Quality Monitoring of Acute Stroke Care in Rhineland-Palatinate, Germany, 2001–2006
Background and Purpose— Quality monitoring projects are useful tools to improve the quality and to assess temporal trends of stroke care in larger populations.
Methods— In Rhineland-Palatinate, Germany, a statewide, hospital-based, acute stroke care quality monitoring project was started in 2001. Initially, participation was mandatory for all hospitals with dedicated stroke units and from 2006 onward was mandatory for all hospitals. Quality monitoring included a structured data assessment and quality indicators for procedural measures.
Results— Between 2001 and 2006, the numbers of patients registered annually (N=6389 vs N=10 610), admission <3 hours after stroke onset (28.2% vs 34.6%), admission via emergency medical systems (38.1% vs 50.3%), and treatment in stroke units (44.3% vs 59.5%) increased significantly (P<0.0001, respectively). In ischemic stroke, use of thrombolytic therapy increased (for patients admitted <3 hours after onset, 6.5% vs 14.1%), whereas therapy with high-dose heparin declined (24.5% vs 6.0%, P<0.0001). Several quality indicators (performance of neuroimaging and Doppler/duplex sonography, neuroimaging <3 hours after admission) showed stable results at a high level; more patients received echocardiography (62.2% vs 74.0%), but fewer patients were rapidly examined by extracranial Doppler/duplex sonography (68.7% vs 62.8%, P<0.0001). Diagnosis and treatment of hypertension and hyperlipidemia, use of aspirin and combined aspirin/dipyridamole, and diagnosis of atrial fibrillation increased (P<0.0001, respectively). Use of oral anticoagulation remained stable at ≈38% of patients with cardioembolism.
Conclusions— Although these results reflect high standards of acute stroke care and improvements regarding early admission, thrombolytic therapy, and several secondary preventive measures, there is still the potential for further improvement regarding thrombolysis, use of oral anticoagulation and statins, and admission to stroke units, for example.
It is an important aim to establish high standards of acute stroke care in accordance with current guidelines, not only in specialized centers but also on a general community level.1 Systematic quality monitoring of acute stroke care is an important tool to ensure such high standards, and quality registries have been implemented in several countries, including Sweden, Germany, Canada, the United States, and Argentina.2–6
In Rhineland-Palatinate, a state in west Germany with ≈4 million inhabitants, quality monitoring of acute stroke care was started in 1996 by neurology departments on a voluntary base. After the first stroke units had been established in 1999, a stroke quality monitoring project was started on January 1, 2001. Quality indicators and a standardized questionnaire were established by an expert group and further developed during the following years. All participating hospital departments received semiannual reports setting their stroke care results into perspective with data from other hospitals. Quality standards were defined, and a so-called structured dialogue was initiated with those hospitals that did not fulfil such standards. In 2006, participation in the monitoring process became mandatory for all hospitals treating stroke patients, and in 2007, quality criteria were changed after a nationwide expert group had developed a new and expanded set of quality criteria.7 Here, we report results of the acute stroke care monitoring project for the years 2001 to 2006.
Subjects and Methods
The Rhineland-Palatinate acute stroke care quality monitoring project is a prospective, multicenter, statewide, hospital-based stroke registry that includes all urban and rural hospitals of any facility type that treat patients with stroke or transient ischemic attack (TIA).
Organization of Stroke Care
Starting in 1999, 6 supraregional stroke units based in neurology departments and an increasing number of regional stroke units, based mostly in medical departments, were established in Rhineland-Palatinate, a German state with a mostly rural population and only a few cities with a population of >100 000 inhabitants. A governmental hospital plan envisioned a statewide establishment of stroke units in Rhineland-Palatinate. According to the German stroke unit model, all stroke units are semi-intensive care units with monitoring of physiological parameters aimed at treatment as soon as possible after stroke.8 Supraregional stroke units have special expertise in stroke neurology; they are equipped to treat very severe strokes, rare stroke etiologies, and juvenile strokes and to implement invasive stroke diagnostics and therapies in their larger catchment area. Regional stroke units deliver all basic stroke care, including intravenous thrombolysis, usually in smaller catchment areas. Patients with stroke or TIA are treated not only on general wards in hospitals without stroke units but also in neurology and internal medicine departments with stroke units if beds on the stroke unit are unavailable or if stroke unit care is deemed inappropriate by the treating physician, for example, in cases of dementia or preexisting severe deficit.
