Hospitalization rates, stroke unit care, and recurrence rates in Austria’s stroke cohort Epidemiologic analysis of 102,107 patients in a nation-wide acute stroke cohort between 2015 and 2019

Abstract

Introduction:

Continuous analysis of key epidemiologic data is irremissible to adapt health-care systems to trends in stroke epidemiology. We present data from 2015 to 2019 on quality indicators of stroke care, including rates on hospitalization, stroke unit care and recurrence rates using medical record-linkage of in-patient routine documentation.

Patients and methods:

We included stroke/TIA patients admitted to Austrian hospitals between 2015 and 2019 aged ⩾20 years using medical record-linkage.

Results:

In our cohort of 102,107 patients with 107,055 treatment episodes, we could show a significant decrease in 1-year cumulative age-adjusted hospitalization rates per 100,000 in TIA (86.3 [95% CI 84.1–88.5] vs 75.4 [95% CI 73.4–77.4], RR 0.87 [95% CI 0.85–0.90]), ischemic stroke (187.3 [95% CI 184.0–190.5] vs 173.4 [95% CI 170.4–176.5], RR 0.93 [95% CI 0.91–0.94]), and intracerebral hemorrhage (28.5 [95% CI 27.3–29.8] vs 22.8 [95% CI 21.7–23.9], RR 0.80 [95% CI 0.76–0.84]). In ischemic stroke the rate of stroke unit care increased significantly (55.7% vs 69.3%; RR 1.14 [95% CI 1.12–1.17]), and acute 1-year recurrences decreased significantly.

Discussion:

We found a decline of the annual age-adjusted cumulative hospitalization rates in stroke/TIA, a higher age of disease manifestation and less severe strokes, which is probably attributable to improved primary and secondary prevention in Austria. The proportion of patients treated at stroke units increased significantly, however a geographical and age-dependency is still evident.

Conclusion:

Age-adjusted hospitalization rates of stroke/TIA patients decreased, and stroke unit care is increasing but the goal of the Stroke Action Plan for Europe is yet to be reached.

Graphical abstract.

Introduction

Acute stroke is considered the second-most common cause of death and a major cause of morbidity as well as disability worldwide. ¹ It has been defined as one of the priorities for the World Health Organization and the United Nations in their actions to reduce the burden of non-communicable diseases. Therefore, actions have to be taken in order to establish global evidence-based prevention and treatment directives, in order to reduce the global burden of stroke. The Stroke Action Plan for Europe (SAP-E) is a pan-European project, which sets targets for the implementation of evidence-based preventive actions, stroke services and treatment targets until 2030. ² A core element is the continuous collection of data and outcomes on multiple levels in order to monitor the effectiveness of measures applied. On a national level, Austria has implemented “Austrian in-patient quality indicators”, constituting national outcome measurements based on in-patient routine documentation. ³ Results are published annually in standardized reports, offering indicator-specific results on a national scale. Special reports and scientific publications elaborate in-depth analysis.

Vigorous monitoring of epidemiologic data and outcomes are important to adapt health-care systems to trends in stroke epidemiology. Recent research showed a decrease in incidence rates of ischemic stroke and intracerebral hemorrhage. ⁴ On the other hand, the SAP-E ² recommends an increase of stroke unit (SU) care for acute stroke patients to 90%, which is a challenging goal. Therefore, the collection of data on stroke unit care, hospitalizations and outcomes is crucial to adapt for future challenges of health-care systems.

Here, we present key epidemiologic data in the time period between 2015 and 2019. Data include hospitalization-rates of stroke and TIA, rates of SU care as well as Intensive Care Unit (ICU) admission, and stroke recurrence rates, using a novel methodological approach of medical record-linkage of Austrian in-patient routine documentation to reconstruct pseudonymized patient-trajectories in Austria’s nation-wide stroke cohort.

