Evidence-Based Clinical Practice Guideline for the Treatment of Calcium Channel Blocker Poisoning

Maude St-Onge

A thesis submitted in conformity with the requirements for the degree of Doctor of Philosophy Institute of Medical Science University of Toronto

Maude St-Onge

Doctor of Philosophy

Institute of Medical Science University of Toronto

2015

Abstract

In the United States, calcium channel blockers were involved in at least 11,764 poisonings and 78 deaths in 2011. Unfortunately, there is an important practice variation and inconsistency in the management of this poisoning that may lead to a possible care gap when clinicians do not adhere to recommendations supported by the evidence. The goal of this thesis is to build an evidence-based guideline for the treatment of calcium channel blocker poisoning endorsed by international organizations. In doing so, this research may substantially improve consistency and adherence to recommendations by adapting the evidence outlined by our systematic review to costs, resources, values and preferences.

Following a review of the literature concerning the mechanisms of toxicity, the clinical presentation, the rational for the use of different interventions, and the theory of

ii guideline development, the research program detailed in this thesis stresses the need for a guideline by demonstrating the burden of illness in a retrospective study, and the important practice variation in a survey of emergency physicians. In order to develop an evidence-based guideline, the information required for decision making is then collected in a systematic review describing the quality of evidence for each intervention, and a cost-effectiveness analysis confirming that cost should not limit access to invasive strategy such as VA-ECMO in centers where the intervention is available.

An international group of experts in emergency medicine, critical care and toxicology developed recommendations for the treatment of calcium channel blocker poisoning based on the very low level of evidence, the balance between risks and benefits, costs and resources, and values and preferences. A rigourous and transparent methodology guided the working group members who were not biased by conflict of interest.

This thesis contributes to the development of recommendations for the in-hospital management of calcium channel blocker poisoning to improve consistency in messaging to the clinicians based on the poor level of evidence. It presents reflections on the process of guideline development in acute care toxicology and suggests future directions to consider.

iii

"In most cases, suicide is a solitary event and yet it has often far-reaching repercussions for many others. It is rather like throwing a stone into a pond; the ripples spread and spread." ALISO N WERTHEIMER, A Special Scar

This thesis is dedicated to my beloved friend, Christian Letendre, who took part of my soul when he took his own life. He taught me that hope is the essence of resilience and that scars can also make you stronger.

Acknowledgments

I would like to thank my thesis supervisor, Laurie Morisson, for her invaluable guidance, advice and feedback. I would also like to thank my thesis committee members, Dr Gordon Rubenfeld, Dr David Juurlink and Dr Tasnim Sinuff, who provided unwavering support and gave me the opportunity to learn from their great expertise.

I owe gratitude to my mentor, Dr René Blais, who has been an excellent role model and helped me always focus on the patients’ best interests. I am also extremely grateful to Mrs Hélène Quevillon, my “coach”, who taught me to go beyond my limits despite difficulties of life.

Numerous other advisors, colleagues and friends, impossible to enumerate here, contributed to the achievement of this thesis. I would like to thank all my colleagues at Sunnybrook Health Sciences Centre, notably Dr Robert Fowler for being an inspiring scientist, and Dr Damon Scales for being so supportive over all these years. I am grateful to Dr Margaret Thompson, my program director, for her patience, comprehension and encouragement. A heartfelt thank you is also extended to Dr Margaret Herridge and Dr Alexis Turgeon who have been a great source of motivation.

In addition, I would like to highlight the contribution of acutely poisoned patients and their families who have been and will always be a constant source of inspiration.

Finally, I owe my deepest gratitude to my family for believing in my dreams and aspirations, and my boyfriend, Miguel, for staying by my side. Thank you for your tireless encouragement.

Contributions

Maude St-Onge (author) solely prepared this thesis. All aspects of this body of work, including the planning, execution and writing of all original research and publications was performed in whole or in part by the author. The following contributions by other individuals are formally and inclusively acknowledged:

Dr Laurie J. Morrison (Supervisor and Thesis Committee Member) - mentorship; guidance and assistance in thesis preparation;

Dr Gordon Rubenfeld (Thesis Committee Member) - mentorship; guidance and assistance in thesis preparation;

Dr David N Juurlink (Thesis Committee Member) - mentorship; guidance and assistance in planning and execution of the systematic review, the recommendations development and manuscripts/thesis preparation;

Dr Tasnim Sinuff (Thesis Committee Member) - mentorship; guidance and participation in the recommendations development, as well as manuscript/thesis preparation;

Jennifer Butters - writing assistance;

Andrea Meeson and Linda Hurrell - assistance with grant application and manuscript writing;

Steve Fan Chun-Po - statistical guidance;

Dr René Blais - assistance in execution of studies for Chapters 4, 5 and 6;

Dr Patrick Archambault - assistance in execution of studies for Chapters 4 and 6;

Dr Natalie LeSage - assistance in execution of studies for Chapters 4 and 6;

Dr Chantal Guimont - assistance in execution of studies for Chapters 4 and 6;

Dr Julien Poitras - assistance in execution of studies for Chapters 4 and 6;

Dr Sophie Gosselin - assistance in execution of study for Chapter 6 and representative of the Canadian Association of Emergency Physicians for the recommendations development;

Dr Bruno Mégarbane - assistance in execution of study for Chapter 7 and representative of the European Society of Intensive Care Medicine for the recommendations development;

Other representatives assisting with recommendations development:

Dr Valéry Lavergne - expert methodologist for the recommendations development;

Dr Benoît Bailey – co-chair

Dr Martin Laliberté - representative of the Canadian Association of Poison Control Centres;

Dr John Muscedere - representative of the Canadian Critical Care Society;

Dr Michael Rieder - representative of the Canadian Paediatric Society;

Dr Éric Lavonas and Russ Kerns - representatives of the American College of Medical Toxicology;

Dr Lee Cantrell - representative of the American Association of Poison Control Centres;

Dr David Jang - representative of the American Academy of Clinical Toxicology;

Dr Ian Gilchrist - representative of the Society of Critical Care Medicine;

Dr Philippe Hantson - representative of the European Association of Poison Centres and Clinical Toxicologists;

Dr Kurt Anseeuw - representative of the European Society of Emergency Medicine;

vii

Dr Reza Afshari and Dr Chen-Chang Yang - representatives of the Asia Pacific Association of Medical Toxicology;

Other co-authors - assisting in execution of study for Chapter 6: Pierre-André Dubé, Dr Jesse Godwin, Dr Jean-Marc Chauny, Anne-Julie Frenette, Martin Darveau, Dr Joanne Provencher.

Other co-authors - assistance in execution of study for Chapters 7: Dr Eddy Fan, Peter Coyte, Rebecca Hancock-Howard.

viii

Table of Contents

ACKNOWLEDGMENTS p. v CONTRIBUTIONS p. vi TABLE OF CONTENTS p. ix LIST OF TABLES p. xiv LIST OF FIGURES p.xv LIST OF APPENDICES p.xvi LIST OF ABBREVIATIONS p.xvii

CHAPTER 1: GENERAL INTRODUCTION p.1 1.1 Preamble p.2 1.2 Thesis organization p.3

CHAPTER 2: LITERATURE REVIEW p.5 2.1 Calcium channel blocker poisoning p.6 2.1.1 Mechanism of toxicity and clinical presentation p.6 2.1.2 Rationale for the use of different interventions p.8 2.1.2.1 Observation and prognostication p.8 2.1.2.2 Decontamination p.9 2.1.2.3 Supportive treatments p.11 2.1.2.4 Antidotes p.11 2.2 Theory of guideline development p.14 2.2.2 Definition, benefits and harms, rigour of development p.14 2.2.2.1 Definition p.14 2.2.2.2 Benefits and harms p.15 2.2.2.3 Rigour of development p.16 2.2.3 Need for a guideline: disease burden, care gap and variation in practice p.18 2.2.4 Scope, analytic framework and key questions p.21 2.2.5 Development of recommendations p.25 2.2.5.1 Committee member selection and conflict of interests p.25 ix

2.2.5.2 The evidence p.29 2.2.5.3 Costs and resource use p.32 2.2.5.4 Values and preferences p.33 2.2.5.5 Consensus development method p.35

2.2.6 Tools and implementation strategies p.39 2.2.7 External review and updates p.41

CHAPTER 3: PROGRAM OF RESEARCH p.45 3.1 Aim p.46 3.2 Program of research p.46

CHAPTER 4: RETROSPECTIVE STUDY - Burden of calcium channel blocker poisoning and care gap p.49 4.1 Introduction p.50 4.2 Methods p.51 4.3 Results p.55 4.4 Discussion p.61 4.5 Conclusion p.64

CHAPTER 5: SURVEY - Variation in opinion and practice p.65 5.1 Introduction p.66 5.2 Methods p.67 5.3 Results p.72 5.4 Discussion p.78 5.5 Conclusion p.81

CHAPTER 6: SYSTEMATIC REVIEW - The evidence p.82 6.1 Introduction p.83 6.2 Methods p.84 6.3 Results p.87 6.3.1 Results of individual studies and risk of bias for medical interventions p.88

6.3.2 Results of individual studies and risk of bias for mechanical interventions p.107 6.3.3 Results of individual studies and risks of bias for interventions for which only small case series, case reports or animal studies are available p.108 6.3.4 Synthesis of results p.109 6.3.5 Risk of bias across studies p.113 6.4 Discussion p.115 6.5 Conclusion p.116

CHAPTER 7: COST-EFFECTIVENESS ANALYSIS - The costs p.117 7.1 Introduction p.118 7.2 Methods p.120 7.2.1 Effects p.122 7.2.2 Costs p.122 7.2.3 Likelihood of events p.127 7.2.4 Statistical analysis p.128 7.3 Results p.129 7.4 Discussion p.131 7.5 Conclusion p.134

CHAPTER 8: RECOMMENDATIONS FOR THE TREATMENT OF CALCIUM CHANNEL BLOCKER POISONING p.135 8.1 Introduction p.136 8.2 Methods p.136 8.2.1 Objective, scope, target users and analytical framework p.136 8.2.2 Recommendations development working group p.138 8.2.3 The evidence p.139 8.2.4 Definitions p.140 8.2.5 The recommendations statements p.141 8.2.6 Values and preferences p.143 8.2.7 Internal and external reviews p.143

8.3 Results p.144 8.3.1 Recommendations p.149 8.3.1.1 Therapy in asymptomatic patients p.149 8.3.1.2 First-line treatments for symptomatic patients p.149 8.3.1.3 Therapy for patients refractory to first-line treatments p.153 8.3.1.4 Therapy for patients in refractory shock or peri-arrest p.155 8.3.1.5 Therapy for patients in cardiac arrest p.156 8.3.1.6 Rationale for not recommending or not suggesting some treatments p.157 8.3.2 Values and preferences p.158 8.3.3 Internal and external reviews p.159 8.3.4 Implementation and applicability p.160 8.4 Discussion p.161 8.5 Conclusion p.163

CHAPTER 9: GENERAL DISCUSSION p.164 9.1 Summary of the research program p.165 9.2 Comparison to other guideline development processes p.168 9.3 Limitations related to these recommendations development p.170 9.4 Remaining issues related to these recommendations development p.174 9.4.1 Recommendations content p.174 9.4.2 Patients and public involvement p.174 9.4.3 Impact of the recommendations p.175 9.4.4 Ethical considerations p.176

CHAPTER 10: CONCLUSION p.179 10.1 Process to achieve the aim of this program of research p.180 10.2 Impact of recommendations development p.181

CHAPTER 11: FUTURE DIRECTIONS p.182 11.1 Future directions in cardiotoxicants poisoning p.183 11.2 Future directions in guideline development in acute care toxicology p.185

xii

11.3 Future directions in guideline implementation in acute care toxicology p.186 11.4 Conclusion p.186

xiii

List of Tables

Table 1 Consensus techniques p.36

Table 2 Baseline characteristics p.57

Table 3 Recommendations not followed p.58

Table 4 Outcomes p.60

Table 5 Characteristics of study participants (n=140) p.72

Table 6 Characteristics of the physicians not present at the meetings p.73

Table 7 Proposed decontamination strategies for Case 2 (less than one hour post-ingestion) p.74

Table 8 Interventions considered depending on training and practice setting p.76

Table 9 Resources reported as being unavailable across practice settings p.77

Table 10 Perceived factors influencing management p.78

Table 11 Results of individual studies and risk of bias p.89

Table 12 Risks of bias across studies for each intervention p.115

Table 13 Assumptions, sources and values for costs used in the analysis p.123

Table 14 Assumptions, sources and values for probabilities used in analysis p.128

Table 15 Participating organizations p.139

Table 16 Level of evidence and strength of recommendation p.142

Table 17 Recommendation table p.146

xiv

List of Figures

Figure 1 Action of calcium, high-dose insulin, lipid emulsion and L-carnitine p.12

Figure 2 Diagram of the guideline development process p.17

Figure 3 Analytic framework used for the development of recommendations for the treatment of calcium channel blocker poisoning. p.22

Figure 4 Possible approaches for considering comorbidities when building clinical practice guidelines. p.24

Figure 5 Voting process for recommendations p.37

Figure 6 Decision-making process on the type and scope of update p.43

Figure 7 Program of research p.48

Figure 8 Quebec Poison Control Centre protocol for the treatment of calcium channel blocker poisoning (2004) p.53

Figure 9 Flow diagram p.87

Figure 10 Simplified schema of the decision model p.121

Figure 11 Tornado diagram - probability variables p.130

Figure 12 Probabilistic sensitivity analysis p.131

Figure 13 Analytical framework for calcium channel blocker poisoning treatment recommendations p.138

