primary hypotheses, r x c tables were constructed for chi-square tests (α ≤5%). Continuous variables such as age, Rankin Scale, and the Barthel index of the activities
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
3. Results
Table 2 Characterization of treatment processes.
Referral by General practitioner Emergency physician Discharge Inpatient (other department) Geriatrics Rehabilitation Nursing home At home
3.1. Patients included
Stroke units
Conventional care
n
%
n
%
P
103 146
27.2 38.6
143 98
43.1 29.5
<.001
19
7.0
19
5.9
<.001
41 57 15 140
15.1 21.0 5.5 51.5
12 95 27 167
3.8 29.7 8.4 52.2
Time stroke onset to admission <3 h 175 Length of hospital stay, mean [d] (SD)
44.5 142 11.22 (5.6)
39.2 13.32 (7.3)
12 330
3.1 85.1
46 117
12.8 32.5
<.001
Extracranial Doppler/duplex none 7 <3 h 220
1.8 56.3
79 30
22.1 8.4
<.001
Transcranial Doppler/duplex none 6 <3 h 222 TTE done 68 TEE done 72 Swallowing assessment 127 Thrombolysis 14
1.5 56.6 18.6 19.8 33.9 3.6
308 11 243 57 111 1
87.7 3.1 69.4 16.4 32.2 0.3
<.001
Physical therapy none Start <24 h
21.5 71.3
124 125
35.2 35.5
<.001
126 2 29 1 30 53 167
59.2 0.9 13.6 0.5 14.1 24.9 78.4
.051 1.000 <.001 .447 .495 .601 .102
CCT none <3 h
84 278
Secondary prevention 1 year after ASS 132 Ticlopidin 2 Clopidogrel 72 Heparine 0 Oral anticoagulation 43 Lipid lowering drugs 71 Antihypertensive drugs 189
stroke onset 50.2 0.8 27.4 0.0 16.3 27.0 71.9
133
.140 .001
<.001 .232 .630 .001
A total of 1484 patients were screened (Fig. 1). Of these, 612 fulfilled the exclusion criteria. For those that were excluded, 255 patients (41.7%) had a time period between symptom onset and hospital admission that exceeded 24 hours, 105 patients (17.2%) had an initial diagnosis of AIS or TIA that was not supported after the complete diagnostic workup, 100 patients (16.3%) had their residence outside of Schleswig-Holstein, 61 patients (10.0%) rejected participation in the study, and 77 patients (12.6%) had a non-ischemic stroke. Another 43 patients (7.0%) were later excluded because of a late refusal or missing written consent. Of the remaining 872 participants, a further 88 patients could not be included in the evaluation because of documentation error. In total, 755 patients were included in the evaluation of the patients' characteristics and treatment processes. These patients were followed up 12 months (± 4 weeks) after AIS or TIA. One month before the planned follow-up, the MDK announced a home visit by physicians or nursing staff. In case of missing feedback, patients were reminded by a phone call a few days before the visit date. When follow-up was refused, an attempt was made to collect data by telephone. Where this was not possible (ill health, complete denial), only the vital status was documented. If a patient was not at home but apparently still alive, another visit was planned 1 to 2 months later. Otherwise, the patient's current address or the date of death was obtained from the respective resident's registration office. Of the 703 contacted patients, 480 could be interviewed personally. The average interval of follow-up was 381 days (SD 48.6). In the period between the end of the AIS or TIA treatment and the planned home visit, 98 patients died. Follow-up could be conducted in 40 cases because of missing information; 52 patients refused an interview. Despite inquiry at the residents' registration office, 27 patients could not be reached. Meanwhile, 6 patients lived outside of Schleswig-Holstein. 3.2. Patient characteristics
CCT indicates cerebral computed tomography; TTE, transthoracal echocardiography; TEE, transesophageal echocardiography; ASS, acetylsalicylate.
of daily living (ADL) were di- or polychotomized using appropriate cutoffs (groups with same size respectively clinical criteria). Intervalscaled variables were analyzed using the t test/Analysis of Variance (ANOVA) test or the U-test/Kruskal–Wallis test as appropriate. To control for possibly confounding variables, multivariate analyses were performed (logistic regression for dichotomous outcomes, Kaplan– Meier estimates with log-rank tests, or Cox regression for time-dependent data). Subgroup analyses were based on age (3 groups), sex, and AIS/TIA status. Our sample size calculation assumed a 16% 1-year mortality rate among patients treated in SI-SU and 32% for patients under CC treatment. Using α = 5% (two-sided), 123 patients per group are required to identify a difference as statistically significant.
