Irrespective of dispatch activation during weeks when the stroke emergency mobile unit (STEMO) was deployed or during control weeks, 78 patients received thrombolysis with tissue plasminogen activator (tPA) within 60 minutes of symptom onset (the golden hour), whereas 451 patients received treatment more than 60 minutes after the onset of symptoms. MRI indicates magnetic resonance imaging.
aOne patient with unknown time of thrombolysis initiation was excluded.
The major difference in proportions between the group treated after deployment of the stroke emergency mobile unit (STEMO) and the group treated with conventional care can be observed within the first 60 minutes of symptom onset.
Ebinger M, Kunz A, Wendt M, Rozanski M, Winter B, Waldschmidt C, Weber J, Villringer K, Fiebach JB, Audebert HJ. Effects of Golden Hour ThrombolysisA Prehospital Acute Neurological Treatment and Optimization of Medical Care in Stroke (PHANTOM-S) Substudy. JAMA Neurol. 2015;72(1):25-30. doi:10.1001/jamaneurol.2014.3188
Copyright 2015 American Medical Association. All Rights Reserved. Applicable FARS/DFARS Restrictions Apply to Government Use.
The effectiveness of intravenous thrombolysis in acute ischemic stroke is time dependent. The effects are likely to be highest if the time from symptom onset to treatment is within 60 minutes, termed the golden hour.
To determine the achievable rate of golden hour thrombolysis in prehospital care and its effect on outcome.
Design, Setting, and Participants
The prospective controlled Prehospital Acute Neurological Treatment and Optimization of Medical Care in Stroke study was conducted in Berlin, Germany, within an established infrastructure for stroke care. Weeks were randomized according to the availability of a specialized ambulance (stroke emergency mobile unit (STEMO) from May 1, 2011, through January 31, 2013. We included 6182 consecutive adult patients for whom a stroke dispatch (44.1% male; mean [SD] age, 73.9 [15.0] years) or regular care (45.0% male; mean [SD] age, 74.2 [14.9] years) were included.
The STEMO was deployed when the dispatchers suspected an acute stroke during emergency calls. If STEMO was not available (during control weeks, when the unit was already in operation, or during maintenance), patients received conventional care. The STEMO is equipped with a computed tomographic scanner plus a point-of-care laboratory and telemedicine connection. The unit is staffed with a neurologist trained in emergency medicine, a paramedic, and a technician. Thrombolysis was started in STEMO if a stroke was confirmed and no contraindication was found.
Main Outcomes and Measures
Rates of golden hour thrombolysis, 7- and 90-day mortality, secondary intracerebral hemorrhage, and discharge home.
Thrombolysis rates in ischemic stroke were 200 of 614 patients (32.6%) when STEMO was deployed and 330 of 1497 patients (22.0%) when conventional care was administered (P < .001). Among all patients who received thrombolysis, the proportion of golden hour thrombolysis was 6-fold higher after STEMO deployment (62 of 200 patients [31.0%] vs 16 of 330 [4.9%]; P < .01). Compared with patients with a longer time from symptom onset to treatment, patients who received golden hour thrombolysis had no higher risks for 7- or 90-day mortality (adjusted odds ratios, 0.38 [95% CI, 0.09-1.70]; P = .21 and 0.69 [95% CI, 0.32-1.53]; P = .36) and were more likely to be discharged home (adjusted odds ratio, 1.93 [95% CI, 1.09-3.41]; P = .02).