Material and methods

Details of the novel methodological approach as well as details on data sources are published separately. ⁵ We use pseudonymized health-related patient data from a standardized national dataset between 2015 to 2019, to establish an unselected Austrian stroke cohort. In brief, acute stroke subtype or transient ischemic attack (TIA) was classified according to the primary ICD-10 discharge diagnosis to TIA (G45), ischemic stroke (IS, I63), intracerebral (I61) or spontaneous subarachnoid hemorrhage (I60). I64 was not used in Austria after 2013. ⁵ Characteristics on hospitalizations included the date of admission and discharge as well as the cause of hospital admission (i.e. primary diagnosis), medical procedure codes (e.g. for SU care), attributes of discharge (e.g. discharge home, transfer to other hospital, in-hospital death) and ICU stays. By using a three-step pseudonymization process including secure-hash algorithm 256 individual patient-trajectories were reconstructed since 2015 in line with rigorous data protection laws.

In Austria, documentation of administrative data as well as medical data is mandatory for each in- and out-patient by federal law. ⁶ In the in-patient setting, each hospital operator is obliged to collect information systematically and completely, which is regarded as “routine documentation” or a “minimum basic data set” (MBDS). Furthermore, each hospital operator is obliged to report these data sets to the data-warehouse of the responsible federal ministry of health. Each data transmission involving the federal ministry of health is encrypted using Secure Copy (SCP) with public key authentication. SCP is a network protocol which uses Secure Shell (SSH) for data transfer and the same mechanisms for authentication. Therefore, it ensures authenticity and confidentiality of the data. Since 2015, a three-step pseudonymization on these data is provided including a non-recalculable pseudonym, a non-recalculable record-ID, and age-groups of 5 years, allowing the reconstruction of individual patient-trajectories. Periodic data audits are being performed, ensuring a high data validity in Austria. ⁵

We identified all individuals hospitalized due to acute stroke or TIA, with multiple hospitalizations per individual equaling the number of treatment episodes over the study period. We assigned treatment episodes of each individual to a unique year by their respective start date. Incident events are reported per 100,000 person-years (py).

The rate of stroke unit care in patients with an ischemic stroke or TIA was computed using the medical procedure code “AA040 – acute stroke-care in a stroke unit” for each treatment episode. The rate of stroke unit and/or intensive-care unit (ICU) care in patients with a hemorrhagic stroke was calculated by using the Simplified Acute Physiology Score 3 (SAPS3) as outlined in the accompanied methodological paper. ⁵ Analysis of stroke unit and/or ICU care across districts in Austria was conducted using the residential postal code available for each patient.

One year recurrence rates after ischemic stroke were calculated in our cohort between 2015 and 2018, ensuring a follow-up period of 365 days, with details described in the methodological paper. ⁵

Statistical analysis

Demographic information of the Austrian population was obtained from the federal statistical office “Statistics Austria.” ⁶ Age-standardized hospitalization-rates were calculated using the direct method and 95% confidence intervals modeled assuming Poisson distribution. Age-adjustment was carried out using the combined Austrian census population of the investigated time periods. For proportional changes in SU care for ischemic events as well as SU and/or ICU care in hemorrhagic events, and changes in recurrence rates, direct age-adjustment was carried out, using the distribution of combined cases by age group as standard population. We used R (version 4.0.3) in the RStudio environment for all calculations. Graphics were created using Microsoft Excel (version 2109 Build 16.0.14430.20256) and OpenLayers (version 6.4) with geographical information from the federal statistical office “Statistics Austria”. ⁷

Results

Between January 1st 2015 and December 31st 2019, 107,055 treatment episodes of 102,107 patients aged 20 years or above with acute stroke or TIA were included for the present analysis. Of these, 49.3% were female (n = 50,360), and 50.7% were male (n = 51,747). A total of 60.9% (n = 65,133) had an acute IS, 27.1% (n = 29,019) a TIA, 3.3% (n = 3488) a subarachnoid hemorrhage (SAH) and 8.8% (n = 9415) an intracerebral hemorrhage (ICH). When excluding patients with TIA, 83.5% had an IS, 12.1% an ICH and 4.5% a SAH. The study period covers 35.2 million person-years (py) at risk.

The 1-year cumulative age-adjusted hospitalization rate of patients with TIA, IS and ICH per 100,000 py of Austria’s general population showed a significant reduction, whilst no significant changes were observed in SAH as depicted in Table 1.

Table 1.

