Twenty years of monitoring acute stroke care in Australia through the national stroke audit programme (1999

Abstract

Background

National organisational surveys and clinical audits to monitor and guide improvements to the delivery of evidence-based acute stroke care have been undertaken in Australia since 1999. This study aimed to determine the association between repeated national audit cycles on stroke service provision and care delivery from 1999 to 2019.

Methods

Cross-sectional study using data from organisational surveys (1999, 2004, 2007–2019) and clinical data from the biennial National Stroke Acute Audit (2007–2019). Age-, sex-, and stroke severity-adjusted proportions were reported for adherence to guideline-recommended care processes. Multivariable, logistic regression models were performed to determine the association between repeated audit cycles and service provision (organisational) and care delivery (clinical).

Results

Overall, 197 hospitals provided organisational survey data (1999–2019), with 24,996 clinical cases from 136 hospitals (around 40 cases per audit) (2007–2019). We found significant improvements in service organisation between 1999 and 2019 for access to stroke units (1999: 42%, 2019: 81%), thrombolysis services (1999: 6%, 2019: 85%), and rapid assessment/management for patients with transient ischaemic attack (1999: 11%, 2019: 61%). Analyses of patient-level audits for 2007 to 2019 found the odds of receiving care processes per audit cycle to have significantly increased for thrombolysis (2007: 3%, 2019: 11%; OR 1.15, 95% CI 1.13, 1.17), stroke unit access (2007: 52%, 2019: 69%; OR 1.15, 95% CI 1.14, 1.17), risk factor advice (2007: 40%, 2019: 63%; OR 1.10, 95% CI 1.09, 1.12), and carer training (2007: 24%, 2019: 51%; OR 1.12, 95% CI 1.10, 1.15).

Conclusions

Between 1999 and 2019, the quality of acute stroke care in Australia has improved in line with best practice evidence. Standardised monitoring of stroke care can inform targeted efforts to reduce identified gaps in best practice, and illustrate the evolution of the health system for stroke.

Keywords

Stroke quality of health care audit and feedback

Introduction

Monitoring the quality of care provided in hospitals is important for patients, clinicians, executives, and funding agencies. Audit and clinical quality registry programmes provide standardised data collection systems to reliably monitor the clinical care of patients against guideline-recommended care processes, standards or frameworks. Evidence for stroke care suggests that adherence to these monitoring and reporting systems improves survival and reduces other adverse outcomes after stroke,^1,2 highlighting their importance.

In Australia, the first national survey of acute hospital services was carried out by the Stroke Foundation in 1999³ (and repeated in 2004). Adapted from a questionnaire used to evaluate stroke in the United Kingdom,⁴ the survey provided a snapshot of stroke services in Australia. This was followed by the publication of the first national clinical guidelines for acute stroke in 2003, updated in 2007, coinciding with the establishment of the National Stroke Acute Audit programme (referred to as the Audit programme hereafter) to monitor adherence to these recommendations. The voluntary Audit programme, based on similar work in the United Kingdom,⁵ has been conducted biennially since. It comprises two components: an organisational survey and a retrospective clinical medical record audit to collect patient-level data on select care processes received during admission, as well as in-hospital outcomes.⁶

This study sought to determine the association between repeated cycles of audit (as a surrogate for time) from 1999 to 2019, access to organisational resources for acute stroke (e.g. workforce, bed numbers, stroke units), and adherence to evidence-based care processes recommended in the national guidelines.

Methods

Cross-sectional study using data from (i) self-reported organisational surveys undertaken in 1999, 2004 and biennially since 2007; and (ii) clinical audit of up to 40 consecutive medical records collected biennially since 2007 (Figure 1). A dedicated coordinator from the Stroke Foundation provided training and support to hospital clinicians collecting these data. Reporting of this study adheres to the Strengthening the Reporting of Observational studies in Epidemiology (STROBE) statement.⁷

Figure 1.

Timelines for the collection of survey and audit data.

Eligibility criteria and hospital participation

Acute hospitals that were invited to participate in the survey and clinical audit varied over cycles, with eligibility based largely on stroke admission numbers (Online Supplement Table S1). Invited hospitals were followed-up by email and phone call by the Stroke Foundation coordinator to improve the response rate. Hospitals choosing not to participate in the clinical audit were still encouraged to complete the organisational survey. As service characteristics in hospitals admitting fewer patients with stroke differ from those treating more patients, recruitment from 2015 onwards primarily included those admitting at least 40–50 patients with stroke annually.

