- Email: [email protected]
327: Retrospective Validation of a Clinical Decision Rule to Safely Rule Out Subarachnoid Hemorrhage in Emergency Department Headache Patients
Research Forum Abstracts intervention was also associated with improvements in compliance with the AAP forward-facing car seat and booster seat recommendations. This suggests that parents can change their behavior as a result of a brief educational session, and that increasing awareness among parents about the AAP recommendations may lead to safer child car safety restraint practices.
324
Risk Factors for Dive Injury
326
Effect of Blood Pressure Hemodynamics on Outcome After Intracerebral Hemorrhage
Enduri S, Bellolio MF, Decker WW, Stead LG/Mayo Clinic College of Medicine, Division of Emergency Medicine Research, Rochester, MN
Beckett A, Kordick MF, Chan SB/Resurrection Medical
Center, Chicago, IL
Introduction: SCUBA diving is an ever-increasing world sport with over 9 million divers and estimated 30 million dives a year performed in the US. SCUBA diving is not a medically regulated sport, thus medical data on diving injuries are limited. Study Objective: Identify risk behaviors, medical conditions, compliance to dive guidelines, and injury patterns of recreational scuba divers. Methods: An Internet-based survey sent to national scuba diving organization websites pertained to risk behaviors and diver safety practices. Questions focused on injuries sustained and risk-related behaviors. Results: Responses were received from 682 of 770 (88.6%) divers logging onto the survey site. 80.6% were certified by a national diving organization. 314/ 550 (51.7%) certified divers reported diving injuries versus 99/132 (75.0%) for non-certified divers (RR⫽1.31; 95%CI: 1.16-1.48; p⬍0.001). Suspected decompression symptoms (DCS) were witnessed by 52.6% of divers but the most self reported injuries were sinus and ENT related (52.4%). 32.7% of certified divers reported medical problems including hypertension, asthma, diabetes, and epilepsy but were 1.31 times less likely to have a diving injury than non-certified divers. No significant differences observed in injuries among the certified divers based on dive frequency (P⫽1.000), medical conditions (P⫽0.750), smoking (P⫽0.545), alcohol (P⫽0.649), or illicit substances use (P⫽0.230). Conclusions: Non-certified divers reported more injuries than certified divers. Although divers witnessed DCS often, sinus and ENT problems were the most common symptoms reported. Among the certified divers, there was a positive association with fewer diving injuries but not with diving frequency, medical condition, smoking, alcohol, or illicit substance use.
325
design for further evaluation studies to assess the safety and efficiency of a structured ED-based TIA clinical pathway.
Implementation of a Transient Ischemic Attack Clinical Pathway: A Pilot Study in Knowledge Translation
Study Objectives: To determine the relationship between blood pressure hemodynamics, heart rate and mortality after emergency department (ED) presentation for hemorrhagic stroke. Methods: This prospective observational study consisted of consecutive patients who presented to our ED with intracerebral hemorrhage within 24 hours of symptoms onset, and had hemodynamic parameters recorded and available for review. Blood pressure (BP) was non-invasively measured at 5 minute intervals for the length of the patient’s ED stay with the Philips M3046A Patient Monitoring System (Philips Medical Systems, Andover, MA). For each of the measures HR, systolic BP (sBP) and diastolic BP (dBP) at baseline and differential (maximum - minimum) were calculated. Death was prospectively ascertained by telephone follow-up, state death certificates and the electronic medical record. Statistical analyses were performed in JMP software, SAS institute, Version 6.0; analysis of variance, t-test and non-parametric tests were used according to the distribution of the data. Results: The overall cohort consisted in 46 patients. The mean age ⫹SD was 69.4 ⫹15.5. There were 26 males 56.5%. The mean heart rate ⫹SD was 75.8 ⫹16.0 bpm; mean sBP ⫹SD of 159.4 ⫹32.6 mm Hg, and a mean dBP ⫹SD of 81.5⫹21.0 mm Hg. The median change is sBP was 35.5 mmHg, the median change in dBP was 23 mmHg, and the median change in HR was 15 bpm. There were a total of 21 deaths (45.7%) in the year after the hemorrhage. There is a statistical difference in the survival between those with a change in the systolic BP ⱖ 36 mmHg versus ⬍ 36 mmHg. (p⫽0.004). There is no statistical difference in the survival between those with a change in the diastolic BP ⬍ 23mm Hg versus ⬍ 23mm Hg. (p⫽0.259). Further, there was no difference in survival by gender (p⫽0.142). The median change in HR between those that died was 20 bpm compared to those alive at one year who had a median change in the HR of 11 bpm. (p⫽0.018).
