Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
1 Department of Anesthesia 2 ANNUAL GARY JOHNSON ANESTHESIOLOGY RESEARCH DAY PROGRAM Friday, May 6, 2016 University of Ottawa – Roger Guindon Hall – 451 Smyth Road 07:30 – 08:00 CONTINENTAL BREAKFAST (Atrium, Roger Guindon Hall) 08:00 – 08:10 OPENING REMARKS (Amphitheatre A, Roger Guindon Hall) Colin McCartney, MBChB, PhD, FRCA, FCARCSI, FRCPC Professor and Chair of Anesthesiology Gregory Bryson, MD, MSc, FRCPC Chair, Gary Johnson Research Day Planning Committee Objectives for Overall Program: 1. Discuss results of anesthesia and perioperative medicine related clinical (RCTs, chart audits and case reports/series), education and bench research that has been performed by residents and fellows affiliated with the University of Ottawa Department of Anesthesiology. 2. Evaluate and discuss the design and execution of randomized clinical trials, quality improvement projects and medical education initiatives. 3. Review the merits of basic statistical tests as they apply to the areas of research mentioned above. Objectives for Sessions One and Two: 4. Examine data and outcomes from your practice environment 5. Critique research design and analysis 6. Appraise research presentation skills 08:10 – 08:45 SESSION ONE: BULLET PRESENTATIONS (Amphitheatre A, Roger Guindon Hall) 3 minute presentations, 2 minutes for questions Epidemiology of New Onset Postoperative Atrial Fibrillation in Cardiac Surgery Patients Jean Abboud, (MS) Meta-analyses of Intrathecal Morphine for Lumbar Spine Surgery Suzanne Todd, MD (Fellow) 3 08:10 – 08:45 SESSION ONE: BULLET PRESENTATIONS (Amphitheatre A, Roger Guindon Hall) 3 minute presentations, 2 minutes for questions Hydroxyethyl Starch in Preclinical Sepsis Models: A Systematic Review and Meta-Analysis Ryan Dong, (MS) ILMA vs Laryngoscopy for Tracheal Intubation in a Clinical Setting: A Systematic Review Sepand Alavifard, (Student) Prevalence and Severity of Apnea in Pediatric Patients Admitted For Post-Anesthetic Respiratory Kamyar Ghaffari, (Research Associate) Case Report: Dexmedetomidine for Difficult to Control Postoperative Pain Sivendiran Mahalingam, MD (PGY3) Impact of Including Regional Anaesthesia in Enhanced Recovery Protocols: A Scoping Review Evan Cole, MD (PGY2) 08:45 – 09:30 SESSION ONE: PODIUM PRESENTATIONS (Amphitheatre A, Roger Guindon Hall) 10 minute presentations, 5 minutes for questions Mesenchymal Stromal Cell Efficacy in Preclinical Models of Myocardial Infarction: A Systematic Review Carly Barron, (MS) The Association of Delayed Emergency Surgery with In-Hospital Mortality Karim Abdulla, MD (PGY5) Pulmonary Artery Pulsatility Index Predicts Need for Inotropic Support Post Cardiac Surgery Asad MirGhassemi, MD (PGY5) 09:30 – 09:50 BREAK (Atrium, Roger Guindon Hall) 4 09:50 – 10:35 SESSION TWO: PODIUM PRESENTATIONS (Amphitheatre A, Roger Guindon Hall) 10 minute presentations, 5 minutes for questions Challenging Authority During An Emergency- The Effect Of A Teaching Intervention Duncan McLuckie, MD (Fellow) The Association between Intraoperative Hypotension and Stroke Following Cardiac Surgery Amy Chung, (MS) Investigating the Effect of Formative Assessment in the Prevention of Peri-Operative Hypothermia Karim Mohamed, MD (PGY3) 10:35 – 11:10 SESSION TWO: BULLET PRESENTATIONS (Amphitheatre A, Roger Guindon Hall) 3 minute presentations, 2 minutes for questions A Systematic Review for a Consensus on Humane Endpoints in Preclinical Acute Lung Injury Ryan McGinn, (MS) Safety of mesenchymal stromal cell therapy for stroke (Safe Cell Stroke): A systematic review and meta-analysis Ryan McGinn, (MS) Paravertebral Block versus Two Step Surgical Block in Breast Surgery Shachar Ben-Zeev, MD (Fellow) Impact of Surgical Special Care Units: A Systematic Review Nick Mendis, MD (PGY2) Effect of Vasopressor Dosing on Microcirculation in Preclinical Models of Sepsis: A Systematic Review Tyler James, (MS) A Systematic Review Examining the Association of Incident Post-Operative Delirium with Mortality Gavin Hamilton, MD (PGY2) Advanced Resuscitation Training and Skill Retention among Healthcare Professionals: A Systematic Review Agata Dzwonek, (MS) 5 11:10 – 11:30 BREAK (Atrium, Roger Guindon Hall) 11:30-12:30 SESSION THREE: VISITING PROFESSOR’S PRESENTATION (Amphitheatre A, Roger Guindon Hall) 45 minute presentations, 15 minutes for questions From Measurement To Improvement: Using Data To Improve Perioperative Care Ramani Moonesinghe, BSc. (Hons) MRCP FRCA FFICM MD(Res) Objectives: 1. To describe the measures which can be used to improve quality in perioperative care 2. To discuss the challenges of using data for quality improvement rather than just measurement 3. To present the approaches being tested in the UK’s national Perioperative Quality Improvement Programme 12:30 - 12:45 CLOSING COMMENTS & AWARDS (Amphitheatre A, Roger Guindon Hall) 6 SESSION ONE Bullet Presentations (3 minute presentation, 2 minutes for questions) 7 Epidemiology of New Onset Postoperative Atrial Fibrillation in Cardiac Surgery Patients Authors: Abboud, J., Rashidian, H., Newton, E., Thommandru, M., Winch, D., and Tran, D.T.T. Purpose: Postoperative atrial fibrillation (POAF) is still a prevalent cardiac surgery complication that is associated with increased risk of important comorbidities and mortality. This study looked at the epidemiology of patients who had no prior history of atrial fibrillation during their hospital stay after cardiac surgery. Methods: Approval was obtained from the University of Ottawa Heart Institute Research Ethics Board. This is a prospective observational cohort study of 1416 adult patients undergoing non-emergent coronary artery bypass grafting (CABG) and/or valve surgery at a single cardiac hospital between 2014 and 2015. Univariate analysis was performed using Chi-square analysis and Student’s t-test to determine risk factors that are predictive of new onset POAF. Results: A total of 486 (34.3%) patients developed new onset POAF. Patients who had POAF were older (69.3 ± 9.7 vs 64.12 ± 11.4 years, p<0.001), had lower creatinine clearance (84.7 ± 33.6 vs 93.7 ± 38.8 mL/hour, p<0.001), underwent valve surgery (47.5 vs 35.6%, p<0.001), and had larger left atrial volumes (34.7 ± 13.0 vs 31.4 ± 11.9 mL/m2, p<0.001) than those who did not develop POAF. The incidence of postoperative complications was significantly higher in the POAF group for readmission to the ICU (6.4 vs 1.1%, p<0.001), reintubation (4.3 vs 1.3%, p<0.001), prolonged intubation (> 48 hours; 5.8 vs 1.6%, p<0.001), cardiogenic pulmonary edema (5.1 vs 1.1%, p<0.001), time spent in ICU (2.96 ± 5.2 vs 1.89 ± 3.7 days, p<0.001), length of hospital stay (12.5 ± 12.4 vs 7.9 ± 7.1 days, p<0.001), acute renal injury (20.4 vs 8.4%, p<0.001), and need for at least one of red blood cells, platelets, or fresh frozen plasma (16.0 vs 11.8%, p<0.05). We did not find a significant difference in stroke (1.4 vs 0.5%, p>0.05), seizure (1.4 vs 0.5%, p>0.05), malignant arrhythmia (4.9 vs 3.0%, p>0.05), gastrointestinal bleeding (1.0 vs 0.5%, p>0.05), heart block (4.5 vs 3.0%, p>0.05), and death (1.2 vs 0.9%, p>0.05). The management of patients postoperatively involved the highest incidence of POAF development on the second day (34%), followed by third (19%), and fourth (15%). Very few POAF developed after one week (average 1%). Among the 486 patients who developed an episode of POAF, 123 (25.3%) had a recurrence within the 14 postoperative days after having returned to regular sinus rhythm. Upon discharge, 49 patients continued to have atrial fibrillation, 192 were on a class III antiarrhythmic agent (Amiodarone), and 14 were on a different antiarrhythmic medication. Conclusion: New onset POAF remains a prevalent complication in the cardiac surgery population, associated with significant perioperative morbidity. This study demonstrated key characteristics of the population of patients undergoing cardiac surgery who develop new onset POAF and the expected complications that can follow. 8 Meta-analyses of Intrathecal Morphine for Lumbar Spine Surgery Authors: Paul, J. (Hamilton Health Sciences (HHS)), Todd, S. (The Ottawa Hospital), Jo, S. (HHS), & Hindle, A. (HHS) Purpose: Intrathecal morphine has the potential to provide prolonged analgesia following surgery but this comes at a risk of opioid side effects including the chance of delayed respiratory depression. Controversy remains as to the optimal dose and indications for intrathecal morphine. Current limitations in literature include variability in study design, dosing regimens and lack of outcomes with adequate sample sizes. The purpose of this meta-analysis of randomized controlled trials was to determine the efficacy of intrathecal morphine in reducing pain scores after lumbar spine surgery and the resulting incidence of opioid side effects and analgesia related adverse events. Methods: REB approval was not required for the study. A literature search of randomized control trials (RCTs) was conducted. The search yielded greater than 600 citations and a total of 10 trials were selected based on an ‘a priori’ inclusion criteria. VAS scores were grouped into six-hour time frames. Results were analyzed in terms of low dose (less than 150 mcg) or high dose (more than 150 mcg) ITM, or combined dose (both less than and more than 150 mcg). The incidence of side effect profiles related to ITM dosing was also analyzed. Odds ratios with corresponding confidence intervals were calculated for all outcomes