Figure 14 Voting process for recommendations p.142

Figure 15 Progression of care for key recommendations p.145

List of Appendices

Appendix 1 Survey of emergency physicians p.220

Appendix 2 Case reports published focusing on an intervention or reporting a side effect p.227

xvi

List of Abbreviations

AACT American Academy of Clinical Toxicology

AAPCC American Association of Poison Control Centres

ACC American College of Cardiology

ACMT American College of Medical Toxicology

AHA American Hearth Association

AGREE Appraisal of Guidelines Research & Evaluation

AMSTAR Assessing the Methodological Quality of Systematic Reviews

ARRIVE Animal Research: Reporting of In Vivo Experiments

CAPCC Canadian Association of Poison Control Centres

CARE Case report guidelines

CCB Calcium Channel Blocker

CCFP(EM) Canadian College of Family Physicians - Certificate of special competence in emergency medicine

CI Confidence Interval

CIHR Canadian Institutes of Health Research

EAPCCT European Association of Poison Centres and Clinical Toxicologists

ECLS Extracorporeal Life Support

ED Emergency Department

ELSO Extracorporeal Life Support Organization

EXTRIP EXtracorporeal Treatments in Poisoning

FRCPC(EM) Fellow of the Royal College of Physicians of Canada in Emergency medicine

xvii

GRADE Grading of Recommendations Assessment, Development and Evaluation

HDI High-Dose Insulin

ICER Incremental Cost-Effectiveness Ratio

ICMJE International Committee of Medical Journal Editors

ICU Intensive Care Unit

LOS Length of Stay

LY Life Year

MOHLTC Ministry of Health and Long-Term Care

NACCT North American Conference on Clinical Toxicology

NRCNA National Research Council of National Academy

OR Odds-Ratios

PAD Pain, Agitation and Delirium

PCC Poison Control Center

PRISMA Preferred Reporting Items for Systematic Reviews and Meta-Analyses

QALY Quality-Adjusted Life Year

SOB Schedule of Benefits

STROBE Strengthening the Reporting of Observational Studies in Epidemiology

ToxIC Toxicology Investigator’s Consortium

VA-ECMO Venoarterial Membrane Oxygenation

VV-EVMO Venovenous Membrane Oxygenation

WBI Whole bowel irrigation

xviii

______

Chapter 1 General Introduction

______

1.1 Preamble

Canadian hospitals report 121-331 poisonings per 100,000 person-years to the poison control centres (Poison and Drug Information Service, 2012; Centre antipoison du

Québec, 2009). American poison control centres report cardiovascular drugs as the substance category with the third-fastest rate of increased exposures (Bronstein et al.,

2012). According to the National Poison Data System in United States, calcium channel blockers were responsible for at least 11,764 exposures and 78 deaths in 2011

(Bronstein et al., 2012). Unfortunately, current recommendations are based only on expert opinions from toxicologists at poison control centres, fail to take costs into account and do not involve knowledge users, such as emergency medicine and critical care physicians, in the decision-making process. Following the Evidence Informed

Decision-Making Process Algorithm (Canadian Institutes of Health Research, 2013), the current practice, the quality of evidence and the costs, and the values and preferences of knowledge users should be considered when elaborating guidelines.

Using calcium channel blocker poisoning as an example, the goal of this thesis is to build an evidence-based guideline in acute care toxicology that will be endorsed by international critical care, emergency medicine and toxicology associations. In doing so, this research will substantially create consistency in messaging to the clinicians and improve the adherence to recommendations by adapting the evidences outlined by our systematic review to costs, resources, values and preferences.

1.2 Thesis organization

This thesis is organized in a multiple paper format rather than a traditional design. This structure reflects the different steps required to build recommendations for the treatment of calcium channel blocker poisoning. Chapter 2 primarily serves to review the literature concerning the mechanisms of toxicity, the clinical presentation and the rationale for the use of different interventions for this poisoning. It also describes the theory and methods of guideline development that were used in this thesis. This will be the framework for the research program detailed in Chapter 3.

Chapters 4,5,6 and 7 are modified from peer-reviewed manuscripts leading to the development of recommendations. Chapters 4 and 5 stress the need for a clinical practice guideline. Chapter 4 is a retrospective study that documents the burden of calcium channel blocker poisoning and illustrates inconsistency with the current recommendations. Chapter 5, a survey of emergency physicians, demonstrates the variation in opinions regarding the management of this poisoning. This chapter also provides information related to resource availability which, when integrated with the evidence documented by the systematic review (Chapter 6) and costs (Chapter 7), constitutes the basis for guideline development. The cost-effectiveness analysis detailed in Chapter 7 focuses on extracorporeal life-support, the most expensive yet not widely available therapy for which there is at least a minimal amount of evidence. It is important to determine the incremental cost-effectiveness ratio for the use of this intervention considering that it could be cited as a reason not to use it in cardiotoxic poisonings.

Chapter 8 constitutes the recommendations for the treatment of calcium channel blocker poisoning built to create consistency in messaging to the clinicians based on the poor level of evidence outlined in the systematic review. The general discussion in Chapter 9 summarizes the process followed to develop those recommendations, compares it to other guidelines, details the limitations and identifies the remaining issues related to knowledge creation and translation in the treatment of calcium channel blocker poisoning. After illustrating the potential impact and contribution of this thesis in Chapter

10 (Conclusion), Chapter 11 describes future studies that will either address the remaining issues or continue to reduce the current gaps in knowledge.

______

Chapter 2 Literature Review

______

Section 2.1 of this chapter is modified from the following: St-Onge M (2014). Chapter:

Calcium channel blocker poisoning in the book "Hantson & Mégarbane, Poisoning with cardiotoxicants" to be published by Springer in 2015.

The following chapter will first review the literature concerning the mechanisms of toxicity, the clinical presentation and the rationale for the use of different interventions for this poisoning (section 2.1). It will then describe the theory and methods of guideline development that were used in this thesis (section 2.2).

2.1 Calcium channel blocker poisoning

2.1.1 Mechanism of toxicity and clinical presentation

Calcium channel blockers were introduced in the 1960s as coronary vasodilators

(Knoch et al., 1963). Verapamil was recognized for the negative inotropic and chronotropic effects that distinguished it from other coronary vasodilators available at that time. By the 1970s, the first cases of calcium channel blocker poisoning had already been reported (Beitzke et al., 1976). Since then, many other calcium channel blockers have been released and used for indications, such as hypertension, atrial fibrillation, and the prevention of cerebral vasospasm prevention and angina.

Calcium channel blockers are divided into two categories: dihydropyridines, such as nifedipine; and non-dihydropyridines, such as diltiazem (a benzothiazepine) and verapamil (a phenylalkylamine). Dihydropyridines preferentially block calcium channels in vascular smooth muscles. Tachycardia may result from reflex sympathetic activation secondary to vasodilation. Non-dihydropyridines generally slow the heart rate by inhibiting the sinoatrial and atrioventricular nodes (Brunton et al., 2008).

The main therapeutic target of calcium channel blockers is the L-type calcium channel.

Other calcium channels, the N-type, T-type, P/Q-type and R-type, play a role mainly in neurons and secretory glands (Tsien et al., 2000). The extent to which these other channels are less receptive to calcium channel blockers is not clear. Dihydropyridines have less affinity for the myocardial calcium channels, but as the dosing increases, the selectivity of the calcium channel blocker declines (Pfeffer et al., 2014). Most calcium channel blockers are substantially bound to plasma proteins and are extensively metabolized. Verapamil and nifedipine are eliminated in the urine; nifedipine is also metabolized to lactic acid. The oral bioavailability, onset of action and plasma half-life vary among the calcium channel blockers (Katzung et al., 2004).

In calcium channel blocker poisoning, the excessive blockade of L-type calcium channels in myocardial and vascular smooth-muscle cell membranes causes hypotension or shock that is secondary to peripheral vasodilation and cardiac dysfunction, such as bradycardia, conduction delay, and negative inotropy (Lheureux et al., 2006). Calcium signaling is also critical to cardiac carbohydrate metabolism. During calcium channel blocker poisoning, the heart switches its energy substrate from the preferred free fatty acids to carbohydrates. The liver subsequently increases glycogenolysis in order to provide enough circulating glucose. However, the calcium channel blocker prevents the pancreatic calcium-mediated insulin release that is necessary for the utilization of glucose by myocardial cells. This results in insulin deficiency, hyperglycaemia and acidemia (Kerns II et al., 2006).

It is difficult to estimate the half-life of calcium channel blockers in an overdose situation, since the pharmacokinetics may be altered by coingestants, the presence of an ileus or decreased gut perfusion. Bezoars have been reported to occur in patients taking calcium channel blockers, notably with extended release nifedipine (Jarry et al., 2008;

Wells et al., 2006).

2.1.2 Rationale for the use of different interventions

2.1.2.1 Observation and prognostication

The evidence underlying the use of each intervention will be detailed in Chapter 6 -

Systematic Review. However, with respect to observing asymptomatic patients, cases with voluntary ingestions have a higher risk of developing toxicity than cases with involuntary ingestions (Truitt et al., 2012). Unfortunately, the criteria for safe observation at home of patients with involuntary ingestion are not well defined.

In order to establish which patients with involuntary ingestions of calcium channel blockers and beta-blockers developed symptoms, Truitt et al (2012) conducted a retrospective review of all American poison control centre charts from January 2007 through December 2009. In the 436 cases reviewed, 10.1% of patients developed symptoms and 7.3% were admitted to hospital. Of note, patients who ingested someone else’s medication were excluded from the analysis.

In terms of prognostication, Levine et al. (2007) conducted a multicentre retrospective study looking at the use of the bedside glucose level as a mortality predictor in adults 8

poisoned with verapamil or diltiazem. For patients who met the composite endpoint

(death during the index hospitalization, need for a temporary pacemaker, need for vasoactive agents), the median initial serum glucose concentration was 10.4 mmol/L

(interquartile range, 7.9–15.0 mmol/L). For those who did not meet the composite endpoint, the value was 7.2 mmol/L (interquartile range, 5.5–8.7 mmol/L; p < 0.0058).

Although their retrospective study was limited by an inability to estimate the time from overdose to emergency department arrival, and by a broad reported range of median glucose increases (Mycyk et al., 2007), the authors demonstrated a clear association between the serum glucose level and the severity of calcium channel blocker poisoning.

In 2011, Mégarbane et al. published an observational study evaluating the ability of clinical and laboratory parameters to predict mortality in verapamil poisoning. A drug concentration higher than 5.0 µM, measured on intensive care unit admission, was the only independent factor associated with mortality (odds of fatality increased 2.76 times; p = 0.01; sensitivity, 100%; specificity, 91%). Unfortunately, serum drug levels are not always available.

2.1.2.2 Decontamination

Decontamination is defined as the removal of a dirty or dangerous substance from a person (Merriam-Webster, 2014). Only gastrointestinal decontamination is relevant to calcium channel blocker poisoning. It includes induced emesis, gastric lavage and activated charcoal, which prevent absorption, or whole bowel irrigation which favours intestinal evacuation (Nelson et al., 2011). Renal replacement therapies, plasma exchange or repeat-dose activated charcoal, for substances known to have

enterohepatic circulation, are not considered decontamination methods. Rather, they are used to enhance elimination of a substance already absorbed (Olson, 2012).

Few decontamination methods have been studied specifically in the context of calcium channel blocker poisoning. Those that have will be detailed in Chapter 6 - Systematic

Review. However, the results from other decontamination studies conducted in participants who were not poisoned may inform the management of poisoned patients.

In healthy volunteers, a study of 32 patients (Laine et al., 1997) showed that the administration of activated charcoal immediately after amlodipine reduced the 72 h urinary excretion by 99% (p < 0.0005). After a delay of 2h, excretion was reduced by

49% (p = 0.001), but after a delay of 6h, the reduction was only 15%.

In addition, Lapatto-Reiniluoto et al. (2000) completed a prospective study in nine healthy volunteers studying the efficacy of activated charcoal and gastric lavage, 30 minutes after ingestion, in preventing the absorption of verapamil. They found a 16% reduction in the peak plasma drug concentration with activated charcoal but no effect with gastric lavage. In 2001, the same group (Lapatto-Reiniluoto et al., 2001) published another study using nine healthy volunteers to evaluate the effect of activated charcoal on the absorption of sustained-release medications (carbamazepine, theophylline and verapamil) ingested one hour earlier. The concentration of verapamil decreased significantly with activated charcoal, by 15 ± 20 ng/mL (p < 0.01), but the combination of activated charcoal with whole bowel irrigation decreased the effect, giving a reduction of only 8.3 ± 5.1 ng/mL (p < 0.001).

2.1.2.3 Supportive treatments

Supportive treatments are used "to sustain the condition of a patient" (Merriam-Webster,

2014) without counteracting the effects of a poison. These treatments can be medical

(fluids, atropine, vasopressors, inotropes) or mechanical (pacemaker, intra-aortic balloon pump, extracorporeal life-support).

In terms of medical interventions, calcium channel blocker poisoning seems to have a variable or transient response to atropine (Ramoska et al., 1993) and the use of vasopressors alone in these patients is controversial (Pfeffer et al., 2014). However, phosphodiesterase inhibitors are positive inotropes, believed to increase levels of cyclic adenosine monophosphate, allowing more calcium to enter cells across membrane channels and resulting in increased muscle contractility (Koury et al., 1996).

Different forms of mechanical interventions, such as pacemakers or intra-aortic balloon pumps, have also been proposed to support the hemodynamics of patients poisoned with a calcium channel blocker. However, these supportive measures do not seem to be sufficient (St-Onge et al., 2014). Therefore, extracorporeal life-support has been identified as a more invasive therapy. Currently extracorporeal life-support has only been offered to European patients, no cases have been reported in North America (St-

Onge et al., 2014).

