View Article PDF

Stroke is now the third most common cause of death and the leading cause of serious adult disability in New Zealand.[[1]] The annual number of people with strokes is expected to rise by a further 40% over the next decade.[[2]]

Following an ischaemic stroke, brain tissue may be rescued if blood flow is restored rapidly. Intravenous thrombolysis (IVT) and stroke clot retrieval (SCR) using stent retrievers significantly improve the odds of disability-free recovery.[[3,4]]

In New Zealand, IVT is provided at all 28 acute stroke hospitals, of which 16 receive remote Telestroke support.[[2]] SCR is provided at three of these stroke centres.[[2]] SCR treatment is limited to a subset of patients with large vessel occlusion (LVO) and consequently more patients qualify for IVT compared to SCR. However, in patients with LVO, SCR offers greater absolute treatment benefit because these are the most severely affected patients and SCR is more effective in clearing large clots than IVT. The pivotal SCR clinical trials all used SCR preceded by IVT treatment, although trials are ongoing to explore SCR without IVT.[[3]]

Stroke thrombolysis rates have varied between centres in New Zealand with key barriers including poor access to experienced stroke physicians in regional settings.[[5]] National and regional initiatives to address these inequalities have included regular audit and reviews, service model changes to reduce treatment delays, public education campaigns to improve community stroke recognition, and the introduction of Telestroke.[[6–8]]

Unfortunately, some population groups in New Zealand appear consistently more vulnerable. Between 1983 and 2003, NZ Europeans have seen a 19% decrease in total stroke incidence, while Pacific patients have seen a 66% increase and Māori a 2% increase.[[9]] Furthermore, while age-adjusted stroke incidence rates in NZ Europeans have declined by 19%, these rates have increased by 16% in Māori and 21% in Pacific peoples.[[9]] The high stroke incidence in Māori and Pacific peoples, combined with the comparatively young onset of strokes in these populations, creates significant burdens on Māori and Pacific stroke survivors, their whānau, communities and society. Reperfusion access by ethnicity has previously been explored in the Northland Region of New Zealand and did not find significant differences.[[10]] Comprehensive data on reperfusion metrics for the entire country have not been previously published.

Women often experience poorer post-stroke outcomes than men due to sex-related differences in presentation and risk factors. Women are more likely to present with strokes at an older age and have worse pre-stroke functionality compared with men.[[11]] Studies have also found that women are less likely to receive IVT than men; the cause of which remains unclear.[[12]] To date, potential sex differences in stroke reperfusion therapy in New Zealand have not been explored.

This study reports on the reperfusion therapy rates in 2019 and 2020 to assess the impact of service improvements on temporal trends, along with the identification of areas that require further improvement. We also explored the association of demographic factors, specifically ethnicity and sex, with reperfusion therapy access and quality metrics.

Design and methods

This is a registry-based prospective observational study including all adult patients (age >16) who received reperfusion therapy in New Zealand between 1 January 2015 and 31 December 2020. All stroke patients who receive reperfusion therapy (IVT and/or SCR) in New Zealand are routinely entered into a national online database using the open source REDCap platform. Data is entered by staff at the primary stroke centre or by a member of the central database administration team. The data in the register was cleaned manually and scrutinised for missing values, data entry errors and notable outliers. Duplicate entries were identified using the National Health Index number. Hospital staff nationwide were contacted for further information where necessary. In cases where important information was missing or missing values could not be reconciled, the patient in question was excluded from both numerator and denominator to complete specific analyses.

The primary patient efficacy end point for the National Stroke Register is day seven vital status with optional reporting of modified Rankin Scale (mRS) at three months. The mRs is a seven point scale with zero normal and six dead, and where independence is defined as a score of 0,1 or 2. The primary safety end point is the rate of symptomatic intracerebral haemorrhage (sICH).[[13]] IVT and SCR rates are calculated using denominator data supplied by the Ministry of Health based on recorded discharge coding. The denominator data includes all patients with ischaemic stroke (ICD 10-AM I63) and strokes “unspecified” (ICD 10-AM I64). Strokes “unspecified” have been included as internal audits have found that most cases represent miscoded ischaemic strokes.

In 2020, we became aware that Ministry of Health data includes multiple duplicates, where patients transferred between hospitals are counted as multiple discharges, despite all discharges referring to a single stroke event. We have now removed all duplicates dating back to 2017. Data for 2015 and 2016 were not available for this reanalysis. This means some of the denominators and intervention rates presented in this paper differ slightly from previously published reports raising historical intervention rates. As we were unable to retrospectivity amend data prior to 2017, this needs to be borne in mind when comparing trends from pre-2017 with post-2017. Due to the slightly inflated denominator for these years the actual thrombolysis rates will have likely been approximately 0.5% higher for thrombolysis and 0.2% higher for stroke clot retrieval than the figures displayed and previously published.[[2]]

Continuous variables were non-normally distributed and were thus analysed and reported as the median value, with additional values denoting IQR of 25[[th]] to 75[[th]] percentiles. Comparative differences were analysed using Wilcoxon rank sum test and presented as p-values. Rate-based results were presented as incidence rates. Comparative differences were analysed using incidence rate comparison and presented with p-values +/- 95% confidence intervals. A p-value less than 0.05 was considered statistically significant. Data analysis was completed in StataIC 16.0

Results

Between 1 January and 31 December 2020, there were 7333 patients with acute ischaemic stroke (n= 7021) or “stroke unspecified” (n= 312). Of this number, 828 (11.3%, 11.8% excluding “stroke unspecified”) were treated with IVT and 404 (5.5%, 5.6% excluding “stroke unspecified”) were treated with SCR. A total of 201 patients were treated with both IVT and SCR (24.3% of all IVT patients). None of these patients were confirmed positive for COVID-19. Baseline characteristics are listed in Table 1.

For IVT in 2020, the median (IQR) door-to-needle time was 61 (41–87) minutes (Table 2). sICH occurred in 24 patients (2.9%). At day seven, 66 (8.1%) patients had died and at three months, 224 (61.7%) of 363 patients where this information was recorded were functionally independent (mRS 0–2). For SCR in 2020, the median (IQR) arrival-to-groin time was 55 (22–113) minutes, sICH rate was 1.2% and 141 of 286 (49.3%), where this information was recorded, were independent at three months (Table 3).

IVT rates have increased from 6.5% in 2015 to 11.3% in 2020 (p=<0.001), but have remained unchanged since 2019 (11.3% both years). The median (IQR) door-to-needle time has reduced from 74 (55–102) in 2015 to 61 (41–87) in 2020, but has remained stable since 2019 when it was 61 (44–84) minutes. The percentage of patients treated with IVT under 60 minutes has also remained stable between 2019 and 2020 (49.1% to 49.7%; p= 0.99). There has been no significant change observed in the number of patients who have died by day seven between 2015 and 2020 (6.7% to 8.1%; p= 0.45). Post-IVT sICH rates have reduced (6.2% in 2015 to 2.9% in 2020; p= 0.01) (Table 2).

SCR rates have increased substantially from 2015 to 2020 (0.5% to 5.5%; p<0.001), but only modestly across the last two years (4.8% in 2019 to 5.5% in 2020; p= 0.08). Arrival-to-groin time remains stable at (median (IQR)) 52 (16–109) minutes in 2019 to 55 (22–113) minutes in 2020 (p=0.45), as has onset-to-reperfusion time; (median (IQR)) 335 (235–522) minutes in 2019 to 365 (245–550) minutes in 2020 (p=0.27). There have been no statistically significant changes between 2019 and 2020 with regard to sICH rates (2.9% to 1.2%; p=0.12), day seven mortality rates (9.1% to 7.5%; p= 0.47), and the number of independent patients at three months (44.6% to 49.3%; p=0.44) (Table 3). NZ European patients residing in DHBs that are SCR centres experienced shorter onset-to-reperfusion times than those in a non-SCR centre DHB of domicile; 286 (206–566) and 403 (295–550) minutes (p<0.001) respectively.

Ethnicity did not have a significant impact on IVT rates in 2020 with 11.7% in NZ Europeans compared with 10.1% in Māori (95% CI 0.92, 1.45; p= 0.20) and 11.6% in Pacific patients (95% CI: 0.75, 1.36; p= 1.00). Variation in SCR rates among different ethnic groups also did not differ with 5.4% of NZ Europeans compared to 5.3% of Māori (95% CI 0.75, 1.43; p= 0.88) and 6.8% of Pacific patients (95% CI: 0.54, 1.19; p= 0.23). Reperfusion rates with IVT and or SCR were 13.3% in NZ Europeans, compared to 12.3% in Māori (95% CI: 0.89, 1.34; p= 0.42), and 13.9% in Pacific patients (95% CI: 0.74, 1.26; p= 0.73).

