Impact of the corona virus (Covid 19)pandemic on stroke code activities and outcome. A retrospective observational study

In this single site, retrospective, observational study patient data were compared for the first seven months from the beginning of the COVID-19 pandemic (1 February to 31 August 2020) with data for the same time frame in 2019 (Pre-COVID-19).

The study was conducted in a 634-bed, tertiary referral hospital in metropolitan Sydney, NSW, Australia. This hospital provides tertiary stroke care (IV thrombolysis, ECR and neurosurgery) to a population of approximately one million in Northern Sydney. Approximately 1200 emergency presentations are triaged to the stroke pathway per year; approximately 800 patients per year are diagnosed with a principal diagnosis of stroke or Transient Ischaemic Attack (TIA). Patient stroke care is provided in the acute stroke unit (ASU), neuroscience intensive care unit (NSICU) and neurosurgery step-down unit (NSDU).

The study sample inclusion criteria comprised: adults (³18 years) with signs and symptoms of acute stroke, who presented to the ED and were treated for acute stroke (commenced on the stroke pathway) and consequently were discharged from hospital with a confirmed diagnosis of stroke (either ischaemic or haemorrhagic) or TIA. Patients who were transferred to the hospital from other referring facilities and admitted directly into the stroke unit were also included. Classification of stroke was defined as acute episodes with a principal International Classification of Diseases 10th Revision (ICD-10) diagnosis of ischaemic or haemorrhagic stroke (I61.0-I61.9, I62.9, I63.0-I63.9 and I64) or TIA (G45.9).

Data were obtained, for the two time periods, from the Health Information Exchange (HIE) data set - a state-wide data repository in NSW Australia - and the hospital stroke service database. Baseline demographics, hospital length of stay, re-admission to the hospital and primary ICD-10 stroke coding were compiled from the HIE. Data examined included: stroke code activation (stroke pathway) and presentations; admission for stroke care; stroke assessment, imaging and reperfusion treatment; length of stay; in-hospital mortality, degree of disability on discharge and at three months post discharge. The Incident Information Management System (IIMS) database was also reviewed to gather adverse events and near miss incidents reported during the two time periods. The primary outcome measures were stroke presentation, admission to stroke care, and reperfusion treatments. Secondary outcome measures were pre-admission and discharge Modified Rankin Score (mRS) and stroke severity as assessed using the National Institutes of Health Stroke Scale (NIHSS) at the time of presentation. Assessment of mRS was conducted by nurses in the ASU, NSICU and NSDU at time of admission and discharge. NIHSS was assessed at the time of presentation by the neurology registrar (a senior neurology trainee). The electronic medical records were reviewed by members of the research team. Patients who experienced a stroke or TIA during their hospital stay (inpatients) were excluded from the study.

Descriptive statistics were used to characterise and summarise the data using frequencies and percentages for categorical data and medians and interquartile ranges for continuous and ordinal data. The Mann-Whitney U-test was used to compare differences in continuous outcomes with skewed distributions and ordinal variables. Differences in categorical variables were assessed using the Pearson’s chi-square test or Fisher’s Exact Test. A value of P<0.05 was considered statistically significant (2-sided). All analyses were conducted using IBM SPSS Statistics v.27.

The study was determined as adhering to the National Health and Medical Research Council definition of quality assurance and evaluation and approved by the Northern Sydney Local Health District Human Research and Ethics Committee. Only de-identified patient data were collected after extraction from the hospital databases and no identifiable data were reported.

The number of adult presentations in the ED decreased from 38,576 in 2019 to 35,172 during the COVID-19 period (seven months study period). There was also a non-significant trend for less stroke presentations in 2020 compared to 2019 (712 vs 742). However, the percentage of stroke patients for all adult ED presentations were similar (P = 0.32) indicating that stroke presentation frequency was impacted less by COVID-19 than non-stroke presentations. Only when comparing the month of April was there a statistically significant difference (P = 0.02) of fewer stroke code activations (- 31.1%) in the COVID-19 period (n = 73) versus the pre-COVID period (n = 106).

Of the 742 stroke codes activated during the pre-COVID-19 period, 428 (57.7%) patients were diagnosed with a stroke or TIA at discharge. This compared to 441 out of 712 (61.9%) in the COVID-19 group (P = 0.10). Patient baseline characteristics were similar for age and gender. Patients with ischaemic strokes constituted approximately two thirds of all strokes and TIAs (P = 0.32). There was a significantly higher proportion of patients with a haemorrhagic stroke in the COVID-19 period throughout May to August (n = 66, 15.4 % vs n =95, 21.5 %, P = 0.02) compared to the Pre-COVID-19 period (Table 1).

The time from ED arrival to ASU, NSICU or NSU admission was significantly shorter during the COVID-19 period (295 vs 260 median minutes, P = 0.03). The proportion of patients admitted to an inpatient stroke bed was similar (P = 0.86). There was no statistical difference between stroke symptom onset and arrival time in hospital (P = 0.53). There was no significant difference in the “door to imaging time” (P = 0.57) (Table 1).

Ischaemic stroke reperfusion treatments involving IV thrombolysis were similar for the two time periods (P = 0.84) and likewise for patients with ischaemic strokes who received ECR (P = 0.30). The median door to needle time was shorter pre-COVID-19 compared to COVID-19 period (60 minutes vs 65 minutes, P = 0.53), although, it was not statistically significant. There was no statistical difference in stroke onset time to administration of IV thrombolysis (P = 0.18).