2.1.2.4 Antidotes

An antidote is defined as: "a remedy to counteract the effects of poison" (Merriam-

Webster, 2014). Intravenous calcium and high-dose insulin are among the antidotes

recommended for calcium channel blocker poisoning (see Figure 1). Intravenous calcium has been widely used. The increased concentration of extracellular calcium provided by this treatment is thought to facilitate cellular calcium entry (Zhou et al.,

2013).

Concerning high-dose insulin, three mechanisms of action have been proposed: 1) it increases inotropy by enhancing coronary blood flow (without increasing oxygen requirements) and by improving calcium handling and glucose transport; 2) it increases intracellular glucose transport by affecting phosphatidylinositide 3-kinase (a major insulin intracellular signaling pathway) and; 3) it decreases vascular resistance by increasing endothelial nitric oxide synthase activity (St-Onge et al., 2013).

Figure 1 Action of calcium, high-dose insulin, lipid emulsion and L-carnitine Reproduced with permission from: St-Onge et al., 2013 12

Glucagon and lipid emulsion are tempting complements to intravenous calcium and high-dose insulin therapy. The proposed use of glucagon in calcium channel blocker poisoning is largely based on its effectiveness in beta-blocker overdose (Stone et al.,

1995). This treatment is thought to activate a secondary messenger, the G-protein, which bypasses the beta receptor (Pfeffer et al., 2014). With respect to the use of lipid emulsion, it is hypothesized that this creates a separate intravascular compartment within which lipophilic drugs are sequestered from their sites of action (Tebbutt et al.,

2006). This treatment also provides fatty acids that may be used by the myocardium and may increase intracellular calcium (Pfeffer et al., 2014).

Other antidotes have been suggested for the treatment of calcium channel blocker poisoning, notably L-carnitine, methylene blue and calcium sensitizers (Levosimendan,

4-aminopyridine, 3,4-DAP, Bay K8644 and CGP 28932) (Graudins et al., 2010;

Magdalan et al., 2003; Tuncok et al., 1998; Plewa et al., 1994; Korstanje et al., 1987;

Gay et al., 1986; Agoston et al., 1984). L-carnitine was initially proposed by Perez et al.

(2011) as a potential antidote for verapamil toxicity. It is thought to decrease insulin resistance, promote intracellular glucose transport, facilitate the metabolism of free fatty acids, and increase calcium channel sensitivity (St-Onge et al., 2013). Others suggest a potential complementary action of high-dose insulin, lipid emulsion and L-carnitine (St-

Onge et al., 2013). Further, methylene blue has been used as a treatment of amlodipine poisoning (Aggarwal et al., 2013; Jang et al., 2011). The use of this antidote is based on its capacity to inhibit the release of nitric oxide, which causes vasodilation.

With all of the possible options considered to counteract toxicity in calcium channel blocker poisoning, it is difficult to prioritize the interventions without a rigourous approach, such as the one recommended by the AGREE Consortium (Brouwers et al.,

2010) or the Institute of Medicine (Graham et al., 2011). The following section of this chapter will detail the methodology of guideline development.

2.2 Theory of guideline development

2.2.2 Definition, benefits and harms and rigour of development

2.2.2.1 Definition

Graham et al. (2006) developed the knowledge-to-action process to represent how knowledge is created, disseminated and implemented to improve the quality of care.

Knowledge creation comprises knowledge inquiry (e.g. primary studies), knowledge synthesis (e.g. systematic reviews) and knowledge tools/products. Knowledge tools, such as guidelines (Davis et al., 2007), are concise and usable messages that are intended to influence behaviors.

Guidelines are defined as "systematically developed statements to assist practitioner and patient decisions about appropriate healthcare for specific clinical circumstances"

(Field et al., 1990). Guidelines are developed with the intent to: improve quality of care and outcomes, reduce inappropriate variation in practice, summarize evidence, favour an efficient resource use, make clinical decisions more transparent, identify areas of knowledge for which research should be prioritized, empower patients, inform public

policy and support quality control activities (Davis et al., 2007). Guideline implementation can provide interesting benefits, but is not without risks.

2.2.2.2 Benefits and harms

Woolf et al. (1999) identified potential benefits of guidelines for patients, healthcare professionals, medical researchers and healthcare systems. For the patients, one important benefit would be the promotion of therapies that have the potential for better outcomes. Clear guidelines can also improve consistency of care and may favour patient and family participation in the decision-making process if versions are available for the lay public. By participating in healthcare decisions, patients can advocate for a better delivery of services and can influence public policy. For healthcare professionals, guidelines clarify which interventions are better supported by the evidence, help to explain the risks and benefits to the healthcare consumer, and can constitute a point of reference for quality improvement activities. For medical researchers, guidelines can emphasize gaps in knowledge that warrant prioritization. Moreover, healthcare systems may benefit from guidelines that highlight cost-effective interventions. Finally, promoting adherence to guidelines may improve public image.

In spite of numerous benefits, if evidence is misleading, misinterpreted or if the patients' needs are not prioritized, guidelines also have the potential to harm patients, healthcare professionals, medical researchers or the healthcare systems (Woolf et al., 1999). First, patients can be harmed if guidelines are inflexible and do not allow clinicians to tailor care to specific circumstances. Furthermore, guidelines are influenced by the opinions and experiences of development committee members. For example, the members may

be distracted from focusing on patients' needs if they prefer to build their recommendations based on costs. On the one hand, healthcare consumer coverage or access to treatments that have been discouraged, even if they are effective, may be difficult. On the other hand, imprudent recommendations for expensive interventions may limit resources for other services of great value. Second, healthcare professionals can be affected if the recommendations are wrong, outdated or in contradiction with other recommendations. This may result in the unfair judgment of clinicians. Even if a guideline is not considered to be a legal document, it may be used by medical experts as citable evidence for malpractice litigation. Third, if further research is inappropriately discouraged this may negatively impact medical researchers as well. Finally, the healthcare system may also be compromised if guideline implementation generates a waste of resources.

2.2.2.3 Rigour of development

Recognizing the potential for harm highlights the importance of following a rigorous, transparent and credible guideline development process. In 2000, Graham et al. compared different practice guideline appraisal instruments and identified 15 containing eight to 142 questions/statements. The authors grouped the statements into 10 important attributes to assess the methodological rigor for guideline development: validity, reliability/reproducibility, clinical applicability, clinical flexibility, multidisciplinary process, clarity, schedule review, dissemination, implementation and evaluation. Each instrument was independently examined. Only two were validated. The Cluzeau's instrument (1999) received the best evaluation and constituted the basis of the

Appraisal of Guidelines, Research and Evaluation in Europe (AGREE II) instrument

(Brouwers et al., 2010). The Shaneyfelt's instrument (1999) was the only other tool validated.

More recently, Schünemann et al., (2013) systematically compiled a comprehensive checklist of items that guideline development tools could consider. This checklist included 18 topics and 146 items to serve as a resource for guideline developers. The topics listed the different development steps and are represented in a diagram (see

Figure 2) to illustrate both how our guideline was built (Chapter 8 - Guideline development) and its limitations (Chapter 9 - General discussion).

Figure 2 Diagram of the guideline development process Reproduced with permission from: Schünemann et al., 2013

The validated AGREE II (Brouwers et al., 2010) and Institute of Medicine (Graham et al., 2011) instruments were used to assess and review the calcium channel blocker poisoning treatment guideline. The following sections in this chapter will describe in greater detail the guideline development methodology in several reports, such as the

Proceedings of the American Thoracic Society Series (Atkins et al, 2012; Boyd et al.,

2012; Burgers et al., 2012; Cluzeau et al., 2012; Fabri et al., 2012; Grimshaw et al.,

2012; Guyatt et al., 2012; Hill et al., 2012; Kelson et al., 2012; Kunz et al., 2012;

Schünemann et al., 2012; Wilson et al., 2012; Yawn et al., 2012), the Implementation

Science Series (Eccles et al., 2012; Woolf et al., 2012; Shekelle et al., 2012), the

Guidelines International Network (Qaseem et al., 2012) and the ADAPTE toolkit

(ADAPTE Collaboration, 2009) for guideline adaptation.

2.2.3 Need for a guideline: disease burden, care gap and variation in practice

In 2012, Atkins et al. conducted a literature review to identify the criteria for whether a topic should be considered for guideline development. Although they found only a few studies that directly addressed this issue, they did identify a substantial consensus about the general factors that should be considered. The Institute of Medicine (Graham et al., 2011) outlined six criteria: prevalence, illness burden, practice variation, potential to improve outcomes (care gap), management costs and cost reduction. Additional criteria were also mentioned, including clinical uncertainty, availability of adequate evidence, practitioner interest and cost for guideline development (Atkins et al., 2012).

To document the prevalence and burden of a given illness, Atkins et al. (2012) recommended the use of national data or review articles. Since not all physicians contact the poison control centre when treating a patient with calcium channel blocker poisoning (St-Onge et al., 2012), a retrospective study based on hospital data from two

Canadian cities was conducted to accurately document the prevalence and burden of this poisoning (St-Onge et al., 2012). A problem in consistency and adherence to recommendations was underlined by the same study.

Atkins et al. (2012) suggested using surveys and expert opinions to evaluate variations in practice, which is an approach we adopted in our recommendations development process. Our survey (St-Onge et al., 2014) highlighted a variation in opinions regarding the management of calcium channel blocker poisonings and specifically assessed the impact of training and practice setting on the management approach chosen in different clinical scenarios. The variation in expert opinion can be observed most notably with the use of extracorporeal life-support in cardiotoxic poisonings that were reported in Europe, but not in North America. Also, a study published in 2013 by Levine et al. stressed that not all toxicologists would choose to initiate early high-dose insulin therapy instead of supporting hemodynamics mainly with vasopressors. Following that publication, Pfeffer et al., (2014) asked: "Should we reprioritize our potpourri of treatment options?" This query certainly reflects the presence of clinical uncertainties.

To document the availability of evidence, Atkins et al. (2012) recommended consideration of existing reviews and to perform preliminary literature searches. To

date, neither has been done for calcium channel blocker poisoning. Although a preliminary search identified some observational studies, these included no information concerning cost. Therefore, our recommendations development process included a systematic review of interventions to be considered for a patient poisoned with a calcium channel blocker and a cost-effectiveness analysis of the intervention that could be perceived to be the most expensive: the extracorporeal life-support.

Atkins et al. (2012) proposed consultation with ad hoc stakeholders to evaluate practitioner interest. The manuscript, published by Pfeffer et al. (2014) in the Annals of

Emergency Medicine, stressed the interest in calcium channel blocker poisoning for the toxicology and emergency medicine communities. For the critical care community, increased use of extracorporeal life-support in critically ill patients is seen as a bridge to recovery, transplant or permanent assist device, which raises the question of its use in poisoned patients (DeLange et al., 2013).

Nonetheless, the question remains whether observational studies, case series, animal studies and case reports are sufficient to develop a clinical practice guideline. In the

EXTRIP (Extracorporeal Treatments in Poisoning) guideline methodology (Lavergne et al., 2012), currently the only guideline methodology published in toxicology, the authors mention:

The guidelines are intended to rely on evidence whenever possible. However, co- chairs acknowledged at the start that the available evidence was likely to be of low quality in most instances. (...) extracorporeal therapies have been recognized as the gold-standard treatment of several specific poisonings (...) element of standard care in the management of poisonings and clinicians support its use overwhelmingly. In such instances a randomized trial would not be feasible or ethical. (...) Rather than await the arrival of new data, the EXTRIP workgroup has elected to proceed with the systematic identification, review, and critique of all 20

available evidence using the GRADE system. (...) When the evidence is incomplete, opinion-based arguments will be proposed following transparent methodology.

The committee developing the recommendations for treatment of calcium channel blocker poisoning faces similar challenges. In such circumstances, the transparency of the consensus process is of primary importance and the benefits of making recommendations (i.e. decreased practice variation, facilitation of future research, decreased costs) should outweigh the risk of harm (i.e. recommending a harmful intervention).

2.2.4 Scope, analytic framework and key questions

The United States Preventive Task Force (Harris et al, 2001) integrated pieces of heterogeneous evidence by first defining the guideline scope and then the analytical framework. The scope of a guideline describes the types of services, the populations of patients and providers and the sites for which the recommendations are intended. For example, the scope of the guideline for the treatment of calcium channel blocker poisoning addresses what type of intervention should be recommended for patients

(adult and pediatric) poisoned with a calcium channel blocker who present to a hospital.

Different options, benefits and harms of all interventions are also considered. Targeted users are bedside treating physicians, consultants (at the bedside or via phone) and the poison control centres.

The analytic framework includes a diagram representing the causal pathways (see

Figure 3). The arrows illustrate questions that evidence must answer, the dotted lines 21

represent associations and the curved arrows indicate adverse events or harms.

Rectangles with rounded corners represent the intermediate outcomes and rectangles with square corners represent health states. Each linkage may be supported by a different level of evidence. Key questions are indicated by numbers. Figure 3 illustrates the analytic framework for the calcium channel blocker treatment recommendations.

Key questions: 1. Is there direct evidence that one (or more than one) intervention reduces mortality, improves functional outcomes, reduces hospital length of stay or reduces intensive care unit length of stay? 2. Does the patient clinical presentation or type of ingestion influence the intervention(s) provided and the outcomes? 3. Does one (or more than one) intervention decrease calcium channel blocker serum concentration, improve hemodynamics or reduce the duration of vasopressor use? 4. Are the intermediate outcomes reliably associated with reduced mortality or improved functional outcomes? 5. Does one (or more than one) intervention(s) result in adverse effects or demonstrate a lack of cost- effectiveness?

Figure 3 Analytic framework used for the development of recommendations for the treatment of calcium channel blocker poisoning Source: St-Onge et al., to be submitted.