Data collection

Details on data collection have been previously published.^6,8 In brief, the survey collected information on the characteristics and resources of the hospitals (e.g. bed and admission numbers) with data gathered by health professionals with a responsibility for stroke management (medical leads and stroke coordinators).

Clinical audit data were collected retrospectively from the medical records of approximately 40 consecutive patients with stroke (ICD-10 codes: I61, I62.9, I63, I64) admitted throughout the previous year. Hospitals with larger annual stroke admissions (e.g. > 400 admissions) were invited to audit more medical records, aiming for 10% of total admissions. Hospitals were de-identified, and no patient-identifying data were collected. Up to five records from each hospital were re-audited by a second staff member to allow monitoring of inter-rater reliability, with a procedure manual and training provided by Stroke Foundation coordinators.

The 1999 and 2004 organisation surveys used paper forms and data were manually entered into a database by project staff. The Audit programme (2007–2019) used a web-based data collection system, with programmed logic checks to minimise erroneous data. Records with missing responses were considered incomplete and excluded. Organisational survey and clinical audit questions were revised over time in response to changes to the Australian Clinical Guidelines for Stroke Management and the Acute Stroke Services Framework.^9,10

Data analyses

Analyses included only questions consistently reported across all audits, and pooled data from hospitals with >40 annual stroke admissions to ensure reliable comparisons over time.

Organisational survey: Analyses used descriptive statistics to summarise survey data for 1999–2019. We ran multivariable logistic regression models to determine the association between repeated audit cycles (indicated by successive years of audit) and availability of acute stroke services; regression models excluded 1999 and 2004 survey data to allow for more robust analysis from similar data collection methods/questions from the Audit programme (2007–2019). Models were multilevel to account for hospital-level variation and included year of audit cycle.

Clinical audit: Only valid responses were included for questions relating to clinical impairments (e.g. arm weakness present). Not documented and unknown responses were assumed to be negative (i.e. the process did not occur) and included in the denominator for analyses of care processes.¹¹ Time metrics were calculated from emergency department (ED) presentation, with the start of the day (00:00:00) imputed if the ED date was known, but time was missing. For relevant time-related care processes with known dates, but missing times, the end of the day (23:59:59) was imputed. Descriptive statistics were used to summarise patient characteristics. The proportion of eligible patients who received each care process determined adherence. Proportions were adjusted for age, sex and inability to walk on admission (a proxy measure for stroke severity),¹² and significant trends in proportions over time were determined.

Multivariable, multilevel logistic regression models were used to determine the association between repeated audit cycles (indicated by successive years of audit) and provision of evidence-based care processes. Models were adjusted for patient variables, such as age, sex, independence prior to admission, stroke type, and validated severity factors (ability to walk, speech disturbance, arm weakness on admission and incontinence).¹³ Secondary analyses were performed to determine adherence to care processes from 2015 to 2019, reflecting more recent relevant changes in care delivery.

Results of multivariable models were reported as odds ratios (ORs) and 95% confidence intervals (CIs). All statistical tests were two sided, and a p-value <0.05 was considered to be statistically significant. Data were analysed using Stata SE 15.0 (www.stata.com).

Ethics

The Heads of Stroke Services of participating hospitals provided consent to participate in the audits. Specific approvals for this project were received from the Stroke Foundation and the Human Research Ethics Committee from Monash University (Project ID 8842).

Results

Overall, 25/129 of the organisational variables, and 73/174 of the clinical variables from the audit cycle in 2019 could be mapped across all the surveys and audits undertaken between 1999 and 2019.

Organisational survey

Between 1999 and 2019, a total of 197 hospitals, each reporting >40 stroke admissions annually, provided data for the organisational survey (Online Supplement Table S2). A total of 149 hospitals completed the 1999 and/or 2004 surveys, and 122 of these also participated in more than one of the subsequent cycles in the Audit programme. Over 90% of hospitals participating in each cycle were public hospitals; the only exception was the 1999 survey, which had a relatively greater number of private hospitals (17%). Between 1999 and 2019, there was a 93% increase in the proportion of hospitals with a stroke unit (2019: 42%, 2019: 81%), a 13-fold increase in hospitals providing intravenous thrombolysis (1999: 6%, 2019: 85%), and a 4-fold increase in hospitals offering rapid assessment and management of people with suspected transient ischaemic attack (1999: 11%, 2019: 61%; Online Supplement Table S3). Access to all allied health professionals for patients with stroke increased from 1999 to 2019 (Online Supplement Table S4). Since 2009, over 77% of hospitals reported access to physiotherapy, occupational therapy, speech pathology, social work, and dietetics.