Stroke Quality Monitoring Project
On January 1, 2001, an official stroke project was started in Rhineland-Palatinate that aimed first at monitoring the quality of acute stroke care and second at obtaining data for administrative planning of stroke care facilities. Initially, participation was mandatory for all hospitals with stroke units and some others in special planning regions, but participation became mandatory for all hospitals in 2006. An executive committee was established to guide and supervise the registry. The project included all patients ≥16 years old with acute first or recurrent ischemic or nontraumatic hemorrhagic stroke or TIA. Exclusion criteria were subarachnoid or subdural hemorrhage, sinus venous thrombosis, and admission >7 days after symptom onset.
Physicians in participating hospitals filled out a structured questionnaire for each patient. A handbook provided definitions of all items. The questionnaire was anonymous and contained data on year of birth; sex; area code; symptom onset; hospital admission; mode of patient referral; neuroradiological, neurosonographic, and cardiac examinations and their timing; acute therapy; in-hospital complications; risk factors (eg, arterial hypertension defined as blood pressure ≥140/90 mm Hg or on antihypertensive treatment; hypercholesterolemia defined as fasting cholesterol ≥200 mg/dL or on lipid-lowering treatment); secondary prevention; length of hospital stay; and mode of discharge. The Barthel Index, the modified Rankin scale, and neurological scores (European Stroke Scale, after 2003 National Institutes of Health Stroke Scale) were assessed <24 hours after admission and at discharge; the modified Rankin scale was also used for the period before stroke/TIA.
First and recurrent ischemic strokes, intracerebral hemorrhages (ICHs), and TIAs were differentiated. The etiology of ischemic stroke/TIA was analyzed according to modified TOAST criteria,9 including a subgroup “presumed local atherothrombotic etiology” in case of atherosclerotic lesions with <50% diameter reduction in the absence of other stroke etiologies.
Questionnaires were sent electronically or by mail to a central quality monitoring institution (SQMed Mainz). Data completeness and plausibility were monitored, and centers were asked for corrections if required. Data were entered into a central data bank, and statistical analyses were performed at the SQMed.
In 2000, the executive committee, including neurologists, internists, and statisticians, developed a set of criteria that intended to monitor the procedural quality of acute stroke care. These quality criteria included (1) performance of cranial computed tomography (CT) or magnetic resonance imaging (MRI) during hospitalization; (2) CT/MRI <3 hours after admission in stroke patients admitted <24 hours after onset; (3) CT/MRI <3 hours in patients with first ischemic or hemorrhagic stroke or TIA, admission to hospital <3 hours, and National Institutes of Health Stroke Scale score ≥4 and ≤25 on admission; extracranial Doppler or duplex sonography (4) during the hospital stay or (5) <24 hours after admission in patients with ischemic stroke/TIA; (6) transcranial Doppler sonography in patients with ischemic stroke/TIA <24 hours after admission on a supraregional stroke unit; and (7) transthoracic or transesophageal echocardiography during hospitalization in patients with ischemic stroke/TIA.
Twice annually, hospitals received detailed reports that allowed comparisons with all hospitals together and with groups of hospitals (supraregional stroke units, regional stroke units, hospitals without stroke units). For all quality indicators, minimally required standards were defined. Hospitals not meeting the required limit at the 95% CI of their individual results were questioned in a so-called structured dialogue. Meetings with all participating hospitals were organized annually to discuss principal results. The project has been performed in accordance with all legal requirements and with the principles of the respective ethics committees.
In descriptive statistics, absolute numbers and percentages, or mean and SD, are given. For comparisons between the 6 years, the χ2 test (nominal scaled data) with respective degrees of freedom and the Kruskal-Wallis test (no normal distribution of metric data) were applied as appropriate. Probability values <0.001 were regarded as significant. Analyses were performed with SPSS, version 16.
The number of participating departments initially ranged between 39 and 49 but rose to 70 in 2006 when participation became mandatory (Table 1). The number of patients registered steadily increased between 2001 (N=6389) and 2006 (N=10 610). About half of the patients were female, with no relevant changes in the male to female ratio over time. The mean age increased from 71.0±12.7 years (2001) to 73.1±12.4 years (2006). Numbers and percentages of patients treated in regional stroke units steeply rose, whereas the proportion of patients treated on general hospital wards declined (P<0.0001). Absolute numbers of all stroke subtypes increased over time. The proportion of ICH was relatively low, ranging between 6.2% and 7.5% (Table 1).
