Stroke clot retrieval (SCR) in patients with large vessel occlusion (LVO) is associated with significantly reduced disability at 90 days compared with intravenous thrombolysis alone.[[1]] The effectiveness of these reperfusion therapies are time critical. In Aotearoa New Zealand, SCR candidates require urgent transfer to one of three centres: Capital & Coast, Te Toka Tumai Auckland or Waitaha Canterbury.

Te Pae Hauora o Ruahine o Tararua, MidCentral serves a population of 186,190 and admits approximately 360 patients with stroke per annum. It provides a 24/7 thrombolysis service and SCR has been provided by Capital & Coast since 2019 via a regional telestroke network. At time of writing, MidCentral did not have perfusion imaging and so based on available evidence, the treatment window for SCR was limited to six hours from stroke onset. Inter-hospital transfer services for MidCentral were provided by the Capital & Coast retrieval team. An audit of patients referred for SCR undertaken in 2020 demonstrated a door-in-door-out (DIDO) time of approximately four hours. This delay limited the number of patients eligible for treatment.

A Search and Rescue Services (SRSL) helicopter is located on Palmerston North Hospital grounds, however, historically its role was limited to prehospital care. Use of the SRSL service could potentially reduce inter-hospital transfer times by at least half, and thus improve SCR accessibility. Here we describe a quality improvement collaboration with SRSL utilising the Taranaki model[[2]] for SCR: In summary, this involves pre-hospital alert, notification to the stroke team, standardised imaging protocol and rapid image transfer, and activation of the helicopter transfer team once LVO is confirmed prior to acceptance by the SCR neurologist.

The purpose of this report is to describe; the tailored application of the Taranaki model at MidCentral, the cohort of patients referred for SCR to date, and ongoing challenges and opportunities.

We considered a rapid improvement in inter-hospital transfer times to be necessary to achieve best patient outcomes. Factors considered most likely to help achieve this goal included: 1) having a narrow, well defined patient population, e.g., including patients with anterior circulation LVO who met SCR criteria and excluding patients with basilar artery occlusion who are at risk of airway compromise and therefore may require the skill set of specialist retrieval services;[[3]] 2) engaging with essential stakeholders; and 3) having a compelling business case.

By 2020, the changes brought in by the National Air Ambulance Sector Reforms required SRSL to have dedicated Critical Care Flight Paramedics. Further, SRSL were willing and able to provide an independent inter-hospital transfer service for time critical patients from hospitals.

Methods

Following a series of discussions and meetings with clinical and service managers, duty nurse managers, SRSL and members of the stroke team, pathways were developed and widely disseminated, and education and training provided. Direct contact numbers for SRSL are displayed prominently in pathways. In August 2020, SRSL started assisting MidCentral with SCR transfers.

For this report, we used a prospective hospital thrombectomy registry with information on demographics, national institutes of health score (NIHSS) at baseline and at 24 hours (scale from 0–42, higher numbers reflecting greater impairment), thrombolysis status and DIDO times. Other data included 90-day functional outcomes using the ordinal modified Rankin Scale, whereby 0–2 represents functional independence, 3–5 progressively increasing dependency, and 6–death. We analysed patients transferred prior (April 2019 to August 2020) and after implementation of the SRSL service (August 2020 to March 2022).

Continuous and categorical variables are presented as mean (SD) or median (IQR) and frequencies. Continuous data were tested using t-test (normally distributed) or Wilcoxon Rank-Sum Test for non-parametric data. Fisher’s exact test was used for categorical data. The relatively small numbers precluded regression analysis. A p-value of ≤0.05 was considered statistically significant. This project was exempt from ethics approval following institutional review. This manuscript was written in accordance with SQUIRE guidelines.[[4]] Data were analysed using STATA/BE 17.

Results

Twenty-four patients (13 female), median age 68.5 years (IQR: 52.5 to 77.5), four NZ Māori and the remainder NZ European, were referred for SCR: Auckland n=3; Canterbury n=1 (during Auckland Level 4 lockdown), and the remainder to Capital & Coast. Seventy-nine percent of patients also received thrombolysis.

Pre SRSL service, eight patients had been transferred, and 16 patient post. With SRSL, DIDO times reduced from 242 to 90.5 minutes (IQR: 69 to 99.5), difference -151.5 minutes [95%CI -109 to -194, p<0.001], (see Figure 1). As an example, our first SCR patient under the SRSL service with an initial NIHSS of 22 had a DIDO time of 64 minutes. The following day post SCR, NIHSS was 2.

Door to groin puncture time reduced from 350 to 197 minutes, difference -153 minutes [95%CI -56 to -251, p<0.01]. The mean admission NIHSS was 14(6), reducing to 7(6) at 24 hours. At three months, 55% of patients were functionally independent and 17% had died. There were no significant differences in the NIHSS change at 24 hours, or proportion of patients at three months, with functional independence pre and post SRSL (p-values >0.05). Proportionally, more patients were sent for SCR following the intervention, though this did not reach statistical significance (OR 1.99 [95%CI 0.83 to 4.75], p=0.14).

Figure 1: Door-in-door-out times.

Discussion

Use of SRSL aeromedical inter-hospital transfer resulted in significant reductions (2.5 hours) in door-in-door-out times and a twofold increase in the numbers able to access stroke clot retrieval. At three months, over half of these patients were functionally independent.

There was no significant difference in outcomes between patients transferred pre and post SRSL. A recent observational study found no difference in outcomes following SCR in patients transferred from Taranaki by air, to patients from Counties Manukau/Waitematā transferred by road, despite a mean 77-minute delay in door to groin time.[[5]] This suggests that equitable outcomes are possible for regional New Zealand, and as in MidCentral, a national drive to reduce transfer times would increase the cohort of patients who could potentially benefit from SCR.  

Strengths of this project are the cost-neutral, simple, intuitive approach. Our patient outcomes are comparable to results from other studies, suggesting appropriate patient selection and use of resources. Finally, we demonstrate generalisability of the Taranaki SCR model. This project has been well received by staff and has encouraged other specialties to consider SRSL for time critical transfers.

Ongoing challenges include the need for frequent reminders and education to the wider hospital staff. In particular, we note increased DIDO times outside normal working hours.

Limitations of this study are the low patient numbers and incomplete data on co-morbidities, which limits our ability to draw further inferences.