Looking at a two-year cohort (2018–2020) Māori and Pacific patients were significantly younger than their NZ European counterparts (60.3 years vs 74 years; p<0.001, and 63 years vs 74 years; p<0.001, respectively). Onset-to-door time was significantly higher in NZ European patients (median (IQR)) 81 (54–123) minutes compared with Māori; 72 (44–112) minutes (p<0.001), and Pacific patients; 70 (48–105) minutes (p= 0.03). Door-to-needle time in 2018–2020 was longer in Māori patients compared with NZ Europeans; (median (IQR)) 66 (48–88) minutes compared to 59 (41–83) minutes (p=0.001). A sensitivity analysis explored whether the difference in slower door-to-needle time for Māori could be due to a higher proportion residing rurally, with previous reports showing slower door-to-needle time in non-urban settings.[[2,9]] Indeed, the sensitivity analysis found that door-to-needle time for NZ Europeans residing in non-urban DHBs was significantly longer than for urban NZ European door-to-needle time: (median (IQR)) 62 (44–86) minutes compared to 57 (40–82) minutes (p= 0.02), respectively. The same trend was observed among Māori. However, this was not statistically significant. Māori non-urban door-to-needle time (median (IQR)) 70 (51–92) minutes compared to Māori urban door-to-needle time (median (IQR)) 63 (42–87) minutes (p=0.11). However, Māori also had longer door-to-needle time in the urban setting when compared to urban NZ Europeans: (median (IQR)) 63 (42–87) minutes compared to 57 (40–82) minutes (p=0.048). Similarly, non-urban Māori had a significantly longer door-to-needle times compared to their non-urban NZ European counterparts; (median (IQR)) 70 (51–92) minutes compared to 62 (44–86) minutes (p= 0.012). There was no significant difference between NZ European and Pacific patients in door-to-needle time (p= 0.28). Overall onset-to-needle time was similar across ethnic groups (NZ European vs Māori; p= 0.16, NZ European vs Pacific; p= 0.37, Māori vs Pacific; p= 0.99).

In 2019, there were overall more men than women who experienced ischaemic events (3661 males versus 3492 females). Female stroke patients were on average 4.4 years older than males upon presentation (73.7 years versus 69.3 years respectively, p<0.001). The rate of IVT performed in females (10.2%) was lower than that in males (12.2%); p= 0.01. However, there was no significant difference in the rate of SCR (female 4.6%; male 5.1%; p= 0.29). The overall reperfusion rate was lower for females than males (12.7% vs 14.9% p= 0.02). Median door-to-needle, onset-to-needle, and onset-to-groin times, sICH and day seven mortality were similar between females and males (Table 4).

Discussion

IVT and SCR rates have risen considerably since 2015. IVT rates have seen a steady rise from 6.5% to 11.3% and SCR rates have risen even more dramatically, from 0.5% to 5.5%. This is likely related to increasing implementation of SCR following the publication of seven pivotal trials on SCR efficacy in 2015 and 2016.[[3]] Since this initial jump, SCR rates have continued to increase more modestly.

The acute stroke reperfusion therapy rate with IVT and or SCR reached 14.0% in 2020 (14.6% when “stroke unspecified” patients are excluded). Eight hundred and twenty-eight patients were treated with IVT and 404 with SCR (201 received both) in a population of 5.1 million people. This equates to an IVT rate of 162 per million people and an SCR rate of 79 per million people. A number of quality improvement initiatives have led to the progress seen in stroke service provision. The implementation of the National Stroke Reperfusion Register in January 2015 has allowed for regular quality-control audits and has provided a focus for annual meetings to discuss national stroke reperfusion data and quality.[[2]] Such registers have also been shown to drive improvement in stroke service provision overseas.[[14]] Other initiatives include expanding Telestroke networks, regular Ministry of Health FAST campaigns, and hospital stroke service improvement projects.[[6–8]]

Our study found that IVT and SCR pre-hospital delays were higher in 2020 compared to 2019. While this was not statistically significant, it does follow an upward trend seen in the last six years. This is presumed to be, at least in part, a paradoxical finding secondary to an increased acceptance of treatment in extended time windows and increased treatment volumes at smaller centres. A drop-in pre-hospital care pathway performance is less likely.

SCR delays continue to improve which is unsurprising given SCR services continue to actively evolve. More work is ongoing under the recently launched Ministry of Health National Stroke Clot Retrieval Service Improvement Programme.[[15]] It is also encouraging to see stable complication rates for both IVT and SCR.

Despite many improvements, areas requiring continued effort have been identified. One such area is the door-to-needle time for IVT, which remains far from the recommended 30-minute target. Further, concerns are raised by the 2019 reduction in reported rate of three-month mRS (0–2) after SCR. Some fluctuation of results may be explained by the inconsistency in mRS data reporting. However, mRS had remained fairly stable between 2011 and 2018. It is possible that genuinely more complications were encountered in the past two years, and this will need to be further explored at the individual SCR centres, although the stable sICH rate provides some reassurance. A potential contributing factor may be that boundaries are being pushed with more patients referred for SCR with poorer baseline health status who do not entirely meet trial criteria. This is supported by the finding that over 50% of SCR patients did not receive IVT, suggesting they had IVT contraindications and may also be borderline SCR cases (“mercy cases”). Finally, with slowly increasing referrals from regional centres it is possible that significant transport delays mean more patients reach the angiography suite too late to benefit from the procedure. This area requires urgent attention and is a focus of the National SCR Programme.[[15]] To monitor SCR time delays and complications more effectively, we have recently introduced additional SCR time metrics and details around procedural complications that will be monitored over time.

The disproportionate burden of stroke and post-stoke complications on Māori and Pacific peoples have long been documented. Previous studies have shown a striking difference in the average age of stroke onset in Māori and Pacific peoples compared with NZ Europeans.[[9]] The results of our study align with existing literature in that Māori and Pacific stroke patients were 13.7 and 11 years younger on average than NZ Europeans.

Pacific people and Māori presented faster to hospital suggesting good stroke awareness and pre-hospital transport access in these populations. However, there were greater in-hospital delays for Māori. A higher proportion of Māori reside in rural areas compared with NZ Europeans and thus slower door-to-needle times at regional hospitals may disproportionally affect Māori.[[2,8,16]] However, a sensitivity analysis exploring the impact of hospital location found that Māori had slower door-to-needle times compared to NZ Europeans even when limiting the analysis to either urban or non-urban settings. While we were unable to adjust for other potential confounders in this study, the recently published REGIONS Care ethnicity analysis was fully adjusted for age, baseline morbidity, risk factors, stroke severity, and geography and identified additional areas of in-hospital stroke care inequities.[[17]] Such findings raise concern about potential racial discrimination within New Zealand stroke services. This will require careful consideration to plan potential interventions and should be monitored going forward.

As part of the 2019 analysis, we also explored sex-related differences in reperfusion therapy. It is well-known in the literature that women tend to present with stroke at an older age than men and with a different profile of risk factors.[[15]] Previous studies have also found that women are less likely to receive IVT despite gaining more net clinical benefit than men, especially at an older age.[[18]] While older age may make some clinicians more reluctant to treat and may thus provide a potential explanation, other studies have also found that suitable female candidates are more likely to be labelled stroke mimics and can inadvertently miss out on treatment.[[18]] This may be due to more frequent atypical presentations or unconscious bias. One study found that when age was corrected for, there were no significant sex-related differences in quality of care.[[18]] As our data were not adjusted for age, it should be viewed as explorative and interpreted with a degree of caution. Further analysis adjusting for age, severity, and baseline morbidity is important.

Our study has several limitations. Firstly, it is difficult to validate all data entered into the clinical register by up to 60 front line clinicians across the country. Therefore, despite the thorough and meticulous system of manual data checking and cleaning that has been implemented to mitigate any errors, it is possible that some mistakes remain. Secondly, the crude outcome measure of “alive or discharged at day seven” does not represent the long road of rehabilitation undertaken by stroke patients. To improve this, reporting of three-month mRS became mandatory in 2019. Despite this, only 50.1% of records included this information in 2020 (down from 54.3% in 2019). To fairly represent the New Zealand post-stroke experience, we must work to increase these numbers. While the main outcomes of interest were intervention rate, sICH and seven-day mortality, we conducted multiple additional analyses. Due to the number of comparisons conducted, there is a possibility that some significant results occurred based on chance alone and these secondary comparisons should thus be interpreted with caution. The inability to update denominator data for 2015 and 2016 will have impacted the precision of historical intervention rate comparisons, however; given the substantial increases in intervention rates over time it is very unlikely that the subtle changes in denominator values will have impacted the overall conclusions. Finally, the analyses by ethnicity and sex were not adjusted for potential confounders such as age, stroke severity, risk factors and time delays and should thus be viewed as primarily exploratory and interpreted with a degree of caution. A sample size was not prospectively determined, and we cannot exclude that significant findings were missed due to the lack of study power.

Despite these limitations, the completeness of data in our register—which includes every patient who underwent reperfusion in New Zealand—strengthens our observations. This minimises selection bias and allows for the accurate analysis of trends in stroke service provision over time.

Conclusion

Reperfusion therapy rates and treatment delays continue to improve although appear to gradually plateau. Complications and mortality have remained stable. The reduction in favourable outcomes following SCR over the last two years is likely related to widening inclusion criteria for patient selection and/or greater delays for more patients due to rising regional transfers. Both require further investigation and mitigation. Ethnic disparities identified include the longer door-to-needle time observed in Māori not explained by geographic factors. Women accessed reperfusion therapy less frequently, potentially explained by older age at presentation. Both require further exploration and action.

View Supplementary Tables.

Summary

Abstract

Aim

This study assessed stroke reperfusion treatments trends in 2019 and 2020 with comparison back to 2015. Additional analyses looked at differences by sex and ethnicity.

Method

The National Stroke Register contains data on all stroke patients who received reperfusion therapies since 2015. Outcomes included treatment rates, delays, mortality and complications by year, sex, and ethnicity. Continuous variables were compared using the Wilcoxon Rank-Sum Test and presented as p-values. Rate-based results were compared using incidence rate comparison and presented as p-values +/- 95% confidence intervals.