Brown MD, Reeves MJ, Glynn T, Majid A, Kothari RU/Grand Rapids MERC/Michigan State University, Grand Rapids, MI; Department of Epidemiology, Michigan State University, East Lansing, MI; Michigan State University Emergency Medicine Residency, Lansing, MI; Department of Neurology and Ophthalmology, Michigan State University, East Lansing, MI; Borgess Research Institute, MSU/Kalamazoo Center for Medical Studies, Kalamazoo, MI
Study Objectives: To assess the feasibility of implementing an emergency department (ED) based transient ischemic attack (TIA) clinical pathway using computer-based clinical support and to evaluate measures of quality, safety and efficiency. Methods: A prospective cohort study of adult patients presenting to a community ED with symptoms consistent with acute TIA. Adherence to the clinical pathway served as a test of feasibility. Compliance with guideline recommendations for anti-thrombotic therapy and vascular imaging were used as process measures of quality. The 90-day risk of recurrent TIA, stroke, or death provided estimates of safety. Efficiency was assessed by measuring the rate of uneventful hospitalization, defined as a hospital admission which did not result in any major medical event or vascular intervention such as endarterectomy or stent. Results: Of the 75 subjects enrolled, physician adherence to the clinical pathway was 85.3% and 35 patients (46.7%) were discharged home from the ED. Antithrombotic agents were prescribed to 68 (90.7%) and vascular imaging was performed in 70 (93.3%). The 90-day risk of recurrent TIA was 7/75 (9.3%; 95% CI: 4.6 to 18.0%), 1 patient experienced stroke (1.3%; 95% CI: 0.2 to 7.2%), and 3 died (4.0%; 95% CI: 1.4 to 11.1%). Uneventful hospitalization occurred in 38/40 (95.0%). Conclusion: Implementation of a clinical pathway for the evaluation and management of TIA using computer-based clinical support is feasible in a community ED setting. This pilot study in knowledge translation provides a framework for the
S102 Annals of Emergency Medicine
Conclusion: A large differential in both systolic blood pressure and heart rate within 24 hours of symptom onset in intracerebral hemorrhage appears to be associated with a higher risk of death at one year.
327
Retrospective Validation of a Clinical Decision Rule to Safely Rule Out Subarachnoid Hemorrhage in Emergency Department Headache Patients
Hart D, Smith SW/Hennepin County Medical Center, Minneapolis, MN
Background: Perry et al. derived a clinical decision rule to identify headache (HA) patients who require computed tomography (CT) and lumbar puncture (LP) to rule out aneurysmal subarachnoid hemorrhage (AnSAH). Among nontransferred, neurologically intact adults (ⱖ 15 years of age) with atraumatic acute HA of ⬍ 1 hour to peak intensity, and no history of recurrent headaches, they found no AnSAH among those who had all of the following: no vomiting (V), diastolic blood pressure (DBP) ⬍ 100, arrival by private transportation, and age ⬍ 45 years. Study Objectives: To retrospectively validate this decision rule and explore the utility of other variables. Specifically, we sought to also investigate those who had a history of recurrent HA but whose presenting HA was different than prior HA, seizure or syncope, and HA onset ⬎ 1 hour.