Age-adjusted 1-year cumulative hospitalization rate per 100,000 person-years (py) with 95% confidence intervals (CI) and relative risk (RR) with 95% CI between 2015 and 2019.

Diagnosis	2015		2019		RR 2019 vs 2015	95% CI
	Age-adjusted hospitalization rate per 100,000 py	95% CI	Age-adjusted hospitalization rate per 100,000 py	95% CI
Transient ischemic attack	86.3	84.1–88.5	75.4	73.4–77.4	0.87	0.85–0.90
Ischemic stroke	187.3	184.0–190.5	173.4	170.4–176.5	0.93	0.91–0.94
Intracerebral hemorrhage	28.5	27.3–29.8	22.8	21.7–23.9	0.80	0.76–0.84
Subarachnoid hemorrhage	9.3	8.6–10.0	9.9	9.2–10.7	1.07	0.99–1.15

CI: confidence interval; RR: relative risk.

One-year cumulative hospitalization showed a decrease in all age-groups (<50 years; 50–64 years; 65–79 years; and ⩾80 years) of TIA and ischemic stroke patients as depicted in Figure 1(a) and (b). This was also the case in ICH, with significant changes in all age-groups except for patients <50 years (Figure 1(c)). Patients with SAH did not show a significant change below the age of 80 years, however, in patients ⩾80 years a significant increase was noted (Figure 1(d)).

Figure 1.

Age-group specific 1-year cumulative hospitalization per 100,000 person-years for (a) transient ischemic attack, (b) ischemic stroke, (c) intracerebral hemorrhage, and (d) subarachnoid hemorrhage between 2015 and 2019.

Bar plots per age-group in 2015 and 2019 with 95% confidence intervals (CI) as well as age-group specific relative risk with 95% CI between 2015 and 2019.

Stroke units in Austria are obligatory integrated in neurological departments and follow national standardized structural criteria. They are comprehensively distributed across Austria to ensure SU access for the majority of inhabitants within less than 45 min ground transport, as shown in Figure 2.

Figure 2.

Austrian stroke unit (SU) masterplan in 2003 and development of SU distribution in Austria between 2005 and 2020.

SU: stroke unit.

Stroke units are shown by dots with adjacent lines representing the served coverage area. Color ranges represent the average access-time in normal traffic by car according to the depicted legends. White fields represent geographical areas with an access-time of <45 min as projected in 2003, and as realized in 2005 and 2020.

Over the entire study period, 61.5% (95% CI 61.1–61.9, n = 38,251) of patients with an IS received SU care, whilst 37.2% (95% CI 36.7–37.9, n = 10,243) with a TIA were treated on a SU. In hemorrhagic events, 71.6% (95% CI 70.8–72.4, n = 8930) received SU and/or ICU care. Patients not treated in specialized services were admitted to general neurology or internal medicine wards.

The probability of patients admitted to one of Austria’s SUs significantly increased between 2015 and 2019 for both TIAs and IS. No change in SU and/or ICU care was noted for SAH, however, a significant increase was seen in patients with ICH (Table 2).

Table 2.

Austrian age-adjusted rates of stroke unit care in ischemic events as well as stroke unit and/or intensive-care unit care in hemorrhagic events with 95% confidence intervals (CI) as well as relative risk (RR) with 95% CI between 2015 and 2019.

Diagnosis	2015		2019		RR 2019 vs 2015	95% CI
	Age-adjusted stroke unit care	95% CI	Age-adjusted stroke unit care	95% CI
Transient ischemic attack	32.7%	31.3–34.2	41.8%	40.0–43.6	1.19	1.15–1.24
Ischemic stroke	55.7%	54.5–57.0	69.3%	67.8–70.8	1.14	1.12–1.17
Diagnosis	2015		2019		RR 2019 vs 2015	95% CI
	Age-adjusted stroke unit and/or Intensive care unit care	95% CI	Age-adjusted stroke unit and/or Intensive care unit care	95% CI
Intracerebral hemorrhage	72.0%	68.2–75.8	82.2%	77.7–86.7	1.08	1.04–1.12
Subarachnoid hemorrhage	64.0%	57.8–70.1	63.5%	57.6–69.5	1.00	0.93–1.07

CI: confidence interval; RR: relative risk.