Conclusions and Relevance
The use of STEMO increases the percentage of patients receiving thrombolysis within the golden hour. Golden hour thrombolysis entails no risk to the patients’ safety and is associated with better short-term outcomes.
clinicaltrials.gov Identifier: NCT01382862
Quiz Ref IDTime to treatment with tissue plasminogen activator (tPA) is crucial to outcomes among patients with acute ischemic stroke.1,2 Numerous attempts have been made to reduce the time from symptom onset to treatment (OTT).3,4 However, many centers struggle to keep the time from arrival at the hospital to initiation of tPA (door-to-needle time) shorter than 60 minutes.5,6 When prehospital times are added to in-hospital delays, an OTT within the first 60 minutes of symptom onset, termed the golden hour, seems out of reach for most patients. In fact, most patients undergoing routine care for stroke receive treatment rather late. In the Safe Implementation of Thrombolysis–Stroke Monitoring Study (SITS-MOST) registry,7 10.6% of 6483 patients were treated within 90 minutes and only 1.4% within 60 minutes. The median OTT in the SITS–International Stroke Thrombolysis Register8 was 145 (interquartile range [IQR], 115-170) minutes. Structured approaches have been successful in increasing thrombolysis rates and shortening door-to-needle times.9- 11 Centers with greater numbers of tPA treatments per year tend to have shorter door-to-needle times compared with smaller centers.5 However, some of the centers with shortened door-to-needle times still have long prehospital times.9,12
Quiz Ref IDA forceful approach to shorten the OTT is prehospital thrombolysis in emergency vehicles equipped with a computed tomographic scanner and a point-of-care laboratory.13- 15 A reduced time from the emergency call to treatment after deployment of such an ambulance compared with regular care was observed in the Prehospital Acute Neurological Treatment and Optimization of Medical Care in Stroke (PHANTOM-S) study.15 We used data from the PHANTOM-S study to evaluate the rate and effectiveness of golden hour thrombolysis.
The study was approved by the Charité Ethics Committee and conducted in accordance with published protocol. Written informed consent was obtained from patients able to communicate and waived for those unable to provide consent, as described in detail previously.15
Methodologic details of the PHANTOM-S study have been described previously.16,17 In brief, an ambulance, the stroke emergency mobile unit (STEMO), was equipped with a computed tomographic scanner and point-of-care laboratory. The unit was staffed with a neurologist trained in emergency medicine (including M.E., A.K., M.W., M.R., B.W., C.W., and J.W.), a paramedic, and a radiology technician. A neuroradiologist was on call to evaluate images acquired on board the STEMO via a teleradiology connection. We evaluated the effects of the STEMO implementation prospectively by comparing weeks with and without STEMO availability. The STEMO was deployed when the dispatchers suspected an acute stroke during emergency calls. If STEMO was not available (during control weeks, while the unit was in operation, or during maintenance), patients received conventional care.
For this post hoc analysis based on data from the PHANTOM-S study, we used the same consecutive patients and baseline variables as in the original study.15 Stroke severity at baseline was assessed according to the National Institutes of Health Stroke Scale (NIHSS).18 In addition, we calculated the proportion of patients treated in 1-hour intervals. The OTT intervals were dichotomized as 60 minutes or less (the golden hour) or longer than 60 minutes. Only 1 STEMO was available within a catchment area of more than 1 million inhabitants. Therefore, the STEMO could not be deployed for all suspected strokes during STEMO weeks when the STEMO was occupied with another emergency. In contrast to the original study,15 we did not compare STEMO weeks with control weeks in the data presented here. Instead we compared tPA treatments occurring after STEMO deployment with tPA treatments during conventional care. Conventional care included tPA treatments during control weeks and tPA treatments during STEMO weeks without STEMO deployment (Figure 1). In a second step, we compared stroke patients who received golden hour thrombolysis with stroke patients who received tPA more than 60 minutes after symptom onset, independent of STEMO deployment or conventional care. Patients with stroke mimics who received tPA were not included in this evaluation of treatment effects.
We calculated unadjusted outcomes for 7- and 90-day mortality, secondary intracerebral hemorrhage, and discharge home among patients with an OTT within 60 minutes compared with those with an OTT longer than 60 minutes. We then performed analyses to achieve the adjusted probability of each outcome.