The percentage of patients admitted <3 hours after symptom onset continuously increased from 28.2% in 2001 to 34.6% in 2006 (P<0.0001, Table 2). In parallel, admission via emergency medical systems steadily increased, from 38.1% in 2001 to 50.3% in 2006, whereas referral by general practitioners declined (P<0.0001). The duration of hospitalization diminished by ≈2 days for ischemic stroke (Table 2) and by ≈2.5 days for TIA (in 2001, 9.8±7.3 days; in 2006, 7.3±4.4 days; P<0.0001). Hospital mortality ranged between 5.6% and 6.6% for ischemic stroke and between 14.0% and 20.0% for ICH, with no relevant temporal trends. Discharge into rehabilitation institutions strongly increased after 2001, reaching a stable proportion of ≈40% (P<0.0001, Table 2).
In ischemic stroke, the use of medium and high doses of heparin strongly declined, from 24.5% in 2001 to 6.5% in 2006 (P<0.0001). The number of departments offering and the number of patients receiving thrombolytic therapy steeply increased. Among patients admitted <3 hours after onset, the rate of thrombolysis increased from 6.5% in 2001 to 14.1% in 2006, and among those who additionally had National Institutes of Health Stroke Scale scores between 4 and 25, the percentage rose from 11.1% in 2003 to 18.4% in 2006 (P<0.0001). The proportion of patients receiving thrombolysis among all patients with ischemic stroke also increased but at a much lower level (1.9% in 2001 vs 4.7% in 2006, Table 3).
Regarding quality indicators, the rate of neuroimaging during the hospital stay was high and stable at ≈97%, as was the proportion of patients receiving CT/MRI within 3 hours after admission (Table 3). In patients admitted early after symptom onset, the rate of CT within 3 hours steadily increased over time (78.6% to 83.7%, P=0.0026). A rather stable proportion of patients (≈88%) was examined by extracranial Doppler or duplex sonography during the hospital stay, and the majority received this test within 24 hours, although percentages slightly declined over time (P<0.0001). The use of echocardiography steadily rose from 62.2% in 2001 to 74.0% in 2006 (P<0.0001, Table 3).
In ischemic stroke, the diagnosis of cardioembolic and large-artery atherothrombotic etiology increased, whereas the diagnosis of lacunar infarcts decreased over time (Table 4). A diagnosis of hypertension, hyperlipidemia, or atrial fibrillation was made almost continuously more often during the observation period (P<0.0001, Table 4). Current or recent smoking initially declined but remained stable within the last 3 years (P<0.0001). For secondary prevention, the proportion of patients receiving aspirin or combined aspirin and dipyridamole increased, whereas administration of clopidogrel declined (P<0.0001, Table 4). Administration or recommendation of oral anticoagulants in ischemic stroke in general or in cardioembolic stroke in particular did not show significant trends over time. The use of antihypertensive and lipid-lowering agents at discharge strongly increased during the observation period (P<0.0001, Table 4).
During the last decade, quality monitoring projects evaluating acute stroke care have been established in several countries,2–6 but only few of these registries included patients of a whole state or country.2 In Germany, several stroke registries were initiated, with the registry in Rhineland-Palatinate being among the earliest. The main objective of these projects is to improve stroke care in larger areas by establishing high quality standards, by disseminating knowledge on stroke care to nonspecialized hospitals, and by providing benchmark results on stroke care performance for individual hospitals. Observing temporal trends is 1 of the tasks of the monitoring system.
According to official hospital statistics, 20 209 (2001), 19 787 (2002), 19 358 (2003), 18 886 (2004), 18 692 (2005), and 18 016 (2006) patients with stroke (International Classification of Diseases codes 10 I61, I63, or I64) or TIA (G45) were treated in hospitals in Rhineland-Palatinate during the observation period. Based on these numbers, the coverage of our quality monitoring projects continuously increased from 31.6% (2001) to 58.9% (2006). However, the aforementioned official numbers should be viewed with some caution, mainly owing to possible diagnostic misclassification. The number of patients diagnosed as having “stroke, not specified as intracerebral hemorrhage or ischemic stroke” (International Classification of Diseases code 10 I64) was relatively high in 2001 (4437 of 20 217, or 21.9%) but steadily decreased until 2006 (972 of 18 016, or 5.4%). Most of these patients probably had not received neuroimaging, which increases the likelihood of misclassification. Therefore, the coverage of patients with “definite stroke” may have been somewhat higher, mainly during the first years, than these rates suggest. The decreasing numbers of patients with unclassified stroke in the statewide hospital statistics themselves reflect an improvement in stroke diagnostics and stroke care during this period.
