Background

In the event of an acute ischemic stroke, fast recanalization of the occluded vessel is crucial because of the rapid death of neurons in ischemic brain tissue [2]. The only evidence-based measures to achieve tissue reperfusion are administration of tissue-type plasminogen activator and endovascular thrombectomy. The Mobile Stroke Unit (MSU) concept was first introduced in 2008 in Homburg, Germany [3]. In 2011 the Stroke Emergency Mobile (STEMO) was implemented in Berlin, Germany and yielded clear improvements regarding thrombolysis rates and onset-to-treatment times [4]. Additionally, ischemic and hemorrhagic stroke patients have been more accurately routed to hospitals with Stroke Units or neurosurgery departments [5]. In the meantime, the MSU concept has been implemented in many cities worldwide, and various groups contribute their experiences to the Pre-hospital Stroke Treatment Organization (PRESTO) [6]. Yet, despite a non-significant trend for disability-free survival at 3 months in favor of the MSU care group in a registry-based analysis [7, 8], definite scientific proof of a positive effect of MSU treatment on functional outcome of stroke patients is still lacking. Thus, we started the B_PROUD study on February 1, 2017 [1]. In the meantime, the STEMO service in Berlin was extended to three simultaneously operating STEMOs covering almost the entire Berlin city area. Since logistical problems occurred during almost the entire first study period, we recognized the necessity of a revised study protocol for an additional evaluation during the period after resolving these challenges. The deviations from the original STEMO setting during the PHANTOM-S trial [4] included reduced accuracy of dispatching for stroke emergencies, fewer STEMO dispatches for severe stroke cases (because general emergency physicians were dispatched to patients with impaired consciousness or instable vital parameters instead of the STEMO team), more frequent cancellations of the STEMO dispatch (thus leading to more cross-overs and in-hospital thrombolyses), and longer distances to scene. These issues led to fewer and more delayed STEMO interventions and thrombolytic treatments. After adjustment of the STEMO dispatch organization, we expect that the use of the dispatch algorithm will be equivalent to its use during the preceding PHANTOM-S trial. Following the recommendation of the B_PROUD Data Safety Monitoring Board and in agreement with the Berlin Department of Internal Affairs, we decided to start a second part of the B_PROUD study - with full reporting of the first evaluation period.

Methods

Aim of the trial

Here, we describe a confirmatory trial to prove the efficacy of the MSU intervention compared to regular care using the modified Rankin Scale (mRS) score 3 months after event.

Study description and study design

B_PROUD (Berlin Prehospital Or Usual Delivery of acute stroke care) is a pragmatic, prospective, multicenter study with blinded outcome assessment (PROBE design). Inclusion of eligible patients is currently carried out in the metropolitan area of Berlin, Germany, in cooperation with all 15 stroke centers of various Berlin hospital owners. The B_PROUD study makes use of the B-SPATIAL registry (Berlin – specific acute therapy in ischemic or hemorrhagic stroke with long-term follow-up, NCT03027453) that collects 3 months follow-up assessment on an opt-out basis. In line with German data-protection legislation and approval of the Berlin data protection representatives, patients are informed beforehand about the planned follow-up and can opt-out at any time before or during the telephone interview or optional questionnaire based assessment.

Arms and intervention

The intervention group consists of patients for whom an MSU is deployed by the dispatch center of the Berlin Fire Department after stroke suspicion during emergency call [9]. Based on the preceding PHANTOM-S trial [4], approximately 45% of the stroke alarms are expected to be handled by regular ambulances because the MSU is already in operation or undergoing service. Compared to regular ambulances offering standard of care, the intervention by MSU includes prehospital neurovascular expertise by a neurologist staffing the MSU, computed tomography (CT) based brain scanning including visualization of large vessel occlusion by CT-angiography, and specific pre-notification to endovascular treatment capable centers. Details of the STEMO equipment, staffing and operational procedures have been published elsewhere [10, 11]. After a decision of the Berlin state government that STEMO care should be available for the entire population of Berlin, two additional STEMOs have been implemented and gone operational in 2017 and 2018 – under the condition of an accompanying scientific evaluation on outcome effects.

Eligibility criteria

All patients calling the emergency services and prompting a stroke alarm [9] at the dispatch center will be screened for eligibility. Only treatment candidates defined by the inclusion and exclusion criteria listed in Table 1 will be included in the primary study population and thus compared for the primary outcome. Thereby, the screening process is operationalized aiming at a minimization of selection bias. Monitoring is carried out continuously for all eligibility criteria. If ascertainment is uncertain, the clinical documentation is submitted to an independent adjudication committee for blinded judgement. Additionally, patients with [1] stroke mimics receiving thrombolysis will be included in a sensitivity analysis together with the primary study population, and patients with [2] intracerebral hemorrhage presenting within 6 h from symptom onset will be analyzed as a companion study population.

