Print page Print page
Switch language
The Capital Region of Denmark - a part of Copenhagen University Hospital

Prehospital management of stroke patients by Emergency Medical Services (EMS) – an evaluation of Emergency Medical Dispatch (EMD) and ambulance personnel

Project: Types of projects

View graph of relations
Approximately 12.000 strokes occur in Denmark annually. The only evi-dence based pharmacological treatment of ischemic stroke is thrombolysis within a time-window of 4.5 hours from symptom onset(1). In 2012, 12% of patients with ischemic stroke in Denmark received thrombolysis treatment. The probability of a favourable 3-month outcome after stroke decreases with time from symptom onset to thrombolysis: the onset to needle time(2). Therefore, it is of the highest importance to identify potential targets for prehospital time reduction, and thereby reduce delay to definitive care.

Prompt recognition of stroke symptoms by both the EMD and ambulance personnel is a crucial step, as it provides access to fast track acute stroke services. In this context, the role of the EMS is essential, as contact to the hospital through EMS opposed to private transport is correlated to de-creased onset-to-door time (3). EMS prenotification of the stroke centre is correlated to shorter door-to-evaluation time, door-to-needle time and onset-to-needle time(4). In Denmark EMS prenotification in acute stroke is well established.
According to Danish guidelines for treatment of acute stroke and transient ischaemic attack (TIA), prehospital diagnosis of stroke is improved by use of validated tests(5). No such is yet implemented in the Capital Region of Denmark.
Recent data shows that the mean time from the ambulance personnel ar-rives at the scene to they start the transport to the hospital (“on scene time”) is 18 minutes in the Capital Region of Denmark (6). American guidelines suggest that on-scene-time should be less than 15 minutes in all cases (unless extenuating circumstances or extrication difficulties are present)(1), however the organisation of stroke care between Scandinavia and USA varies largely.

The aim of this study is:
1. To establish the rate of recognition of stroke symptoms by healthcare personnel in the EMD, ultimately aiming at optimizing performance, and identifying barriers for a fast and correct EMS-response.
2. To assess prehospital time consumption by ambulance personnel on the scene of a stroke, create a baseline for future comparison and generate explorative hypotheses for forthcoming interventions.

The study consists of 3 arms:
1. An epidemiological analysis of prospectively collected data from a 2-year period, in order to establish the rate of identification of stroke symptoms by healthcare personnel in the EMD.
2. A quantification and statistical analysis of on-scene-time in stroke patients, based on prospectively collected data from a registration form filled out by the ambulance personnel after responding to a stroke. The focus of the form is on time spent on predefined tasks aiming at assessing time consumption on the scene, and generating explorative hypotheses for forthcoming interventions.
3. Based on EMS data, we will establish a baseline for on-scene-time and transportation time in patients admitted by fast track stroke service during a 6 months period, for future comparison and ana-lyse data in conjunction with in-hospital delays.

Method and endpoints for each stage:

1. The epidemiological analysis is based on prospectively col-lected data from the Danish Stroke Registry (DSR) and the National Patient Registry (NPR). In the DSR and NPR, ICD-10-codes are linked to the Danish personal identification number, a unique number for every Danish citizen. We shall compare ICD-10 codes from DSR and NPR with dispatch codes from the EMS database, to determine the amount of diagnosed stroke patients recognized during the emergency call. The date of the incident report from the EMS will be matched with the admission date correlated to the discharge ICD-10 code, in case a patient has been admitted more than once.

All patients are included if registered in DSR with an ICD-10 code of stroke (haemorrhagic and ischemic – I61 and I63-64.9), in the NPR with an ICD-10 code of TIA (Transient Ischemic Attack – G45.9) or in the EMS-database with a dispatch code of suspected stroke (A.26.03, A.26.04) from the 2-year period 01-01-2012 – 12-31-2013.

Patients are excluded if the Danish personal identification number is not available in either database. If the patient has been directly referred to the hospital by their GP without the use of a EMS transport, or they have presented themselves in the emergency room, they will not be registered in the EMS database.

Primary endpoint:
• The rate of acute stroke recognized by the EMD during the emer-gency call (sensitivity).

Secondary endpoints:
• Final diagnoses(ICD-10), in patients with a dispatch code of stroke, but another final diagnosis (false positive).
• Dispatch codes, in patients with a final diagnosis of acute stroke, but another dispatch code (false negative).
• The positive predictive values of stroke recognition by healthcare personnel at the EMD.
• The proportion of patients with a final diagnosis of stroke, which accessed the healthcare system through the EMS.

2. To quantify and analyse prehospital time consumption, we will develop a registration form for the ambulance personnel to fill out af-ter responding to a stroke patient.

This registration form will include registration of time intervals spent on:
• Clinical tasks; including prior medical history, clinical examination and measurement of vital parameters
• Communication in relation to prenotification of the stroke centre
• Mobilization of the patient and on-scene conditions
• Patient preparation; including IV-accesses and 12-lead ECG

As well as information regarding:
• Patient preparation; including IV-accesses and 12-lead ECG
• Language/communication barriers
• Specifications of the ambulance (operator, level of education)
• Patient/relative related conditions
• Visitation to acute stroke fast track

Statistical analyses:
Poisson regression analyses will be made with over dispersion taken into account, in order to identify specific time intervals and tasks correlated to extended or decreased total on-scene-time. This model yields distributions similar to previously collected data.
A power calculation has been made, based on this distribution and the as-sumption that a clinically relevant difference, is defined as any time inter-val in cases considered as having a high on-scene-time, (>15 min.) being twice the amount than in cases considered as having a low on-scene-time (≤15 min.). This has shown an estimated demand of 500 registration forms in order to achieve a power of 0,8 with a 5% two-sided confidence interval.