using a random-effects model. Results: VAS scores were available for all 10 trials, 520 patients were randomized and data was available for 453 patients that met our inclusion criteria. Compared to placebo, the pooled results show that both low and high dose ITM groups had significantly lower VAS scores at 0-6 hours [Weighted Mean difference (WMD) of -23.18 (95% Confidence Interval (CI) of -37.66 to -8.70)), 6-12 hours [WMD -11.76 (CI -22.92 to -0.60)], and 12-18 hours [WMD -12.04 (CI -20.60 to -3.49)]. Compared to the control group, the ITM high and low dose groups combined had a significantly higher incidence of pruritis (Odds Ratio with 95% CI of 3.99 (1.78, 8.91)) as well as a significantly lower incidence of sedation [OR 0.35 (0.12, 1.00)]. There was no significant difference between groups regarding nausea, respiratory depression, urinary retention, and post-dural puncture headache. Compared to the control group, the ITM high and low dose groups combined also had significantly lower morphine consumption at 18-24 hours [WMD -10.2 (95% CI -16.6 to -3.8). Conclusion: The addition of intrathecal morphine provides superior post-operative analgesia up to 18 hours, for lumbar spine surgery, with significantly lower morphine consumption. Pruritis is the main side-effect that can be expected regardless of dosing. Despite concern for increased sedation with intrathecal morphine, our study results indicate that the incidence is actually reduced when compared to control group. 9 Hydroxyethyl Starch in Preclinical Sepsis Models: A systematic review and meta-analysis Yuan Yi Dong1; Lauralyn McIntyre2,3; Dean Fergusson2; Noah Kosowski3; Carly Barron1; Osman Ahmed1; Christina Pugliese1; Manoj M Lalu4,5 1Faculty of Medicine, University of Ottawa 2Clinical Epidemiology Program, The Ottawa Hospital Research Institute 3Department of Medicine, Division of Critical Care, The Ottawa Hospital, Centre for Transfusion and Critical Care Research, Canadian Blood Services 4Department of Anesthesiology, The Ottawa Hospital 5Regenerative Medicine Program, The Ottawa Hospital Research Institute Purpose Hydroxyethyl starch (HES) fluid resuscitation has recently been demonstrated to increase in mortality and acute kidney injury in patients with septic shock.1 It is unclear whether an analysis of preclinical HES studies may have helped predict the adverse effects of these fluids. We therefore conducted a preclinical systematic review and meta-analysis to investigate the safety of HES compared to other resuscitation fluids in animal models of sepsis. Here we report on the outcome of mortality. Methods A systematic search of Ovid MEDLINE and Embase was performed in collaboration with an information specialist (inception-01/2015). Citations were screened independently in duplicate. Studies comparing HES vs other resuscitation fluid in preclinical in vivo sepsis models were included. The Cochrane Risk of Bias Assessment Tool was used to assess internal validity of each included study. Construct validity (i.e. clinical generalizability) was assessed using a previously proposed 8 point framework.2 Results are expressed as risk ratios (RR) and 95% confidence intervals (95% CI). Meta-analysis was performed using an inverse variance random effects model. A priori determined outcome ascertainment windows were also analyzed (≤2 days, 2-4 days, ≥4 days). Results Ethics review was not necessary as this is an analysis of published literature.10 articles met eligibility criteria (n=439 animals). Animal models included rat (5 studies), swine (3), and sheep (2). To model disease, studies used IV endotoxin (4), cecal ligation and puncture (3), live bacteria implant (1), live bacteria infusion (1), and fecal peritonitis (1). Comparison fluids included gelatin (5), ringer’s lactate/acetate (5), saline (3), sterofundin (2), albumin (1), and pig plasma (1). Risk of bias was variable: 8 studies reported randomizing but did not describe the method, no studies described allocation concealment. Blinding of personnel and blinding of outcome assessment were performed in 4 and 2 studies, respectively. Studies incorporated a median of 2 (range 1-4) of 8 suggested construct validity criteria to increase clinical relevance (e.g. no studies included animals with comorbidities). Mortality of animals was described in 6 studies and no statistically significant effect of HES on mortality was noted (RR 1.45, 95%CI 0.75-2.75, I2 = 43%). One study reported on animals ≥4 days, with 7/7 animals treated with HES and 0/7 treated with plasma dying, respectively. 10 Conclusion There is a paucity of preclinical evidence regarding the long term safety of HES in animal models of sepsis. Available evidence suffered from variably risk of bias and potentially lower construct validity. Pooled analysis suggested a non-significant trend towards harm with HES. The single study performed with a longer outcome ascertainment window demonstrated harm with HES. The study protocol was prospectively registered on the CAMARADES website http://www.dcn.ed.ac.uk/camarades/research.html#protocols. References 1. Zarychanski R. et al. JAMA 2013; 309: 678-688. 2. François L. et al. Crit. Care Med. 2010; 38: 2401-2408. 11 ILMA vs Laryngoscopy for Tracheal Intubation in a Clinical Setting: A systematic review Alavifard S.1, Blacklock JM.2, Cheng-Boivin OY.2, Cho N.2, Tricco AC.3, Wright E4, Boet S.2 1 Clinical Epidemiology Program, Ottawa Hospital Research Institute 2 Department of Anesthesiology, University of Ottawa 3 La Ka Shing Knowledge Institute of St Michael’s Hospital 4 Health Sciences Library, University of Ottawa Background Endotracheal intubation (ETT) is a life-saving procedure practiced by various healthcare professionals, including: physicians, paramedics, and nurses with advanced training. Although it is a salient procedure ETT may be associated with severe complications such as emergency surgical airway, intensive care admission, or even death1. While direct laryngoscopy (DL) is currently the gold standard for ETT 2, the long learning curve and poor skill retention makes it a challenging skill for non-anesthesiologists to master3. The intubating laryngeal mask airway (ILMA) is a widely spread tool that has reported high success when DL has proven to be difficult4. ILMA has a short learning curve and allows for blind ETT. In addition, it provides protection of the airway; therefore, eliminating the need for laryngoscopy. A systematic assessment of the literature on the effectiveness of ILMA compared to DL for airway management has yet to be performed. This review aims to systematically assess and compare the effectiveness of ILMA to DL for ETT in a clinical setting. Methods REB approval was not required. A systematic search of MEDLINE, EMBASE, CINAHL, Scopus, and Cochrane Database of Systematic Reviews was performed. Screening, data abstraction, and quality assessment were independently conducted in duplicate by pairs of reviewers. Studies were quantitatively pooled using random-effects meta-analysis. Subgroup analysis was used to explore differences in pooled estimates based on operator experience. Results Of the 5059 records retrieved, 24 articles met the selection criteria. Fourteen studies had experienced operators while two had novice operators. Eight studies did not report operator experience. The overall rate of successful intubation was comparable between ILMA and DL (RR 1.01, 95% CI 0.94-1.07). Successful intubation was similar between ILMA and DL for experienced operators (RR 0.98, 95% CI 0.95-1.01), but ILMA was more effective than DL for novices (RR 1.43, 95% CI 1.121.81). Overall, a non-significant trend favored DL for successful intubation on first attempt (RR 0.91, 95% CI 0.83-1.00), which was significant for the experienced operator (RR 0.85, 95% CI 0.76-0.94). With a mean difference of 34 seconds (95% CI 20-49), procedural time for successful intubation in the experienced subgroup favored DL. For novice operators, there was insufficient data to assess success on first attempt and procedural time. 12 Conclusion When experienced operators perform ETT, DL is more effective than ILMA. For novice operators, ILMA may be more effective than DL, but more studies are required. In future studies comparing ILMA and DL, better reporting of operator experience is needed. The full protocol was registered on PROSPERO (CRD42015024747). References 1 Cook TM. Brit J Anaesth 2011:617-642 2 Scott J. Paediatr Anaesth 2009: Suppl 1:24-9 3 Youngquist ST. Acad Emerg Med 2008:1295-1303 4 Combes X. Anesthesiology 2004:1146-1150 5 Messant I. Ann Fr Anesth 2002:622-626 13 Prevalence and Severity of Apnea in Pediatric Patients Admitted for Post-Anesthetic Respiratory Monitoring Splinter W. MD, Associate professor, Department of anesthesiology, CHEO Ghaffari K. MD, Research associate, Department of anesthesiology, CHEO Barrowman N. PhD, Associate professor, Clinical research unit, CHEO Purpose: Post anesthetic apnea is a potentially life-threatening event with many known risk factors.1-3 The typical response to the presence of these risk factors is admission to hospital for overnight monitoring on wards and PICU. Such monitoring has a major impact on hospital resources, family dynamics, and operating room flow. Also, the efficacy of this very intrusive and expensive monitoring is unknown. The objectives of this study were to primarily measure prevalence and severity of a major respiratory event (apnea and/or severe hypopnea) in children admitted