An increase was noted in all age-groups of TIA, IS and ICH patients, however, admission rates were lower in elderly patients (Supplemental Figure 1).

District-wise proportions of SU care were heterogeneous, showing an interquartile range (IQR) of 26.9%–60.7% in TIA (Figure 3(a)), and 57.4%–80.4% in IS (Figure 3(b)). The respective IQR for SU and/or ICU care in ICH was 75.0%–93.8% (Figure 3(c)), and 50.0%–80.0% for SAH (Figure 3(d)).

Figure 3.

Proportions of patients with stroke unit care for (a) transient ischemic attack and (b) ischemic stroke as well as stroke unit and/or intensive-care-unit care for (c) intracerebral hemorrhage and (d) subarachnoid hemorrhage in Austria’s districts in 2019.

District-wise distribution of stroke unit care in Austria for (a) transient ischemic attacks and (b) ischemic stroke as well as stroke unit and/or intensive-care-unit care for (c) intracerebral hemorrhage and (d) subarachnoid hemorrhage in 2019. District-specific rates in Austria’s capital Vienna are extrapolated. Color legend according to proportion ranges (ranges include <25%, 25%, 50%, 50%–60%, 60%–70%, 70%–80%, 80%–90% and ⩾90%). In white districts no cases were registered in 2019. Black dots depict Austrian stroke units.

Recurrence rates of IS within 1 year show a significant decrease between 2015 and 2018 when analyzing stroke recurrences (including IS, ICH and SAH), IS recurrences, as well as IS recurrences additionally including TIA (Table 3).

Table 3.

Age-adjusted 1-year recurrence rate after ischemic stroke with 95% confidence intervals (CI) as well as relative risk (RR) with 95% CI between 2015 and 2018.

Recurrence	2015		2018		RR 2018 vs 2015	95% CI
	Age-adjusted 1-year recurrences	95% CI	Age-adjusted 1-year recurrences	95% CI
Stroke	6.0%	5.9–6.1	4.8%	4.7–4.9	0.81	0.74–0.88
Ischemic stroke	5.7%	5.6–5.8	4.4%	4.3–4.5	0.79	0.72–0.86
Ischemic stroke and TIA	7.1%	7.0–7.3	5.5%	5.4–5.6	0.79	0.73–0.85

CI: confidence interval; RR: relative risk.

Discussion

Herein we present key epidemiologic data and their short-term changes for stroke care of an unselected nation-wide Austrian stroke cohort using novel medical record-linkage of Austrian in-patient routine documentation. ⁵

One key indicator is the hospitalization rates of all stroke patients. In Austria, patients with IS, ICH, and SAH irrespective of age and disability, including individuals from nursing homes and high-level home care, are transmitted to hospital care for stroke. There are no ambulatory or extramural services for TIA management like TIA clinics in Denmark or France.^8,9 This policy is stable over the last decades and will be maintained. Incidence rates are higher than hospitalization rates given the failure of medical record-linkage (8%), patients with stroke symptoms who do not seek medical help, non-hospitalization of stroke patients, and non-consideration of strokes occurring during in-hospital stays. Further details are discussed in the accompanied methodological paper. ⁵

Annual age-adjusted cumulative hospitalization rates decreased from 2015 to 2019 for patients with IS (7%) and TIA (13%). The decline was more pronounced for TIA, and this may be explained by a continuously increased use of magnetic resonance imaging causing a shift from the diagnosis of TIA to the diagnosis of IS (TIA definition changed from a time-based to a tissue-based approach).

We interpret both, the trend in annual cumulative hospitalization of stroke or TIA and the reduced recurrence rates after IS, using the presented stable and robust methodology, as a reliable indicator for the effectiveness of improved primary and secondary prevention in Austria. The Austrian Stroke Unit Registry (ASUR), a governmental quality assurance registry prospectively documenting patient characteristics upon admission and discharge from each of the 37 Austrian SU, provides additional confirmatory data. Data from ASUR showed an increase in patients age with first-ever IS by 4.4 years for men and 3.0 years for women between 2005 and 2018 in Austria. ¹⁰ This is almost double of the increase of the average age in Austria’s general population, which was 2.3 years for men and 1.7 years for women in the same time period. ¹⁰ Furthermore, information on the severity of IS, as measured by the NIHSS, is prospectively documented in ASUR. Bernegger et al. showed that the mean severity of IS significantly declined in Austria over the past 16 years, ¹¹ similar to data from a prospective registry in Japan including over 183,000 patients between 2000 and 2019. ¹² Therefore, these information needs further surveillance as surrogate parameter for the effectiveness of preventive measures and the development of stroke hospitalization in Austria.