Analyses of Audit programme organisational data (collected 2007–2019) found that the odds of hospitals using protocols and providing timely access to care increased significantly per audit cycle (Table 1). For example, the use of ED protocols for rapid triage of patients with stroke increased significantly from 49% in 2007 to 95% in 2019 (OR 1.42 per cycle, 95% CI 1.33, 1.52), as did timely access to carotid Dopplers (OR 1.31 per cycle, 95% CI 1.24, 1.39), and magnetic resonance imaging (OR 1.16 per cycle, 95% CI 1.11, 1.22). Focussing on the 2015–2019 period, we found a two-fold increase in the odds of hospitals offering intravenous thrombolysis 24 hours a day, seven days per week (OR 2.11 per cycle, 95% CI 1.44, 3.10) and of hospitals accessing telehealth for clinical support (OR 2.0 per cycle, 95% CI 1.50, 2.67). Potential ceiling effects with ED protocols and access to timely carotid Dopplers and computerised tomography brain scans were seen (Table 1).

Table 1.

Organisational resources reported for hospitals providing acute care for stroke in Australia, 2007–2019.

	Participating hospitals							2007-2019 OR per cycle (95% CI)^a	2015-2019 OR per cycle (95% CI)^a
Organisational resource	2007 N = 131 n (%)	2009 N = 116 n (%)	2011 N = 112 n (%)	2013 N = 120 n (%)	2015 N = 112 n (%)	2017 N = 117 n (%)	2019 N = 108 n (%)	2007-2019 OR per cycle (95% CI)^a	2015-2019 OR per cycle (95% CI)^a
ED protocols for rapid triage	64 (49)	63 (54)	83 (74)	99 (83)	100 (89)	109 (93)	103 (95)	1.42 (1.33, 1.52)	1.36 (0.95, 1.95)
Access to carotid Doppler	118 (90)	109 (94)	108 (96)	117 (97)	—	—	—	—	—
Within 24 hours	50 (38)	90 (78)	97 (87)	94 (78)	97 (87)	108 (92)	99 (92)	1.31 (1.24, 1.39)	1.45 (0.99, 2.13)
Brain imaging (onsite/offsite)
Access to computerised tomography	125 (95)	113 (97)	112 (100)	119 (99)	—	—	—	—	—
Within 24 hours	121 (92)	112 (97)	110 (98)	119 (99)	108 (96)^b	116 (99)^b	108 (100)^b	—	5.32 (0.71, 39.9)
Access to magnetic resonance imaging	97 (74)	89 (77)	91 (81)	94 (78)	—	—	—	—	—
Within 24 hours	69 (53)	68 (59)	74 (66)	70 (58)	82 (73)	91 (78)	83 (77)	1.16 (1.11, 1.22)	1.29 (0.93, 1.79)
Provides intravenous thrombolysis^d	49 (37)	51 (44)	62 (55)	81 (68)	85 (76)	91 (78)	92 (85)	2.13 (1.70, 2.67)	1.36 (1.04, 1.78)
Offered 24/7	—	40 (34)	53 (47)	69 (58)	77 (69)	82 (70)	86 (80)	1.80 (1.55, 2.08) ^c	2.11 (1.44, 3.10)
Access to telehealth for clinical support	77 (59)	64 (55)	72 (64)	79 (66)	48 (43)^e	58 (50)^e	76 (70)^e	—	2.0 (1.50, 2.67)
Admit all patients with transient ischaemic attack, or access to rapid outpatient assessment clinic	—	51 (44)	53 (47)	65 (54)	54 (48)	60 (51)	66 (61)	1.10 (1.03, 1.16) ^c	1.37 (1.09, 1.74)
Stroke unit	54 (41)	67 (58)	73 (65)	91 (76)	88 (79)	94 (80)	88 (81)	1.54 (1.39, 1.71)	1.35 (0.90, 2.05)

OR: odds ratio; CI: confidence interval; ED: Emergency Department; 24/7: 24 hours a day, 7 days a week.

^ayear of cycle as independent variable, clustered by hospital.