The increase in registered patients was mainly due to greater inclusion of elderly patients and of those with less severe stroke, who are more often treated in medical departments of local hospitals. A higher number of these hospitals started to participate during the later phase of the observation period. These increasing participation rates, on the other hand, limit the comparability of populations in different years. The fraction of ICHs was low, at 6.2 to 7.5% of all patients with stroke or TIA. In the hospital statistics of Rhineland-Palatinate, the rate of ICH differed between 7.9% (2001 and 2003) and 8.7% (2005); however, duplicate registration of single patients by neurological and neurosurgical departments may have led to some overestimation. Recent population-based studies report percentages of ICH among total strokes as low as 6.5%, and the fraction of ICH may currently be lower than estimated in the past.10 Participation of neurosurgery departments is not enforced in our registry. The vast majority of patients with ICH are diagnosed and treated in neurology or medical departments and are referred to neurosurgery if required. However, our registry may not have included the few patients primarily admitted to neurosurgery departments and thus, there may be a slight underestimate of the frequency of and mortality from ICH.
Stroke centers are expected to teach the population about stroke symptoms and to participate in the organization of emergency medical services to enable rapid admission after stroke.1 In line with this aim, the rate of admissions at <3 hours after stroke increased, accompanied by increasing admissions via emergency medical services and lower proportions of patients who had lost time by contacting a general practitioner. In parallel, the numbers and percentages of patients who received thrombolytic therapy steadily increased and reached 4.7% in all ischemic stroke patients and 18.4% in the target population in 2006, percentages that are comparable to5 or higher than6,11,12 those in other studies but lower than in studies based on selected hospitals.13 In line with current guidelines,1 treatment with medium- or high-dose heparin steeply dropped during the observation period, reflecting 1 of the strongest trends in stroke therapy during recent years.
The duration of hospitalization decreased for ischemic stroke and TIA, reflecting a long-term and overall trend in German medicine that was strengthened by introducing reimbursements based on diagnosis-related groups after 2003 and that is not explained by variations in transfer policies to rehabilitation units alone. Hospital mortality did not show a definite trend over time and is comparable with results from other recent surveys.14
Quality indicators can refer to structural, procedural, or outcome measures. The quality indicators used until 2006 reflect procedural measures; they represent prerequisites for therapeutic success, such as rapid neuroimaging, which is required for thrombolytic therapy. In the target population for thrombolysis, the percentage of rapid neuroimaging rose slightly but significantly during recent years. The main result regarding quality indicators is that an initially high standard in diagnostic procedures could be maintained over time along with the increasing number of participating hospitals and patients registered and could even be improved regarding echocardiography. Only the rate of early ultrasound diagnostics slightly declined, as it was probably progressively replaced by early CT and MR angiography. Improving quality of care is reflected by the increasing use of thrombolysis and by reduced administration of intravenous heparin. The scope of quality indicators was limited to procedural aspects. Starting in 2007, an extended set of quality indicators was successfully applied in Rhineland-Palatinate.7 Therefore, we limited ourselves to the years up to 2006.
Cardioembolic stroke and atrial fibrillation were more often diagnosed in this population of increasing age; however, the percentage of oral anticoagulation use prescribed or recommended at discharge for future use did not increase, reflecting the reluctance of many physicians to treat patients after stroke with oral anticoagulants. The rate of oral anticoagulation in ≈40% of cardioembolic stroke is even somewhat higher than recent data from the Swedish Riks-stroke registry15 but certainly requires future improvement. The decreasing rate of lacunar stroke was rather unexpected and may have resulted from diagnostic uncertainty in the differentiation between lacunar and large-artery atherothrombotic stroke. Hypertension was increasingly diagnosed and antihypertensives were more often administered after ischemic stroke. Reluctance to use antihypertensives is certainly influenced by the recommendation to allow higher blood pressure during acute stroke, although improvements in antihypertensive treatment at discharge appear feasible, based on our data. Use of lipid-lowering agents, certainly mostly statins, rapidly increased, reflecting the broader knowledge of their benefit after stroke and recent guidelines.1 However, use of these medications was still relatively low in 2006. Use of aspirin and combined aspirin and dipyridamole greatly increased over time, whereas clopidogrel was less commonly administered, a trend that reflects recommendations of German stroke guidelines.16
The strengths of the present quality monitoring project are the high numbers and the increasing coverage of patients with stroke or TIA during the observation period and the structured assessment of a broad range of data for each patient. Limitations include incomplete acquisition of data for patients treated in neurosurgical departments, the possibility that duplicate registration of a few patients by different hospitals cannot be completely ruled out, and the lack of any on-site data monitoring. Furthermore, the questionnaire that aimed at simplicity of data acquisition by physicians does not allow differentiation between the absence of a diagnosis or procedure and missing information for several parameters.