Hospitalization rates in Austria are similar to age-adjusted cumulative yearly incidence-rates of stroke in different European countries and regions as detailed in the accompanied methodology paper. ⁵ Feigin et al. showed a divergent trend in stroke incidences between high- and low- to middle-income countries from 1970 to 2008. In high-income countries, a 42% decrease of stroke incidence was noted over the past four decades. Low- to middle-income countries showed a 100% increase in incidence rates. ¹³ In the USA, Koton et al. observed that the decrease in stroke incidence is significantly larger in persons above 65 years of age, as compared to a younger population, ¹⁴ however, we observed a decrease in all age-groups.

Annual age-adjusted cumulative hospitalization rates for patients with ICH declined by 20% between 2015 and 2019 in Austria. Gattellari et al. reported declining rates of fatal and nonfatal ICH in New South Wales, Australia, ¹⁵ the GBD Study observed a global decrease in incidence rates. ¹⁶ A possible reason for the decline in hospitalization rates of ICH may be the more frequent use of direct oral anticoagulants (dOACs) instead of Vitamin-K antagonists for the prevention of IS in patients with atrial fibrillation. All major clinical trials of the four different novel anticoagulants showed a substantially lower risk of ICH using dOACs instead of warfarin in the last decade.^17–20 Furthermore, the risk of recurrent ICH within 1 year is low for patients with atrial fibrillation, irrespective of non-exposure to oral anticoagulation (1.5%) or resuming/initiating oral anticoagulation (2.8%). The risk of cerebrovascular ischemic events is also low at 3.2%, irrespective of resuming oral anticoagulation. Also, all-cause death within 1 year was lower in patients with ICH and atrial fibrillation resuming/initiating oral anticoagulation (22%) compared to patients where oral anticoagulation was not recommenced (30.3%). ²¹ Another possible reason might be a better compliance and blood pressure target achievement through early use of combination pills in arterial hypertension ²² instead of initiating antihypertensive medication using monotherapy and consecutive dose escalation.

Annual age-adjusted cumulative hospitalization for patients with SAH did not change between 2015 and 2019 except in patients aged 80 years or above. The main causes of classical and convexity SAH are aneurysms and cerebral amyloid angiopathy. SAH associated with traumatic events should not be classified as vascular events, however, misclassifications take place as shown in the Tyrolean Stroke pathway (data not yet published), which will have to be considered for future analyses. Furthermore, SAH located at the convexity of the brain and superficial siderosis are frequently associated with cerebral amyloid angiopathy in the elderly. Patients typically present with transient sensory and/or motor symptoms and seizures, whereas SAH-typical headaches are rare. ²³ The increase of SAH in patients above 80 years may also be associated with a higher detection rate by a more frequent use of MRI. ²⁴

Notably, we found that the proportion of patients with TIA and IS being treated in Austrian SU significantly increased over time. In 2003, a nation-wide network of SU was planned and subsequently realized (Figure 2) in order to care for over 70% of acute IS patients. ²⁵ Even though SU admission rates were heterogeneous between districts, we show that this goal was achieved. Between 2015 and 2019, we found that the proportion of IS, and TIA patients admitted to SU increased by 14% and 19% respectively. Possible reasons for this increase are structural as well as processual changes, including the use of pre-hospital measures for ambulance services to transfer patients to stroke units, the establishment of stroke units also in peripheral district hospitals, as well as an increase in SU beds in Austria by 17 beds (9.2%) over the study period. However, the new goal to treat >90% of acute stroke patients on SU according to the SAP-E ² is not achieved in any of the 94 districts in Austria. Organized SU care was shown to improve outcome, by reducing the risk of dependency, institutional care and death compared with alternative service, independent of age, sex, stroke type, and stroke severity.^26,27 Our data reveal that the proportion of SU care differs between age groups and regions. However, as patients aged >80 years have similar benefits from SU care, a future focus ought to facilitate SU access for the elderly.

