Original Paper Received: July 26, 2013 Accepted: September 5, 2013 Published online: November 15, 2013
Cerebrovasc Dis 2013;36:383–387 DOI: 10.1159/000355500
Stroke Prenotification Is Associated with Shorter Treatment Times for Warfarin-Associated Intracerebral Hemorrhage Dar Dowlatshahi a Jason K. Wasserman b Ken S. Butcher f Manya L. Bernbaum g A. Adam Cwinn c Antonio Giulivi d Eddy Lang h Man-Chiu Poon i Jessica Tomchishen-Pope e Mukul Sharma j Shelagh B. Coutts g
a Department of Medicine (Neurology), University of Ottawa and Ottawa Hospital Research Institute, b Department of Pathology and Laboratory Medicine, University of Ottawa, c Department of Emergency Medicine, University of Ottawa, d Department of Pathology and Laboratory Medicine, University of Ottawa, and e Faculty of Medicine, University of Ottawa, Ottawa, Ont., f Department of Medicine (Neurology), University of Alberta, Edmonton, Alta., g Departments of Clinical Neurosciences and Radiology, Hotchkiss Brain Institute, University of Calgary, h Department of Emergency Medicine, University of Calgary, and i Departments of Medicine, Pediatrics and Oncology, University of Calgary and Southern Alberta Rare Blood and Bleeding Disorders Comprehensive Care Program, Foothills Hospital, Calgary, Alta., and j Department of Medicine (Neurology) McMaster University, Population Health Research Institute, Hamilton, Ont., Canada
Key Words Stroke prenotification · Warfarin-associated intracerebral hemorrhage · Prothrombin complex concentrates
Abstract Background: Warfarin-associated intracerebral hemorrhage (WAICH) is a devastating disease with increasing incidence. In this setting, treatment with prothrombin complex concentrates (PCC) is essential to correct coagulopathy. Yet despite the availability of coagulopathy correction strategies, significant treatment delays can occur in emergency departments (EDs), which may be overcome using stroke prenotification strategies. To explore this, we compared arrival-totreatment times with PCC for WAICH between two different stroke response systems that used the same international normalized ratio (INR) correction protocol. Methods: We established a registry of consecutive patients presenting with WAICH and treated with PCC presenting to two Canadian
© 2013 S. Karger AG, Basel 1015–9770/13/0366–0383$38.00/0 E-Mail [email protected] www.karger.com/ced
tertiary-care academic stroke centers: one with a stroke prenotification system, and one with a traditional ED assessment, treatment and referral system. In this comparative cohort design, we defined the WAICH diagnosis time as the earliest time point where both INR and CT were available. We compared median times from arrival to treatment, as well as arrival to diagnosis, and diagnosis to treatment. Results: Between 2008 and 2010, we collected data from 123 consecutive patients with intracranial hemorrhage who received PCC for INR correction (79 from ED referral, and 44 prenotification). Onset-to-arrival times, demographics, Glasgow Coma Scale scores, and baseline INR were similar between the two systems. Arrival-to-treatment times were significantly shorter in the prenotification system as compared to the traditional ED referral system (135 vs. 267 min; p = 0.001), which was driven by both decreased arrival-to-diagnosis time (49 vs. 117 min; p = 0.006), as well as decreased diagnosis-to-treatment time (56 vs. 112 min; p < 0.001). Arrival-toscan times and arrival-to-INR times were similarly shorter in
Dar Dowlatshahi C2182 Ottawa Hospital Civic Campus 1053 Carling ave Ottawa, ON, K1Y 4E3 (Canada) E-Mail ddowlat @ toh.on.ca
the prenotification system (68 vs. 118 min and 20.5 vs. 47 min, respectively). Conclusion: Stroke prenotification was associated with shorter arrival-to-treatment times for emergent INR correction in patients with WAICH, which was driven by both faster diagnosis and treatment. Our results are consistent with those seen in ischemic stroke, suggesting that prenotification systems present an opportunity to optimize acute intracerebral hemorrhage therapy. © 2013 S. Karger AG, Basel
Introduction
Warfarin-associated intracerebral hemorrhage (WAICH) is a devastating disease with mortality rates exceeding 60% [1]. Its incidence is projected to increase along with the population age and a 3-fold increase in the prevalence of atrial fibrillation [2–4]. The recent American Heart Association/American Stroke Association guidelines emphasize the need for rapid international normalized ratio (INR) correction for WAICH [5]. This can be achieved within minutes with prothrombin complex concentrates (PCC) [6–8]. Yet despite available INR correction strategies, significant treatment delays can occur in the emergency department (ED) [9, 10]. Prehospital strategies can reduce treatment delays and improve outcomes for time-sensitive emergencies. As there is an inverse correlation between the probability of disability-free survival and onset-to-treatment time in ischemic stroke [11], many stroke centers have reduced door-to-needle times using stroke team prenotification protocols, where ambulance crews alert receiving hospitals that a patient with an acute neurological deficit is en route [12–14]. We hypothesized that prenotification will also result in shorter door-to-needle times and faster INR correction in patients with WAICH. To test this hypothesis, we compared the WAICH treatment times from two tertiary-care stroke centers with identical PCC INR correction protocols: one with a standard ED referral system, and one with stroke team prenotification.