Results

In 2020, 11.3% (828/7333) received intravenous thrombolysis (IVT) and 5.5% (404/7333) underwent stroke clot retrieval (SCR), increasing from 6.5% (389/5963) and 0.5% (30/5963) in 2015, respectively. Among reperfused patients (IVT, SCR, both), 8.3% had died at seven days and 3.0% (29/959) experienced sICH. Door-to-treatment time was stable between 2019 and 2020, with median (IQR) of 61 (44–84) and 61 (41–87) minutes, respectively. Initial presentation to a SCR centre was associated with shorter onset-to-reperfusion time of 286 (206–566) minutes, compared with 403 (295–550) minutes (p<0.001). While onset-to-door time was shorter for Māori (72 (44–112) minutes, p <0.001) and Pacific patients (70 (48–105) minutes, p=0.03) compared with NZ Europeans, door-to-needle time was longer in Māori (66 (48–88) compared to 59 (41–83) minutes (p=0.001). Female (73.7+/15.3 years) patients were on average 4.4 years older than males (69.3+/-14.6 years) and less likely to receive thrombolysis (12.7% vs 14.9%, p=0.02).

Conclusion

Reperfusion therapy rates continue to rise, now driven by increasing rates of SCR. Longer door-to-needle time in Māori and lower reperfusion rates in women require further exploration and attention.

Author Information

Natsuko Fushida-Hardy: BSc and Medical Student, University of Otago, Wellington. Anne Kim: Medical Student, University of Otago, Wellington. Andrew Leighs: MBChB, House Surgeon, Hutt Hospital. Stephanie Thompson: PhD Candidate, University of Otago, Wellington. Alicia Tyson: Stroke Nurse and Registry Co-Ordinator, Wellington Hospital. P Alan Barber: MBChB, PhD, FRACP, Stroke Neurologist and Stroke Lead, Department of Medicine, Auckland University and Department of Neurology, Auckland Hospital. Annemarei Ranta: MD, PhD, FRACP, Stroke Neurologist and Head of Department, Department of Medicine, University of Otago, Wellington and Department of Neurology, Wellington Hospital.

Acknowledgements

We acknowledge the clinicians across New Zealand who have contributed to the stroke register and have been responsive to inquiries during data-cleaning for this report. The Neurological Foundation of New Zealand provided funding support to carry out much of this research during sequential summer studentships.

Correspondence

Prof Anna Ranta MD PhD. FRACP FAHA FAAN, Department of Medicine, University of Otago Wellington, PO Box 7343, Wellington 6242, New Zealand, +64 4 806 1031.

Correspondence Email

anna.ranta@otago.ac.nz

Competing Interests

Nil.

1) Ministry of Health (2018). Mortality and Demographic Data 2015. Retrieved 6 February, 2021, from https://www.health.govt.nz/publication/mortality-2015-data-tables  

2) Hedlund F, Leighs A, Barber P, Lundström E, Wu T, Ranta A. Trends in stroke reperfusion treatment and outcomes in New Zealand. Intern Med J. 2020;50(11):1367-1372. doi:10.1111/imj.14682

3) Goyal M, Menon B, van Zwam W et al. Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials. The Lancet. 2016;387(10029):1723-1731. doi:10.1016/s0140-6736(16)00163-x

4) Stroke Foundation (2021). Clinical Guidelines for Stroke Management. Melbourne Australia. Retrieved 13September, 2021, from https://strokefoundation.org.au/What-we-do/For%20health%20professionals%20and%20researchers/Clinical-guidelines

5) Liu Q, Barber P, Abernethy G, Ranta A. Provision of stroke thrombolysis services in New Zealand: changes between 2011 and 2016. NZMJ. 2017;130(1453):57-62.

6) Gordon C, Bell R, Ranta A. Impact of the national public 'FAST' campaigns. NZMJ. 2019;132(1507):48-56.

7) Wu T, Coleman E, Wright S et al. Helsinki Stroke Model Is Transferrable With “Real-World” Resources and Reduced Stroke Thrombolysis Delay to 34 min in Christchurch. Front Neurol. 2018;9. doi:10.3389/fneur.2018.00290

8) Ranta A, Lanford J, Busch S, Providence C, Iniesta I, Rosemergy I, et al. Impact and implementation of a sustainable regional telestroke network. Intern Med J. 2017; 47(11):1270-1275.

9) Feigin V, McNaughton H, Dyall L. Burden of Stroke in Maori and Pacific Peoples of New Zealand. International Journal of Stroke. 2007;2(3):208-210. doi:10.1111/j.1747-4949.2007.00140.x

10) Samuels I, Wang MTM, Chong KP, et al. Ethnic Differences in Access to Stroke Reperfusion Therapy in Northern New Zealand. Neuroepidemiology. 2020;54(5):427-432. doi:10.1159/000510505

11) Phan HT, Blizzard CL, Reeves MJ, et al. Sex Differences in Long-term Mortality After Stroke in the INSTRUCT (International Stroke Outcomes Study): A meta-analysis of Individual Participant Data. Circ Cardiovasc Qual Outcomes. 2017;10:1-10

12) Reeves MJ, Wilkins T, Lisabeth LD, Schwamm LH. Thrombolysis treatment for acute stroke: issues of efficacy and utilisation in women. Women’s Health. 2011;7:383-390

13) Hacke W, Kaste M, Bluhmki E et al. Thrombolysis with Alteplase 3 to 4.5 Hours after Acute Ischemic Stroke. New England Journal of Medicine. 2008;359(13):1317-1329. doi:10.1056/nejmoa0804656

14) Cadilhac DA, Kim J, Lannin NA, Kapral MK, Schwamm LH, Dennis MS, et al. National stroke registries for monitoring and improving the quality of hospital care: A systematic review. Int J Stroke. 2015;11:28-40.  

15) Ministry of Health. 2020. Stroke Clot Retrieval: A National Service Improvement Programme Action Plan. Wellington: Ministry of Health.

16) Ministry of Health. 2012. Mātātuhi Tuawhenua: Health of Rural Māori 2012. Wellington: Ministry of Health.

17) Thompson S, Barber A, Gommans J, Cadilhac DA, Davis A, Fink J, Harwood M, Levack W, McNaughton H, Feigin V, Abernethy G, Girvan J, Denison H, Corbin M, Wilson A, Douwes J and Ranta A. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Regional Health Western Pacific. 2022;20:100358. DOI:https://doi.org/10.1016/j.lanwpc.2021.100358.

18) Gattringer T, Ferrari J, Knoflach M, et al. Sex-related differences of Acute Stroke Unit Care: Results from the Austrian Stroke Unit Registry. Stroke. 2014;45:1632-1638

For the PDF of this article,
contact nzmj@nzma.org.nz

View Article PDF

Stroke is now the third most common cause of death and the leading cause of serious adult disability in New Zealand.[[1]] The annual number of people with strokes is expected to rise by a further 40% over the next decade.[[2]]

Following an ischaemic stroke, brain tissue may be rescued if blood flow is restored rapidly. Intravenous thrombolysis (IVT) and stroke clot retrieval (SCR) using stent retrievers significantly improve the odds of disability-free recovery.[[3,4]]

In New Zealand, IVT is provided at all 28 acute stroke hospitals, of which 16 receive remote Telestroke support.[[2]] SCR is provided at three of these stroke centres.[[2]] SCR treatment is limited to a subset of patients with large vessel occlusion (LVO) and consequently more patients qualify for IVT compared to SCR. However, in patients with LVO, SCR offers greater absolute treatment benefit because these are the most severely affected patients and SCR is more effective in clearing large clots than IVT. The pivotal SCR clinical trials all used SCR preceded by IVT treatment, although trials are ongoing to explore SCR without IVT.[[3]]

Stroke thrombolysis rates have varied between centres in New Zealand with key barriers including poor access to experienced stroke physicians in regional settings.[[5]] National and regional initiatives to address these inequalities have included regular audit and reviews, service model changes to reduce treatment delays, public education campaigns to improve community stroke recognition, and the introduction of Telestroke.[[6–8]]

Unfortunately, some population groups in New Zealand appear consistently more vulnerable. Between 1983 and 2003, NZ Europeans have seen a 19% decrease in total stroke incidence, while Pacific patients have seen a 66% increase and Māori a 2% increase.[[9]] Furthermore, while age-adjusted stroke incidence rates in NZ Europeans have declined by 19%, these rates have increased by 16% in Māori and 21% in Pacific peoples.[[9]] The high stroke incidence in Māori and Pacific peoples, combined with the comparatively young onset of strokes in these populations, creates significant burdens on Māori and Pacific stroke survivors, their whānau, communities and society. Reperfusion access by ethnicity has previously been explored in the Northland Region of New Zealand and did not find significant differences.[[10]] Comprehensive data on reperfusion metrics for the entire country have not been previously published.

Women often experience poorer post-stroke outcomes than men due to sex-related differences in presentation and risk factors. Women are more likely to present with strokes at an older age and have worse pre-stroke functionality compared with men.[[11]] Studies have also found that women are less likely to receive IVT than men; the cause of which remains unclear.[[12]] To date, potential sex differences in stroke reperfusion therapy in New Zealand have not been explored.

This study reports on the reperfusion therapy rates in 2019 and 2020 to assess the impact of service improvements on temporal trends, along with the identification of areas that require further improvement. We also explored the association of demographic factors, specifically ethnicity and sex, with reperfusion therapy access and quality metrics.

Design and methods

This is a registry-based prospective observational study including all adult patients (age >16) who received reperfusion therapy in New Zealand between 1 January 2015 and 31 December 2020. All stroke patients who receive reperfusion therapy (IVT and/or SCR) in New Zealand are routinely entered into a national online database using the open source REDCap platform. Data is entered by staff at the primary stroke centre or by a member of the central database administration team. The data in the register was cleaned manually and scrutinised for missing values, data entry errors and notable outliers. Duplicate entries were identified using the National Health Index number. Hospital staff nationwide were contacted for further information where necessary. In cases where important information was missing or missing values could not be reconciled, the patient in question was excluded from both numerator and denominator to complete specific analyses.