postoperatively for respiratory monitoring, and secondarily to investigate association of a number of risk factors with a major postoperative event. Methods: Institutional research ethics board was obtained. We retrospectively reviewed patient charts from 2014 and 2015. Patients were admitted for respiratory monitoring. We excluded patients with complex cardiac surgeries. We defined apnea as a pause in airflow for ≥ 10 s and severe hypopnea as a dip in SpO2 to ≤90% or a combination of a rise of PCO2 > 50mmHg and increased respiratory effort.1 A ‘severe’ apnea/hypopnea was associated with a major postoperative event. The following were considered major postoperative events: changes in admission plan, treatment plan and/or monitoring, activation of an ‘acute-crisis’ team, a return to hospital, prolonged admission, permanent anoxic brain injury or death, abnormal O2 administration, manually-assisted or mechanically-assisted ventilation, use of naso–pharyngeal airway, use of crisis medications, e.g. Ventolin, Epinephrine, Naloxone, and delayed extubation or re-intubation. The risk factors of apnea/hypopnea were age, obesity (weight >95th percentile for age), prematurity (gestational age <37wk), procedure e.g. airway surgery, obstructive sleep apnea, respiratory, cardiac and/or neurologic impairment, genetic syndromes and perioperative opioids use. Results: We have reviewed 704 patients who had 821 anesthetic procedures. Apnea and severe hypopnea had a similar impact on severity i.e. an apneic event and hypopneic event had a similar effect on acutecrisis interventions and major postoperative adverse events. There were 34 (4.1%) apnea incidents and 199 (24.2%) severe hypopnea events. Age, prematurity, cardiac and respiratory comorbidities were strongly associated with apnea/hypopnea (p<0.05), but obesity, neurologic comorbidities and genetic syndromes were not. The incidence of apnea/hypopnea was 70% in children <1 yr, while other ages were similar and had only 17 – 24% incidences. Conclusion: Severe hypopnea and apnea are relatively common after routine pediatric anesthesia. Patients with age < 1 yr are especially at increased risk of apnea and hypopnea. References: 1.Thongyam A, et al. Otolaryngol – Head Neck Surg. 2014;151:1046 – 1054. 2. Brown K A, et al. Anesth Analg. 2014;118(6):1157 – 1159. 3. Coté C J, et al. Anesht Analg. 2014;118:1276 – 1283. 14 Case Report: Dexmedetomidine for Difficult to Control Postoperative Pain Authors: Siva Mahalingam, Reva Ramlogan, Naveen Eipe Department of Anesthesiology, The Ottawa Hospital Introduction: Dexmedetomidine is an alpha-2 adrenergic agonist approved for procedural sedation and sedation in ICU patients for up to 24 hours. Emerging evidence supports its use in reducing intraoperative opioid consumption (1,2), prolonging nerve blocks (3) and preventing emergence delirium (4). We report the use of dexmedetomidine for the management of acute postoperative pain. Both patients have consented to the publication of this case series. Case 1 A healthy 35 year-old man was admitted following an ATV accident with a crush injury to his leg. Surgical management was complicated by wound infection, sepsis and rhabdomyolysis. Acute Pain Service (APS) started with IVPCA hydromorphone and a multimodal oral analgesia protocol of acetaminophen, celecoxib, tramadol and pregabalin. He underwent 5 surgeries over 10 days, and by post-op day (POD) 12, he suffered recurrent acute pain crises. In a fully monitored Trauma Unit, he was administered 1mcg/kg dexmedetomidine iv bolus. Pain reduced to 4/10. An infusion of dexmedetomidine was initiated (0.4 mcg/kg/hr) for 24 hours. Oral clonidine was initiated and titrated to 0.2 mg po tid. He had the benefit of an epidural for his 6th surgery; but once removed, a pain crisis resulted. Dexmedetomidine was restarted and continued for 3 days with monitoring (range 0.1- 0.4 mcg/kg/hr). Pain scores and opioid requirements gradually reduced. Pain was stabilized and controlled with multimodal oral analgesics until discharge. Case 2 A 70 year-old woman with chronic pain at multiple foci due to severe rheumatoid arthritis required a revision occipito-cervical fusion. Her baseline pain was 8/10 on a regime of acetaminophen, tramadol, oxycodone, fentanyl CADD IV, pregabalin, memantine, methadone and meloxicam. Postoperatively, she was transferred intubated to the ICU for postoperative care and pain control. APS initiated a ketamine infusion (20mg/hr) and fentanyl infusion (100mcg/hr), in place of her CADD. On POD 1, ICU stopped her propofol sedation to assess for possible extubation. She became agitated due to pain. Propofol was restarted along with a dexmedetomidine infusion at 0.2mcg/kg/hr IV as per APS orders. Later that same afternoon, the patient was assessed by APS and she appeared more comfortable with her eyes opening spontaneously and obeying commands. On POD 2, she remained intubated, but awoke easily to voice and was comfortable. The patient was successfully extubated later that day. Dexmedetomidine was discontinued on POD 3 and the patient transferred out of the ICU. Discussion: These patients’ pain management represents our first APS experiences with dexmedetomidine for the management of severe post-operative pain. Literature review suggests that dexmedetomidine has only been used pre-emptively for intraoperative pain management. Its use for acute painful exacerbations has not been previously reported. Further investigation is needed to explore the anti-nociceptive role of dexmedetomidine for postoperative pain. References 1. Can J Anaesth. 2011 Jun;58(6):540-550 2. Anesth Analg. 2008 Jun;106(6):1741-1748 3. Anesth Analg. 2010 Dec;111(6):1548-1551 4. Anesth Analg. 2010 Oct;111(4):1004-1010 15 Impact of Including Regional Anaesthesia in Enhanced Recovery Protocols: A scoping review McIsaac, DI1,2,3, Cole, ET1 and McCartney, CJL1,2,3. 1Department of Anesthesiology, University of Ottawa, Ottawa, Ontario, Canada 2Department of Anesthesiology, The Ottawa Hospital, Ottawa, Ontario, Canada 3Ottawa Hospital Research Institute, Ottawa, Ontario, Canada Purpose: Regional anesthesia (RA) techniques decrease pain, improve function, and reduce length of stay1. RA is often included in enhanced recovery protocols (ERPs); bundles of perioperative interventions designed to improve outcomes. However, RA techniques are complex interventions, and complexity may reduce ERP compliance. Furthermore, data supporting RA use in ERPs is limited2 and heterogeneous. Therefore, we sought to (1) compile the literature investigating the impact of RA within ERPs using a scoping review, and (2) delineate whether the literature supports a benefit of using RA in ERPs for commonly measured outcomes. Value is a priority for health systems, therefore we further sought to (3) assess whether including RA within ERPs would translate into improved healthcare value as defined by the widely accepted3 domains of the Institute for Healthcare Improvement Triple Aim4 framework (improved population health, better patient experience, lower per capita costs). To guide our analysis we hypothesized a causal pathway where enhanced proximal outcomes would improve intermediary outcomes, thereby improving value (Figure 1). Methods: Research ethics board approval was not required for this study. We conducted a scoping review following best practice methodology5. Studies of adult patients receiving RA within an ERP, compared with any control group not receiving the same RA technique, and reporting at least one outcome related to the Triple Aim were included. MEDLINE, EMBASE, CENTRAL, CDSR, PROSPERO, and the NHS EED were searched up to May 2015. Two reviewers independently assessed studies for inclusion and extracted data related to methodology, population, interventions, comparators and outcomes. Results: Of 695 identified citations, 58 studies were included for analysis. The majority (67%) were controlled prospective trials, primarily in colorectal (53%) and orthopedic surgery (21%). Emergency and ambulatory surgeries were rarely studied. RA techniques varied; predominantly peripheral nerve blockade in orthopedics and epidural analgesia in thoracic/abdominal/pelvic procedures. The majority of studies found that RA within ERPs improved pain and gut function, many demonstrating improved mobility and decreased length of stay. RA was rarely associated with worsened outcomes. Value-based Triple Aim outcomes were seldom reported, and used highly varied metrics. When reported, ERPs with RA appeared to improve population health and cost outcomes, but not patient experience (Figure 1). Conclusion: Existing literature supports a positive impact of RA within ERPs on proximal and intermediary outcomes. Intuitively this should improve healthcare value, however this association has rarely been studied. In order to establish the value of RA in ERPs, anesthesiologists must consistently measure accepted value outcomes (e.g. Triple Aim outcomes) in methodologically rigorous clinical studies. 16 References: 1) Choi S. Reg Anesth Pain Med 2013;38:340–9. 2) Hughes MJ. HPB 2014;16:699–706. 3) IHI Triple Aim Partners (http://www.ihi.org/Engage/Initiatives/TripleAim/Pages/Participants.aspx) 4) Berwick DM. Health Aff 2008;3:759–69. 5) Levac D. Implement Sci 2010;5:1–9. 