^bwithin 3 hours.

^c2009–2019 year analysis.

^dischaemic stroke.

^eused in last 6 months.

Clinical audit

A total of 24,996 cases from 136 hospitals were audited between 2007 and 2019 (126 participated in ≥2 cycles; Online Supplement Table S5). There were significant differences in characteristics of patients between audit cycles (Table 2). Across the cycles, the median age ranged from 75 to 77 years, just over half of the patients in each audit were male, and less than 3% of patients were documented to identify as Aboriginal and/or Torres Strait Islander. Between 2007 and 2019, the proportion of patients with ischaemic stroke increased from 73% to 84% (p < 0.001), while reporting of ‘undetermined’ stroke type decreased from 14% to 4% (p < 0.001). The severity of stroke, defined by the proportions of patients with arm deficits, speech/communication impairments and the inability to walk on admission, or having incontinence within 72 hours of stroke onset, decreased between 2007 and 2019 (Table 2).

Table 2.

Characteristics of patients included in Audit programme cycles in Australia, 2007–2019.

	2007 N = 2663 n (%)	2009 N = 3214 n (%)	2011 N = 3415 n (%)	2013 N = 3576 n (%)	2015 N = 4012 n (%)	2017 N = 4090 n (%)	2019 N = 4026 n (%)	p-value^a
Patient demographics
Median age in years (Q1, Q3)	76 (65, 83)	77 (66, 84)	76 (65, 84)	76 (65, 84)	75 (65, 84)	75 (65, 84)	75 (65, 84)	0.02
Sex, male	1390 (52)	1693 (53)	1815 (53)^d	1952 (55)^d	2203 (55)^d	2249 (55)	2286 (57)^d	<0.001
Patient identifying as Aboriginal and/or Torres Strait Islander	33 (1)	64 (2)^f	97 (3)^f	113 (3)^e	106 (3)^e	96 (2)^e	136 (3)^e	<0.001
Patients with risk factors prior to admission
Atrial fibrillation	668 (28)^f	816 (31)^g	955 (36)^g	968 (33)^g	1022 (29)^g	1002 (27)^f	1009 (26)^e	<0.001
Previous stroke	616 (26)^f	725 (27)^g	775 (28)^g	729 (25)^g	901 (26)^g	873 (23)^f	893 (23)^e	<0.001
Previous TIA	417 (18)^g	457 (18)^g	560 (22)^g	497 (18)^g	542 (16)^g	483 (13)^f	468 (13)^f	<0.001
Diabetes mellitus	577 (24)^f	739 (27)^g	824 (29)^g	901 (30)^g	989 (28)^g	985 (26)^f	1088 (28)^f	0.09
Hypercholesterolemia	888 (40)^g	1174 (45)^g	1372 (51)^g	1413 (48)^g	1489 (43)^g	1571 (42)^f	1565 (41)^f	0.002
Hypertension	1785 (71)^e	2119 (72)^f	2298 (74)^f	2398 (73)^f	2624 (70)^f	2700 (69)^e	2652 (68)^e	<0.001
Ischaemic heart disease	757 (32)^c,g	803 (30)^g	904 (34)^g	899 (31)^g	991 (29)^g	944 (25)^f	961 (25)^f	<0.001
High alcohol consumption	222 (11)^g	283 (13)^g	335 (14)^g	345 (13)^g	396 (13)^g	335 (10)^g	373 (11)^g	0.04
Current smoker	390 (17)^g	493 (20)^g	491 (19)^g	525 (19)^g	532 (17)^g	561 (16)^g	562 (17)^g	<0.001
Pre-stroke function
Independence prior to admission (mRS 0–2)	2128 (82)^e	2002 (68)^f	2273 (73)^f	2388 (73)^f	3210 (80)	3292 (80)	3311 (82)	<0.001
Stroke type
Ischaemic	1914 (73)	2384 (76)	2560 (78)	2809 (80)	3162 (79)	3367 (82)	3376 (84)	<0.001
Intracerebral haemorrhage	334 (13)	500 (16)	512 (15)	419 (12)	541 (13)	503 (12)	478 (12)	<0.001
Undetermined	354 (14)	254 (8)	221 (7)	290 (8)	309 (8)	220 (6)	172 (4)	<0.001
Stroke severity
Arm deficit on admission	1904 (74)^e	2208 (71)^e	2289 (70)^e	2325 (67)^e	2366 (62)^e	2356 (60)^e	2343 (60)^e	<0.001
Speech/communication impairment on admission	1657 (66)^e	2012 (67)^f	2049 (64)^e	2005 (58)^e	22487 (59)^e	2243 (57)^e	2183 (56)^e	<0.001
Incontinence within 72 hours^b	962 (38)^e	1298 (43)^e	1289 (40)^e	1277 (37)^e	1333 (34)^e	1350 (35)^e	1231 (31)^e	<0.001
Inability to walk on admission	1660 (66)^f	2155 (69)^e	2162 (64)^e	2389 (69)^e	2161 (55)^e	2112 (54)^e	2206 (56)^e	<0.001