Patients with suspected stroke where the EMS has been in contact with fast track stroke service at the stroke centre are included.

Exploratory endpoints:
Hypothesis-generating outcome for on-scene-time in relation to:
a. Clinical tasks
b. Communication
c. Logistic/on-scene conditions
d. Thrombolysis preparation
e. Patient/relative related conditions

3. The establishment of on-scene-times and transportation-times will be made by extracting specific time points (e.g. ambulance arri-val at the scene, departure towards hospital and arrival at hospital) from the EMS database during a 6 months period, and analysed in conjunction with data on in-hospital delays.
The stroke centres will provide Danish personal identification numbers for patients referred to the stroke centres as well as door-to-needle times and treatment decisions in order to determine if there is a correlation be-tween recognition and treatment decisions in stroke fast track patients.

As there is no prior comparable registration of this, it is impossible to es-timate a clinically relevant difference in order to make a power calculation and determine the amount of necessary registrations. Thus it has been decided that a 6-month registration will provide a reliable baseline.

All patients with a Danish personal identification number, referred to one of the two regional stroke centres, during a 6-month period will be included. The 6 month period will begin after the above mentioned 500 registration form has been collected, to avoid bias.

Primary endpoints:
• Time from dispatch to ambulance arrival at the scene of accident in stroke fast track patients.
• Time on scene by EMS in stroke fast track patients (on-scene-time).
• Time from departure from the scene of the accident to arrival at the stroke centre in stroke fast track patients.

Secondary endpoints:

• Correlation between prehospital and in-hospital time periods in stroke fast track patients.
• Correlation between recognition and treatment decisions in stroke fast track patients.

The overall aim of this study is, to ensure the best possible prehospital care for all stroke patients. The study will determine, if the EMD is at a high and international level in regards to stroke recognition, and if there is a potential to decrease on-scene time. This will serve as future comparison and have an effect on the structure and future education in the prehospital services, and potentially improve the outcome after acute ischemic stroke.

Ethical considerations:
Approval from the Ethics Committee is not necessary (H-4-2014-FSP). Permission to extract and merge data from the National Patient Registry, the Danish Stroke Registry and the EMS database has been granted from the Danish Data Protection Agency via the Capital Region (2007-58-0015).

Distribution of responsibilities:
Søren Viereck; Medical student, University of Copenhagen; research em-ployee/project manager: Literature search data analysis, development and evaluation of registration form for ambulance personnel, first draft of paper (planned 1. author) and fundraising.

Thea Palsgaard Møller; MD; Research fellow, Emergency Medical Services Copenhagen: Data extraction and supervision on data analysis, critical re-view of results.

Karl Bang Christensen: Associate Professor, Department of Biostatistics, University of Copenhagen: Critical review of protocol, supervision on sta-tistical analyses, critical review of results.

Hanne Christensen; Senior Stroke Neurologist & Associate Professor, MD, PhD, DMSci, Department of Neurology, Bispebjerg Hospital, University of Co-penhagen: Supervision, critical review of protocol, provision of data, supervision of data-analysis, critical review of results

Helle Klingenberg Iversen, MD, Senior Stroke Neurologist & associate professor Department of Neurology, Glostrup Hospital, University of Copenhagen: Supervision, critical review of protocol, provision of data, supervision of data-analysis, critical review of results.

Freddy Lippert; CEO, Emergency Medical Services in the Capital Region of Denmark; Associate professor, University of Copenhagen: Supervision, critical review of protocol, provision of data, supervision of data-analysis, critical review of results and overall clinical and scientific responsibility,


1. Jauch EC, Saver JL, Adams HP Jr, Bruno A, Connors JJB, Demaerschalk BM, et al. Guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke J Cereb Circ. Marts 2013;44(3):870–947.
2. Lees KR, Bluhmki E, von Kummer R, Brott TG, Toni D, Grotta JC, et al. Time to treatment with intravenous alteplase and outcome in stroke: an updated pooled analysis of ECASS, ATLANTIS, NINDS, and EPITHET trials. Lancet. 15. Maj 2010;375(9727):1695–703.
3. Tong D, Reeves MJ, Hernandez AF, Zhao X, Olson DM, Fonarow GC, et al. Times from symptom onset to hospital arrival in the Get with the Guidelines--Stroke Program 2002 to 2009: temporal trends and impli-cations. Stroke J Cereb Circ. Juli 2012;43(7):1912–7.
4. Lin CB, Peterson ED, Smith EE, Saver JL, Liang L, Xian Y, et al. Emergency medical service hospital prenotification is associated with improved evaluation and treatment of acute ischemic stroke. Circ Cardiovasc Qual Outcomes. 1. Juli 2012;5(4):514–22.
5. Dansk Selskab for Apopleksi, Referenceprogram for behandling af pa-tienter med apopleksi, København: 2012.
6. Simonsen S, Andresen M, Michelsen L, Viereck S, Lippert F, Iversen H. Evaluation of pre-hospital transport time of stroke patients to throm-bolytic treatment. Submitted 2014.

ID: 44309408