17 SESSION ONE Podium Presentations (10 minute presentation, 5 minutes for questions) 18 Mesenchymal Stromal Cell Efficacy in Preclinical Models of Myocardial Infarction: A systematic review Barron, C. C.1, Juneau, P.2, McIntyre, L3,4, Fergusson, D.5, Yang, H.2, Liu, P6., Moher, D.3, Mazer, D7., Lalu, M. M2,8 On behalf of the Canadian Perioperative Anesthesia Clinical Trials Group (http://canadianpact.ca/ ) 1 Faculty of Medicine, University of Ottawa Department of Anesthesiology, The Ottawa Hospital 3 Clinical Epidemiology Program, The Ottawa Hospital Research Institute 4 Department of Medicine (Division of Critical Care), University of Ottawa 4 Department of Epidemiology & Community Medicine, University of Ottawa 5 The Ottawa Heart Institute 7 Department of Anesthesia, St. Michael’s Hospital, University of Toronto 8 Regenerative Medicine Program, The Ottawa Hospital Research Institute 2 Purpose Perioperative myocardial infarction (POMI) is a frequent perioperative cardiac complication. Unlike nonoperative myocardial infarction (MI), POMI is suggested to involve a mixture of plaque rupture and supply-demand imbalance. Recent POMI trials have failed to decrease morbidity and mortality. In preclinical models of POMI, exogenous mesenchymal stromal cells (MSCs) display anti-inflammatory and organ protective properties and may represent a promising therapy. We conducted a systematic review to answer the question: “In preclinical models of POMI what effect do MSCs (in comparison to placebo) have on cardiac function, infarct size, inflammation, and death?” Methods Ethics approval was not necessary as this is a review of published preclinical literature. A systematic literature search of MEDLINE, Embase, BIOSIS Previews and Webs of Science was conducted. We identified comparative preclinical studies of MI with animals given MSC therapy versus a vehicle/placebo. Studies had an anesthetic provided before/concurrent to acute MI. The primary outcome left ventricular ejection fraction was grouped according to these time points of outcome ascertainment: 6-24 hrs, >24-72 hrs, >72 hrs-1 wk, >1-3 wks, >3-4wks, >4 wks after administration of MSC/vehicle. Data was pooled using the ratio of weighted means (RoM) method with inverse variance random effects modeling and expressed as odds ratios and 95% confidence intervals (95%CI). Heterogeneity was expressed with the I2 statistic. Results 3216 citations were reviewed and 163 studies met inclusion criteria (representing animals). The majority of studies (139/163) used a small animal preclinical model while 27/163 used large animals. The method used to induce MI was ligation of the left anterior descending artery in 130/163 studies and ischemia-reperfusion in 28/163. Studies that reported anesthesia used volatile agents in 36/132 and injectable agents in 102/132. The majority of experiments used intramyocardial injection of MSCs (128/163), others used intravenous (23/163) or intracoronary (11/163). The tissue source of MSCs was bone marrow (143 studies), umbilical (11) or adipose tissue (13). At early time points following MSC therapy there was no effect on LVEF [6-24 h: RoM 1.011 (95%CI 0.989-1.035), n=6, I2=0%) ; >24-72 h: RoM 0.994 (95%CI 0.944-1.045), n=7, I2=0%]. At later time points a significant improvement in LVEF was noted with MSC therapy [>72 h – 1 week: RoM 1.041 (95%CI 1.018-1.065), n=31, I2=69%) ; >1 week-3 weeks: RoM 1.093 (95%CI 1.056-1.131), n=31, I2=77%; >3-4 weeks: RoM 1.113 (95%CI 1.0981.169), n=78, I2=88%; >4 weeks: RoM 1.072 (95%CI 1.048-1.096), n=29, I2=0%](Fig1). 19 Conclusion This review is the first to provide an estimate of efficacy of MSC therapy in POMI. MSC therapy after MI showed an improvement in LVEF, a predictor of mortality, at time points >72hrs. These results may support a first-in-human evaluation of MSC therapy for POMI and the characteristics of the preclinical studies can be used to inform its design. 20 The Association of Delayed Emergency Surgery with In-Hospital Mortality Abdulla, K. R. 1-2, Forster, A. J. 1-3, McIsaac, D. I.1-3 Affiliations 1. University of Ottawa 2. The Ottawa Hospital 3. Ottawa Hospital Research Institute Purpose Emergency surgery patients often experience delays between presentation and surgical intervention, the causes of which are multifactorial1-3. Surgical delay is associated with mortality in hip fracture patients4. Indication bias likely confounds this association because sicker patients tend to wait longer. The impact of delayed surgery on outcomes after other surgeries is poorly described. To improve emergency operating room (OR) access, our institution introduced a 5-level prioritization system, based on acuity and surgical indication; each level was assigned an acceptable surgical wait time. The objective of this study was to measure the association between surgical delay and in-hospital mortality. Methods Following ethics approval, we performed an observational cohort study at a tertiary-care health sciences center. All adult emergency non-cardiac surgery patients were identified between January 2012 and October 2014. Surgical delay was defined as a wait time from surgical booking to OR entry in excess of priority specific wait time (see table for wait time definitions). Delayed patients were propensity score (PS) matched on a 1:1 basis without replacement to non-delayed patients. We used variables that accounted for characteristics of their hospital admission, preoperative hospitalization, patient characteristics, physiologic instability, and surgical urgency and risk. The relative and absolute association between delay and mortality were calculated. Pre-specified sensitivity analyses, including a generalized additive model to investigate the continuous association between OR wait time and inhospital mortality were used to test the robustness of our primary analysis. Results 15 160 patients were identified; 2 820 (18.6%) had delayed OR access. System factors accounted for 86% of all documented delays. Delayed patients had an in-hospital mortality rate of 4.9% compared to 2.6% for non-delayed patients (crude odds ratio 1.59, 95%CI 1.30-1.93). All delayed patients were successfully PS-matched and balance of covariates was achieved between groups. Within the PSmatched cohort, delay was significantly associated with mortality (odds ratio 1.56, 95%CI 1.18-2.06; absolute risk increase 1.67%; number needed to treat to harm=60). This finding was confirmed in sensitivity analyses. Conclusion In a cohort of emergency surgical patients with known ready for surgery status and detailed patientlevel covariates to account for indication bias, delayed operating room entry after booked emergency surgery was associated with increased risk of in-hospital mortality across surgical specialties. For every 60 delayed surgeries, one extra death occurred; this equates to 80 excess deaths at our hospital during the study period. Importantly, system issues underlie most delays, and will need to be addressed to improve patient flow and outcomes. 21 References 1. Haider A., et. al. Ann Surg. 2015; 262(2):260–6. 2. Anantha RV., et. al. World J Emerg Surg. 2014; 9(1):21. 3. Heng M., et. al. Can J Surg. 2013; 56(3):167–74. 4. Moja L., et. al. PLoS One. 2012; 7(10):e46175. 22 Pulmonary Artery Pulsatility Index Predicts Need for Inotropic Support Post Cardiac Surgery Mir Ghassemi A,1 Ruel M,2 Chung A,1 Bourke M,1 Chen R,1 Sun L1,3 Affiliations: 1. Division of Cardiac Anesthesiology, University of Ottawa Heart Institute 2. Division of Cardiac Surgery, University of Ottawa Heart Institute 3. Institute for Clinical Evaluative Sciences Purpose: The pulmonary artery pulsatility index (PAPI), defined as (pulmonary artery systolic pressure – pulmonary artery diastolic pressure)/central venous pressure,1 is an emerging hemodynamic predictor of right ventricular failure and requirement for inotropic support following left ventricular assist device placement.1-3 To date, the utility of PAPI has not been explored in the general cardiac surgical population. The purpose of this study was to determine whether PAPI predicts inotropic medication use following cardiopulmonary bypass. Methods: Institutional ethics board approval was obtained for this retrospective cohort study of 287 patients undergoing major cardiac surgery requiring cardiopulmonary between January and August 2014 who were monitored intraoperatively with a pulmonary artery catheter (PAC). Excluded were off pump procedures, heart transplants, left ventricular assist device implants, pulmonary thromboembolectomies and lack of intraoperative PAC monitoring. Primary outcome was intraoperative use of one or more inotropic agents post-bypass (i.e., dobutamine, milrinone, epinephrine). Primary predictor was PAPI, calculated as described using the first available pulmonary artery pressure and central venous pressure following induction of anesthesia. Baseline patient characteristics and intraoperative variables were extracted from institutional Perioperative Databases. Hemodynamic variables, and details of number of type of intraoperative inotropes used, were obtained from electronic anesthesia records (CompuRecord, Netherlands). Chi-squared and Wilcoxon rank-sum tests were used where appropriate. We evaluated the unadjusted association between PAPI and log odds of post-bypass inotrope use using restricted cubic splines. This association was further modeled using multivariable logistic regression with adjustment for a priori selected risk factors. Measure of association was odds ratio (95% confidence interval). Analyses were performed using SAS 9.4, with statistical significance defined as a two-tailed P<0.05. Results: The relationship between PAPI and post-bypass inotrope use followed a “U-shaped” distribution (Figure). We therefore categorized PAPI as <0.5, 0.5-1.39 (reference category), and ≥1.4. Overall, 48 (16.7%) of 287 patients required inotropic support post-bypass. Independent predictors of post-bypass inotropic support were PAPI ≥1.4 (OR 2.53, 95% CI 1.11-5.76), combined valve(s) and bypass grafting (OR 8.55, 95% CI 2.99-24.41), thoracic aortic surgery (OR 7.66, 95% CI 1.71-34.33) and preoperative left ventricular ejection fraction <35% (OR 27.25, 95% CI 8.09-91.82). 