Q1: 1^st quartile; Q3: 3^rd quartile; TIA: transient ischaemic attack; mRS: modified Rankin Scale.

^astatistical test for trend 2007 to 2019.

^bfrom stroke onset, except 2013 which was from emergency department presentation.

^cmyocardial infarction.

^d<1% unknown/missing responses.

^e1–5% unknown/missing responses.

^f6–10% unknown/missing responses.

^g11–25% unknown/missing responses.

Figure 2 shows trends in adherence to recommended care processes in each cycle 2007–2019 and Figure 3 the odds for change over time (also Online Supplement Table S6 and Figure S1). For the majority of the recommended care processes, the odds of adherence increased per cycle, for example, stroke unit access (2007: 52%, 2019: 69%; OR 1.15 per cycle, 95% CI 1.14, 1.17), provision of intravenous thrombolysis for patients with ischaemic stroke (2007: 3%, 2019: 11%; OR 1.15 per cycle, 95% CI 1.13, 1.17) and carer training (2007: 24%, 2019: 51%; OR 1.12 per cycle, 95% CI 1.10, 1.15) (Figures 2 and 3). Conversely, the odds of adherence over time fell for prescription of antihypertensives on discharge (OR 0.98 per audit cycle, 95% CI 0.97, 0.99) and there was no difference in the provision of antithrombotic medications on discharge for patients with ischaemic stroke.

Figure 2.

Age-, sex-, and stroke severity-adjusted adherence to evidence-based care processes over time (2007–2019).

Figure 3.

Association between repeated audit cycles and adherence to evidence-based care processes over time (2007–2019).

Restricting analyses to the 2015–2019 period, we found significant increases in the odds for use of swallow screen/assessment prior to food, fluids or oral medication (2015: 58%, 2019: 64%; OR 1.08 per cycle, 95% CI 1.05, 1.11), provision of advice on risk factor modification (2015: 53%, 2019: 63%; OR 1.18 per cycle, 95% CI 1.13, 1.22), and care plan development (2015: 58%, 2019: 68%; OR 1.21 per cycle, 95% CI 1.17, 1.26 (Figure 3, Online Supplement Figure S1). There was no significant difference in stroke unit access or provision of aspirin within 48 hours for those with ischaemic stroke during this period.

Regardless of improvements over time, the 2019 audit findings indicated that three in four patients were not assessed for mood problems during their hospital admission (Online Supplement Table S6), and half of all carers had not received training.

Discussion

This study provides evidence of significant improvements in the organisation of acute stroke care and adherence to most guideline-recommended care processes over the period 1999–2019 in Australia. These included significant improvements in access to stroke units, thrombolysis services, and rapid assessment and management options for patients with transient ischaemic attack. The odds of patients receiving many recommended processes related to acute care, secondary prevention, interdisciplinary assessment and transition from hospital also improved from 2007 to 2019. However, gaps remained in other aspects of evidence-based care, particularly in relation to mood assessments.

Data from the Audit programme provide a snapshot of stroke care and a means for clinicians, health administrators, and government to monitor and guide improvements for acute stroke care. Although changes cannot be solely attributed to participation in the Audit programme, it can be considered an ‘audit and feedback’ intervention, with national and hospital reports distributed by the Stroke Foundation. Therefore, our findings compare favourably to the 4% absolute improvement (+0.5%, +16%) effects of ‘audit and feedback’ interventions reported elsewhere.¹⁴ However, ongoing consideration of methods of feedback delivery is important. The current Stroke Foundation Strategy Plan is focused on how best to facilitate and support targeted quality improvement programmes, particularly in hospitals performing below standard, to maximise potential improvement and change.¹⁵