The primary patient efficacy end point for the National Stroke Register is day seven vital status with optional reporting of modified Rankin Scale (mRS) at three months. The mRs is a seven point scale with zero normal and six dead, and where independence is defined as a score of 0,1 or 2. The primary safety end point is the rate of symptomatic intracerebral haemorrhage (sICH).[[13]] IVT and SCR rates are calculated using denominator data supplied by the Ministry of Health based on recorded discharge coding. The denominator data includes all patients with ischaemic stroke (ICD 10-AM I63) and strokes “unspecified” (ICD 10-AM I64). Strokes “unspecified” have been included as internal audits have found that most cases represent miscoded ischaemic strokes.

In 2020, we became aware that Ministry of Health data includes multiple duplicates, where patients transferred between hospitals are counted as multiple discharges, despite all discharges referring to a single stroke event. We have now removed all duplicates dating back to 2017. Data for 2015 and 2016 were not available for this reanalysis. This means some of the denominators and intervention rates presented in this paper differ slightly from previously published reports raising historical intervention rates. As we were unable to retrospectivity amend data prior to 2017, this needs to be borne in mind when comparing trends from pre-2017 with post-2017. Due to the slightly inflated denominator for these years the actual thrombolysis rates will have likely been approximately 0.5% higher for thrombolysis and 0.2% higher for stroke clot retrieval than the figures displayed and previously published.[[2]]

Continuous variables were non-normally distributed and were thus analysed and reported as the median value, with additional values denoting IQR of 25[[th]] to 75[[th]] percentiles. Comparative differences were analysed using Wilcoxon rank sum test and presented as p-values. Rate-based results were presented as incidence rates. Comparative differences were analysed using incidence rate comparison and presented with p-values +/- 95% confidence intervals. A p-value less than 0.05 was considered statistically significant. Data analysis was completed in StataIC 16.0

Results

Between 1 January and 31 December 2020, there were 7333 patients with acute ischaemic stroke (n= 7021) or “stroke unspecified” (n= 312). Of this number, 828 (11.3%, 11.8% excluding “stroke unspecified”) were treated with IVT and 404 (5.5%, 5.6% excluding “stroke unspecified”) were treated with SCR. A total of 201 patients were treated with both IVT and SCR (24.3% of all IVT patients). None of these patients were confirmed positive for COVID-19. Baseline characteristics are listed in Table 1.

For IVT in 2020, the median (IQR) door-to-needle time was 61 (41–87) minutes (Table 2). sICH occurred in 24 patients (2.9%). At day seven, 66 (8.1%) patients had died and at three months, 224 (61.7%) of 363 patients where this information was recorded were functionally independent (mRS 0–2). For SCR in 2020, the median (IQR) arrival-to-groin time was 55 (22–113) minutes, sICH rate was 1.2% and 141 of 286 (49.3%), where this information was recorded, were independent at three months (Table 3).

IVT rates have increased from 6.5% in 2015 to 11.3% in 2020 (p=<0.001), but have remained unchanged since 2019 (11.3% both years). The median (IQR) door-to-needle time has reduced from 74 (55–102) in 2015 to 61 (41–87) in 2020, but has remained stable since 2019 when it was 61 (44–84) minutes. The percentage of patients treated with IVT under 60 minutes has also remained stable between 2019 and 2020 (49.1% to 49.7%; p= 0.99). There has been no significant change observed in the number of patients who have died by day seven between 2015 and 2020 (6.7% to 8.1%; p= 0.45). Post-IVT sICH rates have reduced (6.2% in 2015 to 2.9% in 2020; p= 0.01) (Table 2).

SCR rates have increased substantially from 2015 to 2020 (0.5% to 5.5%; p<0.001), but only modestly across the last two years (4.8% in 2019 to 5.5% in 2020; p= 0.08). Arrival-to-groin time remains stable at (median (IQR)) 52 (16–109) minutes in 2019 to 55 (22–113) minutes in 2020 (p=0.45), as has onset-to-reperfusion time; (median (IQR)) 335 (235–522) minutes in 2019 to 365 (245–550) minutes in 2020 (p=0.27). There have been no statistically significant changes between 2019 and 2020 with regard to sICH rates (2.9% to 1.2%; p=0.12), day seven mortality rates (9.1% to 7.5%; p= 0.47), and the number of independent patients at three months (44.6% to 49.3%; p=0.44) (Table 3). NZ European patients residing in DHBs that are SCR centres experienced shorter onset-to-reperfusion times than those in a non-SCR centre DHB of domicile; 286 (206–566) and 403 (295–550) minutes (p<0.001) respectively.

Ethnicity did not have a significant impact on IVT rates in 2020 with 11.7% in NZ Europeans compared with 10.1% in Māori (95% CI 0.92, 1.45; p= 0.20) and 11.6% in Pacific patients (95% CI: 0.75, 1.36; p= 1.00). Variation in SCR rates among different ethnic groups also did not differ with 5.4% of NZ Europeans compared to 5.3% of Māori (95% CI 0.75, 1.43; p= 0.88) and 6.8% of Pacific patients (95% CI: 0.54, 1.19; p= 0.23). Reperfusion rates with IVT and or SCR were 13.3% in NZ Europeans, compared to 12.3% in Māori (95% CI: 0.89, 1.34; p= 0.42), and 13.9% in Pacific patients (95% CI: 0.74, 1.26; p= 0.73).

Looking at a two-year cohort (2018–2020) Māori and Pacific patients were significantly younger than their NZ European counterparts (60.3 years vs 74 years; p<0.001, and 63 years vs 74 years; p<0.001, respectively). Onset-to-door time was significantly higher in NZ European patients (median (IQR)) 81 (54–123) minutes compared with Māori; 72 (44–112) minutes (p<0.001), and Pacific patients; 70 (48–105) minutes (p= 0.03). Door-to-needle time in 2018–2020 was longer in Māori patients compared with NZ Europeans; (median (IQR)) 66 (48–88) minutes compared to 59 (41–83) minutes (p=0.001). A sensitivity analysis explored whether the difference in slower door-to-needle time for Māori could be due to a higher proportion residing rurally, with previous reports showing slower door-to-needle time in non-urban settings.[[2,9]] Indeed, the sensitivity analysis found that door-to-needle time for NZ Europeans residing in non-urban DHBs was significantly longer than for urban NZ European door-to-needle time: (median (IQR)) 62 (44–86) minutes compared to 57 (40–82) minutes (p= 0.02), respectively. The same trend was observed among Māori. However, this was not statistically significant. Māori non-urban door-to-needle time (median (IQR)) 70 (51–92) minutes compared to Māori urban door-to-needle time (median (IQR)) 63 (42–87) minutes (p=0.11). However, Māori also had longer door-to-needle time in the urban setting when compared to urban NZ Europeans: (median (IQR)) 63 (42–87) minutes compared to 57 (40–82) minutes (p=0.048). Similarly, non-urban Māori had a significantly longer door-to-needle times compared to their non-urban NZ European counterparts; (median (IQR)) 70 (51–92) minutes compared to 62 (44–86) minutes (p= 0.012). There was no significant difference between NZ European and Pacific patients in door-to-needle time (p= 0.28). Overall onset-to-needle time was similar across ethnic groups (NZ European vs Māori; p= 0.16, NZ European vs Pacific; p= 0.37, Māori vs Pacific; p= 0.99).

In 2019, there were overall more men than women who experienced ischaemic events (3661 males versus 3492 females). Female stroke patients were on average 4.4 years older than males upon presentation (73.7 years versus 69.3 years respectively, p<0.001). The rate of IVT performed in females (10.2%) was lower than that in males (12.2%); p= 0.01. However, there was no significant difference in the rate of SCR (female 4.6%; male 5.1%; p= 0.29). The overall reperfusion rate was lower for females than males (12.7% vs 14.9% p= 0.02). Median door-to-needle, onset-to-needle, and onset-to-groin times, sICH and day seven mortality were similar between females and males (Table 4).

Discussion

IVT and SCR rates have risen considerably since 2015. IVT rates have seen a steady rise from 6.5% to 11.3% and SCR rates have risen even more dramatically, from 0.5% to 5.5%. This is likely related to increasing implementation of SCR following the publication of seven pivotal trials on SCR efficacy in 2015 and 2016.[[3]] Since this initial jump, SCR rates have continued to increase more modestly.

The acute stroke reperfusion therapy rate with IVT and or SCR reached 14.0% in 2020 (14.6% when “stroke unspecified” patients are excluded). Eight hundred and twenty-eight patients were treated with IVT and 404 with SCR (201 received both) in a population of 5.1 million people. This equates to an IVT rate of 162 per million people and an SCR rate of 79 per million people. A number of quality improvement initiatives have led to the progress seen in stroke service provision. The implementation of the National Stroke Reperfusion Register in January 2015 has allowed for regular quality-control audits and has provided a focus for annual meetings to discuss national stroke reperfusion data and quality.[[2]] Such registers have also been shown to drive improvement in stroke service provision overseas.[[14]] Other initiatives include expanding Telestroke networks, regular Ministry of Health FAST campaigns, and hospital stroke service improvement projects.[[6–8]]

Our study found that IVT and SCR pre-hospital delays were higher in 2020 compared to 2019. While this was not statistically significant, it does follow an upward trend seen in the last six years. This is presumed to be, at least in part, a paradoxical finding secondary to an increased acceptance of treatment in extended time windows and increased treatment volumes at smaller centres. A drop-in pre-hospital care pathway performance is less likely.