Methods The Foothills Hospital (Calgary, Alta., Canada) and the Ottawa Hospital (Ottawa, Ont., Canada) are both high-volume tertiary-care stroke centers with identical PCC protocols for INR correction (protocols based on the recommendations from the National Advisory Committee on Blood and Blood Products) [15]. The hospitals differ in their respective ‘code stroke’ protocols: the Ottawa Hospital employs a traditional ED referral system beginning with an emergency
384
Cerebrovasc Dis 2013;36:383–387 DOI: 10.1159/000355500
physician assessment, transportation to a CT scanner by a registered nurse, confirmation of the diagnosis by the emergency physician and initiation of therapy, and subsequent referral to the appropriate neurovascular service. The Foothills Hospital employs a stroke prenotification system where the neurovascular team is alerted prior to the arrival of the ambulance (based on a paramedic prehospital checklist), and assesses the patient immediately upon arrival along with the emergency physician, transports the patient to the CT scanner, and assumes the remainder of care if a stroke is confirmed. In November 2008, both institutions launched a prospective inpatient registry of consecutive patients treated with a PCC (Octaplex®), a plasma-derived concentrate containing all 4 vitamin K-dependent factors (II, VII, IX, X). Upon diagnosis of WAICH, the PCC was prepared and released by the hospital blood bank, and patients were treated and included into the registry. We subsequently collected clinical, imaging and laboratory data from the medical records of patients within the registry. We also collected the times of symptom onset, arrival to the ED, imaging, blood cell count availability, and time of PCC infusion. We defined ‘diagnosis time’ as the earliest time point where both imaging and INR results were available to make a diagnosis of WAICH. In situations where a scan confirming intracranial hemorrhage was done by an outside referring institution, time of diagnosis was considered the earliest point at which INR results were available (where both results were available prior to arrival to a registry hospital, diagnosis time was considered to be the same as arrival time). We compared arrival-to-treatment times between the two institutions using Mann-Whitney U tests. We similarly compared the arrival-to-diagnosis times and diagnosis-to-treatment times between the two institutions. For all statistical analyses, SPSS v.18.0 (Somers, N.Y., USA) was used. The Ottawa Hospital Research Ethics Board (University of Ottawa) and the Conjoint Health Research Ethics Board (University of Calgary) approved the study, and procedures followed institutional guidelines. We obtained waiver of consent from both review boards, and all data were analyzed anonymously.
Results
Between 2008 and 2010, 123 patients with intracranial hemorrhage received PCC for INR correction at the two registry hospitals (table 1). Patients presenting in Ottawa were more likely to have a history of atrial fibrillation, but there were no other statistically significant differences in demographics between the two hospitals. The Glasgow Coma Scale score, baseline INR, and postinfusion INR were not different either (table 2). Clinical outcomes are published elsewhere [6]. It was not possible to establish symptom onset times for 38/79 patients in the ED referral group and for 12/44 patients in the prenotification group. In the ED referral group, arrival time was missing in 10/79 patients, whereas in the prenotification cohort, treatment time was missing in 3/44 patients. Symptom onset-to-arrival times were comparable between the two hospital systems (table 3). However, arrivDowlatshahi et al.