The analytic framework represents complex interrelationships between the interventions, adverse effects, intermediate outcomes and health outcomes. Woolf et al.

(2012) noted that "health outcomes refers to direct measures of health status, including 22

measures of physical morbidity (e.g. dyspnea, blindness, weakness), emotional well- being and mortality (e.g. survival, life expectancy). Health outcomes can be also perceived from a systemic perspective (e.g. length of stay). The same authors define an intermediate outcome as "an effect that leads to a health outcome". This term is distinguished from surrogate outcome, which represents "an effect that is equivalent to a health outcome or can act as its proxy". As such, the analytic framework highlights the important outcomes to be evaluated in order to establish the effectiveness of an intervention, and guides the identification of research priorities to direct scientists' work toward the fundamental questions to be answered for a specific population.

Some authors have raised concerns that guidelines may not properly consider the care of patients with comorbidities (Uhlig et al, 2014; Pronovost, 2013). Therefore, investigators from the Improving Guidelines for Multimorbid Patients Study Group established consensus-based recommendations for each step of guideline development to address this issue.

As a first step, guideline developers should consider how disease-disease, disease- treatment and treatment-treatment interactions impact clinical management and outcomes (see Figure 4). They should then decide whether to (a) focus on an index condition with nuances addressed for specific comorbidities, or (b) choose to address how some marker of overall morbidity affects the management of the index condition.

For example, for the developement of a calcium channel blocker poisoning treatment guideline, stakeholders decided to address how management would change in patients with different comorbidities or coingestions. Other options suggest a focus on: (c)

specific combination of conditions, or (d) multimorbidity. For example, a guideline focusing on polypharmacy would probably adopt the approach illustrated in "d".

Figure 4 Possible approaches for considering comorbidities when building clinical practice guidelines a) Focus on an index condition, and choose to address some number of comorbid conditions; b) Focus on an index condition, and choose to address how some marker of overall morbidity affects the management of the index condition; c) Focus on specific combinations of conditions; and d) Focus on multimorbitidy. Reproduced with permission from: Uhlig et al., 2014

Thereafter, the panel of experts should include stakeholders with experience of managing the relevant patients groups. They should consider the impact of relevant coexisting conditions in the formulation of all key questions and components, or in the selection of outcomes. The search strategy should include studies done with subjects having a combination of conditions, or a specific search for relevant subgroups should

be performed. The stakeholders should then assess how the inclusion of comorbidities may modify the directness of evidence. Finally, when building recommendations, the panel should consider to what degree coexisting conditions will influence the balance between the risks and benefits of an intervention. They may then modify the strength of a recommendation for specific subgroups.

In calcium channel blocker poisonings, a significant proportion of patients have comorbidities and coingestions. Many patients are known to have hypertension, cardiac disease, or suffer from a mental illness (St-Onge et al., 2012). The National Poison Data

System reports that, within the last decade, the percentage of exposures suspected to be suicides ranged from 30.3 to 53.9% (Mowry et al., 2013). In calcium channel blocker poisoning, 80% of cases were reported as being voluntary ingestions (St-Onge et al.,

2012). Coingestions were frequent – notably with benzodiazepines (39%), other antihypertensive drugs (32%), ethanol (30%), antidepressants (24%) or beta-blockers

(13%) (St-Onge et al., 2012). Therefore, for each recommendation, the committee to develop the calcium channel blocker treatment recommendations decided to consider how comorbidities or coingestions would change the statement.

2.2.5 Development of recommendations

2.2.5.1 Committee member selection and conflict of interests

The scope, analytic framework and key questions are central to the guideline development process. However, guideline credibility depends largely on the composition of the expert panel and whether they would be affected by potential conflicts of interest. 25

The Canadian Medical Association (Davis et al., 2007) and the Institute of Medicine

(Graham et al., 2011) suggests including in the working group a facilitator, representatives from professional groups that would be significantly affected by the guideline, clinicians, methodological experts, patient representatives, policy makers and an administrative assistant. However, Kunz et al., (2012) suggest a target of 15 members or less to avoid endless debates and complicated decision-making processes.

The guideline organizers should try to find a balance between members with expertise, who might be more likely to be subject to bias, and non experts, who may be less biased but also less knowledgeable (Davis et al., 2007). Funders should have no role in the guideline development and chairs or co-chairs should not have any conflict of interest (Graham et al., 2011).

Boyd et al. (2012) define a conflict of interest as "when an individual’s personal interests

(e.g., direct and indirect financial or intellectual) have the potential to compete with or influence behavior related to the individual’s professional interests or obligations (i.e., evaluating the evidence and drafting recommendations for clinical practice guidelines)".

Unfortunately, guidelines are not always as transparent as they should be. Choudhry et al. (2002) published the results of a cross-sectional survey of 192 authors of 44 guidelines endorsed by North American and European societies between 1991 and July

1999. On average, 81% of the guidelines authors had some form of interaction with drug manufacturers. Fifty-nine percent had relationships with companies that commercialized one of the drugs being considered for the guideline. Although only 7% of authors perceived that their interactions influenced the recommendations, 19% of their co- authors thought it did. Moreover, Taylor and Giles (2005) reported that half of the

guidelines in the US National Guideline Clearinghouse database reported no information regarding funding sources or financial conflict of interest.

Consequently, Guyatt et al. (2010) proposed an approach to address conflicts of interest. First, they suggest that equal emphasis be placed on intellectual and financial conflicts and explicit criteria should be provided for both. Second, they recommend that the methodologists in charge of each question or subject should not be conflicted.

Finally, they emphasize that experts with important financial or intellectual conflicts of interest can collect and interpret evidence, but not develop a recommendation for a specific question. Boyd et al. (2012) proposed a matrix for considering perception related to both commercial sponsorship and conflicts of interest. When there is both commercial sponsorship and conflicted members, the risk is perceived as high and that member should be excluded from the process. When the same conflicted member is involved without commercial sponsorship, the risk is considered to be moderate. They suggest that all processes of guideline development be documented carefully to ensure that other committee members will balance the conflicted member's views. When there are no conflicted members, but commercial involvement is present, the risk is still moderate, but identification of alternative funding sources is recommended. The development of a guideline for the treatment of calcium channel blocker poisoning did not involve commercial sponsorship or conflicted members, which is considered ideal.

To address conflicts of interest (direct and indirect, financial and intellectual), Boyd et al.

(2012) suggest that each committee member provide a self-disclosure at the beginning of the collaboration. The organizers can then decide if the member can participate with

or without direct involvement in building the recommendations. During guideline development, the organizers should make the process transparent and for each recommendation consider whether the individuals should abstain from discussion and/or voting. Further, Boyd et al. (2012) proposed that a policy for conflict of interest management should be in place and known by the committee members. For the guideline development for the treatment of calcium channel blocker poisoning, all stakeholders completed the International Committee of Medical Journal Editors (ICMJE) conflict of interest form, which is requested by more than 500 journals (ICMJE, 2014;

Boyd et al., 2012). The information was available to all committee members. The organizers ensured that the stakeholders involved in the process would build the recommendations and vote without being conflicted.

Lenzer et al. (2013) identified red flags for knowledge users that should raise suspicion of bias that could potentially be introduced as a result of commercial sponsorship or conflict of interest:

• Sponsor(s) is a professional society that receives substantial industry funding; • Sponsor is a proprietary company, or is undeclared or hidden; • Committee chair(s) have any financial conflict; • Multiple panel members have any financial conflict; • Any suggestion of committee stacking that would pre-ordain a recommendation regarding a controversial topic; • No or limited involvement of an expert in methodology in the evaluation of evidence; • No external review; • No inclusion of non-physician experts/patient representative/community stakeholders.

These red flags were also considered when knowledge users assessed the recommendations for the treatment of calcium channel blocker poisoning.

2.2.5.2 The evidence

After defining the scope, guideline purpose and specification of the modalities related to stakeholders involvement, the AGREE collaboration focused on the rigor of development (Brouwers et al., 2010). Within that domain, the authors stressed the importance of using systematic methods to search for evidence related to each key question of the analytical framework, clearly described the criteria for selecting the evidence, and provided a detailed description of the strengths and limitations of the body of evidence. If a systematic review has already been published, the guideline developers should appraise its quality based on the Assessment of Multiple Systematic

Reviews (AMSTAR) (Shea et al., 2007). They can also assess reporting using the

Preferred Reporting Items for Systematic reviews and Meta-analysis (PRISMA) (Moher et al., 2009). In circumstances where there are no available systematic reviews, where published studies do not directly address the key question, or where the quality of available systematic reviews is too poor, guideline developers should perform a new synthesis of the evidence.

When a new systematic review is required, the American Heart Association Clinical

Practice Guideline Methodology Group (Jacob et al., 2013) recommends incorporating the Grading of Recommendations Assessment, Development and Evaluation (GRADE) methodology to evaluate the quality of evidence for each outcome across studies

(Guyatt et al., 2011). The quality can be evaluated as: a) high, when there is confidence that the true effect lies close to that of the estimate effect (as often occurs for randomized trials); b) moderate, when the true effect is likely to be close to the estimate of the effect but there is a possibility that it is substantially different; c) low, when the

true effect may be substantially different (as is often the case for observational studies); and d) very low, when there is little confidence in the effect estimate (Balshem et al.,

2011). Ratings are modified downward when there are study limitations, imprecision, result inconsistencies, indirectness of evidence or when publication bias is likely.

Ratings are modified upward when there is a large magnitude of effect, a dose response effect or there are confounders that are likely to minimize the effect (Balshem et al.,

2011).

The evaluation of the risk of bias is therefore crucial to assess the evidence quality.

GRADE suggests using the Cochrane risk of bias tool for randomized trials and observational studies (Guyatt et al., 2011). The STROBE checklist (Von Elm et al.,

2008) can also be used to evaluate the quality of observational study reporting as can the Thomas' tool built by the National Collaborating Centre for Methods and Tools

(2008). Although the Newcastle Ottawa Scale can evaluate the quality, Hartling et al.

(2013) raised concerns about the variable agreement and the lack of evidence that this tool can identify in studies with biased results.

Some committees may consider not including case series in their synthesis of evidence, but Chambers et al. (2009) emphasized that case series can strengthen the credibility of a review for emerging interventions. They indicate the importance of case series as a useful source of evidence concerning safety. In their systematic review of the use of radiofrequency catheter ablation to treat atrial fibrillation, adverse events and complications were rarely reported in controlled trials. They recognized the biases inherent in this study design, but reinforced the fact that non-RCT evidence is often

required to ensure clinical credibility of the review. In recognition of this reality, the

Institute of Health Economics developed a quality appraisal tool for case series studies using a modified Delphi technique (Moga et al., 2012). The authors reported a moderate

(kappa = 0.552) to substantial (kappa = 0.806) interrater agreement. A study with 14 or more yes responses (70% or higher) was considered acceptable quality. This tool was used for the critical appraisal of case series in the systematic review of treatments for calcium channel blocker poisonings (St-Onge et al., 2014).

Concerning the inclusion of animal studies in guideline development, Lamontagne et al.

(2010) recognized that animal research may inform clinical practice. When assessing literature reviews of therapeutic interventions for sepsis, they found that 27% of articles assumed that data from preclinical studies could apply to human patients. However, most reviews did not appraise the risk of bias or the clinical relevance of the included animal studies. In 2013, Krauth et al. published a systematic review looking at instruments for assessing risk of bias in animal studies in which they identified 30 distinct tools. The instrument suggested by Hobbs et al. (2005) had been tested for reliability, but reported only 3/13 criteria. The tool from Sena et al. (2007) was validated and reported up to 8/13 criteria, but was specifically developed for strokes. The Animals in Research: Reporting In Vivo Experiments ARRIVE) checklist (Kilkenny et al., 2010) was the only instrument reporting seven or more of the 13 criteria that were not developed for a specific disease. However, the Gold Standard Publication Checklist

(Hooijmans et al., 2011) has been suggested for use instead of the ARRIVE checklist.

The former documents inclusion and exclusion criteria, provides more details for each item, but does not consider conflict of interest reporting. To cover a maximum of criteria,

the systematic review looking at interventions for the treatment of calcium channel blocker poisonings (St-Onge et al., 2014) used both the ARRIVE guidelines and the

National Research Council Institute for Laboratory Animal Research tool (2011), which increased the number of criteria evaluated to 9/13.

Finally, Gagnier et al. (2014) proposed use of Case Report (CARE) guidelines to improve completeness and transparency of published case reports to provide a better gauge for harm or effectiveness. Systematic reviews have included case reports to identify complications or adverse events (Roberts et al., 2014). Some authors may not perceive a benefit of looking for an article with a weak level of evidence, but Guyatt et al. (2014) mentioned:

Nevertheless, reasons to consider a structured approach to formulating weak recommendations are: 1) for physicians who would appreciate guidance, providing that guidance; 2) specifying that a recommendation is weak will reduce the likelihood that it will be inappropriately considered as a measure of performance or quality of care; 3) signaling clinicians that patient values and preferences are likely to play a large role in these instances; and 4) helping in the justification for the research community of further research.

Based on that argument and considering 28% of statements documented in toxicology textbooks are based on case reports (Chuang and Heard, 2011), those types of articles were included in the systematic review for the treatment of calcium channel blocker poisonings.