Volume , . : September
Research Forum Abstracts Methods: We searched the ED database from January 1, 1996 through June 30, 2006 for a primary or secondary diagnosis (Dx) of “subarachnoid hemorrhage” and of “headache” and for a chief complaint (CC) of “headache.” We also searched the hospital ICD-9 code for patients with AnSAH. We then reviewed all identified AnSAH charts for: age, sex, presenting complaint, transfer-in status, neurological exam, vomiting, DBP, ambulance transport, and rapidity of HA onset. We also searched for history of previous similar HA, Dx by CT vs. LP, and history of syncope or seizure prior to arrival. Results: There were 12,472 ED patients with a CC of HA and 16,471 with final ED Dx of HA. 169 patients with new non-traumatic AnSAH were identified: 118 presented with altered mental status, 6 had a focal neurologic deficit. Of 7 neurologically intact patients, 3 had seizure activity (none would be missed by the Perry rule), and 4 had syncope prior to arrival (2 would have been missed by the rule, one due to HA onset ⬎ 1 hour). 1 patient had a history of similar recurrent headaches but would have been identified by the Perry rule, and 6 other patients had onset to peak HA intensity ⬎ 1 hour, 1 of whom would have been missed by the Perry rule. Thus, 37 (0.30% of all HA CC) patients had non-obvious SAH, 30 (81%) of whom would have been identified by Perry’s decision rule, only if they excluded those with seizure or syncope. The rule missed 6 patients by requiring onset ⬍ 1 hour; if the rule were also applied to these 6, 35 of 37 would be detected. Three with final Dx of SAH had a CC other than HA. 4 of 31 patients with onset ⬍ 1 hour, and 3 of 6 with onset 1 hour, required LP for Dx (p⫽.07, Fisher exact test). Conclusions: Perry’s rule missed 8 of 169 AnSAH; excluding those with seizure or syncope, it missed 6. The rule was 81% sensitive for non-obvious AnSAH in our cohort. However, if patients with HA onset ⬎ 1 hour are included in the rule, it identified 35 of 37 non-obvious AnSAH (sensitivity, 95%). The specificity could not be determined. LP may be more important for patients with onset ⬎ 1 hour than those with onset ⬍ 1 hour.
Table 1: Parameter Estimates for the Change in the Edema/Hematoma Volume Ratio Over 1 Hour or 20 Hours as a Function of Hemodynamic Parameters
329 328
Perihematomal Edema Growth is not Associated With Hemodynamic Variables
Jauch EC, Lindsell CJ, Adeoye O/University of Cincinnati, Cincinnati, OH
Study Objectives: Perihematomal edema formation in spontaneous intracerebral hemorrhage (ICH) is associated with worse clinical outcome. Identifying hemodynamic variables associated with edema growth could provide therapeutic targets. This study investigated the potential role of hemodynamic factors in perihematomal edema development. Methods: This was a post hoc analysis of clinical and computed tomography data from a prospective observational study of ICH patients presenting within 3 hours from symptom onset. Edema volumes were measured at hospital arrival, 1 hour and 20 hours from presentation. Blood pressure and heart rate, recorded at 19 time points between presentation and 20 hours, were used to derive hemodynamic variables (peak and mean systolic and diastolic blood pressures, heart rate, mean arterial pressure, pulse pressure). For assessing the relationship between hemodynamic parameters and edema growth, generalized linear modeling was used with edema/hematoma volume ratio as the primary dependent variable, adjusting for history of hypertension, aspirin or nonsteroidal antiinflammatory use, and, at 20 hours only, antihypertensive medication use. Separate models were constructed for each hemodynamic parameter. Results: Of 103 patients enrolled, 76 were eligible for analysis of edema growth at 1 hour after removing those with incomplete data and early surgery, 56 were included for analysis of 20 hour edema growth. No single hemodynamic variable was found to be associated with 1 or 20 hour perihematomal growth (table 1). Conclusion: Our data suggest that hemodynamic parameters do not influence the growth of perihematomal edema.
