Cutting the Prehospital On-Scene Time of Stroke Thrombolysis in Helsinki
A Prospective Interventional Study
Background and Purpose—Significant portion of the prehospital delay consists of minutes spent on the scene with the patient. We implemented a training program for the emergency medical services personnel with the aim to optimize the on-scene time (OST) and to study the impact of different elements of prehospital practice to the OST duration.
Methods—In this prospective interventional study, key operational emergency medical service performance variables were analyzed from all thrombolysis candidates transported to the Helsinki University Hospital emergency department. The catchment period was 4 months before and 4 months after the implementation.
Results—One hundred and forty-one patients were managed as thrombolysis candidates before and 148 patients after the training program implementation. The OST duration for the groups was 25 (20.5–31) and 22.5 (18–28.5) minutes, respectively (P<0.001). Physician consultations via telephone were associated with a longer (odds ratio 0.546 [0.333–0.893]) and advanced life support training with a shorter OST (odds ration 1.760 [1.070–2.895]).
Conclusions—Implementation of the emergency medical services training program successfully decreased the OST of thrombolysis candidates by 10%. Higher expertise level of the ambulance crew was associated with shorter OST, and decisions to consult a physician via telephone were reflected by longer OST.
Significant progress has been made in Helsinki stroke chain of survival in streamlining the in-hospital management of stroke thrombolysis patients, but similar expedition has proved difficult to achieve in the prehospital management steps.1 In selected comprehensive stroke centers, the door-to-treatment time for intravenous thrombolysis is already well <30 minutes, but the prehospital onset-to-door time has remained stagnant above the 60-minute mark through almost 2 decades.2–4 Although patient-dependent deliberations still dominate the prehospital bottleneck, problems do exist within the organized chain of recovery.5,6 Although early symptom recognition, ambulance dispatch using the stroke code, and transport using high priority have received considerable attention in numerous reports, similar honing has not occurred in the workflow of the ambulance crew operating on the scene.
In our first 3-year report of the sequential prehospital delays of thrombolysed patients, the on-scene time (OST) varied between 18 and 23 minutes depending on symptom severity.5 To our surprise, our recent analysis revealed that the median OST had increased to 24 minutes—almost 10 minutes longer than the current American Stroke Association guidelines recommend (≤15 minutes).6 This led us to perform a quality assurance program, including a lecture-based training package for the emergency medical system (EMS) personnel, aiming to shorten the on-scene operation and streamline the workflow. This study aimed to decrease the median OST of thrombolysis candidates to ≤20 minutes by enhancing the performance of the personnel with a focused training program.
The planning of this prospective interventional study took advantage of existing continuously gathered prehospital EMS data. The study plan was approved by the Departments of Emergency Medicine and Neurology of the Helsinki University Hospital. Because the study was register based, no separate ethical review board approval was required. The study setting, existing prehospital stroke protocol, hospital emergency department (ED) process, and data collection principles have been described elsewhere.5,6
During a 3-month training period, all ambulance crews (both advanced and basic life support certified) in the Helsinki University Hospital area attended a 45-minute training session and participated in interactive follow-up group sessions (online-only Data Supplement).
The key practice points targeted by the training program were to (1) increase general time awareness of the EMS staff (ie, adding an OST timer function to the paramedics’ laptop computers), (2) limit the overall number of emergency on-scene procedures (ie, ≤3 allowed attempts in intravenous cannulation), (3) transport the patient to the ambulance using a carrying chair instead of an ill-maneuverable stretcher, and (4) transform the general management workflow into a typical load and go setup to automate fast assessment and immediate transport.
All patients transported by ambulances to Helsinki University Hospital ED using the stroke code 4 months before and after the training period were registered, and their prehospital electronic patient reports were retrieved from the Merlot Medi (CGI Inc) database. Transportations from other healthcare institutions were not included. Key operational variables describing EMS performance were compared before and after implementing the training program. A dichotomized analysis was made to short and long OST duration groups using the sample median. Variables with >90% complete data sets and P<0.2 were selected for regression analysis to determine associations with a short OST. The statistical analysis was conducted using the SPSS 21 statistical package. Significance was considered at P<0.05.
During the 8-month study period, the EMS transported a total of 879 suspected stroke patients to Helsinki University Hospital, of which 20 were regarded as interfacility transports. Of the remaining 859 patients, 289 were identified as potential thrombolysis candidates by the emergency medical dispatchers and managed accordingly by the EMS using high priority ambulance transport, the stroke code, and a prenotification to the ED. One hundred and forty-one patients were registered before and 148 after the implementation of the training package. The groups did not differ in terms of patient age, sex, frequency of physician’s telephone consultations, or number of advanced life support–trained crews (P>0.1), but the training program was found to be followed by a decrease in the OST by 10% (Figure). However, this did not translate into net savings in the overall dispatch-to-hospital time, which remained at 45 minutes (P>0.1).