2.2.5.3 Costs and resource use

Guyatt et al. (2011) recommended that decisions on the direction (for/against) and grade strength (strong/weak) of the recommendation be made after considering the quality of evidence, and the balance between desirable and undesirable outcomes,

values and preferences. The guideline development committee should consider whether any revision of direction or strength is necessary after considering resource use or costs. The US Preventive Services Task Force (Harris et al., 2001) recommends that costs be considered, but not as a first priority. They also stress that statements should not be modified to accommodate concerns about insurance coverage. In that regard Hill et al. (2012) proposed to label recommendations that are particularly sensitive to resource considerations and evaluate them based on systematic reviews of randomized controlled trials reporting resource utilization with direct comparisons between interventions of interest. They recognized that cost-effectiveness analysis or description of acquisition cost may be an alternative. If a product is much more expensive than alternatives but adds no benefit, a valid formal analysis is likely to favour the alternatives. However, cost-effectiveness analysis results may be difficult to interpret unless bench-marking estimates of cost-effectiveness ratios have been previously decided. Values and preferences certainly have a role in the establishment of the willingness to pay.

2.2.5.4 Values and preferences

To ensure that values and preferences are considered, the Institute of Medicine

(Graham et al., 2011) recommends involving a former patient, patient advocate or patient organization representative in the guideline development process. Their participation may contribute to make the recommendations more acceptable to the patient population and, thus, guideline implementation more likely (Kelson et al., 2012).

However, including public views is often challenging, due to a tendency to make hypothetical judgments that can conflict with the interests of patients and caregivers.

Montori et al. (2013) stressed that a committee development panel should avoid making strong recommendations when the best course of action heavily depends on the patient's values and preferences.

In order to facilitate appropriate participation, the steering committee for patients and public involvement of the Guidelines International Network (Knaapen et al., 2012) propose to first identify whether the goal is to consult patients to consider their values and preferences, to directly solicit their participation or efficiently communicate recommendations, so they can provide their input for the decision-making. For each goal, Knaapen et al. (2012) proposed specific strategies:

Consultation: (...) This can include methods such as surveys, focus groups, individual interviews, online consultation, the use of primary research on patients’ needs and expectations, or the use of a systematic review of studies on patients’ and the public’s perspective. Participation: (...) This can be done through participation of patient and public representatives on guideline development groups and other methods. Communication: (...) This can include the production of plain language versions of clinical practice guidelines or the development of patient decision aids or education material.

Thus far, the integration of the values and preferences of patients and families in guideline development in critical care and toxicology has been limited. However, the majority of calcium channel blocker poisoning has resulted from voluntary ingestion (St-

Onge et al., 2012), and studies have raised the possibility that clinicians may have a greater perceived stigmatization of suicide than the general population (Scocco et al.,

2012). Moreover, nearly half of the patients described in a Canadian retrospective study are known to have psychiatric disease or substance abuse (St-Onge et al., 2012), and studies have suggested that physicians may be reluctant to provide care to this population (McLaughlin D et al., 2006). For example, a national specialist survey in 34

Canada revealed that among those who responded, the majority of physicians would not provide Hepatitis C treatment to a current IV drug user who uses needle exchange on a regular basis (Myles et al., 2011). Therefore, it was of primary importance to have patients and their relatives involved in the development of the clinical practice guidelines.

Consideration of the values and preferences of other stakeholders, such as care deliverers, healthcare managers, governments, health insurers, employers or manufacturers may also be relevant (Cluzeau et al., 2012). This is particularly the case if they have a direct interest in the guideline, and particularly if their participation may impact the implementation of the recommendations. The development committee for the recommendations concerning the treatment of calcium channel blocker poisoning recognized the importance of values and preferences of bedside clinicians, nurses, pharmacists and poison control centre representatives. Therefore, they contributed to the external review process to facilitate implementation. A blog was also created to encourage the involvement of guidelines users, decision makers, patients and their relatives in the process of developing recommendations

(http//:poisoningsguidelines.com).

2.2.5.5 Consensus development method

The variety of different methods that can be used to formulate recommendations to integrate the evidence, costs, use of resources, values and preferences for a given treatment are presented in Table 1.

Table 1 Consensus techniques

Consensus Mailed Private Formal Face-to- Structured Aggregation Method Questionnaires Decisions Feedback Face Interaction Method Elicited of Group Contact Choices Informal –- –- –- + –- Implicit Delphi + + + –- + Explicit Method Nominal –- + + + + Explicit group Technique Consensus –- –- –- + –- Implicit development conference Reproduced with permission from: Kunz et al., 2012

When building recommendations based on a low level of evidence, as is often the case in toxicology, the transparency and rigour of the consensus process are of primary importance (Lavergne et al., 2012). Kunz et al. (2012) favoured a formal consensus process that provides an opportunity for active and equal involvement of all participants, notably because this strategy controls articulate members and reduces the power of strong individuals. A formal consensus also allows panel members to retract any firmly stated opinions without losing face. However, this process requires substantial administrative support, skilled and trained facilitators, as well as a representative panel that is limited in size to avoid inflexibility that may arise if the panel is poorly moderated.

Nair et al. (2011) compared several formal methods of consensus development: the

RAND/UCLA Appropriateness Method, the Delphi method, the nominal group technique, the National Institutes of Health consensus development conference and the

Grading of Recommendations Assessment, Development and Evaluation (GRADE) system. They suggest the use of GRADE, RAND/UCLA Appropriateness Method or the 36

Nominal group technique for guideline development. They noted that a consensus conference is probably more appropriate when there is controversy or a high degree of public interest with a potentially important impact on health care cost. The recommendations development committee for the treatment of calcium channel blocker poisoning applied GRADE and the RAND/UCLA Appropriateness Method (Fitch et al.,

2011) similar to that used by the EXtracorporeal TReatments In Poisoning (EXTRIP) workgroup (Lavergne et al., 2012).

The objective of a formal consensus method is to identify a central tendency among the panel and grade the level of agreement (Nair et al., 2011). Figure 5 illustrates the voting process described by the EXTRIP workgroup (Lavergne et al., 2012) and used by the recommendations committee for the treatment of calcium channel blocker poisoning.

The disagreement index is defined as the Inter-percentile range divided by the Inter- percentile range adjusted for symmetry, and describes the dispersion of ratings more effectively than the mean absolute deviation from the median. A value less or equal than one indicates agreement.

Figure 5 Voting process Reproduced with permission from: Lavergne et al., 2012

Once agreement is reached, each recommendation should provide a clear description of the potential benefits and harms, a summary of relevant available evidence (quantity and quality) and evidentiary gaps, an explanation of the part played by values, opinions, theory and clinical experience, a rating of both the level of confidence and the strength of the recommendations and a description of any difference in opinion regarding the recommendation (Graham et al., 2011). The template followed by the committee developing the recommendations for the treatment of calcium channel blocker poisoning was as follows:

• Description of quantity (number of studies from each type), quality (risk of

bias) and consistency of evidence (maximum 50 words);

• Explanation of potential benefits, harms, costs and use of resources

(maximum 100 words);

• Description of the rationales in favour and against the use of the intervention.

Reasons why a rationale was valued more than the other (maximum 200

words);

• Statement proposal formulated as: "In (adult/pediatric) patients poisoned with

a calcium channel blocker (asymptomatic at arrival / symptomatic at arrival /

in severe shock refractory to standard care / in cardiac arrest, “we

recommend” (for level 1 recommendations); or "we suggest" (for level 2

recommendation) or "it would be reasonable" (for level 3 recommendations)

the use of X (dose, way of administration, duration of treatment) to improve Y

(outcome)." Level of evidence: _____ Grade of recommendation: _____

(maximum 50 words); and

• Description of the circumstances in which the recommendations would

change (e.g. co-morbidities, co-ingestions, practice setting, etc.) (maximum

100 words).

This statement formulation was then modified to provide recommendations in a content, format and vocabulary that would facilitate implementation (Graham et al., 2011). As recommended by the AGREE II collaboration (Brouwers et al., 2009), guideline developers must ensure clarity of presentation and applicability for end users given that an international survey of emergency care stakeholders found that 66% of the respondents were unfamiliar with the GRADE system (Aboulsoud et al., 2012).

2.2.6 Tools and implementation strategies

Guideline formats and content are important determinants of guideline usability.

Therefore, Gagliardi et al. (2011) developed a conceptual framework of implementability based on a meta-narrative approach. Eight implementability domains emerged from the

18 studies reviewed: adaptability, usability, validity, applicability, communicability, accommodation, implementation and evaluation. The only study identified by Gagliardi et al. (2011) performed with critical care practitioners that focused on guideline formats and content, indicated that guidelines should be easily accessible, easily applicable, have concise recommendations with accompanying tools, such as checklists, and be supported by strong evidence (Jones et al., 2007). The intensive care unit PAD (pain, agitation and delirium) care bundle represents one example of a successful format (Barr et al., 2013). In emergency care, the preferred format was clinical protocols that incorporated recommendations into workflow with clear identification of key recommendations (Aboulsoud et al., 2012).

In terms of implementation strategies, Sinuff et al. (2013) published a systematic review that looked at knowledge translation interventions for critically ill patients and found that protocols with or without training were associated with the greatest improvements in critical care processes. In a prospective study, Sinuff et al. (2013) also demonstrated an impact of the implementation of 14 guideline recommendations with multifaceted intervention on ventilator-associated pneumonia. The intervention targeted the entire

ICU team, but was modified for each user group. Standardized presentation in electronic, paper and web-based formats, web-base self-assessment quizzes, reminders with bedside rounds checklist, monthly newsletters, involvement of local opinion leaders and a multidisciplinary guideline implementation team, as well as a one- day workshop, were among the strategies used.

Finally, the comprehensive approach proposed by Pronovost (2013) outlined five strategies to increase guideline adherence: 1) include an unambiguous checklist with interventions linked in time and space; 2) identify and mitigate barriers to guideline use and share successful implementation strategies; 3) integrate guidelines for conditions that commonly coexist; 4) rely on systems, rather than the action of individual clinicians, to insure that patients receive recommended therapies; and 5) create transdisciplinary teams and pool expertise from clinical epidemiology (evidence synthesis), implementation science and systems engineering to develop scholarly guidelines with practice strategies. Operating on the assumption that guidelines are knowledge tools that are part of the knowledge-to-action process (Davis et al., 2007), after the assessment of barriers and implementation of interventions the above steps would

remain in place to monitor knowledge use, evaluate the outcomes and sustain ongoing knowledge use (Grimshaw et al., 2012).

2.2.7 External review and updates

To ensure a rigourous methodology is followed in the process of guideline development an external review is needed. Critical appraisal tools were created for that purpose. The ones most commonly adopted are The AGREE (Brouwers et al., 2009) and the Institute of Medicine (Graham et al., 2011) critical appraisal tools, which were used to appraise the recommendations for the treatment of calcium channel poisoning.

The AGREE instrument (Brouwers et al., 2009) evaluates 23 key items organized in six domains: 1) Scope and purpose; 2) Stakeholder involvement; 3) Rigour of development;

4) Clarity of presentation; 5) Applicability; and 6) Editorial independence. Each item is rated on a seven-point scale and a user's manual provides guidance on how to rate each item. A scale domain score is then obtained that allows guideline weakness to be identified.

The Institute of Medicine (Graham et al., 2011) also provides clinical practice guideline appraisal tools that are divided into seven attributes, which include a number of dimensions. The attributes are: clinical applicability, clinical flexibility, reliability/reproducibility, validity, clarity, schedule review and multidisciplinary process.

A "yes/no" score is assigned to each dimension. For each "yes", an associated rating of satisfactory, conditionally satisfactory or unsatisfactory is given. For each "no", a judgment of whether the omission is unimportant, minor or major is made.

The Institute of Medicine (Graham et al., 2011) states that external reviewers should comprise a full spectrum of relevant stakeholders, including scientific and clinical experts, organizations, patients and representatives of the public. A draft of the guideline should also be made available to the general public for comments, notably by press release or through the use the social media. For example, the Virtual AIDS office of Hong Kong sent a press release for public consultation when a draft of guidelines for

HIV infection was created (Government of Hong Kong, 2013). The documents were available on the Hong Kong Advisory Council on AIDS website and the lay public could provide comments by email.

The guideline development committee should consider all comments and keep a written record of the rationale for modifying, or not, the guideline. For the development of the recommendations for the treatment of calcium channel blocker poisoning, scientific and clinical external reviewers participated throughout the process and an anonymous external review by organizations, patients and representatives of the public was done at the end of the process. The drafts of the documents were also available on the blog of the committee (http://poisoningsguidelines.com), the Canadian Association of Poison

Control Centres website and other websites of associations participating in the recommendations development.

Finally, guideline updates should be considered when there is a significant change in evidence, in outcomes considered to be important, in available interventions, and in values placed on outcomes or available health care resources (Graham et al., 2011).

External information provided by the guideline development committee and continuous literature searches may help to detect such changes. Becker et al. (2014) also considered a systematic update every three to five years, but admit that partial updates are often sufficient. The recommendations development committee for the treatment of calcium channel blocker poisoning adopted the algorithm for decision-making processes of the type and scope of update suggested by Becker et al. (2014) (see Figure 6).

Figure 6 Decision-making process for the type and scope of guideline update. Reproduced with permission from: Becker et al., 2014

In conclusion, this chapter briefly summarizes the literature concerning the mechanisms of toxicity, the clinical presentation and the rationale for the use of interventions to treat calcium channel blocker poisoning. An explanation of the theory behind guideline development is also presented, and provides the basis for the research program

described in the following chapter. The evidence underlying each treatment strategy will be detailed in Chapter 6 - Systematic review.

______

Chapter 3 Program of Research

______

Chapter 3 details the research program that links chapters 4 to 8 (retrospective study, survey, systematic review, cost-effectiveness analysis, recommendations development).

It first underlines the aim of the research process that is reached in three steps: 1) identification of a need for a guideline; 2) documentation of the information required for decision-making; and 3) recommendations development.

3.1 Aim

The aim of this program of research is to build an evidence-based guideline endorsed by international critical care, emergency medicine and toxicology associations. In doing so, the goal is to improve consistency and adherence to recommendations by adapting the evidence outlined by the systematic review to costs, values and preferences.