SCR delays continue to improve which is unsurprising given SCR services continue to actively evolve. More work is ongoing under the recently launched Ministry of Health National Stroke Clot Retrieval Service Improvement Programme.[[15]] It is also encouraging to see stable complication rates for both IVT and SCR.

Despite many improvements, areas requiring continued effort have been identified. One such area is the door-to-needle time for IVT, which remains far from the recommended 30-minute target. Further, concerns are raised by the 2019 reduction in reported rate of three-month mRS (0–2) after SCR. Some fluctuation of results may be explained by the inconsistency in mRS data reporting. However, mRS had remained fairly stable between 2011 and 2018. It is possible that genuinely more complications were encountered in the past two years, and this will need to be further explored at the individual SCR centres, although the stable sICH rate provides some reassurance. A potential contributing factor may be that boundaries are being pushed with more patients referred for SCR with poorer baseline health status who do not entirely meet trial criteria. This is supported by the finding that over 50% of SCR patients did not receive IVT, suggesting they had IVT contraindications and may also be borderline SCR cases (“mercy cases”). Finally, with slowly increasing referrals from regional centres it is possible that significant transport delays mean more patients reach the angiography suite too late to benefit from the procedure. This area requires urgent attention and is a focus of the National SCR Programme.[[15]] To monitor SCR time delays and complications more effectively, we have recently introduced additional SCR time metrics and details around procedural complications that will be monitored over time.

The disproportionate burden of stroke and post-stoke complications on Māori and Pacific peoples have long been documented. Previous studies have shown a striking difference in the average age of stroke onset in Māori and Pacific peoples compared with NZ Europeans.[[9]] The results of our study align with existing literature in that Māori and Pacific stroke patients were 13.7 and 11 years younger on average than NZ Europeans.

Pacific people and Māori presented faster to hospital suggesting good stroke awareness and pre-hospital transport access in these populations. However, there were greater in-hospital delays for Māori. A higher proportion of Māori reside in rural areas compared with NZ Europeans and thus slower door-to-needle times at regional hospitals may disproportionally affect Māori.[[2,8,16]] However, a sensitivity analysis exploring the impact of hospital location found that Māori had slower door-to-needle times compared to NZ Europeans even when limiting the analysis to either urban or non-urban settings. While we were unable to adjust for other potential confounders in this study, the recently published REGIONS Care ethnicity analysis was fully adjusted for age, baseline morbidity, risk factors, stroke severity, and geography and identified additional areas of in-hospital stroke care inequities.[[17]] Such findings raise concern about potential racial discrimination within New Zealand stroke services. This will require careful consideration to plan potential interventions and should be monitored going forward.

As part of the 2019 analysis, we also explored sex-related differences in reperfusion therapy. It is well-known in the literature that women tend to present with stroke at an older age than men and with a different profile of risk factors.[[15]] Previous studies have also found that women are less likely to receive IVT despite gaining more net clinical benefit than men, especially at an older age.[[18]] While older age may make some clinicians more reluctant to treat and may thus provide a potential explanation, other studies have also found that suitable female candidates are more likely to be labelled stroke mimics and can inadvertently miss out on treatment.[[18]] This may be due to more frequent atypical presentations or unconscious bias. One study found that when age was corrected for, there were no significant sex-related differences in quality of care.[[18]] As our data were not adjusted for age, it should be viewed as explorative and interpreted with a degree of caution. Further analysis adjusting for age, severity, and baseline morbidity is important.

Our study has several limitations. Firstly, it is difficult to validate all data entered into the clinical register by up to 60 front line clinicians across the country. Therefore, despite the thorough and meticulous system of manual data checking and cleaning that has been implemented to mitigate any errors, it is possible that some mistakes remain. Secondly, the crude outcome measure of “alive or discharged at day seven” does not represent the long road of rehabilitation undertaken by stroke patients. To improve this, reporting of three-month mRS became mandatory in 2019. Despite this, only 50.1% of records included this information in 2020 (down from 54.3% in 2019). To fairly represent the New Zealand post-stroke experience, we must work to increase these numbers. While the main outcomes of interest were intervention rate, sICH and seven-day mortality, we conducted multiple additional analyses. Due to the number of comparisons conducted, there is a possibility that some significant results occurred based on chance alone and these secondary comparisons should thus be interpreted with caution. The inability to update denominator data for 2015 and 2016 will have impacted the precision of historical intervention rate comparisons, however; given the substantial increases in intervention rates over time it is very unlikely that the subtle changes in denominator values will have impacted the overall conclusions. Finally, the analyses by ethnicity and sex were not adjusted for potential confounders such as age, stroke severity, risk factors and time delays and should thus be viewed as primarily exploratory and interpreted with a degree of caution. A sample size was not prospectively determined, and we cannot exclude that significant findings were missed due to the lack of study power.

Despite these limitations, the completeness of data in our register—which includes every patient who underwent reperfusion in New Zealand—strengthens our observations. This minimises selection bias and allows for the accurate analysis of trends in stroke service provision over time.

Conclusion

Reperfusion therapy rates and treatment delays continue to improve although appear to gradually plateau. Complications and mortality have remained stable. The reduction in favourable outcomes following SCR over the last two years is likely related to widening inclusion criteria for patient selection and/or greater delays for more patients due to rising regional transfers. Both require further investigation and mitigation. Ethnic disparities identified include the longer door-to-needle time observed in Māori not explained by geographic factors. Women accessed reperfusion therapy less frequently, potentially explained by older age at presentation. Both require further exploration and action.

View Supplementary Tables.

Summary

Abstract

Aim

This study assessed stroke reperfusion treatments trends in 2019 and 2020 with comparison back to 2015. Additional analyses looked at differences by sex and ethnicity.

Method

The National Stroke Register contains data on all stroke patients who received reperfusion therapies since 2015. Outcomes included treatment rates, delays, mortality and complications by year, sex, and ethnicity. Continuous variables were compared using the Wilcoxon Rank-Sum Test and presented as p-values. Rate-based results were compared using incidence rate comparison and presented as p-values +/- 95% confidence intervals.

Results

In 2020, 11.3% (828/7333) received intravenous thrombolysis (IVT) and 5.5% (404/7333) underwent stroke clot retrieval (SCR), increasing from 6.5% (389/5963) and 0.5% (30/5963) in 2015, respectively. Among reperfused patients (IVT, SCR, both), 8.3% had died at seven days and 3.0% (29/959) experienced sICH. Door-to-treatment time was stable between 2019 and 2020, with median (IQR) of 61 (44–84) and 61 (41–87) minutes, respectively. Initial presentation to a SCR centre was associated with shorter onset-to-reperfusion time of 286 (206–566) minutes, compared with 403 (295–550) minutes (p<0.001). While onset-to-door time was shorter for Māori (72 (44–112) minutes, p <0.001) and Pacific patients (70 (48–105) minutes, p=0.03) compared with NZ Europeans, door-to-needle time was longer in Māori (66 (48–88) compared to 59 (41–83) minutes (p=0.001). Female (73.7+/15.3 years) patients were on average 4.4 years older than males (69.3+/-14.6 years) and less likely to receive thrombolysis (12.7% vs 14.9%, p=0.02).

Conclusion

Reperfusion therapy rates continue to rise, now driven by increasing rates of SCR. Longer door-to-needle time in Māori and lower reperfusion rates in women require further exploration and attention.

Author Information

Natsuko Fushida-Hardy: BSc and Medical Student, University of Otago, Wellington. Anne Kim: Medical Student, University of Otago, Wellington. Andrew Leighs: MBChB, House Surgeon, Hutt Hospital. Stephanie Thompson: PhD Candidate, University of Otago, Wellington. Alicia Tyson: Stroke Nurse and Registry Co-Ordinator, Wellington Hospital. P Alan Barber: MBChB, PhD, FRACP, Stroke Neurologist and Stroke Lead, Department of Medicine, Auckland University and Department of Neurology, Auckland Hospital. Annemarei Ranta: MD, PhD, FRACP, Stroke Neurologist and Head of Department, Department of Medicine, University of Otago, Wellington and Department of Neurology, Wellington Hospital.

Acknowledgements

We acknowledge the clinicians across New Zealand who have contributed to the stroke register and have been responsive to inquiries during data-cleaning for this report. The Neurological Foundation of New Zealand provided funding support to carry out much of this research during sequential summer studentships.

Correspondence

Prof Anna Ranta MD PhD. FRACP FAHA FAAN, Department of Medicine, University of Otago Wellington, PO Box 7343, Wellington 6242, New Zealand, +64 4 806 1031.

Correspondence Email

anna.ranta@otago.ac.nz

Competing Interests

Nil.

1) Ministry of Health (2018). Mortality and Demographic Data 2015. Retrieved 6 February, 2021, from https://www.health.govt.nz/publication/mortality-2015-data-tables  

2) Hedlund F, Leighs A, Barber P, Lundström E, Wu T, Ranta A. Trends in stroke reperfusion treatment and outcomes in New Zealand. Intern Med J. 2020;50(11):1367-1372. doi:10.1111/imj.14682

3) Goyal M, Menon B, van Zwam W et al. Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials. The Lancet. 2016;387(10029):1723-1731. doi:10.1016/s0140-6736(16)00163-x

4) Stroke Foundation (2021). Clinical Guidelines for Stroke Management. Melbourne Australia. Retrieved 13September, 2021, from https://strokefoundation.org.au/What-we-do/For%20health%20professionals%20and%20researchers/Clinical-guidelines

5) Liu Q, Barber P, Abernethy G, Ranta A. Provision of stroke thrombolysis services in New Zealand: changes between 2011 and 2016. NZMJ. 2017;130(1453):57-62.