In our cohort, the risk of recurrent stroke after an IS was 5.3% and IS recurrences occurred in 5.0% after an IS over the investigated period, ⁵ with a significant 19% relative risk reduction for stroke recurrences, and 21% relative risk reduction for IS recurrences between 2015 and 2018. Declining trends of recurrence after stroke have been shown in a meta-analysis including study periods from 1961 to 2006. ²⁸ A more recent meta-analysis did not find a decrease in 1-year recurrences after ischemic stroke over an investigated period from 1997 to 2019. ²⁹ More recent trials reported 1-year recurrence rates of 5.1% after TIA and minor stroke, ³⁰ and 6.7% after acute IS in four European countries. ³¹ In the recent large-scale STROKE-CARD randomized controlled trial conducted in two centers in Austria, which included 2149 stroke and TIA patients, 1-year recurrence rates were similar with 6.4%.^32,33 We interpret the reported reduced risk of recurrence again as a surrogate parameter for improved secondary prevention. Our recurrence rates may be lower than those reported in the literature, as very early recurrences occurring within the same hospitalization may not be recorded due to the national definition of primary diagnosis as explained in the accompanied methodological paper. ⁵

In Austria, a nation-wide structured, multifaceted post-stroke disease management program according to the STROKE-CARD concept^32,33 has been implemented and will be remunerated as of 2022, in order to provide a standardized 3-month follow-up of patients with TIA or IS. This initiative will collect reliable information on 3-month death and outcome. Furthermore, rates of unscheduled re-admissions and recurrences can be calculated using the presented methodology. ⁵ Hence, we will be able to analyze that is, rates of post-stroke cardiovascular re-admissions, pneumonia, and fractures before and after the implementation of this 3-month follow-up. Therefore, the presented methodology will aid in generating real-world data of interventions on a national scale.

As shown, Austria has not yet reached the challenging goals of the SAP-E. Consequently, a comprehensive and in-depth analysis of the status-quo of Austria’s stroke care is necessary, to inform evidence-based improvements of the national strategy for integrated stroke care. ³⁴ Within this strategy, multiple measures on multiple levels of stroke care are continuously being defined under the premises of patient-orientation, cross-sectoral continuity as well as evidence-based medicine, therapies and care. In light of the needed improvement in stroke unit care, as mentioned above, structural improvements include, for example, the regional implementation of Austria’s national integrated stroke strategy, and centralization of specific stroke treatments (e.g. endovascular treatment). For emergency measures, emergency ambulance services are advised to utilize regionally adapted standardized chains of rescue, including, for example, the Austrian Prehospital Stroke Score, ³⁵ and direct transport to the nearest adequate stroke center, which have been shown to accelerate treatment courses, increase thrombolysis-rates and improve 3-month outcomes. ³⁶ Nationally, the number of dedicated SU beds is currently being increased. These measures ought to contribute to reach up to the goals of the SAP-E. The evaluation of quality-standards is based on the scientific analysis of outcome parameters as well as on the degree of implementation of national recommendations. Considering the scientific analysis, Austria has implemented annual national outcome measurements based on in-patient routine documentation – “Austrian in-patient quality indicators” (A-IQI). ³ Results are published annually in standardized reports, offering indicator-specific results on a national scale. In-depth analyses supplement available quality indicators and are reviewed concerning their integration into annual analyses within the A-IQI committees. Results on national, regional and hospital levels are regularly submitted to the involved stakeholders after review and plausibility checks within the A-IQI steering-committee. Statistical outliers trigger a peer-review within the respective hospital and department.

Conclusion

In Austria, annual age-adjusted cumulative hospitalization rates are decreasing for TIA, IS and ICH probably due to a decline of incidence rates by improved primary and secondary prevention. Furthermore, the age at disease manifestation is increasing and the proportions of patients with TIA, IS and ICH admitted to SU or ICU is increasing but the goal of the Stroke Action Plan for Europe is yet to be reached.