3.2 Program of research

Figure 7 illustrates the program of research following the theory of guideline development previously described. As a first step, a retrospective study conducted in two Canadian cities (described in Chapter 4) and a survey of emergency physicians

(detailed in Chapter 5) identified a demonstrable need for a guideline. The second step was to summarize the quality of evidence with a systematic review of interventions for the treatment of calcium channel blocker poisoning (presented in Chapter 6). Following this, information related to the cost-effectiveness of the use of extracorporeal life- support in cardiotoxic poisonings (discussed in Chapter 7) was also conducted, considering that cost could be cited as a reason not to use this intervention despite

being associated with an improvement in survival in studies of low level of evidence.

The systematic review and the cost-effectiveness analysis contributed to the decision- making process of the recommendations development (third step described in Chapter

8). The fourth and final step represents the future directions that will be taken after the completion of this thesis: dissemination, implementation, evaluation of use and frequent updates.

STEP 1: Need for a guideline

Retrospective study Survey (Burden of CCB poisoning and care gap) (Practice variation)

STEP 2: Information required for decision-making

Systematic review Cost-effectiveness analysis (The evidence) (Costs and resources)

STEP 3: DEVELOPMENT OF RECOMMENDATIONS (Adapting the evidence to costs, values and preferences)

STEP 4: Future directions Dissemination/implementation Evaluation of use

Update

Figure 7 Program of research

______

Chapter 4 Retrospective Study:

Burden of calcium channel blocker

poisoning and care gap

______

This chapter is modified with permission from the following: St-Onge M, Archambault P, Lesage N, Guimont C, Poitras J, Blais R (2012). Adherence to calcium channel blocker poisoning treatment recommendations in two Canadian cities. Clin Toxicol 50:424-30.

The retrospective study detailed in this chapter is part of the first step of this program of research. It identifies the need for a guideline by documenting the burden of calcium channel blocker poisoning and by ascertaining the presence of inconsistency in care.

4.1 Introduction

Poison control centres' databases may not reflect the true burden of disease because they do not include data of patients, whose cases medical staff do not report to the poison control centre. In addition, many details are often not reported by the attending team when they call the poison control centre, which contributes to the incomplete nature of the available data. Health outcomes, such as survival, are not always available in the current poison control centre data set so calcium channel blocker poisoning mortality or morbidity rates are unknown. Moreover, no study has documented whether physicians call the poison control centre for a particular poisoning or if the interventions suggested by the poison control centre are actually applied.

The primary objective of this study was to assess the rate of treatment compliance with the poison control centres' recommendations for the treatment of calcium channel blocker poisoning. The secondary objective was to document the burden of the disease by describing the characteristics and outcomes of patients. The targeted outcomes included the duration of vasopressor use, length of stay in the critical care unit, length of stay in the hospital, morbidity and mortality.

4.2 Methods

Assessing adherence to recommendations and burden of disease was captured by a retrospective chart review of adults poisoned with a calcium channel blocker who were admitted to one of the 23 primary, secondary and tertiary hospitals in the greater

Quebec City and Montreal areas. Psychiatric and pediatric hospitals were not included because of the very low likelihood of finding patients meeting the inclusion criteria determined for this study. An average of 20 emergency physicians were registered per hospital (range: 9 to 38). Thirteen centres had physicians who had completed a five- year training program in emergency medicine. The busiest emergency department evaluated 34,560 non-ambulatory patients per year and the least busy evaluated 3,748 non-ambulatory patients per year. Approval from all of the Directors of Professional

Services and required Research Ethics Boards was obtained.

The inclusion criteria were defined a priori. All adult cases of calcium channel blocker poisoning in Quebec City and Montreal regions between January 2004 and November

2007 were included. Poisoning was defined as taking a higher dose than prescribed or taking a calcium channel blocker that was not prescribed. The patients who were hospitalized for another reason and then poisoned themselves during their hospital stay were excluded.

To identify the relevant cases, medical record databases from all of the hospitals were scrutinized using poisonings or adverse effects as keywords. The specific cases involving a calcium channel blocker were then manually identified. A transfer from one hospital to another was considered as a single intoxication event.

The following diagnostic and therapeutic interventions ordered by the attending physician for the calcium channel blocker poisoning management were evaluated during this chart review: Monitoring, decontamination (activated charcoal, whole bowel irrigation, etc.), support measures (atropine, fluids, vasopressors), calcium, glucagon, high-dose insulin, phosphodiesterase III inhibitors, mechanical circulatory support

(transcutaneous or transvenous pacemaker, intra-aortic balloon, extracorporeal circulation) and any other interventions made by the treating physician.

Data was collected by one observer (MSO) using two data abstraction forms previously approved by a toxicologist (RB) and two emergency physicians (GL, HS). The first form recorded the basic characteristics, therapeutic interventions, vital signs, signs of hypoperfusion (altered mental status, anuria or oliguria, metabolic acidosis and renal failure), outcome measures and whether the poison control centre was called or not.

The second form abstracted the steps taken one by one, in order to conclude whether the recommendations (Figure 8) had been followed or not.

i) The patient who was asymptomatic during his/her stay was classified in the

non-compliant group if he/she was not monitored or decontaminated properly.

ii) The patient who was symptomatic at any point during his/her stay was

classified in the non-compliant group if he/she was not monitored or

decontaminated appropriately, if therapies were used but not indicated or if the

recommended interventions were not administered. The recommendations were

considered as followed if the patient was monitored and all the interventions,

including decontamination, were done as recommended until the patient

demonstrated no signs of toxicity.

CLINICAL SIGNS OF TOXICITY YES NO ↓ ↓ O2, cardiac monitoring, IVs O2, cardiac monitoring, IVs (central line if possible), intubation PRN ↓ ↓ Decontamination if indicated Decontamination if indicated ↓ ↓ Atropine for symptomatic bradycardia Observation in ED or ICU IV fluids ↓ Vasopressors Dopamine (first line treatment) ______CCB ingestion BB ingestion ↓ ↓ Calcium Chloride 10% 10-20 ml Glucagon 2-10 mg (50-150 µg/kg) (0,2 ml/kg) or Calcium Gluconate 10% Repeat Q10 min PRN 30-60 ml (0,6 ml/kg) IV in 5 min Consider infusion Repeat Q10-20 min PRN for 3-4 doses Consider infusion ↓ ↓ High-dose insulin: Insulin bolus (1 U/kg) with 25-50 ml of Dextrose 50% IV followed by an insulin infusion at 1 U/kg/h and a Dextrose infusion at 0,5 g/kg/h. Check glycemia Q15-30 min. ↓ ↓ CCB: Glucagon as per BB ingestion BB: Calcium as per CCB ingestion ↓ ↓ Milrinone: bolus 50 µg/kg in 10 min followed by an infusion of 20 mg (20 ml) in 80 ml of Dextrose 5% or NaCl 0,9% (200 µg/m), infusion 0,375-0,75 µg/kg/min (may not be used if hypotension) ↓ Mechanical support: pacemaker, intra-aortic balloon, extracorporeal circulation (if signs of shock despite all interventions)

Figure 8 Quebec Poison Control Center protocol for the treatment of calcium channel blocker poisoning (2004)

For all therapeutic interventions, the abstractor recorded the heart rate, systolic/diastolic blood pressure and all signs of hypoperfusion before and after having initiated the therapeutic intervention. The treatment effect variable was documented after the known peak effect of the medication. Satisfying responsiveness to treatment was defined as a

normal heart rate, normal blood pressure and no present signs of hypoperfusion. It was assumed that the physician should move on to the next step in the protocol if the heart rate or the blood pressure stayed persistently abnormal, or if the patient showed persistent signs of hypoperfusion.

The observer first looked at the nurses’ notes to identify the exact treatment administration time and the vital signs at that moment. If this was not possible, the physicians’ orders and notes were then considered. If the information was unavailable, or if the patient’s vital signs were taken more than an hour after initiating the treatment, the data were considered missing. A blinded second observer collected the data for 8% of the charts reviewed from the first observer in order to confirm that the process of chart reviewing had no ambiguities.

The duration of vasopressor administration (from the first vasopressor to the discontinuation of all vasopressors) and the length of stay in intensive care were calculated from the nurses’ chart documentation. Mortality within the first five days and the length of stay in hospital were documented with the hospital discharge summary.

Morbidity was defined as any new health problem or exacerbation of a known health disease occurring in the first five days following the poisoning. Mortality and morbidity outcomes were documented within the first five days of hospitalisation because they were more likely to be directly related to the poisoning. Cerebral anoxia, acute renal failure, acute pulmonary oedema and ileus were recorded as morbidity, if diagnosed in any part of the chart by the physician. Rhabdomyolysis was recorded if it was diagnosed by the physician or if CK was higher than 10 000 IU/L (recognized as a typical elevation

in the literature). Myocardial ischemia was defined as a minimum of 1mm depression in the ST segment or T waves inversion on the ECG or if diagnosed in any part of the chart by the physician. Finally, metabolic acidosis was defined as a blood pH and bicarbonate level lower than normal laboratory values.

4.3 Results

A total of 680 charts in Quebec City and 1,085 in Montreal were reviewed. 103 adult calcium channel blocker poisonings were identified and all were included in the study

(34 in Quebec City and 69 in Montreal). The majority (78%) were intentional poisonings.

Most were women (59%) in their fifties (average age = 56) with a past medical history of hypertension (56%), psychiatric or drug abuse disorders (47%) and heart diseases

(37%). Most had ingested non-dihydropyridine calcium channel blockers (67%) and 39 had ingested a dihydropyridine (38%). Coingestions were frequent: benzodiazepines

(39%), other hypotensive agents (32%, including thiazide diuretics, ACE inhibitors, angiotensin II receptor blockers, spironolactone, clonidine or nitrates) and alcohol

(30%). A beta-blocker was involved in 13% of cases. None of the cases reported a significant salicylate or acetaminophen overdose requiring treatment.

There was a greater frequency of women, cardiac disease, diabetes and voluntary ingestions in patients for whom the treating physicians did not follow the poison control centre protocol (the non-compliant group). The non-compliant group also included more frequent ingestions of non-dihydropyridine, benzodiazepines and anti-hypertensive

medication compared to patients for whom physicians followed poison control centre advice (the compliant group). Patients in the non-compliant group presented earlier, had higher glycemia and lower blood pressure with more frequent atrioventricular block or junctional rhythm. Table 2 describes the baseline characteristics of the patients included in the sample.

Table! 2 Baseline characteristics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`N*:'(/*&.!,<='<>!3*N>2),!/1!%&'()*'+6! 3C87!!G!+./0!456! 3G87!G!+./0!456! 3$87!C!+./0!456! 3G87!G!+./0!456! 3G!86! 3C87!G!+./0!456! 3?87!$!+./0!456! 49),&;)!+=+'/.(:!2.//-!%,)++N,)!3>>P;6! GGG! G$H! GCH! GEE! GBE! GEH! GGC! ]!GEE!>>P;!3*N>2),!/1!%&'()*'+6! B! E! B! $G! E! $G! $A! ^!GBE!>>P;!3*N>2),!/1!%&'()*'+6! E! ?! ?! H! C! B! GG! 49),&;)!-(&+'/.(:!2.//-!%,)++N,)!3>>P;6! BH! AA! AG! ??! D$! ?B! B$! ]!BE!>>P;!3*N>2),!/1!%&'()*'+6! H! $! A! $G! E! $G! $D! ^!GEE!>>P;!3*N>2),!/1!%&'()*'+6! E! $! $! G! E! G! H! _(;*+!/1!<=%/%),1N+(/*!3*N>2),!/1!%&'()*'+6! H! E! H! $$! E! $$! $A! 49),&;)!;.=:)>(&!3>>/.@a6! D7C! B7F! A7A! G$7$! A7E! GG7D! GE7C! ]!H!>>/.@a!3*N>2),!/1!%&'()*'+6! E! E! E! G! E! G! G! ^!GE!>>/.@a!3*N>2),!/1!%&'()*'+6! ?! C! A! CH! E! CH! $G! "!#!$!%&'()*'+!%,)+)*')-!(*!+./0!&',(&.!1(2,(..&'(/*!3+./0!4567!*/!-&'&!386!&,)!&9&(.&2.)!:/*:),*(*;!'<)!<)&,'!,<='<>!&'!%,)+)*'&'(/*!1/,!?!%&'()*'+

The poison control centre was called in 74% of the cases (81% of cases in the compliant group, 68% of cases in the non-compliant group). The protocol was followed in 43 patients (42%) and not followed in 60 (58%). No improvement in compliance to the recommendations was observed from one year to another. The reasons why recommendations were considered as “not followed” were most commonly related to an omission or underdosage of high-dose insulin (20 cases); glucagon being given even if not indicated (14 cases) or due to inappropriate decontamination (10 cases). In six cases, whole bowel irrigation was tried even when the patient was hemodynamically unstable. In four cases, gastric lavage was performed when not indicated. Vasopressors were not used in 10 cases even though they had been indicated by the poison control centre. Mechanical interventions (pacemaker, intra-aortic balloon, extracorporeal life support) should have been attempted in nine cases, and none of these received appropriate high-dose insulin. Eight patients were not monitored despite the fact that the time of presentation and the history of ingestion dictated this need. Table 3 describes the reasons why the recommendations were classified as not followed.

