Berlin prehospital or usual delivery of acute stroke care

Protocol

Berlin prehospital or usual delivery of acute stroke care – Study protocol

International Journal of Stroke 0(0) 1–6 ! 2017 World Stroke Organization Reprints and permissions: sagepub.co.uk/journalsPermissions.nav DOI: 10.1177/1747493017700152 wso.sagepub.com

Martin Ebinger1,2, Peter Harmel1, Christian H Nolte1,2,3, Ulrike Grittner1,4, Bob Siegerink1 and Heinrich J Audebert1,2

Abstract Rationale: Prehospital stroke care in specialized ambulances increases thrombolysis rates, reduces alarm-to-treatment times, and improves the prehospital triage. Preliminary analyses suggest cost-effectiveness. However, scientific proof of better functional outcome compared to usual care is still lacking. Aim: To prove better functional outcomes after deployment of the Stroke Emergency Mobile compared to regular ambulances. Sample size estimates: A sample size of 686 patients will be required in each arm (Stroke Emergency Mobile group vs. regular care) to detect a difference regarding the primary outcome with 80% power at a two-sided significance level of 0.05. Methods and design: This is a pragmatic, prospective study with blinded outcome assessment. Primary outcome will be functional status as defined by modified Rankin Scale score three months after the incident event. We will include cerebral ischemia patients within a predefined catchment area in Berlin, Germany. The study population consists of patients who might be candidates for acute recanalizing treatments, with onset-to-alarm time 4 h, symptoms not resolved at time of ambulance arrival, and able to walk without assistance prior to the qualifying incident. About 45% of Stroke Emergency Mobile dispatches are expected to be handled by regular ambulances, since Stroke Emergency Mobile will be already in operation creating the control group. Primary outcome: Functional outcome after three months measured by the modified Rankin Scale over the entire range. Discussion: The results will inform decision makers on the effectiveness of Stroke Emergency Mobile. Keywords Stroke, prehospital, telemedicine, thrombolysis, thrombectomy, endovascular treatment, functional outcome Received: 28 August 2016; accepted: 2 December 2017

Introduction and rationale Stroke is a leading cause of disability and mortality.1 Intravenous thrombolysis and endovascular thrombectomy (ET) are the only evidence-based causal treatments of acute ischemic stroke.2 The effects of both treatments are time dependent.3,4 Shifting thrombolysis into the prehospital phase on the Berlin Stroke Emergency Mobile (STEMO) is safe, increases thrombolysis rates, and reduces time delay to treatment.5,6 Deployment of the STEMO resulted in improved triage and a 10-fold increase of thrombolysis within the so-called golden hour.7,8 Data from post-hoc analysis7 and registry-based comparisons9 suggest a positive effect on outcome by prehospital

start of thrombolysis. Preliminary data suggest costeffectiveness.10 We seek to determine, whether or not deployment of STEMO for acute stroke patients is associated with better outcomes three months later, 1 Center for Stroke Research Berlin, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany 2 Department for Neurology, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany 3 Berlin Institute of Health (BIH), Berlin, Germany 4 Department of Biostatistics and Clinical Epidemiology, Charite´ – Universita¨tsmedizin Berlin, Berlin, Germany

Corresponding author: Peter Harmel, Charite´ – Universita¨tsmedizin Berlin, Charite´platz 1, 10117 Berlin, Germany. Email: [email protected]

International Journal of Stroke, 0(0)

2

International Journal of Stroke 0(0)

similar to another ongoing study11 (clinicaltrials.gov NCT02190500).

Methods Design B_PROUD (Berlin Prehospital Or Usual Delivery of acute stroke care) is a pragmatic, prospective study with blinded outcome assessment (see Figure 1). B_PROUD is designed to determine whether the deployment of STEMO (intervention) leads to better functional outcome at three months (primary endpoint) of stroke patients eligible for acute treatments compared to regular emergency medical service (EMS). The protocol is registered in http://clinicaltrials.gov (NCT02869386) and has been approved by the ethical review committee of the Charite´ – Universita¨tsmedizin Berlin (EA4/109/15).

Patient population Patients with suspected stroke, aged 18 years, causing an emergency call from the catchment area, and prompting STEMO alarm by the central dispatch center12 will be included in the study. For definition of the primary study population (treatment candidates) see Table 1. Of note, patients with contraindications against intravenous thrombolysis but possible indication for ET will be included (see exclusion criteria 5).

In case of uncertainty, a blinded adjudication committee will decide whether patients were eligible or not.

Treatment allocation B_PROUD will be conducted as a non-randomized intervention study or natural experimental study where the treatment allocation is based on STEMO availability. This design that mimics a randomized trial was chosen because (1) STEMO care has become part of regular care in Berlin, and therefore, randomization at patient level is ethically not acceptable; (2) informed consent for study participation cannot be retrieved over the phone in Germany. A key element of STEMO care is that the suspicion of stroke following the stroke identification algorithm12 prompts a stroke alarm in the dispatch center. However, previous experience showed that in about 45% of stroke alarms, STEMO could not be deployed because it was already in operation, in service, or otherwise unavailable.6 Whether or not STEMO is available usually remains unknown to the dispatcher when prompting the alarm, thus keeping the treatment allocation concealed. By this way, limited availability emulates a natural comparison group with all key elements of a randomized clinical trial (i.e. random and concealed treatment allocation) where STEMO is compared to standard of care. Patients will be allocated to one of two study arms for the analyses (STEMO available or STEMO not

Figure 1. B_PROUD study flow.


Ebinger et al.

3

Table 1. Inclusion and exclusion criteria for study population Inclusion criteria: 1. Suspected acute stroke according to the dispatcher stroke identification algorithm12 during STEMO hours and within the STEMO catchment area 2. Age 18 years Inclusion criteria for primary study population: 3. Final diagnosis of ischemic stroke (ICD10: I63) or TIA (G45 except G45.4) 4. Prestroke mRS 3 (able to ambulate without assistance) 5. Confirmed onset-to-alarm time 4 h Exclusion criteria: 1. Symptom remission until arrival of EMS (according to conventional EMS documentation) 2. Malignant or other severe primary disease with life expectancy 1), as well as in patients >80 years living at home with help or living in an institution before stroke (mRS2 vs. mRS>2). Additional outcomes analyzed in the whole study population (not restricted to the primary study population, i.e. treatment candidates with or without need of assistance prestroke) are: discharge status, in-hospital mortality, survival analysis, and delivery rates to facilities with capability of ET in cases of large vessel occlusions or neurosurgery in cases of intracerebral hemorrhage as well as cost-effectiveness.

Primary outcome

Data safety monitoring board

The primary endpoint is the modified Rankin Scale (mRS) after three months in the primary study population (Figure 1, Table 1) assessed with a standardized telephone interview questionnaire which will be recorded for treatment-blinded rating.

Even though our study tests two standard practices for prehospital stroke care, a data safety monitoring board (DSMB) has been established for issues of patient safety, but especially treatment efficacy data, protocol adherence, and advice whether to continue, modify or

Intervention


4

International Journal of Stroke 0(0)

stop the study. Members of the DSMB are from institutions outside of Germany. The following stopping rules will be suggested to the DSMB: In case of >10 deaths within seven days (or at discharge, whatever comes first) or >10 symptomatic secondary hemorrhages per 100 treatment candidates in the STEMO group, or if interim analysis after a total of 300 treatment candidates reveals a worse outcome in STEMO (p