23 Conclusion: PAPI ≥1.4 is a predictor of need for inotropic support following cardiopulmonary bypass. Given the high frequency of PAC use in cardiac surgical patients, this simple hemodynamic marker is an easy and promising tool for identifying those who are more likely to benefit from use of inotropic agents in the perioperative setting. Log Odds of Post-Bypass Inotrope Use References 1. Kang G. J Heart Lung Transplant. 2014;33:S86. 2. Morine KJ. J Card Fail. 2013;19:S17. 3. Morine KJ. Circulation. 2014;130:A19195. Figure. Spline transformation of the unadjusted association between pulmonary artery pulsatility index (PAPI) and the log odds of post-bypass inotrope use. The red outer curves represent 95% confidence intervals. PAPI 24 SESSION TWO Podium Presentations (10 minute presentation, 5 minutes for questions) 25 Challenging Authority during an Emergency- the Effect of a Teaching Intervention Duncan McLuckie MBBCh FRCPC,3 Zeev Friedman MD,1 Sev Perelman FRCPC,2 Meghan Andrews FRCPC,4 Laura MK Noble R.R.T,5 Archana Malvade,6 and M. Dylan Bould MBChB MEd FRCA7 1. Associate Professor and Staff Anesthesiologist, Mount Sinai Hospital, Toronto, ON, Canada 2. Assistant Professor and Staff Emergency Physician, Mount Sinai Hospital, Toronto, ON, Canada 3. Fellow, Children’s Hospital of Eastern Ontario, University of Ottawa, Ottawa, ON, Canada 4. Staff anesthesiologist, The Montfort Hospital, Ottawa, ON, Canada 5. Anesthesia Assistant and Respiratory Therapist, Mount Sinai Hospital, Toronto, ON, Canada 6. Anesthesia Assistant and Respiratory Therapist, Mount Sinai Hospital, Toronto, ON, Canada 7. Associate Professor and Staff Anesthesiologist, Children’s Hospital of Eastern Ontario, University of Ottawa, Ottawa, ON, Canada Purpose Effective communication is crucial during crisis situations. Hierarchy is deeply engrained in the culture of medicine and can cause a failure in effective communication. Previous research has shown that residents were unable to effectively challenge a superior’s wrong decision during a crisis situation,1-3 which can contribute to preventable mortality.4 The objective of the study was to assess whether a teaching intervention enabled residents to effectively challenge wrong clinical decisions made by their staff during a simulated life threatening airway emergency. Methods Following local ethics board approval, second year residents were randomized to receive a teaching intervention targeting the cognitive and interpersonal skills needed to challenge a superior’s decision, or a control group which received a generic crisis management training. Two weeks later, subjects participated in a simulated crisis where they were provided with opportunities to challenge a staff on a wrong decision in a can’t intubate can’t ventilate scenario. Deliberate deception was used in order to preserve the “natural” hierarchy gradient and residents were told that staff and resident were being evaluated to compare behavior in simulated settings to clinical settings. Performances were video recorded and later assessed and scored in random order by two trained independent raters, blinded to group allocation and unfamiliar with the subjects, using the previously validated modified AdvocacyInquiry Score (mAIS). Results 51 residents were recruited and fifty completed the study. One video was excluded because of technical issues. The inter-rater reliability of the mAIS scores among raters (ICC=0.87, 95% CI: (0.70, 0.94)) was excellent. The median (IQR [range]) of the maximal mAIS (our primary outcome) across all challenging opportunities and averaged out across raters was significantly better in the intervention group 5.0 (4.50-5.62 [4-6]) than in the control group 3.5 (3.0-4.75 [3-6]) (Cohen’s D effect size 2.0, Mann-Whitney U=52.5, p<0.001) (figure 1). The median of the total number of challenges was 7.5 (5.610.0 [3.0-14.5]) in the control group vs. 12.8 (10.0-16.0 [8.5-20.0]) in the intervention group (Cohen’s D =1.6, U=79, p<0.001). Figure 1: box plot of mAIS in the control vs intervention group, box is IQR, horizontal line is median, whiskers are range 26 Conclusion These results demonstrate that a short teaching intervention was effective in improving residents’ ability to challenge a wrong decision by a superior. This suggests that residents are not given the proper tools to challenge authority during a crisis situation. Incorporating a similar intervention into the residency curriculum may address this problem and improve communication and patient safety during crisis management. References: 1. Sydor D. Br J Anaesth. 2013 ;110 :463-71. 2. Friedman Z. Anaesthesia. 2015; 70: 1119-29. 3. Bould MD. Can J Anaesth. 2015; 62: 576-86 4. Bromiley M. R Coll Anaesth Bull 2008; 48: 2442-5 27 The Association between Intraoperative Hypotension and Stroke Following Cardiac Surgery Chung AM,1 Wells G,2,3 Hossain A,3 Bourke M,4 Sun LY4,5 Affiliations: 1. Faculty of Medicine, University of Ottawa 2. School of Epidemiology, Public Health and Preventative Medicine, University of Ottawa 3. Cardiovascular Research Methods Center, University of Ottawa Heart Institute 4. Division of Cardiac Anesthesiology and Critical Care, University of Ottawa Heart Institute 5. Institute for Clinical Evaluative Sciences Purpose: Cerebrovascular accidents (CVAs) occur in 1.8-9.7% of patients undergoing major cardiac surgery and represent a serious postoperative complication.1-3 While intraoperative hypotension (IOH) is thought to play a role, no model to date has addressed the combined effect of IOH during distinct periods of cardiac surgery. We investigated whether varying magnitudes and durations of IOH pre-, during and post-cardiopulmonary bypass (CPB) were associated with postoperative CVA. Methods: Following research ethics board approval, we conducted a retrospective cohort study of 7692 patients undergoing major cardiac surgery at the University of Ottawa Heart Institute requiring CPB between November 2009 and June 2014. Patients undergoing off pump procedures were excluded. The primary exposures were, separately, total durations of MAP < 55, 65 and 75 mm Hg, pre-, during and postCPB. The primary outcome was postoperative ischemic CVA, defined as new focal or global neurologic deficit of cerebrovascular origin lasting ≥ 24h and non-hemorrhagic in nature. The diagnosis of CVA was verified by reviewing brain CT or MRI reports. Intraoperative invasive blood pressure measurements were recorded every 15 seconds in an electronic patient record, with any artifacts removed using an automated algorithm. The relationship between hypotension and CVA was modeled using logistic regression with propensity score adjustment. Independent CVA risk factors were identified through a non-parsimonious logistic regression model. Measure of association was OR (95% CI). All analyses were conducted using SAS 9.4, with statistical significance defined by a 2-tailed P < 0.05. Results: CVAs occurred in 147 patients (1.9%) and were associated with any duration of MAP < 75 mmHg during and post CPB, in a dose-response fashion according to duration and severity of IOH (Table). Pre-CPB IOH was not associated with CVA (Table). Additional independent CVA risk factors included older age, combined valve and bypass surgery, surgery on the thoracic aorta, preoperative shock, cooling while on bypass, hemodynamic instability post bypass, new onset postoperative atrial arrhythmias, and reopening following surgery. Conclusion: MAP < 75 mmHg during and post-CPB was associated with CVAs, with evidence of a dose-response relationship with increasing severity and duration of hypotension. The ability to define critical thresholds and durations of hypotension associated with ischemic brain injury may lead to prompt preventative interventions. This study provides impetus for future research to develop a personalized goal-directed therapy for high-risk cardiac surgical patients. 28 References: 1. Gottesman RF. Stroke 2006; 37:2306-2311. 2. McKhann GM. Stroke 2006; 37:562-571. 