Table 1. Baseline characteristics of the ED referral and prenotification cohorts
Age, mean ± SD, years Male, n (%) Prior medical history, n (%) Stroke MI/CAD DVT/PE ICH SAH SDH Hypertension Diabetes Atrial fibrillation CHF Smoker Hemorrhage type, n (%) ICH SDH EDH SAH IVH Baseline cerebral hemorrhage volume, mean ± SD
ED referral (n = 79)
Prenotification (n = 44)
p value
76.6±10.3 47 (59.5)
74.4±12.3 28 (63.6)
0.29 0.70
25 (31.6) 26 (32.9) 11 (13.9) 0 (0) 3 (3.8) 1 (1.3) 39 (49.3) 23 (29.1) 65 (82.3) 18 (22.8) 1 (1.3)
12 (27.3) 12 (27.3) 9 (20.5) 1 (2.3) 1 (2.3) 1 (2.3) 30 (68.2) 11 (25.0) 12 (27.3) 6 (13.6) 3 (6.8)
0.68 0.55 0.45 0.36 1.00 1.00 0.06 0.68
Cerebrovasc Dis 2013;36:383–387 DOI: 10.1159/000355500
Stroke Prenotification Is Associated with Shorter Treatment Times for Warfarin-Associated Intracerebral Hemorrhage Dar Dowlatshahi a Jason K. Wasserman b Ken S. Butcher f Manya L. Bernbaum g A. Adam Cwinn c Antonio Giulivi d Eddy Lang h Man-Chiu Poon i Jessica Tomchishen-Pope e Mukul Sharma j Shelagh B. Coutts g
a Department of Medicine (Neurology), University of Ottawa and Ottawa Hospital Research Institute, b Department of Pathology and Laboratory Medicine, University of Ottawa, c Department of Emergency Medicine, University of Ottawa, d Department of Pathology and Laboratory Medicine, University of Ottawa, and e Faculty of Medicine, University of Ottawa, Ottawa, Ont., f Department of Medicine (Neurology), University of Alberta, Edmonton, Alta., g Departments of Clinical Neurosciences and Radiology, Hotchkiss Brain Institute, University of Calgary, h Department of Emergency Medicine, University of Calgary, and i Departments of Medicine, Pediatrics and Oncology, University of Calgary and Southern Alberta Rare Blood and Bleeding Disorders Comprehensive Care Program, Foothills Hospital, Calgary, Alta., and j Department of Medicine (Neurology) McMaster University, Population Health Research Institute, Hamilton, Ont., Canada
Key Words Stroke prenotification · Warfarin-associated intracerebral hemorrhage · Prothrombin complex concentrates
Abstract Background: Warfarin-associated intracerebral hemorrhage (WAICH) is a devastating disease with increasing incidence. In this setting, treatment with prothrombin complex concentrates (PCC) is essential to correct coagulopathy. Yet despite the availability of coagulopathy correction strategies, significant treatment delays can occur in emergency departments (EDs), which may be overcome using stroke prenotification strategies. To explore this, we compared arrival-totreatment times with PCC for WAICH between two different stroke response systems that used the same international normalized ratio (INR) correction protocol. Methods: We established a registry of consecutive patients presenting with WAICH and treated with PCC presenting to two Canadian
© 2013 S. Karger AG, Basel 1015–9770/13/0366–0383$38.00/0 E-Mail [email protected] www.karger.com/ced
tertiary-care academic stroke centers: one with a stroke prenotification system, and one with a traditional ED assessment, treatment and referral system. In this comparative cohort design, we defined the WAICH diagnosis time as the earliest time point where both INR and CT were available. We compared median times from arrival to treatment, as well as arrival to diagnosis, and diagnosis to treatment. Results: Between 2008 and 2010, we collected data from 123 consecutive patients with intracranial hemorrhage who received PCC for INR correction (79 from ED referral, and 44 prenotification). Onset-to-arrival times, demographics, Glasgow Coma Scale scores, and baseline INR were similar between the two systems. Arrival-to-treatment times were significantly shorter in the prenotification system as compared to the traditional ED referral system (135 vs. 267 min; p = 0.001), which was driven by both decreased arrival-to-diagnosis time (49 vs. 117 min; p = 0.006), as well as decreased diagnosis-to-treatment time (56 vs. 112 min; p < 0.001). Arrival-toscan times and arrival-to-INR times were similarly shorter in
Dar Dowlatshahi C2182 Ottawa Hospital Civic Campus 1053 Carling ave Ottawa, ON, K1Y 4E3 (Canada) E-Mail ddowlat @ toh.on.ca
the prenotification system (68 vs. 118 min and 20.5 vs. 47 min, respectively). Conclusion: Stroke prenotification was associated with shorter arrival-to-treatment times for emergent INR correction in patients with WAICH, which was driven by both faster diagnosis and treatment. Our results are consistent with those seen in ischemic stroke, suggesting that prenotification systems present an opportunity to optimize acute intracerebral hemorrhage therapy. © 2013 S. Karger AG, Basel
Introduction
Warfarin-associated intracerebral hemorrhage (WAICH) is a devastating disease with mortality rates exceeding 60% [1]. Its incidence is projected to increase along with the population age and a 3-fold increase in the prevalence of atrial fibrillation [2–4]. The recent American Heart Association/American Stroke Association guidelines emphasize the need for rapid international normalized ratio (INR) correction for WAICH [5]. This can be achieved within minutes with prothrombin complex concentrates (PCC) [6–8]. Yet despite available INR correction strategies, significant treatment delays can occur in the emergency department (ED) [9, 10]. Prehospital strategies can reduce treatment delays and improve outcomes for time-sensitive emergencies. As there is an inverse correlation between the probability of disability-free survival and onset-to-treatment time in ischemic stroke [11], many stroke centers have reduced door-to-needle times using stroke team prenotification protocols, where ambulance crews alert receiving hospitals that a patient with an acute neurological deficit is en route [12–14]. We hypothesized that prenotification will also result in shorter door-to-needle times and faster INR correction in patients with WAICH. To test this hypothesis, we compared the WAICH treatment times from two tertiary-care stroke centers with identical PCC INR correction protocols: one with a standard ED referral system, and one with stroke team prenotification.