Volume , . : September
Automated Activation of a Stroke Team and Radiology Reduces Time to Evaluation and Brain Imaging
Baer A, Singer AJ, Byers AM, Perkins C/Stony Brook University, Stony Brook, NY
Study Objectives: National guidelines recommend door-to-brain imaging times of less than 25 minutes and door-to-needle times for administering thrombolytics of less than 60 minutes. We performed a before and after trial to determine the impact of institutional implementation of a brain attack protocol (BAT) on stroke quality of care measures. Methods: Study Design: Before and after trial. Setting-Suburban academic ED with Stroke Center. Subjects: Consecutive patients presenting with acute stroke within 3 hours of symptom onset. Interventions: Automated group paging system implemented in 4/05 alerting the stroke team to report to the ED within 15 minutes, and radiology to make the CT scanner immediately available. A prospective institutional stroke registry was implemented in 7/04. Outcomes: Data were collected on times from arrival in the ED to initial CT scan, door to neurologist assessment, and door to thrombolytic administration. Analysis: Outcomes before and after implementing the BAT were compared with t-tests and chi-square tests. Results: There were 40 patients before and 146 patients after BAT implementation. Their mean age was 68.3⫹/⫺16.2 years; 42% were female. Groups were similar in age and gender. Mean NIH stroke scores were higher in the after period (8.5⫹/⫺8.2 vs. 5.0⫹/⫺5.7; P⫽0.008). The mean time to arrival of the stroke team was significantly reduced (mean difference 73.3 min; 95% CI, 54.3-92.3 min) after BAT (4.8⫹/⫺4.7 vs. 78.1⫹/⫺116.9 min). The mean time to brain imaging was significantly reduced (mean difference 50.8 minutes; 95% CI, 37.5-64.1) after BAT (29.5⫹/⫺28.1 min vs. 80.3⫹/⫺64.5 min). TPA use before and after BAT were 8/9 and 41/41 eligible patients respectively (P⫽0.40). Mean time to TPA was not reduced after BAT; mean difference 5.7 min, 95% CI, ⫺42.0-30.6 (92.3⫹/⫺49.5 min vs. before 98.0⫹/⫺25.4). Conclusion: Automated activation of a stroke team and radiology reduced the time to evaluation by a neurologist and the time to brain imaging but not time to TPA.
Annals of Emergency Medicine S103
324
Risk Factors for Dive Injury
326
Effect of Blood Pressure Hemodynamics on Outcome After Intracerebral Hemorrhage
Enduri S, Bellolio MF, Decker WW, Stead LG/Mayo Clinic College of Medicine, Division of Emergency Medicine Research, Rochester, MN
Beckett A, Kordick MF, Chan SB/Resurrection Medical
Center, Chicago, IL
Introduction: SCUBA diving is an ever-increasing world sport with over 9 million divers and estimated 30 million dives a year performed in the US. SCUBA diving is not a medically regulated sport, thus medical data on diving injuries are limited. Study Objective: Identify risk behaviors, medical conditions, compliance to dive guidelines, and injury patterns of recreational scuba divers. Methods: An Internet-based survey sent to national scuba diving organization websites pertained to risk behaviors and diver safety practices. Questions focused on injuries sustained and risk-related behaviors. Results: Responses were received from 682 of 770 (88.6%) divers logging onto the survey site. 80.6% were certified by a national diving organization. 314/ 550 (51.7%) certified divers reported diving injuries versus 99/132 (75.0%) for non-certified divers (RR⫽1.31; 95%CI: 1.16-1.48; p⬍0.001). Suspected decompression symptoms (DCS) were witnessed by 52.6% of divers but the most self reported injuries were sinus and ENT related (52.4%). 32.7% of certified divers reported medical problems including hypertension, asthma, diabetes, and epilepsy but were 1.31 times less likely to have a diving injury than non-certified divers. No significant differences observed in injuries among the certified divers based on dive frequency (P⫽1.000), medical conditions (P⫽0.750), smoking (P⫽0.545), alcohol (P⫽0.649), or illicit substances use (P⫽0.230). Conclusions: Non-certified divers reported more injuries than certified divers. Although divers witnessed DCS often, sinus and ENT problems were the most common symptoms reported. Among the certified divers, there was a positive association with fewer diving injuries but not with diving frequency, medical condition, smoking, alcohol, or illicit substance use.