At admission, demographic characteristics and health status differed between the 2 groups (Table 1). Significant differences existed with regard to sex, age, diagnosis (TIA/AIS), dysphagia, degree of disability before the event, and the presence of risk factors such as hypertension, atrial fibrillation, other cardiac diseases, and smoking status. 3.3. Characterization of the treatment processes SI-SU and CC patients were admitted by a general practitioner in 27% and 43% of the cases, respectively, and by an emergency physician in 39% and 30% of cases, respectively (Table 2). The rate of admission within 3 hours after symptom onset was higher in the SI-SU group than in the CC group (45% vs. 39%). A CCT was carried out more often for SI-SU patients (97% vs. 87%) and more frequently within 3 h (85% vs. 33%). A similar situation was found with the ultrasound examination of the brain-supplying arteries and with physiotherapy. Only transthoracic echocardiography was carried out more often in the CC hospitals (69% vs. 19%). This was not the case for transoesophageal echocardiography (16.4 % vs. 19.8%). At the time of the follow-up observation 1 year after stroke, no significant differences between the SI-SU and CC patients regarding relevant groups of secondary prevention strategies (Table 2) were measurable. 3.4. Outcomes
2.5. Data protection, ethics The Ethics Committee of the Medical Faculty of the University of Schleswig-Holstein, Campus Lübeck examined and approved the study protocol (May 15, 2001).
3.4.1. All patients Univariate analyses (n[mortality] = 715, n[Rankin] = 480) show a 1-year mortality in the SI-SU group of 16.3% (n = 60) and, in the CC group, of 26.0% (n = 90, P = .001). The percentage of surviving patients with a
134
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
Fig. 1. Participation flow chart.
high disability rate (Rankin Scale 3–5) at 12 months was 32.1% for SI-SU patients (n = 85) and 40.5% for CC patients (n = 87; P = .057). As shown in Table 1, a complete model combining all relevant factors was chosen for the multivariate analysis. The factors included were chosen with respect to their clinical relevance, measurability, and distribution. The model was retained for all outcomes and subgroup analyses. The following factors significantly influence mortality after 1 year (Fig. 2): old age, paresis, dysphagia, and a high degree of disability (Rankin 4–5) before admission or during hospital stay. Atrial fibrillation and other cardiac diseases significantly increase the risk of dying within 1 year after stroke. Controlling for these factors, we found that strokeunit care does not significantly reduce the 1-year mortality risk (odds ratio [OR] 0.67, 95% CI 0.41–1.10) in patients with AIS or TIA. The following independent factors were shown to increase the risk of moderate to very severe disability (Rankin 3–5) of the surviving patients 1 year after stroke: very old age (>84 years), a high degree of disability at admission, paresis, dysarthria or dysphasia, and previous stroke. Treatment of AIS or TIA patients in the SI-SU significantly reduced the risk of disability, irrespective of the other factors (OR 0.53, 95% CI 0.31–0.90) (Fig. 2). 3.4.2. Exploratory subgroup analyses Multivariate analysis of subgroups with respect to differential effects of SI-SU care on 1-year mortality shows that patients with ischemic stroke (OR 0.47; 95% CI 0.27–0.83), male patients (OR 0.34; 95% CI 0.16–
0.72), and patients 85 years of age or older (OR 0.16; 95% CI 0.03–0.79) benefit significantly from treatment in SI-SU as compared with CC care. In comparison with conventional care (Fig. 3), patients with TIA who were treated in SI-SU have a nonsignificant higher risk of dying within 1 year after incident (95% CI 0.9–8.24). SI-SU care showed a significant beneficial effect on disability level 1 year after stroke in patients with ischemic strokes (OR 0.44; 95% CI 0.22–0.87) and in female patients (OR 0.42; 95% CI 0.20–0.91). Adjusted (for sex, age, paresis, dysphagia, dysarthria/dysphasia, mRS before stroke onset and on admission, and comorbidity) survival curves show a significantly lower mortality for ischemic stroke patients treated in SI-SU. A difference in mortality rates between the groups of about 10% was reached after approximately 2 to 3 months (Fig. 4).