While our findings for care improvements align with those reported in China,¹⁶ they are more modest compared to similar programmes in the USA such as the Get-With-The-Guidelines-Stroke (GWTG-Stroke)¹⁷ or the Paul Coverdell Stroke Program¹⁸, or the Sentinel Stroke National Audit Programme in the United Kingdom.¹⁹ However, cross-country comparisons are challenging because of differences in nature and scope of programmes.²⁰ For example, GWTG-Stroke also involves a series of quality improvement cycles and collaborative workshops to develop hospital protocols and processes.¹⁷ Multinational collaborations have been formed to develop a common set of process or performance measures to allow for cross-country comparison of stroke care quality.²¹

Stroke unit care is the most universally applicable recommendation for improving stroke outcomes.²² Patients treated in a stroke unit are more often provided with evidence-based care,²³ and have a reduced odds of death and disability of approximately 20%.²² We found an increase in the odds of patients being treated in a stroke unit from 2007 to 2019, but adherence appears to have stalled in recent years, with only two in three patients accessing stroke unit care 2019. There is a need to ensure that there are systems, infrastructure, and resources in place for all patients with stroke to be treated in a stroke unit as recommended in the clinical guidelines and Acute Stroke Services Framework for Australia.¹⁰

We found improvements in the delivery of intravenous thrombolysis, a time-critical therapy that reduces death and disability from ischaemic stroke,²⁴ over time. This improvement may have been influenced by increased use of telemedicine, shown to have increased thrombolysis rates in rural hospitals in Victoria (Australia).²⁵ Yet, only one in ten patients with ischaemic stroke received intravenous thrombolysis in the 2019 audit, and only a third received this within 60 minutes of arrival at hospital,²⁶ which is considerably lower than what has been reported in other countries.^17,27 Wider improvements are thus needed, with the introduction of the Mobile Stroke Unit²⁸ promising to increase access to relevant interventions.

Observed improvements in stroke care may be explained by concurrent changes in the organisation of hospital services and resources with, for example, diagnostic scans now provided to almost all patients with stroke, enabling identification of eligible cases for reperfusion therapies, and minimising the proportion of undetermined stroke type. Also, the introduction of triage protocols for patients with stroke within EDs and wider availability of telehealth modalities all increased the chances of patients receiving time-sensitive reperfusion therapies. Other aspects of acute care delivery that have improved since 2007 include access to timely allied health, involvement of the patient and family in management decisions, education for risk factor management, and support for patients and carers returning to the community. Continued efforts are required to maintain improvements in the organisation of stroke care and ensure continued monitoring of performance in future audits.

Strengths and limitations

This study was able to draw on a large comprehensive national dataset including organisational and clinical care processes developed through analytical and consensus methods and evaluated using a standardised approach over many years. The audit has been relatively adaptable, with questions/response options changed and updated to reflect clinical practice recommendations at the time. While this meant that we were unable to investigate all potentially important care processes from recent guideline recommendations, the care processes included in this study cover most important areas of acute stroke care.

We previously reported between-hospital variations in stroke care according to state, rurality, and size, finding access to stroke unit care to be more likely in hospitals in urban areas and in hospitals with a larger number (350+) of annual stroke admissions.¹¹ While the introduction of telemedicine options has led to more equitable access to care including thrombolysis in regional/rural locations,²⁶ there remains a need for continued efforts to reduce disparities to ensure equitable care for patients with stroke.

Our analysis was limited to hospitals participating in the Audit programme and we cannot assume that the cases audited are representative of all patients with stroke. Contextual information was not available on hospitals that did not participate in the Audit programme-although over 80% of eligible hospitals identified participated in each cycle.²⁶ Missing data and other forms of bias associated with the quality of information may have influenced the data quality used in the analysis. To improve reporting practices, hospital staff received training prior to completion of the audits and had access to ongoing support from Stroke Foundation staff. In-built logic checks in the data collection tool also assisted in minimising data collection error, with the inter-reliability of responses between different auditors shown to be substantial or better in prior research.²⁹ Finally, a small number of care processes analysed (e.g. antihypertensive medications) appeared to follow a non-linear trend, which was not accounted for in our analysis of change over time. Further understanding the nuances as to why adherence to some care processes may have fluctuated over cycles is important.