Table 3 Recommendations not followed Related to high-dose insulin (HDI) 20 HDI indicated but not tried 8 HDI with no bolus and an inappropriate infusion 5 HDI with an inappropriate infusion 4 HDI with an inappropriate bolus and infusion dosage 3 Glucagon not indicated 14 Related to decontamination 10 Whole bowel irrigation contraindicated 6 Gastric lavage not indicated 4 Pressors indicated but not tried 10 Mechanical interventions indicated but not tried 9 No monitoring 8 No response to fluid and no other intervention tried 7 Related to pacemaker 7 Transvenous pacemaker not indicated 4 Transcutaneous pacemaker not indicated 3 Calcium indicated but not tried 6 Extracorporeal life-support indicated but not tried 6 Interventions indicated but not tried 5

For the entire sample, the results showed a mean of eight days of hospitalization, 47 hours of intensive care, 11 hours of vasopressors usage, a morbidity of 50% and a mortality of 6%. Acute renal failure (35%), metabolic acidosis (25%), acute pulmonary oedema (15%), aspiration pneumonia (15%), rhabdomyolysis (8%), myocardial ischemia (7%), abnormal liver function tests (AST/ALT) (6%), cerebral anoxia (4%) and ileus (3%) were among the most frequent complications. Table 4 details the outcomes.

Table 4 Outcomes

OUTCOMES TOTAL Protocol followed (43/103) Protocol not followed (60/103) (103 patients) Total (43) Contact No contact Total (60) Contact No contact with PCC with PCC with PCC with PCC (35/43) (8/43) (40/60) (20/60) Hospital LOS Mean: 8 Mean: 5 Mean: 6 Mean: 4 Mean: 9 Mean: 6 Mean: 16 (days) Median: 2 Median: 1 Median: 1 Median: 2 Median: 5 Median: 3 Median: 7 SD: 16 SD: 20 ICU LOS Mean: 47 Mean: 21 Mean: 20 Mean: 24 Mean: 67 Mean: 68 Mean: 76 (hours) Median: Median: 0 Median: 0 Median: 21 Median: Median: Median: 45 18 SD: 54 34 34 SD: 98 Duration of Mean: 11 Mean: 3 Mean: 4 Mean: 1 Mean: 17 Mean: 16 Mean: 17 vasopressor Median: 0 Median: 0 Median: 0 Median: 0 Median: 0 Median: 0 Median: 10 use (hours) SD: 9 SD: 25

Mortality 6 (6%) 0 (0%) 0 (0%) 0 (0%) 6 (10%) 3 (8%) 3 (15%)

Morbidity 51 (50%) 11 (26%) 6 (14%) 5 (63%) 40 (67%) 26 (65%) 14 (70%)

Acute kidney 36 (35%) 5 (12%) 4 (11%) 1 (13%) 31 (52%) 20 (50%) 11 (55%) injury Metabolic 26 (25%) 3 (7%) 2 (6%) 1 (13%) 23 (38%) 15 (38%) 8 (40%) acidosis Acute 15 (15%) 1 (2%) 1 (3%) 0 (0%) 14 (23%) 10 (25%) 4 (20%) pulmonary oedema Aspiration 15 (15%) 1 (2%) 1 (3%) 0 (0%) 14 (23%) 9 (23%) 5 (25%) pneumonia Rhabdomyoly 8 (8%) 0 (0%) 0 (0%) 0 (0%) 8 (13%) 7 (18%) 1 (5%) sis Abnormal liver 6 (6%) 1 (2%) 1 (3%) 0 (0%) 5 (8%) 1 (3%) 4 (20%) function tests Myocardial 7 (7%) 0 (0%) 0 (0%) 0 (0%) 7 (12%) 5 (13%) 2 (10%) ischemia Ileus 3 (3%) 1 (2%) 1 (3%) 0 (0%) 2 (3%) 2 (5%) 0 (0%) Cerebral 4 (4%) 0 (0%) 0 (0%) 0 (0%) 4 (7%) 2 (5%) 2 (10%) anoxia

4.4 Discussion

The primary objective of this retrospective study was to evaluate whether physicians call and comply with Quebec Poison Control Centre's recommendations. Although most physicians called the poison control centre, the majority did not subsequently follow the centre's recommendations for the treatment of calcium channel blocker poisoning. This may reflect a problem with the evidence base in toxicology and knowledge translation strategies used by the poison control centres.

As outlined by Chuang and Heard (2010), most of the current calcium channel blocker poisoning treatment recommendations in major toxicology textbooks are based on case reports (40%), animal studies (22%) or are not referenced (11%). The Quebec Poison

Control Centre's current recommendations for the treatment of calcium channel blocker

(Figure 8) are based on a review article published by two toxicologists (DeWitt and

Waksman, 2004). A lack of confidence in the validity of the conclusions made by the poison control centre may partly explain why some physicians decided not to follow the protocol.

As described in the Ottawa Model of Research Use (Logan, 1999), other potential barriers explaining a lack in recommendations adherence may be related to the potential adopters: concerns about the risk of certain interventions (like hypoglycemia with the high-dose insulin as mentioned by Greene et al. in 2007), awareness or knowledge of these recommendations or misunderstanding of the protocol. The non- compliance may also be related to the practice environment, economic or structural 61

factors. These may include inconsistency in the recommendations from one Canadian province to another (like for the use of glucagon), unavailability of extracorporeal life- support or of critical care monitoring, uncontrolled events, such as not being able to have a central IV access, and cultural factors, such as not believing in the benefit of saving patients who attempted suicide. Moreover, it is possible that recommendations are not followed because evidence-based innovations are not considered in the actual protocol (e.g. lipid emulsion).

In terms of knowledge translation strategies, the protocol illustrated in Figure 8 is explained by the Quebec Poison Control Centre's nurses on the phone and faxed to the physicians who placed the call to the Poison Control Centre. Although a follow up is conducted by the Poison Control Centre's nurses for most of the calcium channel blocker poisoning cases, this is not always done directly with the treating physician; it may be done with the bedside nurse or the resident. Toxicologists are available if more details are required, but they are not always directly involved. As illustrated by

Grimshaw (2006), a more effective dissemination and implementation of the recommendations using frequent reminders, feedback and contact with local leaders may improve protocol adherence.

The secondary objectives of this study were to describe the characteristics and outcomes of the patients. These were previously presented and described in Tables 2 and 4. However, the observation that the non-compliance (protocol not followed) group had poorer outcomes must be interpreted with caution. At the time of admission, this group was apparently already in a more compromised state of health (higher glycemia,

lower blood pressure with more frequent atrioventricular block or junctional rhythm) than their counterparts in the compliance group. In fact, it may be more difficult to call the poison control centre and follow the recommendations when the patient is unstable.

This study has some limitations. First, it was carried out in urban regions with adult patients, limiting its validity outside this population group. Nevertheless, the charts picked out by the medical record archivists included all cases listed in the Quebec

Poison Control Center's database. This indicates that the search strategy employed by the medical record archivists was exhaustive.

Second, the results are a reflection of charting quality. Fortunately, most of the data were available because the information required was well-documented (prescription, medication administration, vital signs, length of stay, mortality). It was however difficult to confirm hypoperfusion because not all patients had a urinary catheter or had blood gas testing done and the Glasgow Coma Scale was not systematically recorded. As such, it is possible that a patient with a normal heart rate and a normal blood pressure was considered stable because the physician did not look for, or record, signs of hypoperfusion. These patients could have been judged as being less sick than they were in reality.

Third, the observer was not blinded to the objectives of this chart review. To decrease this potential bias, the data abstraction forms used were detailed and allowed for little subjectivity. A second observer (archivist) blinded to the research objectives collected the data for 8% of the charts. Both abstractors found similar results in the data collected.

4.5 Conclusion

In conclusion, the majority of physicians did not follow the Quebec Poison Control

Center's recommendations for the treatment of calcium channel blocker poisoning, although most of them called the poison control centre. This could be related to a lack of confidence in the validity of the actual protocol, practice environment factors and potential adopter factors. Lack of adherence can be interpreted as a care gap, but given that the current recommendations of the Quebec Poison Control Centre are not evidence-based, it can also be regarded as a reflection of practice variation. The following chapter explores this issue further by assessing whether there is a variation in opinions among emergency physicians concerning the treatment of calcium channel blocker poisoning, and if the physicians' training and practice settings explain part of the variation.

______

Chapter 5 Survey:

Variation in opinions and practice

______

This chapter is modified from the following: St-Onge M, Chun-Po SF, Blais R (2014). Impact of training and practice setting on the physician's opinion regarding the treatment of a type of poisoning. Submitted to BMC Pharmacology and Toxicology.

Chapter 4 documented the disease burden and a lack of adherence to recommendations in the treatment of calcium channel blocker poisoning. Considering that the current recommendations are not the product of a rigourous evidence-based evaluation process, the lack of adherence to may be only a reflection of an inconsistency and variation in opinions and practice. Chapter 5 examines whether there is a variation in opinions concerning the treatment of calcium channel blocker poisoning and evaluates the impact of physicians' training and practice setting on this variation.

This study adds further evidence documenting the manifest need for a guideline (Step 1 of recommendations development process).

5.1 Introduction

Given the lack of adherence to recommendations underlined by the retrospective study previously detailed in Chapter 4, it was hypothesized that there was a variation in opinions regarding the treatment of calcium channel blocker poisoning and that part of that variation may be explained by previous toxicology training. The limited availability of resources in some rural regions may also influence the care strategy proposed by physicians. No study thus far has assessed at the impact of training or practice setting on the physician's opinion regarding the management strategy of poisonings.

In Canada, most clinicians providing care in emergency departments are family physicians practicing without a one year certificate of special competency in emergency medicine (CCFP(EM)). The results of a 2005 survey of 32 community emergency

departments in southwestern Ontario revealed that 70.1% of the clinicians had no formal emergency medicine training (Bhiman et al., 2007; Ovens et al., 1993). In the

Province of Quebec, unpublished data collected for the purpose of this study showed that approximately 1552 physicians work in the 104 emergency departments, and of these, only 129 had completed a five-year Emergency Physicians Board Certificate

(FRCPC(EM)) with the Royal College of Physicians and Surgeons of Canada, which includes at least a one-month toxicology rotation in most of the programs. Very little training in toxicology is included in the family medicine program and the exposition to toxicology during the CCFP(EM) program varies depending on the university and the candidate. Moreover, most of family physicians work in secondary or community/primary centres where fewer resources are available.

The objectives of this study were to: 1) Evaluate if there is variation in the emergency physicians' opinion regarding the treatment of poisonings (more specifically calcium channel blocker poisoning) and; 2) assess if that variation is influenced by their training

(family physician, one year CCFP(EM) certification or five-year FRCPC(EM) certification) and practice setting.

5.2 Methods

As stated by the Association québécoise d'établissements de santé et de services sociaux and the Province of Quebec Health Ministry, a primary or community centre is defined as an emergency department with less than 10,000 patients on stretcher/year and 24h access to a general medicine physician, an anesthesiologist and a general

surgeon. A secondary centre is defined as an emergency department with 10,000 to

20,000 patients on stretcher/year and 24h access to an internal medicine physician, a radiologist and an orthopedist. Finally, a tertiary or academic centre is defined as an emergency department with more than 20,000 patients on stretcher/year and 24h access to some subspecialties, such as nephrology, neurosurgery, and cardiac or thoracic surgery.

The survey (an anonymous written questionnaire, with multi-option variables and unstructured response format) was conducted between September 2008 and 2011 among emergency physicians working in the Province of Quebec. Physicians who were aware of the study details were excluded from the sample. To ensure a proportional participation of physicians from community/primary, secondary and academic/tertiary centres, a cluster weight-based group sampling method was used to identify the emergency departments. Emergency physicians were then invited to participate. The weight of each emergency department was established based on the number of emergency physicians working in that department and the estimated proportion that would attend their emergency department meeting where the survey would be administered. A targeted sample size of 34 emergency departments (13 community/primary centres, 11 secondary centres, 10 academic/tertiary centres) were randomly selected by SAS software (version 9.3) among the 104 hospitals in the

Province to detect a 10% absolute difference with 80% power at a level of significance of 0.05.

The chiefs of the 34 emergency departments were contacted to obtain authorization to administer a 30-minute written questionnaire during one of the department meetings. A minimum of three emails and/or phone calls was made before the absence of response was considered as a refusal. The subject of the survey was not revealed until the meeting where the physicians' participation was subsequently solicited by the study investigators. After consenting to participate, the emergency physicians were asked to select their management for each of the following six diltiazem (immediate release, unknown quantity) poisoning clinical scenarios: (1) stable, asymptomatic, more than one hour post-ingestion; (2) stable, asymptomatic, less than one hour post-ingestion; (3) unstable, at arrival to the emergency department; (4) unstable, did not respond to support measures defined as fluid, atropine and vasopressors; (5) unstable, did not respond to support measures and calcium; (6) unstable, did not respond to support measures, calcium and high-dose insulin. Item generation and reduction was based on the results of a retrospective study conducted in the same province (St-Onge et al.,

2012) in order to represent the severity spectrum of this type of poisoning.

All cases presented to the emergency physicians were 35-year-old men without previous medical history who presented after a suicidal ingestion. All unstable cases had a heart rate of 36 beats per minute (first-degree atrioventricular block), a blood pressure of 80mmHg systolic, a blood glucose level of 11mmol/L and a Glasgow Coma

Scale (GCS) score of 14/15. The participants were told to consider all scenarios as being non-related to avoid any confusion. After proposing strategies to the clinical cases in the questionnaire, the physicians then identified which resources would be available at their hospital and which factors could have influenced their management of a calcium

channel blocker poisoning. The questionnaire was pre-tested in a secondary centre with

12 emergency physicians and was available in French or English (see appendix 1). The research ethics board at the Institut universitaire de cardiologie et de pneumologie de

Québec approved the conducting of the study.

Each of the key independent variables (type of training, practice setting) and outcome variables (interventions considered) were detailed for each case. The reported resources available and the factors perceived as influencing the calcium channel blocker poisoning management were also described.