3. Roach GW. N Engl J Med 1996; 335:1857-1863. Table. Association of total and longest durations of mean arterial pressure < 55, 65 and 75 mmHg and cerebrovascular accident. IOH Severity Propensity-adjusted OR (95% CI) per 10 min of Total IOH duration MAP < 55 mmHg Pre-Bypass 0.97 (0.83 – 1.13) During Bypass 1.21 (1.15 – 1.28) Post Bypass 1.07 (1.01 – 1.13) MAP < 65 mmHg Pre-Bypass 0.98 (0.90 – 1.06) During Bypass 1.13 (1.09 – 1.17) Post Bypass 1.05 (1.01 – 1.09) MAP < 75 mmHg Pre-Bypass 0.94 (0.89 – 1.00) During Bypass 1.09 (1.06 – 1.13) Post Bypass 1.05 (1.01 – 1.08) MAP = mean arterial pressure; IOH = intraoperative hypotension; OR = odds ratio; CI = confidence interval 29 Investigating the Effect of Formative Assessment in The Prevention of Peri-Operative Hypothermia Mohamed, K.; Brehaut, J.; Bryson, G.; Clavel, N.; Forster, A.; Grimshaw, J.; McIsaac, D.; Pysyk, C.; Taljaard, M.; Boet, S. Purpose: Perioperative hypothermia remains common and is associated with increased operating costs and poor postoperative outcomes[1-4]. Formative assessment (i.e. feedback) can improve physicians’ practice[5]; however, the most effective type of feedback in temperature management is unknown. This study aims to measure the impact of targeted feedback on anesthesiologists’ perioperative temperature management. Methods: Research Ethics Board approval was obtained (#20140117-01H). This study was a three-arm cluster randomized controlled trial. Based on a pilot study, pre determined power determinations, and sample size calculations, it was determined that at least 14 anesthesiologists were needed in each arm. Participants’ perioperative temperature management performance was audited prior to any intervention (six-month baseline phase). Participants’ were then randomized to one of three groups: 1. No feedback (control), 2. Benchmarked feedback, or 3. Ranked feedback. Feedback was then provided according to group allocation (six-month intervention phase) and performance in each group was assessed. Comparisons between pre and post-intervention and between groups were made. Physician performance was analyzed using % hypothermic patients. Statistical analysis was performed using a generalized linear mixed model with logit link and binomial distribution, accounting for clustering of patients by provider and for correlation in repeated measures on the same provider over time. Results: Forty-five staff anesthesiologists at The Ottawa Hospital participated. Our preliminary analysis demonstrates no difference between the three groups in the proportion of hypothermic patients during the intervention phase, controlling for the pre-intervention phase and campus (p=0.27). Neither benchmarked or ranked feedback showed a statistically significant difference in the proportion of hypothermic patients from control during the intervention phase (p=0.46 and p=0.38). Conclusion: Despite room for improvement in temp management, preliminary data suggest no evidence that formative assessment using benchmarked or ranked feedback is more effective than no feedback in decreasing the proportion of hypothermic patients after surgery. It may be necessary for future interventions to bundle multiple feedback methods in order to measure a significant change. References: 1. Sessler DI. Anesthesiology 2001; 95: 531 2. Lenhardt R, Marker E, Goll V, et al. Anesthesiology 1997; 87: 1318 3. Sun LS, Adams DC, Delphin E, et al: Crit Care Med 1997; 25: 1990 4. Frank SM, Fleisher LA, Breslow MJ, et al:. JAMA 1997; 277: 1127 5. Hysong, Sylvia J. Medical Care 47.3 (2009): 356-363. Web. 30 SESSION TWO Bullet Presentations (3 minute presentation, 2 minutes for questions) 31 A Systematic Review for a Consensus on Humane Endpoints in Preclinical Acute Lung Injury McGinn R1, Stewart DJ2, McIntyre L3, Griffin G4, Fergusson DA4, Barron CC1, Lalu MM2 1 Faculty of Medicine, University of Ottawa; 2 Regenerative Medicine Program, The Ottawa Hospital Research Institute; 3 Department of Medicine, Division of Critical Care, The Ottawa Hospital; 4 Clinical Epidemiology Program, The Ottawa Hospital Research Institute Introduction: Acute respiratory distress syndrome is a devastating condition with a high rate of mortality. However, the mechanisms of organ failure and death remain unknown. Therefore, rigorous preclinical studies of acute lung injury (ALI), with a focus on severe and late stages of disease, are needed to evaluate novel therapies. This results in unresolved animal welfare issues, since true death is not routinely accepted as an experimental endpoint. Instead, surrogate humane endpoints of death are used as they are thought to minimize suffering while allowing a valid assessment. To date, there is no consensus regarding surrogate humane endpoints of death in preclinical ALI. Prior to engaging key stakeholders in a consensus process, we undertook a systematic search to identify existing guidance for surrogate humane endpoints of death in small animal models of ALI. Methods: A systematic search of Medline and Embase was performed in collaboration with an information specialist (inception–09/2015). Retrieved citations were screened and general study characteristics were extracted independently in duplicate. Primary studies, reviews, and editorials providing guidance for humane endpoints in small animal models of ALI and critical illness (e.g. sepsis) were included. Results: Our search retrieved 1744 citations and 10 articles met eligibility criteria; 7 were primary studies and 3 were review articles (including 1 consensus statement for animal models of sepsis). These articles examined ALI (7 articles) and other models of critical illness (3) in small (6) and large animal models (4). Four articles identified body temperature as a surrogate humane endpoint (surface [<28.8°C], oral [<32.0°C] or rectal [<32.0°C]). Similarly, 3 articles reported the percentage of body weight lost (>20%) and/or loss of appetite as useful endpoints. Fur cleaning behaviour (e.g. presence of piloerection or rough hair coat) and hunched posture were identified as surrogate endpoints by 3 and 2 articles, respectively. Finally, 2 articles suggested severe dyspnea or altered breath sounds as humane endpoints. Additional endpoints suggested by one article each included blood pressure (exact pressure not specified), animal activity, closed eyelids when stimulated, the presence of biochemical evidence of organ failure (biomarkers not defined) and an 8-point composite score. Eight of 10 endpoints lacked a quantitative method for objective assessment and none were prospectively validated. Conclusions: This is the first systematic review of the humane endpoints for preclinical ALI. Although several surrogate humane endpoints were identified, none have been prospectively validated. This highlights the need for consensus guidelines to develop humane endpoints that can be applied in ALI studies across laboratories in Canada. Decreasing the variability in endpoints may improve generalizability of preclinical studies which, in turn, may increase preclinical to clinical translation of novel clinical therapies. Funding source: Canadian Institutes of Health Research. 32 Safety of Mesenchymal Stromal Cell Therapy for Stroke (Safe Cell Stroke): A systematic review and meta-analysis McGinn R1, Juneau P2, Fergusson DA2, Stewart DJ3,4, Dowlatshahi D5,6,7, Corbett D4, Lalu MM3,8 1 Faculty of Medicine, University of Ottawa; 2Clinical Epidemiology Program, The Ottawa Hospital Research Institute; 3Regenerative Medicine Program, The Ottawa Hospital Research Institute; 4 Department of Cellular and Molecular Medicine, University of Ottawa; 5Department of Medicine, Neurology, The Ottawa Hospital; 6Neuroscience Program, The Ottawa Hospital Research Institute; 7 Department of Medicine, Epidemiology and Community Medicine, University of Ottawa; 8Department of Anesthesiology, The Ottawa Hospital. Introduction Stroke is the second most common cause of death worldwide and the incidence of perioperative stroke is up to 7% in non-cardiac surgery. Following acute treatment, no options are available for long-term rehabilitation. Preclinical evidence suggests that mesenchymal stromal cells (MSCs) may be therapeutic; however, no systematic synthesis has been performed regarding the safety of MSCs for stroke patients. Therefore, in preparation for a clinical trial to treat chronic stroke, we conducted a systematic review to assess MSC safety compared to the standard of care in stroke patients. Methods A systematic search of Medline, Embase and Cochrane Central Register of Controlled Trials was performed with an information specialist (inception – 01/2016). Retrieved citations were screened with study characteristics and safety outcomes (immediate and long-term adverse events) extracted independently in duplicate. Controlled-comparison (randomized and non-randomized), case series and cohort studies examining MSCs as treatment for stroke in humans were included. Completeness of adverse event reporting was assessed using the CONSORT approach to harm reporting as a guide. Adverse events were pooled and described using odds ratio (OR) and 95% confidence intervals (95% CI). An OR <1 favors MSC therapy. Meta-analysis was performed using Peto’s method with correction of zero-count cells. Results Our search retrieved 1672 citations with 6 eligible studies: 5 interventional (1 randomized- and 1 nonrandomized control trial with 3 case series) and 1 observational (cohort). Adverse events were defined completely in 1 study and partially in 3 studies. A total of 106 stroke patients were included (98 ischemic, 2 hemorrhagic, 6 unreported, all subacute or chronic); 45 received MSCs (12 allogeneic, 28 autologous, 5 unreported) and 61 received standard of care. MSC dosing ranging from 2-18 x 107 was delivered by one (4 studies) or multiple injections (2 studies) via intravenous (4 studies), intra-arterial (1 study) or intracerebral (1 study) within 3-260 weeks following stroke. Patients were followed for a median of 52 weeks (range: 24-118). MSCs were not associated with increased death [OR 0.37 (95%CI 0.13-1.02); I2=0, n=3 studies]; in 3 case series, 1/18 patients died. Similarly, MSCs were not associated with neurological complications [OR 1.32 (95% CI 0.32-5.43); I2=0, n=2 studies]; in 3 case series, 0/18 patients any neurological symptoms. Despite a theoretical risk of MSCs causing cancer, the present data showed no association [OR: 0.758 (95% CI 0.11-4.86); I2=0, n=3 studies]; in 3 case series, 0/14 patients developed cancer. MSCs were not associated with increased risk of hematological, gastrointestinal, or renal adverse events. 