6) Gordon C, Bell R, Ranta A. Impact of the national public 'FAST' campaigns. NZMJ. 2019;132(1507):48-56.

7) Wu T, Coleman E, Wright S et al. Helsinki Stroke Model Is Transferrable With “Real-World” Resources and Reduced Stroke Thrombolysis Delay to 34 min in Christchurch. Front Neurol. 2018;9. doi:10.3389/fneur.2018.00290

8) Ranta A, Lanford J, Busch S, Providence C, Iniesta I, Rosemergy I, et al. Impact and implementation of a sustainable regional telestroke network. Intern Med J. 2017; 47(11):1270-1275.

9) Feigin V, McNaughton H, Dyall L. Burden of Stroke in Maori and Pacific Peoples of New Zealand. International Journal of Stroke. 2007;2(3):208-210. doi:10.1111/j.1747-4949.2007.00140.x

10) Samuels I, Wang MTM, Chong KP, et al. Ethnic Differences in Access to Stroke Reperfusion Therapy in Northern New Zealand. Neuroepidemiology. 2020;54(5):427-432. doi:10.1159/000510505

11) Phan HT, Blizzard CL, Reeves MJ, et al. Sex Differences in Long-term Mortality After Stroke in the INSTRUCT (International Stroke Outcomes Study): A meta-analysis of Individual Participant Data. Circ Cardiovasc Qual Outcomes. 2017;10:1-10

12) Reeves MJ, Wilkins T, Lisabeth LD, Schwamm LH. Thrombolysis treatment for acute stroke: issues of efficacy and utilisation in women. Women’s Health. 2011;7:383-390

13) Hacke W, Kaste M, Bluhmki E et al. Thrombolysis with Alteplase 3 to 4.5 Hours after Acute Ischemic Stroke. New England Journal of Medicine. 2008;359(13):1317-1329. doi:10.1056/nejmoa0804656

14) Cadilhac DA, Kim J, Lannin NA, Kapral MK, Schwamm LH, Dennis MS, et al. National stroke registries for monitoring and improving the quality of hospital care: A systematic review. Int J Stroke. 2015;11:28-40.  

15) Ministry of Health. 2020. Stroke Clot Retrieval: A National Service Improvement Programme Action Plan. Wellington: Ministry of Health.

16) Ministry of Health. 2012. Mātātuhi Tuawhenua: Health of Rural Māori 2012. Wellington: Ministry of Health.

17) Thompson S, Barber A, Gommans J, Cadilhac DA, Davis A, Fink J, Harwood M, Levack W, McNaughton H, Feigin V, Abernethy G, Girvan J, Denison H, Corbin M, Wilson A, Douwes J and Ranta A. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Regional Health Western Pacific. 2022;20:100358. DOI:https://doi.org/10.1016/j.lanwpc.2021.100358.

18) Gattringer T, Ferrari J, Knoflach M, et al. Sex-related differences of Acute Stroke Unit Care: Results from the Austrian Stroke Unit Registry. Stroke. 2014;45:1632-1638

For the PDF of this article,
contact nzmj@nzma.org.nz

View Article PDF

Stroke is now the third most common cause of death and the leading cause of serious adult disability in New Zealand.[[1]] The annual number of people with strokes is expected to rise by a further 40% over the next decade.[[2]]

Following an ischaemic stroke, brain tissue may be rescued if blood flow is restored rapidly. Intravenous thrombolysis (IVT) and stroke clot retrieval (SCR) using stent retrievers significantly improve the odds of disability-free recovery.[[3,4]]

In New Zealand, IVT is provided at all 28 acute stroke hospitals, of which 16 receive remote Telestroke support.[[2]] SCR is provided at three of these stroke centres.[[2]] SCR treatment is limited to a subset of patients with large vessel occlusion (LVO) and consequently more patients qualify for IVT compared to SCR. However, in patients with LVO, SCR offers greater absolute treatment benefit because these are the most severely affected patients and SCR is more effective in clearing large clots than IVT. The pivotal SCR clinical trials all used SCR preceded by IVT treatment, although trials are ongoing to explore SCR without IVT.[[3]]

Stroke thrombolysis rates have varied between centres in New Zealand with key barriers including poor access to experienced stroke physicians in regional settings.[[5]] National and regional initiatives to address these inequalities have included regular audit and reviews, service model changes to reduce treatment delays, public education campaigns to improve community stroke recognition, and the introduction of Telestroke.[[6–8]]

Unfortunately, some population groups in New Zealand appear consistently more vulnerable. Between 1983 and 2003, NZ Europeans have seen a 19% decrease in total stroke incidence, while Pacific patients have seen a 66% increase and Māori a 2% increase.[[9]] Furthermore, while age-adjusted stroke incidence rates in NZ Europeans have declined by 19%, these rates have increased by 16% in Māori and 21% in Pacific peoples.[[9]] The high stroke incidence in Māori and Pacific peoples, combined with the comparatively young onset of strokes in these populations, creates significant burdens on Māori and Pacific stroke survivors, their whānau, communities and society. Reperfusion access by ethnicity has previously been explored in the Northland Region of New Zealand and did not find significant differences.[[10]] Comprehensive data on reperfusion metrics for the entire country have not been previously published.

Women often experience poorer post-stroke outcomes than men due to sex-related differences in presentation and risk factors. Women are more likely to present with strokes at an older age and have worse pre-stroke functionality compared with men.[[11]] Studies have also found that women are less likely to receive IVT than men; the cause of which remains unclear.[[12]] To date, potential sex differences in stroke reperfusion therapy in New Zealand have not been explored.

This study reports on the reperfusion therapy rates in 2019 and 2020 to assess the impact of service improvements on temporal trends, along with the identification of areas that require further improvement. We also explored the association of demographic factors, specifically ethnicity and sex, with reperfusion therapy access and quality metrics.

Design and methods

This is a registry-based prospective observational study including all adult patients (age >16) who received reperfusion therapy in New Zealand between 1 January 2015 and 31 December 2020. All stroke patients who receive reperfusion therapy (IVT and/or SCR) in New Zealand are routinely entered into a national online database using the open source REDCap platform. Data is entered by staff at the primary stroke centre or by a member of the central database administration team. The data in the register was cleaned manually and scrutinised for missing values, data entry errors and notable outliers. Duplicate entries were identified using the National Health Index number. Hospital staff nationwide were contacted for further information where necessary. In cases where important information was missing or missing values could not be reconciled, the patient in question was excluded from both numerator and denominator to complete specific analyses.

The primary patient efficacy end point for the National Stroke Register is day seven vital status with optional reporting of modified Rankin Scale (mRS) at three months. The mRs is a seven point scale with zero normal and six dead, and where independence is defined as a score of 0,1 or 2. The primary safety end point is the rate of symptomatic intracerebral haemorrhage (sICH).[[13]] IVT and SCR rates are calculated using denominator data supplied by the Ministry of Health based on recorded discharge coding. The denominator data includes all patients with ischaemic stroke (ICD 10-AM I63) and strokes “unspecified” (ICD 10-AM I64). Strokes “unspecified” have been included as internal audits have found that most cases represent miscoded ischaemic strokes.

In 2020, we became aware that Ministry of Health data includes multiple duplicates, where patients transferred between hospitals are counted as multiple discharges, despite all discharges referring to a single stroke event. We have now removed all duplicates dating back to 2017. Data for 2015 and 2016 were not available for this reanalysis. This means some of the denominators and intervention rates presented in this paper differ slightly from previously published reports raising historical intervention rates. As we were unable to retrospectivity amend data prior to 2017, this needs to be borne in mind when comparing trends from pre-2017 with post-2017. Due to the slightly inflated denominator for these years the actual thrombolysis rates will have likely been approximately 0.5% higher for thrombolysis and 0.2% higher for stroke clot retrieval than the figures displayed and previously published.[[2]]

Continuous variables were non-normally distributed and were thus analysed and reported as the median value, with additional values denoting IQR of 25[[th]] to 75[[th]] percentiles. Comparative differences were analysed using Wilcoxon rank sum test and presented as p-values. Rate-based results were presented as incidence rates. Comparative differences were analysed using incidence rate comparison and presented with p-values +/- 95% confidence intervals. A p-value less than 0.05 was considered statistically significant. Data analysis was completed in StataIC 16.0

Results

Between 1 January and 31 December 2020, there were 7333 patients with acute ischaemic stroke (n= 7021) or “stroke unspecified” (n= 312). Of this number, 828 (11.3%, 11.8% excluding “stroke unspecified”) were treated with IVT and 404 (5.5%, 5.6% excluding “stroke unspecified”) were treated with SCR. A total of 201 patients were treated with both IVT and SCR (24.3% of all IVT patients). None of these patients were confirmed positive for COVID-19. Baseline characteristics are listed in Table 1.

For IVT in 2020, the median (IQR) door-to-needle time was 61 (41–87) minutes (Table 2). sICH occurred in 24 patients (2.9%). At day seven, 66 (8.1%) patients had died and at three months, 224 (61.7%) of 363 patients where this information was recorded were functionally independent (mRS 0–2). For SCR in 2020, the median (IQR) arrival-to-groin time was 55 (22–113) minutes, sICH rate was 1.2% and 141 of 286 (49.3%), where this information was recorded, were independent at three months (Table 3).

IVT rates have increased from 6.5% in 2015 to 11.3% in 2020 (p=<0.001), but have remained unchanged since 2019 (11.3% both years). The median (IQR) door-to-needle time has reduced from 74 (55–102) in 2015 to 61 (41–87) in 2020, but has remained stable since 2019 when it was 61 (44–84) minutes. The percentage of patients treated with IVT under 60 minutes has also remained stable between 2019 and 2020 (49.1% to 49.7%; p= 0.99). There has been no significant change observed in the number of patients who have died by day seven between 2015 and 2020 (6.7% to 8.1%; p= 0.45). Post-IVT sICH rates have reduced (6.2% in 2015 to 2.9% in 2020; p= 0.01) (Table 2).