325
design for further evaluation studies to assess the safety and efficiency of a structured ED-based TIA clinical pathway.
Implementation of a Transient Ischemic Attack Clinical Pathway: A Pilot Study in Knowledge Translation
Study Objectives: To determine the relationship between blood pressure hemodynamics, heart rate and mortality after emergency department (ED) presentation for hemorrhagic stroke. Methods: This prospective observational study consisted of consecutive patients who presented to our ED with intracerebral hemorrhage within 24 hours of symptoms onset, and had hemodynamic parameters recorded and available for review. Blood pressure (BP) was non-invasively measured at 5 minute intervals for the length of the patient’s ED stay with the Philips M3046A Patient Monitoring System (Philips Medical Systems, Andover, MA). For each of the measures HR, systolic BP (sBP) and diastolic BP (dBP) at baseline and differential (maximum - minimum) were calculated. Death was prospectively ascertained by telephone follow-up, state death certificates and the electronic medical record. Statistical analyses were performed in JMP software, SAS institute, Version 6.0; analysis of variance, t-test and non-parametric tests were used according to the distribution of the data. Results: The overall cohort consisted in 46 patients. The mean age ⫹SD was 69.4 ⫹15.5. There were 26 males 56.5%. The mean heart rate ⫹SD was 75.8 ⫹16.0 bpm; mean sBP ⫹SD of 159.4 ⫹32.6 mm Hg, and a mean dBP ⫹SD of 81.5⫹21.0 mm Hg. The median change is sBP was 35.5 mmHg, the median change in dBP was 23 mmHg, and the median change in HR was 15 bpm. There were a total of 21 deaths (45.7%) in the year after the hemorrhage. There is a statistical difference in the survival between those with a change in the systolic BP ⱖ 36 mmHg versus ⬍ 36 mmHg. (p⫽0.004). There is no statistical difference in the survival between those with a change in the diastolic BP ⬍ 23mm Hg versus ⬍ 23mm Hg. (p⫽0.259). Further, there was no difference in survival by gender (p⫽0.142). The median change in HR between those that died was 20 bpm compared to those alive at one year who had a median change in the HR of 11 bpm. (p⫽0.018).
Brown MD, Reeves MJ, Glynn T, Majid A, Kothari RU/Grand Rapids MERC/Michigan State University, Grand Rapids, MI; Department of Epidemiology, Michigan State University, East Lansing, MI; Michigan State University Emergency Medicine Residency, Lansing, MI; Department of Neurology and Ophthalmology, Michigan State University, East Lansing, MI; Borgess Research Institute, MSU/Kalamazoo Center for Medical Studies, Kalamazoo, MI
Study Objectives: To assess the feasibility of implementing an emergency department (ED) based transient ischemic attack (TIA) clinical pathway using computer-based clinical support and to evaluate measures of quality, safety and efficiency. Methods: A prospective cohort study of adult patients presenting to a community ED with symptoms consistent with acute TIA. Adherence to the clinical pathway served as a test of feasibility. Compliance with guideline recommendations for anti-thrombotic therapy and vascular imaging were used as process measures of quality. The 90-day risk of recurrent TIA, stroke, or death provided estimates of safety. Efficiency was assessed by measuring the rate of uneventful hospitalization, defined as a hospital admission which did not result in any major medical event or vascular intervention such as endarterectomy or stent. Results: Of the 75 subjects enrolled, physician adherence to the clinical pathway was 85.3% and 35 patients (46.7%) were discharged home from the ED. Antithrombotic agents were prescribed to 68 (90.7%) and vascular imaging was performed in 70 (93.3%). The 90-day risk of recurrent TIA was 7/75 (9.3%; 95% CI: 4.6 to 18.0%), 1 patient experienced stroke (1.3%; 95% CI: 0.2 to 7.2%), and 3 died (4.0%; 95% CI: 1.4 to 11.1%). Uneventful hospitalization occurred in 38/40 (95.0%). Conclusion: Implementation of a clinical pathway for the evaluation and management of TIA using computer-based clinical support is feasible in a community ED setting. This pilot study in knowledge translation provides a framework for the
S102 Annals of Emergency Medicine
Conclusion: A large differential in both systolic blood pressure and heart rate within 24 hours of symptom onset in intracerebral hemorrhage appears to be associated with a higher risk of death at one year.