4. Discussion For the first time in Germany, this study compared the rates of survival and severe disability 1 year after acute cerebral ischemia onset between patients treated under the SI-SU concept of the DGN/ESO or standard conventional care. We were able to show that the SI-SU concept implied a reduction of disability in patients with an ischemic stroke or TIA. With respect to 1-year mortality, our data show a significant advantage of SI-SU care for patients with acute ischemic strokes but not for TIA patients.
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
135
Fig. 2. Adjusted regression analysis of factors influencing death and disability 1 year after stroke onset (adjusted for influence factors shown in Table 1 with the exception of the Barthel–ADL-Index).
Contrary to the traditional rehabilitative-stroke-unit concept, [30,31] the SI-SU concept is characterized by both a multidisciplinary and semi-intensive care emphasis in diagnosis, monitoring, treatment, and early rehabilitation. Process descriptions show that the concept was put into practice. For the CC therapy, our crude rate for 1-year mortality (26.0%) matches those from other studies. These are summarized in a recent Cochrane Review from the Stroke Units Trialists' Collaboration [9] and
resulted in a standardized mortality rate of 25.2% for general medical wards after 1 year. In our study, the mortality rate of the SI-SU patients was substantially lower (16.3%) than the Cochrane Review's rate, with a weighted rate of 22.2%. However, results from individual studies varied between 5.7% and 45.2% for groups of between 29 and 279 cases. In our study, we refrained from using a composite outcome. In several international studies, [14,30,32,33] mortality was combined with either severe disability or level of dependence and institutional
Fig. 3. Adjusted regression analysis of SI-SU care compared to CC treatment* regarding death and disability 1 year after stroke onset (adjusted for influence factors shown in Table 1 with the exception of the Barthel-ADL-Index).
136
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137
Fig. 4. Adjusted survival curve* for ischemic stroke patients treated in the CC or SI-SU concept (adjusted for influence factors shown in Table 1 with the exception of the BarthelADL-Index).
beginning of physiotherapy and logopedic therapy, and the use of gastric tubes. Between the CC clinics, there was a great variability in the use of almost all diagnostic measures and physiotherapy, as well as logopedics. The diagnostic and therapeutic profile seemed to depend more on the availability of resources than on a unified concept or existent guidelines [42]. Therefore, caution is required when speaking of the SI-SU and the CC care. Despite some limitations, this study showed that the DGN strokeunit concept, which contain the ESO recommendations for stroke centers [21], offer a beneficial effect with respect to 1-year mortality and disability, as well as length of hospital stay, in patients suffering from ischemic stroke. Our results are in line with a recent metaanalysis comparing models of inpatient stroke care [43] and strength the notion that the beneficial effect of acute stroke unit treatment on outcome is statistically significant. These findings support the notion that stroke care should be focused on larger centres. Further studies are necessary to evaluate the effects of the SI-SU on subgroups of TIA patients. Acknowledgments and Funding
care. If we had combined mortality and disability (Barthel index 0–70), our study would have resulted in a significant benefit for all SI-SU patients (OR 0.55; 95% CI 0.35–0.87). An extension of our regression model to thrombolysis, led to no substantial change of the results. Concerning mortality of ischemic stroke patients the beneficial effect of the SI-SU-treatment did not change (OR 0.48; 95% CI 0.27–0.85) and thrombolysis did not show a significant influence on outcome (OR 0.70; 95% CI 0.14–3.42). The consideration of this item violates our approach of the program evaluation. Nevertheless we conducted this analysis, since the thrombolysis represents a substantial element of the SI-SU-treatment. The effects of the SI-SU cannot obvious be reduced alone to thrombolysis. Others elements of semi intensive care were not evaluated because this study focus on the outcome effects of the whole concept and not on isolated factors. Our study has several limitations. Participating clinics were to report every suspected stroke to the coordinating center. Based on routine administrative data (Section 301 of the German Social Security Code V) from each hospital, we could estimate the proportion of all stroke victims that were