Data from the biennial Audit programme have informed important clinical and policy initiatives, and quality improvement programmes in Australia. However, the establishment of the Australian Stroke Clinical Registry (AuSCR) in 2009 has facilitated prospective monitoring of the quality of acute stroke care beyond the standard reporting periods used for the Audit programme. While not all hospitals participate in the AuSCR, at an individual level, the AuSCR data are considered to be more representative for care processes that are provided in only a subset of patients such as thrombolysis and endovascular clot retrieval. As data for the Audit programme and the AuSCR are collected within the same integrated data management system, the Australian Stroke Data Tool,³⁰ data collection burden is minimised for clinicians regarding core variables (e.g. age, sex, type of stroke, access to stroke unit care). Overall, data from the Audit programme and the AuSCR are considered complimentary. The Audit contributes organisational details and a comprehensive snapshot of the care provided to patients with stroke, while the AuSCR provides ongoing, routinely collected minimum data, and additional information on patient outcomes post-discharge (mortality and health-related quality of life).¹² In future, it is important that there is continued monitoring of acute care for patients with stroke and that proactive efforts are made to reduce gaps identified in best-practice care using data from these two collection programmes.

Conclusion

Using data from the National Stroke Audit programme, we were able to reliably demonstrate the evolution of acute stroke services in Australia. Many aspects of stroke service provision and care delivery have improved over the period 1999–2019, but continued efforts are required to ensure further development and equity in access to best practice care. Investment in standardised monitoring of stroke care can provides insight for targeted efforts to reduce identified gaps in best-practice, and can enable the assessment of progress over time.

Supplemental Material

Supplemental Material - Twenty years of monitoring acute stroke care in Australia through the national stroke audit programme (1999–2019): A cross-sectional study

Supplemental Material for Twenty years of monitoring acute stroke care in Australia through the national stroke audit program (1999–2019): A cross-sectional study by Tara Purvis, Dominique A Cadilhac, Kelvin Hill, Megan Reyneke, Muideen T Olaiya, Lachlan L Dalli, Joosup Kim, Lisa Murphy, Bruce CV Campbell and Monique F Kilkenny in Journal of Health Services Research & Policy.

Footnotes

Acknowledgements

We thank the hospital clinicians and patients who contributed data to the audits. We acknowledge Justine Watkins, Patrick Young, Elizabeth Freeman, Jo Maxwell and Dawn Harris for their contributions as coordinators of the Audit programme over the years. In addition, we recognise the contributions of Toni Aslett in her role as the Executive Director of Stroke Services at the Stroke Foundation, Sharon McGowan as Chief Executive Officer, and Erin Lalor as prior Chief Executive Officer of the Stroke Foundation, and the Clinical Council members.

Author contributions

MFK, TP, DAC, and KH were involved in various aspects of the study design. TP, MTO, LLD, JK, MFK, and DAC contributed to the statistical analysis approach, data analysis and interpretation. TP drafted the paper, and all authors (DAC, KH, MR, MTO, LLD, JK, LM, BCVC, MFK) critically revised successive drafts of the manuscript. All authors have read and approved the final version of the manuscript.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Monash University researchers were commissioned by the Stroke Foundation to independently produce a report based on 1999–2017 data, which is publicly available (). This paper provides a more contemporary understanding as it includes the 2019 audit cycle, with updated analysis methods included. DAC acknowledges research fellowships from the National Health and Medical Research Council (NMHRC) (1063761 co-funded by Heart Foundation; 1154273). MFK acknowledges support from an Early Career Fellowship from the NHMRC (1109426).

Ethical approval

ORCID iDs

Tara Purvis

Dominique A Cadilhac

Supplemental Material

Supplemental material for this article is available online.

References

Abilleira

Ribera

Permanyer-Miralda

, et al. Noncompliance with certain quality indicators is associated with risk-adjusted mortality after stroke. Stroke 2012; 43: 1094–1100.

Cadilhac

Andrew

Lannin

, et al. Quality of acute care and long-term quality of life and survival: the Australian stroke clinical registry. Stroke 2017; 48: 1026–1032.

Smarreli

Cadilhac

Pearce

. Access to organised stroke care in Australia: results of a national survey of acute stroke treating hospitals. Abstract - European Stroke Conference. Cerebrovasc Dis 2001; 11: 93.

Ebrahim

Redfern

. Stroke Care - a matter of chance. A national survey of stroke services. London: Stroke Association, 1999.

Rudd

Irwin

Rutledge

, et al. The national sentinel audit for stroke: a tool for raising standards of care. J R Coll Physicians Lond 1999; 33: 460–464.