327
Retrospective Validation of a Clinical Decision Rule to Safely Rule Out Subarachnoid Hemorrhage in Emergency Department Headache Patients
Hart D, Smith SW/Hennepin County Medical Center, Minneapolis, MN
Background: Perry et al. derived a clinical decision rule to identify headache (HA) patients who require computed tomography (CT) and lumbar puncture (LP) to rule out aneurysmal subarachnoid hemorrhage (AnSAH). Among nontransferred, neurologically intact adults (ⱖ 15 years of age) with atraumatic acute HA of ⬍ 1 hour to peak intensity, and no history of recurrent headaches, they found no AnSAH among those who had all of the following: no vomiting (V), diastolic blood pressure (DBP) ⬍ 100, arrival by private transportation, and age ⬍ 45 years. Study Objectives: To retrospectively validate this decision rule and explore the utility of other variables. Specifically, we sought to also investigate those who had a history of recurrent HA but whose presenting HA was different than prior HA, seizure or syncope, and HA onset ⬎ 1 hour.
Volume , . : September
Research Forum Abstracts Methods: We searched the ED database from January 1, 1996 through June 30, 2006 for a primary or secondary diagnosis (Dx) of “subarachnoid hemorrhage” and of “headache” and for a chief complaint (CC) of “headache.” We also searched the hospital ICD-9 code for patients with AnSAH. We then reviewed all identified AnSAH charts for: age, sex, presenting complaint, transfer-in status, neurological exam, vomiting, DBP, ambulance transport, and rapidity of HA onset. We also searched for history of previous similar HA, Dx by CT vs. LP, and history of syncope or seizure prior to arrival. Results: There were 12,472 ED patients with a CC of HA and 16,471 with final ED Dx of HA. 169 patients with new non-traumatic AnSAH were identified: 118 presented with altered mental status, 6 had a focal neurologic deficit. Of 7 neurologically intact patients, 3 had seizure activity (none would be missed by the Perry rule), and 4 had syncope prior to arrival (2 would have been missed by the rule, one due to HA onset ⬎ 1 hour). 1 patient had a history of similar recurrent headaches but would have been identified by the Perry rule, and 6 other patients had onset to peak HA intensity ⬎ 1 hour, 1 of whom would have been missed by the Perry rule. Thus, 37 (0.30% of all HA CC) patients had non-obvious SAH, 30 (81%) of whom would have been identified by Perry’s decision rule, only if they excluded those with seizure or syncope. The rule missed 6 patients by requiring onset ⬍ 1 hour; if the rule were also applied to these 6, 35 of 37 would be detected. Three with final Dx of SAH had a CC other than HA. 4 of 31 patients with onset ⬍ 1 hour, and 3 of 6 with onset 1 hour, required LP for Dx (p⫽.07, Fisher exact test). Conclusions: Perry’s rule missed 8 of 169 AnSAH; excluding those with seizure or syncope, it missed 6. The rule was 81% sensitive for non-obvious AnSAH in our cohort. However, if patients with HA onset ⬎ 1 hour are included in the rule, it identified 35 of 37 non-obvious AnSAH (sensitivity, 95%). The specificity could not be determined. LP may be more important for patients with onset ⬎ 1 hour than those with onset ⬍ 1 hour.