Table 1 Inclusion and exclusion criteria for the study population
Full size table

Outcome measures

The primary outcome is the modified Rankin score (mRS) 3 months after the acute event which is the most common outcome measure in stroke trials [12]. Since the type of intervention does not allow blinding of patients, the structured telephone interviews after 3 months are recorded and subsequently assessed by stroke experts who are unaware of the treatment arm allocation. For those patients who remain unreachable via phone or mail, we use information from registration offices regarding vital and residential status including living address. This information allows assessment of the co-primary outcome consisting of the following three categories: ‘1’ able to ambulate (mRS 0–3) or (if mRS not available) living at home, ‘2’ living with severe disability (mRS 4–5) or living in institutional care, and ‘3’ death (mRS 6). We introduced this novel outcome definition within a study protocol amendment during the first part of the B_PROUD trial in order to use all available information for the purpose of informing parties involved in acute stroke care organization on the effectiveness of the MSU concept. Only if the lost-to-follow-up rate is higher than 9% in this community nested trial, this outcome will be used as a co-primary outcome measure. .

Secondary outcomes include performance measures like thrombolysis and endovascular treatment rates and process times as well as clinical outcomes such as quality of life and dichotomized three-months mRS. All outcome measures are listed in in detail in Table 2.

Table 2 Outcome measures
Full size table

A continuous reporting system for serious adverse events of special interests (SAESI) is implemented to check rates of symptomatic secondary intracerebral hemorrhages (sICH) or deaths. In case of more than 10 deaths within 7 days (or at discharge, whatever comes first), or more than 10 symptomatic sICH per 100 treatment candidates in the MSU group, the study has to be stopped after recommendation by an external data safety monitoring board. To address the intention-to-treat approach for outcome analyses, we rigorously consider STEMO availability as determinative and, for example, count cancellations of STEMO dispatches to the MSU intervention group.

Sample size estimation

The proposed sample size to be analyzed is 1372 patients – equal to the sample size of B_PROUD part 1. Sample size calculation was originally performed in October 2015 based on outcomes seen in a registry-based comparison of the first 193 patients with pre-hospital thrombolysis on STEMO and 615 consecutive patients with in-hospital thrombolysis of the Charité thrombolysis registry. Inclusion of 1500 patients will be necessary, considering 9% lost-to-follow-up. For the primary outcome, we expect the following differences (STEMO vs. control group): mRS 0: 21/21%; 1: 21/15%; 2: 7/9%; 3: 20/12%; 4: 11/14%; 5: 5/9%; 6: 15/20%. The Mann-Whitney test with two-sided significance level of 0.05 has 80% power to detect such a group difference in at least 686 patients per group.

In case of more than 9% lost-to-follow-up, the study will be seen as successful only if both co-primary endpoints (mRS, mRS in three categories: 0–3, 4–5, 6) show significant better outcome for the intervention using a two-sided significance level of α = 0.05.

The power calculation was conducted with the R-Package sample size [13]. Since there is scarce information on the possible effect size, an interim analysis for a blinded sample size re-estimation is planned after primary outcome assessment of 300 patients.

Contacts

B_PROUD is sponsored by Charité – University Medicine Berlin, Berlin, Germany via the Center for Stroke Research Berlin and the excellence cluster NeuroCure, Berlin, Germany, and funded by German Research Foundation. The B_PROUD MSU evaluation is conducted in close collaboration with the Berlin Fire Department and its medical lead of EMS as well as with the 15 Berlin stroke centers.

Perspective

While scientific evaluations of the Berlin MSU implementation so far have shown that pre-hospital thrombolysis is safe and associated with a substantial shortening in time to treatment [4, 7], a definite proof of clinical benefit has not yet been accomplished. The B_PROUD trial was therefore designed to provide confirmatory evidence that earlier (pre-hospital) stroke work-up and treatment leads to better functional outcomes. Recruitment to B_PROUD part 1 was started in February 2017 and is planned to be completed in 2019, concurrent with a similar trial, the BEST-MSU trial [14]. Despite similar inclusion criteria of the latter also ensuring the comparison of only treatment candidates, our study design allows follow-up of the patients without written informed consent which may to some extend limit external validity by introducing a selection bias.

In addition to the primary outcome, the B_PROUD study will provide valuable information on effects of earlier treatment (blood pressure lowering and anticoagulation reversal) on hematoma volumes in intracerebral hemorrhages.

While the time saving approach of the prehospital MSU concept is intriguing, stakeholders of stroke care need reliable information on the effect size of clinical outcomes and cost-effectiveness. The expected benefits in outcome and potentially reduced costs for hospital and long-term care need to be weighed against additional costs of implementation and running not only a single but three STEMOs. B_PROUD and parallel studies such as BEST-MSU [14] are designed to provide the needed data and will therefore support future decision making.