In conclusion, our registry emphasizes the concept that stroke quality monitoring is feasible on a state- or country-wide basis and that it is helpful to improve and maintain stroke care quality in larger populations. The increasing rates of patients using emergency medical services and arriving early at the hospital and the increasing use of thrombolytic therapy are among the most successful aspects of the quality monitoring project.
- Received February 17, 2010.
- Accepted February 27, 2010.
Kapral MK, Laupacis A, Phillips SJ, Silver FL, Hill MD, Fang J, Richards J, Tu JV. Investigators of the Registry of the Canadian Stroke Network. Stroke care delivery in institutions participating in the Registry of the Canadian Stroke Network. Stroke. 2004; 35: 1756–1762.
Heuschmann PU, Kolominsky-Rabas PL, Misselwitz B, Hermanek P, Leffmann C, Janzen RWC, Röther J, Buecker-Nott HJ, Berger K, for the German Stroke Registers Study Group. Predictors of in-hospital mortality and attributable risks of death after ischemic stroke. Arch Intern Med. 2004; 164: 1761–1768.
The Paul Coverdell Prototype Registries Writing Group. Acute stroke care in the US: results from 4 pilot prototypes of the Paul Coverdell National Acute Stroke registry. Stroke. 2005; 36: 1232–1240.
Sposato LA, Esnaola MM, Zamora R, Zurrú MC, Fustinoni O, Saposnik G, on behalf of ReNACer Investigators and the Argentinian Neurological Society. Quality of ischemic stroke care in emerging countries: the Argentinian National Stroke Registry (ReNACer). Stroke. 2008; 39: 3036–3041.
Heuschmann PU, Biegler MK, Busse O, Elsner S, Grau A, Hasenbein U, Hermanek P, Janzen RWC, Kolominsky-Rabas PL, Kraywinkel K, Lowitzsch K, Misselwitz B, Nabavi DG, Otten K, Pientka L, von Reutern GM, Ringelstein GB, Sander D, Wagner M, Berger K. Development and implementation of evidence-based indicators for measuring quality of acute stroke care. The quality indicator board of the German Stroke Registers Study Group (ADSR). Stroke. 2006; 37: 2573–2578.
Adams HP Jr, Bendixen BH, Kappelle LJ, Biller J, Love BB, Gordon DL, Marsh EE III. Classification of subtype of acute ischemic stroke: definitions for use in a multicenter clinical trial. Trial of Org 10172 in Acute Stroke Treatment. Stroke. 1993; 24: 35–41.
Deng YZ, Reeves MJ, Jacobs BS, Birbeck GL, Kothari RU, Hickenbottom SL, Mullard AJ, Wehner S, Maddox K, Majid A, for the Paul Coverdell National Acute Stroke Registry Michigan Prototype Investigators. IV tissue plasminogen activator use in acute stroke: experience from a statewide registry. Neurology. 2006; 66: 306–312.
Kapral MK, Fang J, Hill MD, Silver F, Richards J, Jaigobin C, Cheung AM; for the Investigators of the Registry of the Canadian Stroke Network. Sex differences in stroke care and outcomes: results from the registry of the Canadian Stroke Network. Stroke. 2005; 36: 809–814.
Glader EL, Stegmayr B, Norrving B, Terént A, Hulter-Asberg K, Wester PO, Asplund K. Riks-Stroke Collaboration: large variations in the use of oral anticoagulants in ischemic stroke patients with atrial fibrillation: a Swedish national perspective. J Intern Med. 2004: 255: 22–32.
Diener HC, Allenberg JR, Bode C, Busse O, Forsting M, Grau AJ, Haberl RL, Hacke W, Hamann GF, Hennerici M, Grond M, Ringelstein EB, Ringleb PA. Leitlinie Primär- und Sekundärprävention der zerebralen Ischämie. III ed. Stuttgart/New York; Thieme: 2005.