Methods The Foothills Hospital (Calgary, Alta., Canada) and the Ottawa Hospital (Ottawa, Ont., Canada) are both high-volume tertiary-care stroke centers with identical PCC protocols for INR correction (protocols based on the recommendations from the National Advisory Committee on Blood and Blood Products) [15]. The hospitals differ in their respective ‘code stroke’ protocols: the Ottawa Hospital employs a traditional ED referral system beginning with an emergency
384
Cerebrovasc Dis 2013;36:383–387 DOI: 10.1159/000355500
physician assessment, transportation to a CT scanner by a registered nurse, confirmation of the diagnosis by the emergency physician and initiation of therapy, and subsequent referral to the appropriate neurovascular service. The Foothills Hospital employs a stroke prenotification system where the neurovascular team is alerted prior to the arrival of the ambulance (based on a paramedic prehospital checklist), and assesses the patient immediately upon arrival along with the emergency physician, transports the patient to the CT scanner, and assumes the remainder of care if a stroke is confirmed. In November 2008, both institutions launched a prospective inpatient registry of consecutive patients treated with a PCC (Octaplex®), a plasma-derived concentrate containing all 4 vitamin K-dependent factors (II, VII, IX, X). Upon diagnosis of WAICH, the PCC was prepared and released by the hospital blood bank, and patients were treated and included into the registry. We subsequently collected clinical, imaging and laboratory data from the medical records of patients within the registry. We also collected the times of symptom onset, arrival to the ED, imaging, blood cell count availability, and time of PCC infusion. We defined ‘diagnosis time’ as the earliest time point where both imaging and INR results were available to make a diagnosis of WAICH. In situations where a scan confirming intracranial hemorrhage was done by an outside referring institution, time of diagnosis was considered the earliest point at which INR results were available (where both results were available prior to arrival to a registry hospital, diagnosis time was considered to be the same as arrival time). We compared arrival-to-treatment times between the two institutions using Mann-Whitney U tests. We similarly compared the arrival-to-diagnosis times and diagnosis-to-treatment times between the two institutions. For all statistical analyses, SPSS v.18.0 (Somers, N.Y., USA) was used. The Ottawa Hospital Research Ethics Board (University of Ottawa) and the Conjoint Health Research Ethics Board (University of Calgary) approved the study, and procedures followed institutional guidelines. We obtained waiver of consent from both review boards, and all data were analyzed anonymously.
Results
Between 2008 and 2010, 123 patients with intracranial hemorrhage received PCC for INR correction at the two registry hospitals (table 1). Patients presenting in Ottawa were more likely to have a history of atrial fibrillation, but there were no other statistically significant differences in demographics between the two hospitals. The Glasgow Coma Scale score, baseline INR, and postinfusion INR were not different either (table 2). Clinical outcomes are published elsewhere [6]. It was not possible to establish symptom onset times for 38/79 patients in the ED referral group and for 12/44 patients in the prenotification group. In the ED referral group, arrival time was missing in 10/79 patients, whereas in the prenotification cohort, treatment time was missing in 3/44 patients. Symptom onset-to-arrival times were comparable between the two hospital systems (table 3). However, arrivDowlatshahi et al.
Table 1. Baseline characteristics of the ED referral and prenotification cohorts
Age, mean ± SD, years Male, n (%) Prior medical history, n (%) Stroke MI/CAD DVT/PE ICH SAH SDH Hypertension Diabetes Atrial fibrillation CHF Smoker Hemorrhage type, n (%) ICH SDH EDH SAH IVH Baseline cerebral hemorrhage volume, mean ± SD
ED referral (n = 79)
Prenotification (n = 44)
p value
76.6±10.3 47 (59.5)
74.4±12.3 28 (63.6)
0.29 0.70
25 (31.6) 26 (32.9) 11 (13.9) 0 (0) 3 (3.8) 1 (1.3) 39 (49.3) 23 (29.1) 65 (82.3) 18 (22.8) 1 (1.3)
12 (27.3) 12 (27.3) 9 (20.5) 1 (2.3) 1 (2.3) 1 (2.3) 30 (68.2) 11 (25.0) 12 (27.3) 6 (13.6) 3 (6.8)
0.68 0.55 0.45 0.36 1.00 1.00 0.06 0.68