successfully recruited. It varied between 28% and 64% of all eligible patients for the CC group. For the 2 SI-SUs, it amounted to 63% and 100%, respectively. This may have led to a certain and, unfortunately, uncontrollable degree of selection bias. Follow-up information was not available from all patients. Mortality and disability data were missing in 5.3% (n = 40) and 11.3% (n = 85) of all patients, respectively. So we performed a best-worst case scenario analysis. In the worst case scenario for AIS patients the beneficial effects of SI-SU was not longer present (outcome 1 year after stroke onset for mortality: OR 0.639, 95% CI 0.39–1.06 and for disability OR 0.61, 95% CI 0.34–1.11). The evidence for TIA patients for both conditions (worse and best) and the beneficial effect for AIS patients in the best condition scenario did not change. Because of organizational reasons, we were not able to conduct a randomized-controlled trial [32,34–40]. Furthermore, knowing the advantages of stroke units as described in existing studies, it appears to be ethically rather questionable whether to admit patients to what is presumably a less effective facility merely on the basis of randomization [41]. Therefore, we used a quasi-experimental approach capitalizing on 2 concurrent treatment concepts associated with different hospitals with non-overlapping catchment areas. The two groups of participating hospitals were requested to coordinate the key points of their therapeutic approach. However, an analysis of the process data revealed significant differences in certain areas within the groups. Between the stroke units, these included, in particular, type of aftercare (geriatrics vs. rehabilitation), ultrasound examinations of the brain-supplying arteries, utilization of MRT, the
The quality assurance program for stroke treatment (QuSS) and this embedded study were financed by the federal government, the statutory health insurance, and the medical advisory board of the statutory health insurance of Schleswig-Holstein (MDK). The following departments participated in the QuSS program: Department of Neurology, University of Schleswig-Holstein, Campus Lübeck (head: Professor Dr. Kömpf, MD); Clinic for Neurology at the Medical Center in Itzehoe (head: Prof. Dr. Thie, MD); Department of Internal Medicine at the Ostholstein Medical Centers in Oldenburg/Holstein (head: Dr. Jakobeit, MD) and Eutin (head: Dr. Gützkow, MD); Department of Internal Medicine at the Husum Clinic (head: Dr. Haacke, MD); and the Department of Internal Medicine at the District Hospital of Plön in Preetz (head: Dr. Holst, MD); Department of Internal Medicine at the Malteser Hospital, St. Franziskus, Flensburg (head: Professor Dr. Saal, MD); Clinic for Neurosurgery (head: Dr. Schmidt, MD) and Internal Medicine (head: Professor Dr. Machraoui, MD), Diakonissen Hospital, Flensburg; Clinic for Internal Medicine at the District Hospital of Rendsburg (head: Professor Dr. Herrlinger, MD); Clinic for Internal Medicine (head: Dr. Keck, MD) and Rehablitation and Geriatric Medicine (head: Dr. Stamm, MD) at the Medical Center in Heide (Westküstenklinikum); Clinic for Internal Medicine (head: Professor Dr. Bethge, MD), Clinic for Geriatric Medicine (head: Dr. Naurath, MD) and Clinic for Neurology and Psychiatry at the Medical Center in Neumünster (head: Professor Dr. Hansen, MD), Friedrich-Ebert Krankenhaus; Department of Neurology, University of SchleswigHolstein, Campus Kiel (head: Professor Dr. Deuschl, MD); Department of Neurology, Clinic for Psychiatry, Neurology und Rehabilitation, Schleswig (head: Dr. Leonhardt, MD). The authors thank the medical advisory board of statutory health insurance of Schleswig-Holstein (MDK) for supporting the QuSS program and especially Dr. Buss, MD and Dr. Erben, PhD, for their assistance in preparing the study files and the statistical workup. References [1] Evers SM, Struijs JN, Ament AJ, van Genugten ML, Jager JH, van den Bos GA. International comparison of stroke cost studies. Stroke 2004;35:1209–15. [2] Feinberg WM. Primary and secondary stroke prevention. Curr Opin Neurol 1996;9: 46–52. [3] Ghatnekar O, Persson U, Glader EL, Terent A. Cost of stroke in Sweden: an incidence estimate. Int J Technol Assess Health Care 2004;20:375–80. [4] Humphrey PR. Management of transient ischaemic attacks and stroke. Postgrad Med J 1995;71:577–84. [5] Taylor TN, Davis PH, Torner JC, Holmes J, Meyer JW, Jacobson MF. Lifetime cost of stroke in the United States. Stroke 1996;27:1459–66. [6] Weimar C, Lüngen M, Wagner M, Kraywinkel K, Evers T, Busse O. Cost of stroke care in Germany. An analysis of the stroke data bank of the German Foundation Stroke-Aid. Aktuelle Neurologie 2002;29:181–90.