The percentage of complications during hospital admission was similar (P = 0.95); in-hospital mortality was 9.3% (n =40) [pre-COVID-19] compared to 12.5% (n = 55) in the COVID-19 period (P = 0.16). There were three incidents (one patient with complications resulting from a fall, one with skin graft issue, and one patient with respiration pneumonia) with a Severity Assessment Code (SAC) of 2 (moderate harm attributed to health care provision) reported in the pre-COVID-19 time period and five (three patients with peripheral intravenous cannula site infections, and two complications from a fall) in the COVID-19 time period.

Length of stay was comparable for the patients’ hospital stay (median 6 vs 6 days, P = 0.61). The median pre-admission mRS was zero. At the time of presentation to hospital during the COVID-19 period, patients had a higher stroke severity than before the pandemic (median NIHSS 3 vs 4, P = 0.049). There were no differences in stroke outcomes (degree of disability) measured by the mRS on discharge from hospital or at three-month follow-up (Table 2).

This retrospective study shows that only at the beginning of COVID-19 community transmissions in Sydney during 2020, was there a transient period of fewer stroke presentations. This finding is similar to the reported healthcare activity in NSW (Sutherland et al.). During April and the first half of May 2020, restrictions on community movement and interactions were strictly implemented in Sydney. These measures may have helped reduce community transmissions however, they may have caused patients with acute stroke symptoms to avoid calling an ambulance or presenting to ED for fear of contracting the infection in hospital.

With regard to the type of stroke, more patients presented to the ED with haemorrhagic stroke during the COVID-19 pandemic than the year before the pandemic. This might explain the higher baseline stroke severity scores during the COVID-19 pandemic period as well as the marginal longer hospital length of stay. However, there was no significant change in the total number of confirmed stroke or TIA presentations, rate of reperfusion therapy, time to present to hospital or receive therapy or any other measured clinical outcome. This reflects the strength of the organisation and commitment of the ED and stroke teams who continue to focus on providing the best care possible to stroke patients during the pandemic.

In addition, while there was no statistically significant difference in overall access to stroke unit care, the median time to acute stroke unit admission was significantly shorter during the pandemic. The reported changes in the delivery of acute stroke services as a result of the COVID-19 pandemic has differed significantly worldwide. While some hospital systems have been significantly restructured (including the redeployment of senior clinicians to other areas, the reallocation of acute stroke units to care for COVID-19 patients, modified stroke pathways and changes to triage and transportation protocols), other hospitals have not undergone those changes (Bersano et al., 2020; Cadilhac et al., 2021; Hoyer et al., 2020; Markus & Brainin, 2020). Stroke services in our organisation were not reallocated so continuation of usual clinical care was possible.

Internationally, there have been reports of delays in hyperacute stroke treatment delivery due to changes in workflow processes of the emergency and radiology departments in order to prioritise infection preventative strategies (Cadilhac et al., 2021; Hoyer et al., 2020; Markus & Brainin, 2020). In contrast, our study demonstrated that the provision of reperfusion therapy was not adversely impacted for ED stroke presentations. This may be attributable to the lower case numbers of COVID-19 reported in Australia in 2020.

There was a significantly shorter transfer time to the acute stroke unit /stroke bed during the COVID-19 period. A possible explanation is that during COVID-19 transfer from ED to inpatient beds was expedited to minimise possible exposure to COVID 19 in ED. Additionally, there may have been a decrease in workload in the stroke unit relative to the ED workload given new COVID-19 workflow modifications. Of concern was the higher number of haemorrhagic stroke presentations during the pandemic and this remains unexplained. Further investigation of confounding patient variables such as risk factor profiles and/or medication control and compliance is needed.

We acknowledge that the data from one single centre on Australia’s east coast is not representative of true population-based data in a global pandemic and that the numbers were too small to detect minor but potentially important clinical differences reliably. An additional limitation is the lack of complete data for some of the variables (the COVID-19 time period contained more complete data for all measures), which may have introduced sample bias. Finally, the omission of time-based metrics for ECR (onset-to-groin or door-to-groin time) was also a limitation, as it is well known that timeliness of ECR improves functional outcomes (Bersano et al., 2020).

Of concern to all health professionals was that the COVID-19 pandemic may delay presentations to ED. It remains unclear if the pandemic altered behaviour so that people chose not to present to ED despite significant symptomatology. In this study while overall adult presentations were lower, stroke presentations remained unchanged. While stroke severity on presentation was marginally higher in the COVID-19 time period (median NIHSS of 4 in COVID-19 group compared with NIHSS of 3 pre-pandemic group), this may be related to the severity of haemorrhagic strokes as reported in an earlier study (Andersen, Olsen, Dehlendorff, & Kammersgaard, 2009).

Future studies could involve a larger cohort and include all patients who present to hospital with stroke symptoms, not just those who were admitted following ED stroke code activation (for stroke treatment) as presented in this study. Further COVID-19 studies could examine the long-term outcomes of patients presenting during the pandemic. Of specific importance is the outcome of stroke patients with COVID-19 infection, as early data indicates an increased risk of ischaemic stroke in people with COVID-19, as a result of SARS-COV-2 associated hypercoagulability and thrombo-inflammation (Connors & Levy, 2020).