Table 1: Parameter Estimates for the Change in the Edema/Hematoma Volume Ratio Over 1 Hour or 20 Hours as a Function of Hemodynamic Parameters
329 328
Perihematomal Edema Growth is not Associated With Hemodynamic Variables
Jauch EC, Lindsell CJ, Adeoye O/University of Cincinnati, Cincinnati, OH
Study Objectives: Perihematomal edema formation in spontaneous intracerebral hemorrhage (ICH) is associated with worse clinical outcome. Identifying hemodynamic variables associated with edema growth could provide therapeutic targets. This study investigated the potential role of hemodynamic factors in perihematomal edema development. Methods: This was a post hoc analysis of clinical and computed tomography data from a prospective observational study of ICH patients presenting within 3 hours from symptom onset. Edema volumes were measured at hospital arrival, 1 hour and 20 hours from presentation. Blood pressure and heart rate, recorded at 19 time points between presentation and 20 hours, were used to derive hemodynamic variables (peak and mean systolic and diastolic blood pressures, heart rate, mean arterial pressure, pulse pressure). For assessing the relationship between hemodynamic parameters and edema growth, generalized linear modeling was used with edema/hematoma volume ratio as the primary dependent variable, adjusting for history of hypertension, aspirin or nonsteroidal antiinflammatory use, and, at 20 hours only, antihypertensive medication use. Separate models were constructed for each hemodynamic parameter. Results: Of 103 patients enrolled, 76 were eligible for analysis of edema growth at 1 hour after removing those with incomplete data and early surgery, 56 were included for analysis of 20 hour edema growth. No single hemodynamic variable was found to be associated with 1 or 20 hour perihematomal growth (table 1). Conclusion: Our data suggest that hemodynamic parameters do not influence the growth of perihematomal edema.
Volume , . : September
Automated Activation of a Stroke Team and Radiology Reduces Time to Evaluation and Brain Imaging
Baer A, Singer AJ, Byers AM, Perkins C/Stony Brook University, Stony Brook, NY
Study Objectives: National guidelines recommend door-to-brain imaging times of less than 25 minutes and door-to-needle times for administering thrombolytics of less than 60 minutes. We performed a before and after trial to determine the impact of institutional implementation of a brain attack protocol (BAT) on stroke quality of care measures. Methods: Study Design: Before and after trial. Setting-Suburban academic ED with Stroke Center. Subjects: Consecutive patients presenting with acute stroke within 3 hours of symptom onset. Interventions: Automated group paging system implemented in 4/05 alerting the stroke team to report to the ED within 15 minutes, and radiology to make the CT scanner immediately available. A prospective institutional stroke registry was implemented in 7/04. Outcomes: Data were collected on times from arrival in the ED to initial CT scan, door to neurologist assessment, and door to thrombolytic administration. Analysis: Outcomes before and after implementing the BAT were compared with t-tests and chi-square tests. Results: There were 40 patients before and 146 patients after BAT implementation. Their mean age was 68.3⫹/⫺16.2 years; 42% were female. Groups were similar in age and gender. Mean NIH stroke scores were higher in the after period (8.5⫹/⫺8.2 vs. 5.0⫹/⫺5.7; P⫽0.008). The mean time to arrival of the stroke team was significantly reduced (mean difference 73.3 min; 95% CI, 54.3-92.3 min) after BAT (4.8⫹/⫺4.7 vs. 78.1⫹/⫺116.9 min). The mean time to brain imaging was significantly reduced (mean difference 50.8 minutes; 95% CI, 37.5-64.1) after BAT (29.5⫹/⫺28.1 min vs. 80.3⫹/⫺64.5 min). TPA use before and after BAT were 8/9 and 41/41 eligible patients respectively (P⫽0.40). Mean time to TPA was not reduced after BAT; mean difference 5.7 min, 95% CI, ⫺42.0-30.6 (92.3⫹/⫺49.5 min vs. before 98.0⫹/⫺25.4). Conclusion: Automated activation of a stroke team and radiology reduced the time to evaluation by a neurologist and the time to brain imaging but not time to TPA.
Annals of Emergency Medicine S103