SCR rates have increased substantially from 2015 to 2020 (0.5% to 5.5%; p<0.001), but only modestly across the last two years (4.8% in 2019 to 5.5% in 2020; p= 0.08). Arrival-to-groin time remains stable at (median (IQR)) 52 (16–109) minutes in 2019 to 55 (22–113) minutes in 2020 (p=0.45), as has onset-to-reperfusion time; (median (IQR)) 335 (235–522) minutes in 2019 to 365 (245–550) minutes in 2020 (p=0.27). There have been no statistically significant changes between 2019 and 2020 with regard to sICH rates (2.9% to 1.2%; p=0.12), day seven mortality rates (9.1% to 7.5%; p= 0.47), and the number of independent patients at three months (44.6% to 49.3%; p=0.44) (Table 3). NZ European patients residing in DHBs that are SCR centres experienced shorter onset-to-reperfusion times than those in a non-SCR centre DHB of domicile; 286 (206–566) and 403 (295–550) minutes (p<0.001) respectively.

Ethnicity did not have a significant impact on IVT rates in 2020 with 11.7% in NZ Europeans compared with 10.1% in Māori (95% CI 0.92, 1.45; p= 0.20) and 11.6% in Pacific patients (95% CI: 0.75, 1.36; p= 1.00). Variation in SCR rates among different ethnic groups also did not differ with 5.4% of NZ Europeans compared to 5.3% of Māori (95% CI 0.75, 1.43; p= 0.88) and 6.8% of Pacific patients (95% CI: 0.54, 1.19; p= 0.23). Reperfusion rates with IVT and or SCR were 13.3% in NZ Europeans, compared to 12.3% in Māori (95% CI: 0.89, 1.34; p= 0.42), and 13.9% in Pacific patients (95% CI: 0.74, 1.26; p= 0.73).

Looking at a two-year cohort (2018–2020) Māori and Pacific patients were significantly younger than their NZ European counterparts (60.3 years vs 74 years; p<0.001, and 63 years vs 74 years; p<0.001, respectively). Onset-to-door time was significantly higher in NZ European patients (median (IQR)) 81 (54–123) minutes compared with Māori; 72 (44–112) minutes (p<0.001), and Pacific patients; 70 (48–105) minutes (p= 0.03). Door-to-needle time in 2018–2020 was longer in Māori patients compared with NZ Europeans; (median (IQR)) 66 (48–88) minutes compared to 59 (41–83) minutes (p=0.001). A sensitivity analysis explored whether the difference in slower door-to-needle time for Māori could be due to a higher proportion residing rurally, with previous reports showing slower door-to-needle time in non-urban settings.[[2,9]] Indeed, the sensitivity analysis found that door-to-needle time for NZ Europeans residing in non-urban DHBs was significantly longer than for urban NZ European door-to-needle time: (median (IQR)) 62 (44–86) minutes compared to 57 (40–82) minutes (p= 0.02), respectively. The same trend was observed among Māori. However, this was not statistically significant. Māori non-urban door-to-needle time (median (IQR)) 70 (51–92) minutes compared to Māori urban door-to-needle time (median (IQR)) 63 (42–87) minutes (p=0.11). However, Māori also had longer door-to-needle time in the urban setting when compared to urban NZ Europeans: (median (IQR)) 63 (42–87) minutes compared to 57 (40–82) minutes (p=0.048). Similarly, non-urban Māori had a significantly longer door-to-needle times compared to their non-urban NZ European counterparts; (median (IQR)) 70 (51–92) minutes compared to 62 (44–86) minutes (p= 0.012). There was no significant difference between NZ European and Pacific patients in door-to-needle time (p= 0.28). Overall onset-to-needle time was similar across ethnic groups (NZ European vs Māori; p= 0.16, NZ European vs Pacific; p= 0.37, Māori vs Pacific; p= 0.99).

In 2019, there were overall more men than women who experienced ischaemic events (3661 males versus 3492 females). Female stroke patients were on average 4.4 years older than males upon presentation (73.7 years versus 69.3 years respectively, p<0.001). The rate of IVT performed in females (10.2%) was lower than that in males (12.2%); p= 0.01. However, there was no significant difference in the rate of SCR (female 4.6%; male 5.1%; p= 0.29). The overall reperfusion rate was lower for females than males (12.7% vs 14.9% p= 0.02). Median door-to-needle, onset-to-needle, and onset-to-groin times, sICH and day seven mortality were similar between females and males (Table 4).

Discussion

IVT and SCR rates have risen considerably since 2015. IVT rates have seen a steady rise from 6.5% to 11.3% and SCR rates have risen even more dramatically, from 0.5% to 5.5%. This is likely related to increasing implementation of SCR following the publication of seven pivotal trials on SCR efficacy in 2015 and 2016.[[3]] Since this initial jump, SCR rates have continued to increase more modestly.

The acute stroke reperfusion therapy rate with IVT and or SCR reached 14.0% in 2020 (14.6% when “stroke unspecified” patients are excluded). Eight hundred and twenty-eight patients were treated with IVT and 404 with SCR (201 received both) in a population of 5.1 million people. This equates to an IVT rate of 162 per million people and an SCR rate of 79 per million people. A number of quality improvement initiatives have led to the progress seen in stroke service provision. The implementation of the National Stroke Reperfusion Register in January 2015 has allowed for regular quality-control audits and has provided a focus for annual meetings to discuss national stroke reperfusion data and quality.[[2]] Such registers have also been shown to drive improvement in stroke service provision overseas.[[14]] Other initiatives include expanding Telestroke networks, regular Ministry of Health FAST campaigns, and hospital stroke service improvement projects.[[6–8]]

Our study found that IVT and SCR pre-hospital delays were higher in 2020 compared to 2019. While this was not statistically significant, it does follow an upward trend seen in the last six years. This is presumed to be, at least in part, a paradoxical finding secondary to an increased acceptance of treatment in extended time windows and increased treatment volumes at smaller centres. A drop-in pre-hospital care pathway performance is less likely.

SCR delays continue to improve which is unsurprising given SCR services continue to actively evolve. More work is ongoing under the recently launched Ministry of Health National Stroke Clot Retrieval Service Improvement Programme.[[15]] It is also encouraging to see stable complication rates for both IVT and SCR.

Despite many improvements, areas requiring continued effort have been identified. One such area is the door-to-needle time for IVT, which remains far from the recommended 30-minute target. Further, concerns are raised by the 2019 reduction in reported rate of three-month mRS (0–2) after SCR. Some fluctuation of results may be explained by the inconsistency in mRS data reporting. However, mRS had remained fairly stable between 2011 and 2018. It is possible that genuinely more complications were encountered in the past two years, and this will need to be further explored at the individual SCR centres, although the stable sICH rate provides some reassurance. A potential contributing factor may be that boundaries are being pushed with more patients referred for SCR with poorer baseline health status who do not entirely meet trial criteria. This is supported by the finding that over 50% of SCR patients did not receive IVT, suggesting they had IVT contraindications and may also be borderline SCR cases (“mercy cases”). Finally, with slowly increasing referrals from regional centres it is possible that significant transport delays mean more patients reach the angiography suite too late to benefit from the procedure. This area requires urgent attention and is a focus of the National SCR Programme.[[15]] To monitor SCR time delays and complications more effectively, we have recently introduced additional SCR time metrics and details around procedural complications that will be monitored over time.

The disproportionate burden of stroke and post-stoke complications on Māori and Pacific peoples have long been documented. Previous studies have shown a striking difference in the average age of stroke onset in Māori and Pacific peoples compared with NZ Europeans.[[9]] The results of our study align with existing literature in that Māori and Pacific stroke patients were 13.7 and 11 years younger on average than NZ Europeans.

Pacific people and Māori presented faster to hospital suggesting good stroke awareness and pre-hospital transport access in these populations. However, there were greater in-hospital delays for Māori. A higher proportion of Māori reside in rural areas compared with NZ Europeans and thus slower door-to-needle times at regional hospitals may disproportionally affect Māori.[[2,8,16]] However, a sensitivity analysis exploring the impact of hospital location found that Māori had slower door-to-needle times compared to NZ Europeans even when limiting the analysis to either urban or non-urban settings. While we were unable to adjust for other potential confounders in this study, the recently published REGIONS Care ethnicity analysis was fully adjusted for age, baseline morbidity, risk factors, stroke severity, and geography and identified additional areas of in-hospital stroke care inequities.[[17]] Such findings raise concern about potential racial discrimination within New Zealand stroke services. This will require careful consideration to plan potential interventions and should be monitored going forward.

As part of the 2019 analysis, we also explored sex-related differences in reperfusion therapy. It is well-known in the literature that women tend to present with stroke at an older age than men and with a different profile of risk factors.[[15]] Previous studies have also found that women are less likely to receive IVT despite gaining more net clinical benefit than men, especially at an older age.[[18]] While older age may make some clinicians more reluctant to treat and may thus provide a potential explanation, other studies have also found that suitable female candidates are more likely to be labelled stroke mimics and can inadvertently miss out on treatment.[[18]] This may be due to more frequent atypical presentations or unconscious bias. One study found that when age was corrected for, there were no significant sex-related differences in quality of care.[[18]] As our data were not adjusted for age, it should be viewed as explorative and interpreted with a degree of caution. Further analysis adjusting for age, severity, and baseline morbidity is important.