33 Conclusion Few studies to date have investigated MSCs as a treatment for stroke. Available studies incompletely defined and reported adverse events. The limited data available suggests that MSC therapy may be safe for stroke patients. This protocol was registered in PROSPERO (CRD42016033503). 34 Paravertebral Block versus Two Step Surgical Block in Breast Surgery Dr. Shachar Ben-Zeev, Dr. Michelle Chiu, Dr. Anne Lui, Dr. Dan McIsaac Introduction Recovery from breast surgery is often complicated by pain, nausea, sedation and impaired functional recovery1. Compared to placebo, paravertebral blockade (PVB) decreases postoperative pain and nausea2. However, PVB is not without risk2; evidence regarding its superiority to other analgesic techniques is limited by factors including small sample size, inadequate blinding and lack of active comparator2. This study’s objective was examining the impact of PVB compared to surgical field block (SFB) on postoperative quality of recovery (QoR), pain, and shoulder function. Methods Following ethical approval we conducted secondary analysis of data collected for a published double blinded randomized trial (NCT01089933)3. Breast cancer surgery patients were randomized to PVB or SFB. The PVB group had T1-T6 PVB with 5ml 0.5% Ropivicaine per level and saline injected by surgeon into the wound and in the drain. The SFB group had saline subcutaneous injections at T1-T6 and 0.5% Ropivicaine injected by surgeon into the wound (10ml) and in the drain (20ml). Differences in QoR on postoperative day (POD) 2 were assessed using a Wilcoxon test. Differences in the proportion of patients with pain numeric rating scale (NRS) >3 on POD2 were assessed using a Chi-square test. Differences in Constant score (a validated 100-point scale to measure shoulder function) at first follow up were assessed using a t-test. Generalized linear models were used for QoR and NRS repeated measures across the first 7 PODs. P-values <0.05 were considered significant. Results 129 patients were recruited, 65 were randomized in a concealed manner to the PVB group, 64 to the SFB group. Characteristics between groups were similar. There was no difference in POD 2 QoR (PVB: 18, IQR 17-18; SFB: 17, IQR 17-18, P= 0.83), proportion of patients with NRS>3 on POD 2 (PVB: 16.4%; SFB: 14.1%, P=0.81), or Constant score (PVB:75 in PVB; SFB: 69, P=0.68) between the groups. Differences in QoR and NRS were not significant in repeated measures analysis (P=0.76, P=0.25 respectively). Discussion To our knowledge, this is the first adequately blinded randomized trial evaluating the impact of PVB versus an active comparator on recovery, pain and function after oncologic breast surgery. There does not appear to be significant benefit from PVB compared to SFB on any of our measured outcomes in the early postoperative period, while previous analysis of this data shows no benefit in terms of chronic post-surgical pain3. We hypothesize that SFB may be superior to surgical local anesthetic infiltration, and may explain the divergence of our findings from previous studies of PVB. These findings should be considered when weighing the risks and benefits of neural versus surgical field block in similar patient populations. 35 1. Acta Anaesthesiol Scand. 1999;43:770 2. Anesth Analg. 2010;110:1740 3. Ann Surg Oncol. 2014;21:795 36 IMPACT OF SURGICAL SPECIAL CARE UNITS: A SYSTEMATIC REVIEW Nicholas Mendis1, MD; Gavin M. Hamilton1, MD, MSc; Daniel I. McIsaac1,2,10, MD, MPH; Dean A. Fergusson3,10, PhD; Hannah Wunsch4, MD, MSc; Daniel Dubois2, MD; Homer Yang1,2,10, MD; Colin McCartney1,2,10, MBChB, PhD; Lauralyn McIntyre5,10, MD, MSc; Michaël Chassé10, MD, MSc; Heather McDonald6, MD; Alexis Turgeon7, MD, MSc; Sonia Sampson8, MD; André Denault9, MD, PhD; Manoj M. Lalu2,10*, MD, PhD On behalf of the Canadian Perioperative Anesthesia Clinical Trials Group (PACT) 1 Department of Anesthesiology, University of Ottawa Department of Anesthesiology, The Ottawa Hospital 3 Department of Medicine, Surgery & Epidemiology and Department of Community Medicine, The University of Ottawa 4 Department of Critical Care Medicine, Sunnybrook Health Sciences Centre, and Department of Anesthesia, University of Toronto 5 Department of Critical Care, The Ottawa Hospital, University of Ottawa 6 Department of Anesthesia, University of Manitoba 7 Department of Anesthesia and Department of Critical Care, Université Laval 8 Department of Anesthesia, Memorial University of Newfoundland 9 Department of Anesthesia and Department of Critical Care, Montreal Heart Institute 10 The Ottawa Hospital Research Institute 2 Purpose: Perioperative intermediate care units (termed surgical special care unit, or SSCU) may improve surveillance and identification of at-risk surgical patients. Institution of an SSCU may lead to global improvements across patient outcomes, as well as reduce the workload and financial burden at a systems level. We conducted a systematic review in order to investigate the effects of a 3-level model of care delivery (i.e. ward, SSCU, ICU) compared to a 2-level model of care (i.e. ward, ICU) on postoperative mortality, morbidity, and healthcare resource utilization. Methods: As this was a systematic review of published literature, REB approval was not required. Randomized controlled trials (RCTs) and non-randomized comparator studies (NRCTs) that compared a three vs two level model of care of perioperative non-cardiac surgery patients were included. A systematic search of Medline, CINAHL, Embase, and the Cochrane library was performed (inception-01/2015). Retrieved citations were screened and data extracted independently in duplicate. Data were extracted for mortality (primary outcome) as well as serious adverse events (SAEs), length of stay, and hospital costs (secondary outcomes). We planned pooling data (relative risk) using random effect models with the DerSimonian and Laird method, if applicable. Results: 1868 citations were retrieved by our search and 21 studies met eligibility criteria (2 RCTs, 19 NRCTs, 44134 patients). SSCUs were characterized by continuous monitoring (12 studies), absence of mechanical ventilation (8 studies), nursing:patient ratios (range 1:2-1:4), and number of beds (5, 3-33; median, range). Thirteen studies reported on mortality, three of which reported overall in-hospital mortality in a 2 vs. 3-level model of care. Significant methodological heterogeneity precluded pooled analysis, however two of the three studies demonstrated no difference in overall hospital mortality, and one demonstrated increased mortality after institution of an SSCU. Four studies reported ICU-specific mortality, two of which demonstrated an increased ICU mortality after an SSCU was established. Four 37 studies compared total in-hospital costs, two of which demonstrated reductions with a 3-level model of care. Nine studies reported on hospital length of stay and demonstrated no significant difference. Four studies reported SAE data, however heterogeneity in reporting precluded analysis. Conclusions: In this first systematic review of SSCUs, we observed significant heterogeneity in SSCU design and reporting of outcomes. Available data suggests creating an SSCU may increase in-ICU mortality with no difference in overall in-hospital mortality. This may reflect a ‘decanting’ of lower acuity patients from the ICU to the SSCU in a 3-level model of care. The potential effects of SSCUs on hospitalization costs warrants further investigation. The protocol for this study was prospectively registered with PROSPERO (CRD 20154025155). 38 Effect of Vasopressor Dosing on Microcirculation in Preclinical Models of Sepsis: A systematic review James, TE1’ ; Barron, CC1; Lamontagne, F2; Belley-Cote, E3; Yang, H1,4; Fergusson, D1,5; D’Aragon, F3; McIntyre, L5,6; Stewart, DJ7.8; Lalu, MM4,7 1 Faculty of Medicine, University of Ottawa Department of Medicine, Universite de Sherbrooke 3 Department of Clinical Epidemiology and Biostatistics, McMaster University 4 Department of Anesthesiology, The Ottawa Hospital 5 Clinical Epidemiology Program, The Ottawa Hospital Research Institute 6 Deparment of Medicine, Division of Critical Care, The Ottawa Hospital, University of Ottawa 7 Regenerative Medicine Program, The Ottawa Hospital Research Institute 8 Department of Cell and Molecular Medicine, University of Ottawa 2 Purpose: Surviving Sepsis Guidelines suggest titrating vasopressors to a mean arterial pressure of at least 65 mm Hg in sepsis; however this recommendation has weak supporting evidence (Grade C)1. Vasopressor-induced vasoconstriction could potentially restrict blood flow to vital organs and cause harm. The largest clinical study2 and systematic review3 of vasopressor dosing did not demonstrate mortality differences between high versus low doses in patients. Thus a summary of preclinical evidence may provide needed insight into the effects of vasopressors in septic shock. This systematic review investigates the question, “In preclinical studies using animal models of sepsis, what is the effect of different vasopressor dosing strategies on microcirculatory flow and mortality?” Methods: Ethics review was not required as this is a review of existing literature. We searched MEDLINE, BIOSIS, and Embase for preclinical studies of sepsis comparing low versus high