A. Walter et al. / Journal of the Neurological Sciences 287 (2009) 131–137 [7] Kolominsky-Rabas PL, Heuschmann PU, Marschall D, Emmert M, Baltzer N, Neundorfer B, Schoffski O, Krobot KJ. Lifetime cost of ischemic stroke in Germany: results and national projections from a population-based stroke registry: the Erlangen Stroke Project. Stroke 2006;37:1179–83. [8] Collaborative systematic review of the randomised trials of organised inpatient (stroke unit) care after stroke. Stroke Unit Trialists' Collaboration. BMJ 1997;314:1151–59. [9] Organised inpatient (stroke unit) care for stroke. Stroke Unit Trialists' Collaboration. Cochrane Database Syst Rev. 2004. [10] Langhorne P, Williams BO, Gilchrist W, Howie K. Do stroke units save lives? Lancet 1993;342:395–8. [11] Agency for Health Care Policy and Research (AHCPR). “Post Stroke Rehabilitation”. [No. 16]. 1995. AHCPR Publication No. 95-0062. Clinical Practice guideline. [12] Fritze J. Die Evaluation von Stroke Units als medizinische Technologie. Schwartz FW, Köbberling J, Raspe HH, editors. [15]. 1999. Nomos Verlagsgesellschaft BadenBaden. Health Technology Assessment. [13] Schroeder A, Heiderhoff M, Köbberling J. Stroke Units: Update des HTA-Berichts “Die Evaluation von Stroke Units als medizinische Technologie”. Deutsche Agentur für Health Technology Assessment des Deutschen Instituts für Medizinische Dokumentation und Information, editor. [Band 11]. 2004. Health Technology Assessment. [14] Candelise L, Gattinoni M, Bersano A, Micieli G, Sterzi R, Morabito A, PROSIT Study Group. Stroke-unit care for acute stroke patients: an observational follow-up study. Lancet 2007;369:299–305. [15] Scottish Intercollegiate Guidelines Network (SIGN). National Clinical Guideline “Management of Patients with Stroke, Part I-IV” recommend for use in Scotland by the Scottish Intercollegiate Guidelines Network (SIGN). 1998. [16] Berlit P, Busse O, Diener C, Ringelstein EB. Guidelines for the implementation of stroke units from the Committee on Stroke Units of the Deutsche Gesellschaft für Neurologie. Aktuelle Neurologie 1996;23:171–5. [17] Berlit P, Popescu O, Klotzsch C, Diehl RR, Berg-Dammer E. Treatment of acute stroke on the stroke unit. Initial experiences with an acute stroke unit in Germany. Nervenarzt 1997;68:122–8. [18] Ringelstein EB. Recommendation for organization of special stroke departments (stroke units). 1997 implementation. Nervenarzt 1998;69:180–5. [19] Ringelstein EB. Concept of interregional and regional stroke treatment in Germany. Report of the recommendation of Commission 1.06 “Stroke Units and Acute Stroke Therapy” of the German Society for Neurology. Nervenarzt 2000;71:322–4. [20] European Stroke Initiative recommendations for stroke management. European Stroke Council, European Neurological Society and European Federation of Neurological Societies. Cerebrovasc Dis 2000;10:335–51. [21] European Stroke Organisation (ESO) Executive Committee. ESO Writing Committee. Guidelines for management of ischaemic stroke and transient ischaemic attack 2008. Cerebrovasc Dis 2008;25:457–507. [22] Busse O. Stroke units and stroke services in Germany. Cerebrovasc Dis 2003;15 (Suppl 1):8–10. [23] Sachverständigenrat für die Konzertierte Aktion im Gesundheitswesen. Bedarfsgerechtigkeit und Wirtschaftlichkeit. Band III: Über-, Unter- und Fehlversorgung. 2001. Gutachten 2000/2001.