Our study has several limitations. Firstly, it is difficult to validate all data entered into the clinical register by up to 60 front line clinicians across the country. Therefore, despite the thorough and meticulous system of manual data checking and cleaning that has been implemented to mitigate any errors, it is possible that some mistakes remain. Secondly, the crude outcome measure of “alive or discharged at day seven” does not represent the long road of rehabilitation undertaken by stroke patients. To improve this, reporting of three-month mRS became mandatory in 2019. Despite this, only 50.1% of records included this information in 2020 (down from 54.3% in 2019). To fairly represent the New Zealand post-stroke experience, we must work to increase these numbers. While the main outcomes of interest were intervention rate, sICH and seven-day mortality, we conducted multiple additional analyses. Due to the number of comparisons conducted, there is a possibility that some significant results occurred based on chance alone and these secondary comparisons should thus be interpreted with caution. The inability to update denominator data for 2015 and 2016 will have impacted the precision of historical intervention rate comparisons, however; given the substantial increases in intervention rates over time it is very unlikely that the subtle changes in denominator values will have impacted the overall conclusions. Finally, the analyses by ethnicity and sex were not adjusted for potential confounders such as age, stroke severity, risk factors and time delays and should thus be viewed as primarily exploratory and interpreted with a degree of caution. A sample size was not prospectively determined, and we cannot exclude that significant findings were missed due to the lack of study power.

Despite these limitations, the completeness of data in our register—which includes every patient who underwent reperfusion in New Zealand—strengthens our observations. This minimises selection bias and allows for the accurate analysis of trends in stroke service provision over time.

Conclusion

Reperfusion therapy rates and treatment delays continue to improve although appear to gradually plateau. Complications and mortality have remained stable. The reduction in favourable outcomes following SCR over the last two years is likely related to widening inclusion criteria for patient selection and/or greater delays for more patients due to rising regional transfers. Both require further investigation and mitigation. Ethnic disparities identified include the longer door-to-needle time observed in Māori not explained by geographic factors. Women accessed reperfusion therapy less frequently, potentially explained by older age at presentation. Both require further exploration and action.

View Supplementary Tables.

Summary

Abstract

Aim

This study assessed stroke reperfusion treatments trends in 2019 and 2020 with comparison back to 2015. Additional analyses looked at differences by sex and ethnicity.

Method

The National Stroke Register contains data on all stroke patients who received reperfusion therapies since 2015. Outcomes included treatment rates, delays, mortality and complications by year, sex, and ethnicity. Continuous variables were compared using the Wilcoxon Rank-Sum Test and presented as p-values. Rate-based results were compared using incidence rate comparison and presented as p-values +/- 95% confidence intervals.

Results

In 2020, 11.3% (828/7333) received intravenous thrombolysis (IVT) and 5.5% (404/7333) underwent stroke clot retrieval (SCR), increasing from 6.5% (389/5963) and 0.5% (30/5963) in 2015, respectively. Among reperfused patients (IVT, SCR, both), 8.3% had died at seven days and 3.0% (29/959) experienced sICH. Door-to-treatment time was stable between 2019 and 2020, with median (IQR) of 61 (44–84) and 61 (41–87) minutes, respectively. Initial presentation to a SCR centre was associated with shorter onset-to-reperfusion time of 286 (206–566) minutes, compared with 403 (295–550) minutes (p<0.001). While onset-to-door time was shorter for Māori (72 (44–112) minutes, p <0.001) and Pacific patients (70 (48–105) minutes, p=0.03) compared with NZ Europeans, door-to-needle time was longer in Māori (66 (48–88) compared to 59 (41–83) minutes (p=0.001). Female (73.7+/15.3 years) patients were on average 4.4 years older than males (69.3+/-14.6 years) and less likely to receive thrombolysis (12.7% vs 14.9%, p=0.02).

Conclusion

Reperfusion therapy rates continue to rise, now driven by increasing rates of SCR. Longer door-to-needle time in Māori and lower reperfusion rates in women require further exploration and attention.

Author Information

Natsuko Fushida-Hardy: BSc and Medical Student, University of Otago, Wellington. Anne Kim: Medical Student, University of Otago, Wellington. Andrew Leighs: MBChB, House Surgeon, Hutt Hospital. Stephanie Thompson: PhD Candidate, University of Otago, Wellington. Alicia Tyson: Stroke Nurse and Registry Co-Ordinator, Wellington Hospital. P Alan Barber: MBChB, PhD, FRACP, Stroke Neurologist and Stroke Lead, Department of Medicine, Auckland University and Department of Neurology, Auckland Hospital. Annemarei Ranta: MD, PhD, FRACP, Stroke Neurologist and Head of Department, Department of Medicine, University of Otago, Wellington and Department of Neurology, Wellington Hospital.

Acknowledgements

We acknowledge the clinicians across New Zealand who have contributed to the stroke register and have been responsive to inquiries during data-cleaning for this report. The Neurological Foundation of New Zealand provided funding support to carry out much of this research during sequential summer studentships.

Correspondence

Prof Anna Ranta MD PhD. FRACP FAHA FAAN, Department of Medicine, University of Otago Wellington, PO Box 7343, Wellington 6242, New Zealand, +64 4 806 1031.

Correspondence Email

anna.ranta@otago.ac.nz

Competing Interests

Nil.

1) Ministry of Health (2018). Mortality and Demographic Data 2015. Retrieved 6 February, 2021, from https://www.health.govt.nz/publication/mortality-2015-data-tables  

2) Hedlund F, Leighs A, Barber P, Lundström E, Wu T, Ranta A. Trends in stroke reperfusion treatment and outcomes in New Zealand. Intern Med J. 2020;50(11):1367-1372. doi:10.1111/imj.14682

3) Goyal M, Menon B, van Zwam W et al. Endovascular thrombectomy after large-vessel ischaemic stroke: a meta-analysis of individual patient data from five randomised trials. The Lancet. 2016;387(10029):1723-1731. doi:10.1016/s0140-6736(16)00163-x

4) Stroke Foundation (2021). Clinical Guidelines for Stroke Management. Melbourne Australia. Retrieved 13September, 2021, from https://strokefoundation.org.au/What-we-do/For%20health%20professionals%20and%20researchers/Clinical-guidelines

5) Liu Q, Barber P, Abernethy G, Ranta A. Provision of stroke thrombolysis services in New Zealand: changes between 2011 and 2016. NZMJ. 2017;130(1453):57-62.

6) Gordon C, Bell R, Ranta A. Impact of the national public 'FAST' campaigns. NZMJ. 2019;132(1507):48-56.

7) Wu T, Coleman E, Wright S et al. Helsinki Stroke Model Is Transferrable With “Real-World” Resources and Reduced Stroke Thrombolysis Delay to 34 min in Christchurch. Front Neurol. 2018;9. doi:10.3389/fneur.2018.00290

8) Ranta A, Lanford J, Busch S, Providence C, Iniesta I, Rosemergy I, et al. Impact and implementation of a sustainable regional telestroke network. Intern Med J. 2017; 47(11):1270-1275.

9) Feigin V, McNaughton H, Dyall L. Burden of Stroke in Maori and Pacific Peoples of New Zealand. International Journal of Stroke. 2007;2(3):208-210. doi:10.1111/j.1747-4949.2007.00140.x

10) Samuels I, Wang MTM, Chong KP, et al. Ethnic Differences in Access to Stroke Reperfusion Therapy in Northern New Zealand. Neuroepidemiology. 2020;54(5):427-432. doi:10.1159/000510505

11) Phan HT, Blizzard CL, Reeves MJ, et al. Sex Differences in Long-term Mortality After Stroke in the INSTRUCT (International Stroke Outcomes Study): A meta-analysis of Individual Participant Data. Circ Cardiovasc Qual Outcomes. 2017;10:1-10

12) Reeves MJ, Wilkins T, Lisabeth LD, Schwamm LH. Thrombolysis treatment for acute stroke: issues of efficacy and utilisation in women. Women’s Health. 2011;7:383-390

13) Hacke W, Kaste M, Bluhmki E et al. Thrombolysis with Alteplase 3 to 4.5 Hours after Acute Ischemic Stroke. New England Journal of Medicine. 2008;359(13):1317-1329. doi:10.1056/nejmoa0804656

14) Cadilhac DA, Kim J, Lannin NA, Kapral MK, Schwamm LH, Dennis MS, et al. National stroke registries for monitoring and improving the quality of hospital care: A systematic review. Int J Stroke. 2015;11:28-40.  

15) Ministry of Health. 2020. Stroke Clot Retrieval: A National Service Improvement Programme Action Plan. Wellington: Ministry of Health.

16) Ministry of Health. 2012. Mātātuhi Tuawhenua: Health of Rural Māori 2012. Wellington: Ministry of Health.

17) Thompson S, Barber A, Gommans J, Cadilhac DA, Davis A, Fink J, Harwood M, Levack W, McNaughton H, Feigin V, Abernethy G, Girvan J, Denison H, Corbin M, Wilson A, Douwes J and Ranta A. The impact of ethnicity on stroke care access and patient outcomes: a New Zealand nationwide observational study. Lancet Regional Health Western Pacific. 2022;20:100358. DOI:https://doi.org/10.1016/j.lanwpc.2021.100358.

18) Gattringer T, Ferrari J, Knoflach M, et al. Sex-related differences of Acute Stroke Unit Care: Results from the Austrian Stroke Unit Registry. Stroke. 2014;45:1632-1638

Contact diana@nzma.org.nz
for the PDF of this article

Subscriber Content

The full contents of this pages only available to subscribers.
Login, subscribe or email nzmj@nzma.org.nz to purchase this article.

LOGINSUBSCRIBE