Harris

Cadilhac

Hankey

, et al. National Stroke Audit: The Australian Experience. Clinical Audit 2010; 2: 25–31.

von Elm

Altman

Egger

, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: Guidelines for Reporting Observational Studies. Plos Med 2007; 4: e296.

Cadilhac

Lalor

Pearce

, et al. Access to stroke care units in Australian public hospitals: facts and temporal progress. Intern Med J 2006; 36: 700–704.

Stroke Foundation . Clinical guidelines for stroke management 2017. Melbourne: Stroke Foundation, 2017.

10.

National Stroke Foundation . National acute stroke services Framework 2015. Melbourne: National Stroke Foundation, 2015.

11.

Cadilhac

Purvis

Reyneke

, et al. Evaluation of the national stroke audit Program: 20-year report. Melbourne: Stroke Foundation, 2019.

12.

Cadilhac

Lannin

Anderson

, et al. Protocol and pilot data for establishing the Australian Stroke Clinical Registry. Int J Stroke 2010; 5: 217–226.

13.

Counsell

Dennis

McDowall

, et al. Predicting outcome after acute and subacute stroke: development and validation of new prognostic models. Stroke 2002; 33: 1041–1047.

14.

Ivers

Jamtvedt

Flottorp

, et al. Audit and feedback: effects on professional practice and healthcare outcomes. Cochrane Database Syst Rev 2012; 6: CD000259.

15.

Stroke Foundation . Stroke Strategy 2024: empowering, supporting, connecting. Melbourne: Stroke Foundation, 2022.

16.

Wang

Zhao

, et al. Substantial Progress Yet Significant Opportunity for Improvement in Stroke Care in China. Stroke 2016; 47: 2843–2849.

17.

Ormseth

Sheth

Saver

, et al. The american heart association's get with the Guidelines (GWTG)- Stroke development and impact on stroke care. Stroke Vasc Neurol 2017; 2: 94–105.

18.

Overwyk

Yin

Tong

, et al. Defect-free care trends in the Paul Coverdell National Acute Stroke Program, 2008-2018. Am Heart J 2021; 232: 177–184.

19.

Sentinel Stroke National Audit Programme , School of Population Health and Environmental Sciences, King’s College London, on behalf of the Intercollegiate Stroke Working Party. Sentinel Stroke National Audit Programme Annual Public Report 2018. London: SSNAP, 2019.

20.

Cadilhac

Kim

Lannin

, et al. National stroke registries for monitoring and improving the quality of hospital care: a systematic review. Int J Stroke 2016; 11: 28–40.

21.

Norrving

Bray

Asplund

, et al. Cross-national key performance measures of the quality of acute stroke care in western europe. Stroke 2015; 46: 2891–2895.

22.

Stroke Unit Trialists' Collaboration . Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev 2013; 2013(9): CD000197.

23.

Cadilhac

Ibrahim

Pearce

, et al. Multicenter comparison of processes of care between Stroke Units and conventional care wards in Australia. Stroke 2004; 35: 1035–1040.

24.

Wardlaw

Murray

Berge

, et al. Thrombolysis for acute ischaemic stroke. Cochrane Database Syst Rev 2014; 2014(7): CD000213.

25.

Bladin

Moloczij

Ermel

, et al. Victorian Stroke Telemedicine Project: implementation of a new model of translational stroke care for Australia. Intern Med J 2015; 45: 951–956.

26.

Stroke Foundation . National stroke audit. Acute services report 2019. Melbourne: Stroke Foundation, 2019.

27.

Stanley

Bhalla

James

, et al. Moving the dial of stroke care. In: The Sixth SSNAP Annual Report. United Kingdom: Healthcare Quality Improvement Partnership, 2020.

28.

Zhao

Coote

Easton

, et al. Melbourne mobile stroke unit and reperfusion therapy. greater clinical impact of thrombectomy than thrombolysis. Stroke 2020; 51: 922–930.

29.

Stroke Foundation . National stroke audit acute services report 2021. Supplementary report. Melbourne: Stroke Foundation, 2021.

30.

Ryan

Ghuliani

Grabsch

, et al. Development, Implementation and Evaluation of the Australian Stroke Data Tool (AuSDaT): comprehensive data capturing for multiple uses. Health Inf Manag 2022; 1: 183335832211171. DOI: 10.1177/18333583221117184.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.71 MB