137
[24] Kolominsky-Rabas PL, Sarti C, Heuschmann PU, Graf C, Siemonsen S, Neundoerfer B, Katalinic A, Lang E, Gassmann KG, von Stockert TR. A prospective communitybased study of stroke in Germany—the Erlangen Stroke Project (ESPro): incidence and case fatality at 1, 3, and 12 months. Stroke 1998;29:2501–6. [25] Stroke–1989. Recommendations on stroke prevention, diagnosis, and therapy. Report of the WHO Task Force on Stroke and other Cerebrovascular Disorders. Stroke 1989;20:1407-31. [26] Sulter G, Steen C, De Keyser J. Use of the Barthel index and modified Rankin scale in acute stroke trials. Stroke 1999;30:1538–41. [27] van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Gijn J. Interobserver agreement for the assessment of handicap in stroke patients. Stroke 1988;19:604–7. [28] Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, Marsh III EE. Classification of subtype of acute ischemic stroke. Definitions for use in a multicenter clinical trial. TOAST. Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24:35–41. [29] Ringelstein EB, Zeumer H, Schneider R. Contribution of computer tomography of the brain to differential typology and differential therapy of ischemic cerebral infarct. Fortschr Neurol Psychiatr 1985;53:315–36. [30] Kalra L, Evans A, Perez I, Knapp M, Donaldson N, Swift CG. Alternative strategies for stroke care: a prospective randomised controlled trial. Lancet 2000;356:894–9. [31] Ronning OM, Guldvog B. Stroke units versus general medical wards, I: twelve- and eighteen-month survival: a randomized, controlled trial. Stroke 1998;29:58–62. [32] Silva Y, Puigdemont M, Castellanos M, Serena J, Suner RM, Garcia MM, Davalos A. Semi-intensive monitoring in acute stroke and long-term outcome. Cerebrovasc Dis 2005;19:23–30. [33] Cavallini A, Micieli G, Marcheselli S, Quaglini S. Role of monitoring in management of acute ischemic stroke patients. Stroke 2003;34:2599–603. [34] Sulter G, Elting JW, Langedijk M, Maurits NM, De Keyser J. Admitting acute ischemic stroke patients to a stroke care monitoring unit versus a conventional stroke unit: a randomized pilot study. Stroke 2003;34:101–4. [35] Garraway WM, Akhtar AJ, Hockey L, Prescott RJ. Management of acute stroke in the elderly: follow-up of a controlled trial. Br Med J 1980;281:827–9. [36] Kaste M, Palomaki H, Sarna S. Where and how should elderly stroke patients be treated? A randomized trial. Stroke 1995;26:249–53. [37] Sivenius J, Pyorala K, Heinonen OP, Salonen JT, Riekkinen P. The significance of intensity of rehabilitation of stroke–a controlled trial. Stroke 1985;16:928–31. [38] Wood-Dauphinee S, Shapiro S, Bass E, Fletcher C, Georges P, Hensby V, Mendelsohn B. A randomized trial of team care following stroke. Stroke 1984;15:864–72. [39] Lincoln NB, Willis D, Philips SA, Juby LC, Berman P. Comparison of rehabilitation practice on hospital wards for stroke patients. Stroke 1996;27:18–23. [40] Indredavik B, Bakke F, Solberg R, Rokseth R, Haaheim LL, Holme I. Benefit of a stroke unit: a randomized controlled trial. Stroke 1991;22:1026–31. [41] Hacke W. A late step in the right direction of stroke care. Lancet 2000;356:869–70. [42] Dodel RC, Haacke C, Zamzow K, Paweilik S, Spottke A, Rethfeldt M, Siebert U, Oertel WH, Schoffski O, Back T. Resource utilization and costs of stroke unit care in Germany. Value Health 2004;7:144–52. [43] Foley N, Salter K, Teasell R. Specialized stroke services: a meta-analysis comparing three models of care. Cerebrovasc Dis 2007;23:194–202.