With relatively few patients who experienced a secondary hemorrhage (n = 29) or died within 7 days (n = 24), we had to restrict the adjustment to 2 variables and adjusted for age (in decades) and stroke severity (NIHSS score per point). For the outcomes of death within 90 days (n = 75) and discharge home (n = 239), we adjusted for age (in decades), sex, atrial fibrillation, and NIHSS score categories according to the Third International Stroke Trial (NIHSS scores, 0-5, 6-10, 11-15, 16-20, and ≥21).19 We considered P < .05 to be a statistically significant difference.
All data of this post hoc analysis presented herein were recalculated on the basis of the PHANTOM-S data set. During the 21 months from May 1, 2011, through January 31, 2013, there were 3213 emergency calls for suspected stroke during STEMO weeks and 2969 during control weeks (Figure 1). Of the 1804 patients with STEMO deployment (44.1% male; mean [SD] age, 73.9 [15.0] years), 200 received tPA treatment. Of those, tPA infusion was started in 17 patients after hospital arrival.15 Of 4378 patients in conventional care (45.0% male; mean [SD] age, 74.2 [14.9] years), 330 were treated with tPA (110 of those in STEMO weeks and 220 in control weeks). Quiz Ref IDOverall, 200 of 614 patients with ischemic stroke (32.6%) received thrombolysis when the STEMO was deployed and 330 of 1497 (22.0%) received thrombolysis in conventional care (P < .001). The mean NIHSS score was higher among patients with STEMO deployment compared with patients receiving conventional care (10.5 vs 9.1 [P = .02]). Median OTT was 24.5 minutes shorter after STEMO deployment compared with conventional care (80.5 [IQR, 54-126] vs 105.0 [IQR, 82-146] minutes; P < .01). In all ischemic strokes, the rate of golden hour thrombolysis increased from 16 of 1497 patients (1.1%) during conventional care to 62 of 614 (10.1%) after STEMO deployment (Figure 2). Among all patients who received thrombolysis, the proportion of golden hour thrombolysis was 6-fold higher after STEMO deployment (62 patients [31.0%] vs 16 [4.9%]; P < .01). Of all patients who received golden hour thrombolysis, 66 (84.6%) received tPA during a STEMO week and 12 (15.4%) during a control week.
Median OTT was 50.0 (IQR, 43-55) minutes in golden hour thrombolysis vs 105.0 (IQR, 85-155) minutes in all other thrombolysis (P < .001). In patients who received golden hour thrombolysis, median NIHSS score was higher than that in patients who received tPA more than 60 minutes after symptom onset (12 [IQR, 5-18] vs 7 [IQR, 4-13]; P = .006) (additional baseline characteristics are given in Table 1). In unadjusted univariate analysis of the outcomes mortality within 7 and 90 days, discharge home, and hemorrhagic complications, we detected no significant differences between golden hour thrombolysis and later thrombolysis (Table 2). In adjusted analysis, patients with golden hour thrombolysis had no higher risks for 7- or 90-day mortality (adjusted odds ratios, 0.38 [95% CI, 0.09-1.70]; P = .21 and 0.69 [95% CI, 0.32-1.53]; P = .36) compared with patients with longer OTT and were more likely to be discharged home (adjusted odds ratio, 1.93 [95% CI, 1.09-3.41]; P = .02) (Table 3).
Quiz Ref IDPrehospital stroke treatment within the STEMO concept increased the number of tPA treatments within the golden hour almost 10-fold. Golden hour thrombolysis was associated with improved patient outcomes regarding the discharge destination from acute in-hospital care. We observed no increased risk for hemorrhagic complications or mortality in patients undergoing golden hour thrombolysis. Deployment of STEMO led to a significant proportion of patients with extremely early treatment, even in a setting with already short median door-to-needle times and OTT in the control group (36.0 [IQR, 28-51] and 105.0 [IQR, 81-145] minutes, respectively). Our results regarding safety and discharge status after golden hour thrombolysis are in line with previous analyses of the association of improved outcomes with shorter OTT.1,2,20
Thrombolytic treatment within the first 60 minutes has rarely been reported. Only 4 of 3670 patients with time to treatment within 1 hour were included in the randomized stroke thrombolysis trials according to the last pooled data set.21 Nevertheless, the term golden hour has become used frequently in acute stroke management.3,4 The rapid death of brain cells after blockage of cerebral blood supply makes immediate treatment to dissolve or remove clots indispensable.