dose vasopressors. Only active comparison studies were considered; for incremental dosing/cross-over designs only the initial dose was considered for analysis. Internal validity was assessed by a modified Cochrane Risk of Bias Assessment Tool4. Construct validity (i.e. “clinical relevance”) was assessed using previously published criteria5. Microcirculation and mortality results were pooled using a random effects model and expressed as either ratio of means (RoM) or odds ratios (OR) and 95% confidence intervals (CI). Ratios <1 favor low dose vasopressor. Heterogeneity was evaluated with the I2 statistic. Results: 2785 articles were screened and 8 articles met eligibility criteria (n=323 animals). Study characteristics are shown in Table 1. Risk of bias was unclear in all studies for randomization, allocation concealment, and blinding of personnel; and three studies were at high risk of bias for selective outcome reporting. Studies incorporated 4-8 (median=6) of 9 suggested construct validity criteria (e.g. no studies included animals with comorbidities; seven studies included fluid resuscitation). Three experiments from two studies examined microcirculation and seven experiments in four studies assessed mortality. Low versus high vasopressor dosing had no effect on microcirculatory flow (RoM 1.01, 95%CI 0.77-1.32, I2=86%, n=3 experiments); defined subgroup analysis of experiments with fluid resuscitation also demonstrated no effect on microcirculation (RoM 0.91, 95%CI 0.82-1.01, I2=0%, n=2). Low versus high vasopressor dosing had no effect on overall mortality (OR 0.54, 95%CI 0.19-1.58, I2=0%, n=7). 39 Conclusions: A limited number of preclinical studies have investigated vasopressor dosing strategies in sepsis. Given this limitation, high versus low vasopressor dosing had no apparent effect on either microcirculatory flow or mortality, similar to clinical work done to date. The study protocol was prospectively registered on CAMARADES website: (http://www.dcn.ed.ac.uk/camarades/research.html#protocols). References: 1. Dellinger RP. Crit Care Med 2013;41(2):580-637. 2. Asfar P. N Engl J Med 2014;370(17):1583-93. 3. D’Aragon F. Shock 2015;43(6):530-9. 4. Higgins JPT. The Cochrane Collaboration 2011. 5. Lamontagne F. Crit Care Med 2010;38(12):2401-8. Table 1. Study Characteristics Author Year Animal Model of disease Vasopressor Dose N Ishikawa 2012 Sheep Live E. coli Terlipressin Fixed 13 Santos 2011 Hamster LPS Dopamine Fixed 67 Li 2009 Rat LPS Norepinephrine Fixed 232 Lange 2007 Sheep LPS Terlipressin Fixed 27 Chin 2002 Swine LPS Dopamine Fixed 19 Lange 2011 Sheep LPS Terlipressin Fixed 27 Minneci 2004 Dog E. coli Epinephrine Fixed 78 Norepinephrine Fixed Vasopressin Fixed Norepinephrine Titrate to mean arterial pressure Correa 2013 Swine Feces E. coli – Escherichia coli, LPS – Lipopolysaccharide 24 40 A Systematic Review Examining the Association of Incident Post-Operative Delirium with Mortality Authors: Hamilton, GM 1,4, MD MSc; Wheeler, KE 2,4, BSc; Di Michelle, J 2,4 BHSc; Lalu, MM 1,3,4, MD PhD; McIsaac, DI 1,3,4, MD, MPH Department of Anesthesiology1, Faculty of Medicine2, The Ottawa Hospital Research Institute3, The University of Ottawa, Ottawa, Ontario4 Purpose: Delirium is a fluctuating disorder that represents a decompensation of cerebral function and can result in acute and reversible cognitive decline. Delirium has been associated with adverse post-operative outcomes, but controversy exists as to whether delirium is an independent predictor of mortality, especially in perioperative patients. Our objective was to assess the association between incident postoperative delirium and mortality in adult non-cardiac surgery patients. Methods: Ethical approval was not required for this project. The review was conducted in accordance with Metaanalysis Of Observational Studies in Epidemiology (MOOSE) guidelines. A systematic search of studies published between January 1981 and September 2015 was conducted using Cochrane, Medline/PubMed, CINAHL, and Embase databases. Eligible studies included randomized controlled trials, cohort studies and case-control studies with a validated measure of incident postoperative delirium, and mortality as an outcome. No language restrictions were applied. Screening was conducted by two independent reviewers. Study design, demographic data, exposure and outcome data were extracted. Pooled effect estimates were calculated with random-effects models using Stata 10.0 and expressed as odds ratios (OR) and 95% confidence intervals (95%CI). Heterogeneity was evaluated with the I2 index. Risk of bias was assessed using the Cochrane Risk of Bias Tool for NonRandomized Studies. We pre-specified a primary analysis which included only studies that adequately controlled for specified confounders (age, gender, comorbidity, previous cognitive impairment and surgery type/urgency), as well as a pooled analysis of unadjusted event rates. Results: 4626 citations were identified; 38 studies were included. No studies were judged at low risk of bias. Pooled analysis of unadjusted event rates (29 studies; n=6402) suggested that delirium was associated with a four-fold increase in the odds of death (OR= 4.13, 95%CI 3.50 - 4.87; I2=37.4%). A formal pooled analysis of adjusted outcomes was not possible due to heterogeneity of effect measures reported. However, in studies that controlled for pre-specified confounders, none found a statistically significant association between incident postoperative delirium and mortality (4 studies in hip surgery; n=1014) after an average follow-up of 31 months (Figure 1). Overall, as study quality improved, the association between delirium and mortality decreased. Conclusion: Few high quality studies are available to estimate the impact of incident postoperative delirium on mortality. Studies that controlled for key confounders did not demonstrate a significant independent association of delirium on postoperative mortality. Larger effect sizes noted when confounding bias was clearly present suggest that delirium may be an indicator of underlying factors that pre-dispose a patient to an increased risk of death rather than a true independent risk factor. The systematic review protocol was registered with PROSPERO (#CRD42015029805). 41 References: 1. Inouye SK. N Engl J Med 2006;354:1157–1165. 2. Young J. BMJ 2007;334:842–6 3. Witlox J. JAMA 2010;304:443–51 42 Advanced Resuscitation Training and Skill Retention Among Healthcare Professionals: A Systematic Review Agata Dzwonek MD2017 / Presenting Author Sylvain Boet, Kelly Au, Nitan Garg, Andrew E. Beck, Luc Tremblay, M. Dylan Bould Purpose Acute life-threatening emergencies requiring resuscitation represent a huge burden for society in terms of mortality, disability, and cost.1 Healthcare providers are expected to provide high-quality resuscitation regardless of the time since their last training. Skill retention however, declines as early as 3 to 6 months after training.2 Despite rigorous, standardized skills training and initial demonstration of competence, poor quality resuscitation is still commonly observed in actual cardiac arrests.3 This study aimed to perform a systematic review and meta-analysis of literature pertaining to the effectiveness, variability and key features of various advanced structured resuscitation training (SRT) programs designed for healthcare professionals (eg. ACLS, NRP). Methods REB was not required for this systematic review. Protocol registered with the PROSPERO database (CRD42014013851). We searched eligible articles through MEDLINE, EMBASE, CENTRAL, CINAHL, PsycINFO, ERIC, Scopus, Northern Light, and reference lists of relevant articles and systematic reviews. We included randomized controlled trials that examined skill retention of healthcare professionals following an advanced SRT program designed for healthcare providers. Only studies published in English, regardless of the publication date, were included. Two reviewers independently applied the eligibility criteria, extracted data, and assessed study quality. We assessed the risk of bias and quality of studies using the Cochrane’s risk of bias tool and the Medical Education Research Study Quality Instrument (MERSQI). Results A total of 3,471 citations were identified using the search strategies. After removing duplicates, 1,981 titles and abstracts were screened for their relevance in this review and 1,138 were selected for full-text screening. Following this, 14 met our eligibility criteria and were included in this review. Data extraction is currently underway. For quantitative synthesis, we will group studies according to the advanced SRT program investigated and pool the results of studies comparing skill retention. We will use a randomeffects model represented with forest plots, with risk ratios for dichotomous outcomes and standardised mean differences for continuous outcomes. If a meta-analysis is not appropriate, we will use a critical synthesis approach to examine the articles for common themes and summarize these findings. Conclusion This systematic review will provide a complete overview of the effectiveness, variability and key features of a range of different clinical advanced SRT programs designed for healthcare professionals. The results will help to suggest the optimal design features to improve resuscitation skill retention. References 1. Merchant R. Crit Care Med 2011; 39(11): 2401-2406 2. Wik L. JAMA 2005; 293(3):299-304. 3. Smith K. Resuscitation 2008; 78(1):59-65.