Consultation with a prehospital emergency physician or a neurologist via telephone was associated with longer on-scene durations in the univariate comparisons and multivariable regression model, whereas higher training level of the EMS personnel promoted a shorter OST. Participation in the training program showed a strong trend toward shorter OST durations (Tables 1–2).
This study describes novel steps in the continuous optimizing and honing of the stroke chain of recovery in the comprehensive stroke center in Helsinki.2,3,5,6 Although several components of the prehospital chain of thrombolysis candidates are dictated by the surroundings (ie, rush hour traffic), the workflow of the on-scene management is modifiable and includes adjustable performance task variables. This report highlights the possibility to decrease the OST significantly with a dedicated training program targeting the existing, experienced EMS personnel. This study also demonstrates that the training of ambulance crews and their decisions about physician consultations influence the composition of the OST and the prehospital phase. Fewer consultations and less time on-scene did not lead to premature rushing of patients into ED, because 10% more patients reached the hospital diagnosis of cerebral ischemia (P<0.1) to receive recanalization therapy with at least the same frequency (intravenous thrombolysis alone 14.3% and thrombolysis+thrombectomy 5.4%; online-only Data Supplement).
Earlier reports from the United States and Denmark have described shorter OSTs of between 15 and 18 minutes.7,8 However, these reports lacked valuable information about the convergence of prehospital and in-hospital phases to expedite in-hospital management. Because different EMS systems can have fundamental differences, direct comparisons are difficult, and the ideal OST must be determined locally. The key is to do as little as possible on the scene but not to omit tasks that delay the ED process (ie, intravenous cannulation). Because emergency procedures in a moving vehicle can jeopardize both the patient and the EMS personnel, they should be avoided. However, postponing the hospital prenotification and some written reporting to the transport phase can sometimes be beneficial.
The study raises novel modifiable elements in the prehospital chain. A low-cost education program had a measurable influence on OST reduction, which was associated with the education level of the crew (Tables 1–2). Therefore, educating the personnel of imperatives stemming from the time-dependent ischemic pathophysiology may create a time-saving opportunity. Regardless of the geographical position, rural or metropolitan, the OST contains a considerable modifiable time-saving potential.
The study was based on a single-center sample, which limited its size. This might also explain why the obvious decrease in the OST did not directly translate into reduced dispatch-to-hospital time, although the latter did associate with reduced OST (Table 1). Repetitive training is challenging to include in the training scheme of EMS personnel who every day manage a capricious palette of injured and critically ill. In Helsinki, suspected acute stroke comprise only 3% of the task load of the EMS. Therefore, the overall results are reassuring, even though decreasing system-based delays must be implemented as a continuous dynamic process.
In conclusion, the EMS training package successfully decreased the OST of thrombolysis candidates by 10%, from 25 to 22.5 minutes. Higher advanced life support training of the ambulance crew was associated with a shorter OST and the decision to consult a physician via telephone with a longer OST. Despite the fact that the prehospital chain of care is largely dependent on the incidental physical setting, several minutes can still be spared by a focused training program targeting on how to operate on the scene of an acute stroke.
Sources of Funding
This study was supported by Helsinki University Hospital research funds (EVO), Finska Läkaresällskapet, Finnish Medical Foundation, Laerdal Foundation for Acute Medicine, and the Paavo Nurmi Foundation.
Presented in part at the Emergency Medical Services 2016 Congress in Copenhagen, Denmark, June 2016.
The online-only Data Supplement is available with this article at http://stroke.ahajournals.org/lookup/suppl/doi:10.1161/STROKEAHA.116.014531/-/DC1.
- Received July 6, 2016.
- Revision received September 13, 2016.
- Accepted September 27, 2016.
- © 2016 American Heart Association, Inc.
- Lindsberg PJ,
- Soinne L,
- Roine RO,
- Salonen O,
- Tatlisumak T,
- Kallela M,
- et al
- Fonarow GC,
- Zhao X,
- Smith EE,
- Saver JL,
- Reeves MJ,
- Bhatt DL,
- et al
- Puolakka T,
- Strbian D,
- Harve H,
- Kuisma M,
- Lindsberg PJ
- Patel MD,
- Brice JH,
- Moss C,
- Suchindran CM,
- Evenson KR,
- Rose KM,
- et al