Starting treatment at the scene has become a realistic scenario. The high percentage of early treatments also raises hopes for other potentially effective treatments, such as neuroprotective substances.
This study has limitations. The significance and generalizability of the results presented here are limited owing to the nature of a post hoc analysis. These explorative results need to be confirmed using a prospective approach. The PHANTOM-S study was originally designed to detect a reduced time from the emergency call to treatment and was not powered to detect functional outcome differences. The use of discharge destination from acute in-hospital care can only be interpreted as a surrogate variable. The higher frequency of patients discharged home was only shown for patients treated within the golden hour and not for the entire cohort of patients who received STEMO care. Randomization was not performed at the patient level, but weeks were allocated to usual stroke care or additional STEMO availability. Finally, bias of spontaneous recovery remains a possibility when comparing extremely early vs late thrombolysis.
The question of generalizability of the prehospital thrombolysis concept warrants further studies. Not all countries are accustomed to emergency physicians in the field, which may be required for adaptation of the STEMO concept. The OTT found in our study after STEMO deployment (median OTT, 80.5 minutes) and during conventional care (median OTT, 105.0 minutes) were much shorter than in published stroke thrombolysis registries (median OTTs, 144 minutes in Saver et al2 and 140 minutes in Wahlgren et al7) or even in best-practice hospital systems (median OTT, 119 minutes for Meretoja et al9). Apart from a selection bias by restriction to patients with ambulance care, this difference may be explained at least in part by a well-established metropolitan stroke care system with trained dispatchers and paramedics, short distances to stroke units, and rather optimized in-hospital procedures. During the study period, the dispatchers identified patients with typical stroke symptoms during the emergency call and notified the emergency medical services, which may have raised awareness of the entire rescue chain.
The concept of prehospital thrombolysis is still relatively new, and experience is limited to few groups. Further improvements in time reduction may be expected with growing routine.
After analyzing data from the Get With the Guidelines Stroke Program, Saver and colleagues2 concluded that every effort should be made to accelerate thrombolytic treatment in patients with stroke. Our post hoc analysis supports this time-is-brain concept. Quiz Ref IDGolden hour thrombolysis was associated with better short-term outcomes.
Accepted for Publication: September 4, 2014.
Corresponding Author: Martin Ebinger, MD, Department of Neurology, Charité–Universitätsmedizin Berlin, Charitéplatz 1, 10117 Berlin, Germany (email@example.com).
Published Online: November 17, 2014. doi:10.1001/jamaneurol.2014.3188.
Author Contributions: Drs Ebinger and Audebert had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Ebinger, Audebert.
Acquisition, analysis, or interpretation of data: All authors.
Drafting of the manuscript: Ebinger, Audebert.
Critical revision of the manuscript for important intellectual content: Kunz, Wendt, Rozanski, Winter, Waldschmidt, Weber, Villringer, Fiebach, Audebert.
Statistical analysis: Audebert.
Obtained funding: Audebert.
Administrative, technical, or material support: Ebinger, Kunz, Wendt, Waldschmidt, Weber, Villringer.
Study supervision: Ebinger, Fiebach, Audebert.
Conflict of Interest Disclosures: Dr Audebert received speaker honoraria from Boehringer Ingelheim, the manufacturer of alteplase (not involved in this trial), and speaker and consultancy honoraria from Lundbeck A/S, which has conducted a recent acute stroke thrombolysis trial. No other disclosures were reported.
Funding/Support: The PHANTOM-S study was supported by the Zukunftsfonds Berlin and the Technology Foundation Berlin with cofinancing by the European Regional Development Fund.
Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.
Additional Contributions: Ulrike Grittner, PhD, Center for Stroke Research Berlin, Charité–Universitätsmedizin Berlin, provided statistical advice. He received no compensation for this contribution.