Multivariate logistic regression was used to analyze choices of medical intervention separately because participants were allowed to select multiple interventions (antidotes or not) in the aforementioned clinical Scenario 3 - 5. Outcome variables under consideration were the use of atropine, external pacemakers, IV calcium and high-dose insulin in Scenario 3 (unstable at emergency department arrival); the use of external pacemakers, IV calcium and high-dose insulin in Scenario 4 (unstable, did not respond to support measures defined as fluid, atropine and vasopressors); the use of external pacemakers and high-dose insulin in Scenario 5 (unstable, did not respond to support measures and calcium).

In the logistic regression models, the choice of a medical intervention in relation to independent variables was quantified in terms of odds-ratios (ORs). OR>1 suggests an increase in the likelihood of the use of a medical intervention. The key independent variables were type of training and practice setting as categorical variables. Family

physicians without a one year CCFP(EM) certificate and community/primary emergency departments, respectively, were the reference groups. Participants’ age, gender and clinical experiences were included as control variables in the models and 95% confidence intervals (CIs) and p-values were assessed using Wald’s statistics.

In addition to the logistic regression models assessing the choices of medical interventions, descriptive data analysis was used to examine resource unavailability reported by the participants. We expected strong correlation in responses among participants working at the same emergency department. Hence, the data analysis was conducted in a two-step manner. At the first step, the number and the proportions of the participants who reported a medical resource as unavailable were summarized by the emergency department. A medical resource was deemed unavailable within the emergency department if ≥50% of participants at the same emergency department reported its unavailability. Using this criterion, the numbers and proportions of emergency departments with unavailable medical resources were summarized by practice setting at the second step. Considering the small number of participating emergency departments, Fisher’s exact tests were used to evaluate differences in resources unavailability among different types of emergency departments. In addition, we also conducted sensitivity analysis for this analysis by changing the threshold value to 40% and 80%.

5.3 Results

Nineteen emergency departments (6 primary, 6 secondary and 7 tertiary centres), which included a total of 140 emergency physicians, accepted the invitation to participate in the study. Table 5 describes the characteristics of the participating clinicians.

Table 5 Participants characteristics (n=140)

Overall Practice setting Training2 Primary Secondary Tertiary Family CCFP(EM) FRCPC(EM) centres centres centres physicians

Age (mean; 40 (8) 39 (7) 40 (7) 41 (9) 39 (8) 38 (8) 49 (8) std) Years of 12 (9) 10 (8) 12 (8) 13 (9) 11 (8) 10 (8) 20 (9) clinical experience (mean; std) N % N % N % N % N % N % N % N 140 33 42 65 83 39 15 Male1 78 57% 18 59% 24 59% 36 55% 43 53% 20 53% 13 87% Previous 29 21% 4 17% 7 17% 18 28% 10 12% 15 39% 4 27% experience or training concerning CCB poisoning Practice setting Primary 33 24% ______31 37% 2 5% 0 0% centres Secondary 42 30% ______28 34% 10 26% 3 20% centres Tertiary 65 46% ______24 29% 27 69% 12 80% centres Training2 Family 83 61% 31 94% 28 68% 24 38% ______physicians CCFP(EM) 39 29% 2 6% 10 24% 27 43% ______(one year) FRCPC(EM) 15 11% 0 0% 3 7% 12 19% ______(5 year) Notes: 1) Two participants (1 family physician in a primary centre and 1 CCFP(EM) in a secondary centre) did not report their sex. 2) Three participants (1 in a secondary centre and 2 in a tertiary centre) did not report the types of training.

Two academic/tertiary centres, six secondary centres and seven community/primary centres did not participate because they could not find a time for the survey to be administered before the end of the study period or never answered the emails and phone calls. Thirteen of the 19 participating emergency departments (5 community/primary centres, 4 secondary centres, 4 academic/tertiary centres) agreed to provide information on the physicians who were not present at the meeting. The characteristics of the absent clinicians (Table 6) were generally similar to those who participated in the study, but a greater proportion of males were present at the meetings. Only one clinician who was present at an emergency department meeting refused to participate. All other physicians (a total of 140) eligible for the study agreed to complete the survey.

Table 6 Characteristics of the physicians not present at the meetings

Characteristics Primary centers Secondary Tertiary centers Total (n=22) centers (n=44) (n=33) (n=99) 60% present 48% present 66% present Sex (% of male) 48% (56%) 50% (59%) 48% (55%) 49% (57%) Years of clinical 9 (10) 13 (12) 15 (13) 13 (12) experience (mean) Training, n (%) Family physicians 91% (94%) 77% (68%) 55% (38%) 71% (61%) CCFP(EM) 9% (6%) 23% (24%) 24% (43%) 21% (29%) FRCP(EM) 0% (0%) 0% (7%) 21% (19%) 7% (11%)

Only the second scenario (Case 2) had a patient presenting less than one hour post- ingestion. Most physicians (134/140) decided to decontaminate the second case, but the responses for the chosen method varied. Some used more than one method and/or considered multiple-dose charcoal. Table 7 describes the decontamination method(s) or the enhanced elimination therapies considered depending on the practice setting,

clinical experience and training. For the other cases, the majority did not decontaminate the patient.

Table 7 Proposed decontamination strategies for Case 2 (presentation less than one hour post-ingestion) No decon- Method of decontamination Method of enhanced tamination elimination Ipecac Gastric Activated Whole Multiple- Dialysis syrup lavage charcoal bowel dose irrigation charcoal Practice setting Primary 1 (3%) 1 (3%) 5 (15%) 21 (64%) 1 (3%) 2 (6%) 0 (0%) centers (n=33) Secondary 2 (5%) 0 (0%) 7 (17%) 26 (62%) 3 (7%) 8 (19%) 0 (0%) centers (n=42) Tertiary 3 (5%) 1 (2%) 19 (29%) 42 (65%) 1 (2%) 13 (20%) 0 (0%) centers (n=65) Training FP (n=83) 1 (1%) 2 (2%) 21 (25%) 56 (67%) 0 (0%) 17 (20%) 0 (0%) CCFP(EM) 1 (3%) 0 (0%) 6 (15%) 25 (64%) 4 (10%) 7 (18%) 0 (0%) (n=39) FRCPC(EM) 2 (13%) 0 (0%) 4 (27%) 8 (53%) 1 (7%) 4 (27%) 0 (0%) (n=15)

In terms of intervention, fluid management was consistent among the different groups.

Most clinicians (105/140) elected to give fluid to the third case, but gave less and less as the clinical scenarios showed the patient was not responding to therapies (47/140 for the last case). Nearly two-thirds (64%) of the physicians considered giving calcium upon emergency department arrival, while 54% proposed atropine and 30% suggested starting vasopressors.

Glucagon was ordered by 44-61% of the clinicians, but the high-dose insulin was initiated only in 50% of the sickest patients (27 to 50%, depending on the scenario).

Milrinone, internal pacemaker, intralipids and extracorporeal life support were rarely identified as potential therapies for calcium channel blocker poisonings.

Table 8 details the proportion of clinicians who considered specific therapies for calcium channel blocker poisoning (calcium, high-dose insulin) depending on their training and practice setting. Antidote use was more frequent in FRCPC(EM) vs. family physicians and external pacemaker was used even less frequently. Physicians with a five-year

FRCPC(EM) certificate were more likely to order high-dose insulin in Scenario 3, 4 and

5 than family physicians (p-values of 0.0047, 0.0301 and 0.0695, respectively). The same observation can be made for the order of calcium in FRCPC(EM) and CCFP(EM) vs. family physicians in Scenario 4, although the evidence was not as strong as the use of high-dose insulin (p-value= 0.0543).

Table 8 Interventions considered

Intervention HDI Calcium External pacemaker Atropine Scenario 3 4 5 3 4 3 4 5 3 Level of training CCFM vs FP 1.26 1.16 2.40 1.39 3.28 0.54 0.67 0.59 1.07 unadjusted OR (0.46- (0.48- (1.03- (0.57- (1.21- (0.22- (0.28- (0.25- (0.46- 3.35) 2.76) 5.81) 3.50) 10.09) 1.27) 1.59) 1.40) 2.56)

1.08 1.03 2.35 1.08 2.43 0.68 0.75 0.64 0.95 adjusted OR (0.36- (0.39- (0.93- (0.42- (0.85- (0.26- (0.29- (0.25- (0.37- 3.11) 2.65) 6.09) 2.87) 7.81) 1.71) 1.88) 1.61) 2.43) FRCPC vs FP 7.25 3.69 2.68 1.43 8.36 0.12 0.18 0.32 0.84 unadjusted OR (2.07- (1.10- (0.81- (0.41- (1.45- (0.02- (0.03- (0.09- (0.24- 28.62) 13.90) 9.95) 5.85) 159.41) 0.52) 0.67) 1.06) 2.78)

adjusted OR 9.40 5.97 3.83 1.16 7.62 0.13 0.17 0.40 1.35 (2.26- (1.48- (0.97- (0.29- (1.15- (0.02- (0.03- (0.10- (0.34- 45.68) 28.17) 17.33) 5.27) 153.91) 0.61) 0.74) 1.50) 5.28) Type of emergency department Secondary vs primary 1.32 2.17 1.63 1.50 1.04 0.61 0.36 0.65 0.34 (0.42- (0.79- (0.62- (0.58- (0.38- (0.23- (0.13- (0.24- (0.11- unadjusted OR 4.45) 6.33) 4.35) 3.94) 2.86) 1.59) 0.96) 1.68) 0.96)

adjusted OR 1.14 2.25 1.69 1.37 0.93 0.66 0.35 0.64 0.30 (0.34- (0.79- (0.62- (0.51- (0.33- (0.24- (0.13- (0.23- (0.09- 4.06) 6.80) 4.68) 3.68) 2.66) 1.75) 0.93) 1.69) 0.89) Tertiary vs primary 1.24 1.61 1.33 1.96 1.01 1.51 1.07 1.33 0.19 (0.39- (0.57- (0.50- (0.74- (0.36- (0.57- (0.39- (0.49- (0.06- unadjusted OR 4.25) 4.79) 3.57) 5.30) 2.81) 4.06) 2.92) 3.65) 0.54)

1.14 1.57 1.29 1.88 0.90 1.68 1.06 1.48 0.14 adjusted OR (0.34- (0.53- (0.46- (0.68- (0.30- (0.63- (0.39- (0.53- (0.04- 4.06) 4.87) 3.63) 5.26) 2.60) 4.63) 2.85) 4.20) 0.43) Notes: - Data are adjusted for age, gender, years of clinical experience and previous experience with CCB poisoning - OR>1 indicates an increase in likelihood of using the intervention - P-values were evaluated based on Type 3 statistics using likelihood ratios - Values in bold are statistically significant

The resources reported as being unavailable (or unknown) varied across practice settings. Table 9 shows variations across practice setting were statistically significant for internal pacemakers, extracorporeal life-support and levosimendan. Of note, a

noticeable proportion of clinicians identified milrinone (68%) and lipid emulsion (63%) as unavailable therapies, regardless of practice setting.

Table 9 Resources reported as being unavailable across practice settings

Practice setting Total Primary Secondary Tertiary P-value n=19 centers centers centers n=6 n=6 n=7 N % N % N % N % IV calcium 0 0% 0 0% 0 0% 0 0% HDI 0 0% 0 0% 0 0% 0 0% External 0 0% 0 0% 0 0% 0 0% pacemaker Glucagon 1 5% 1 17% 0 0% 0 0% 0.6316 Internal 5 26% 4 57% 1 17% 0 0% 0.0183 pacemaker Lipid emulsion 12 63% 4 67% 3 50% 5 71% 0.8437 Milrinone 13 68% 6 100% 3 50% 4 57% 0.2067 ECLS 15 79% 6 100% 6 100% 3 43% 0.0168 Levosimendan 16 84% 6 100% 3 50% 7 100% 0.0413

Resource unavailability Resource Sufficient 1 5% 1 17% 0 0% 0 0% 0.6316 human resources HDI: high-dose insulin ECLS: extracorporeal life-support

Finally, FRCPC(EM) trained clinicians mentioned access to critical care units as influencing the management of calcium-channel blocker poisoning. Family physicians without CCFP(EM) were the only ones who mentioned the poison control centre as influencing the management of their poisoned patients. Family physicians with or without CCFPM(EM) also underlined guidelines and protocols as having an influence on their strategy. Table 10 details the perceived influencing factors on the management of calcium channel blocker poisonings.

Table 10 Perceived influencing factors on management

Influencing Practice setting Training TOTAL factors Primary Secondary Tertiary FP CCFP(EM) FRCPC(EM) (n=140) centers centers centers (n=83) (n=39) (n=15) (n=33) (n=42) (n=65) Cases' frequency 0% 0% 3% 1% 0% 7% 1% Physicians' 2% 14% 9% 8% 13% 7% 11% experience and knowledge Nurses' 0% 0% 2% 0% 3% 0% 1% experience and knowledge Past medical 7% 5% 3% 5% 5% 0% 4% history Co-morbidities 0% 0% 0% 0% 0% 0% 0% Co-ingestions 12% 5% 11% 10% 10% 7% 9% Time of ingestion 15% 7% 11% 11% 13% 7% 11% CCB formulation 11% 5% 5% 7% 3% 7% 6% Clinical state 21% 17% 5% 20% 26% 27% 24% Response to 0% 0% 3% 0% 3% 7% 1% therapy Access to 21% 17% 5% 14% 8% 13% 12% antidote Access to 9% 0% 3% 4% 0% 7% 3% invasive treatment Access to 12% 3% 6% 4% 13% 7% 6% consultant Access to critical 3% 14% 9% 4% 15% 27% 9% care Access to human 9% 5% 9% 7% 10% 7% 8% resources Access to poison 12% 9% 6% 10% 0% 0% 6% control center Access to 7% 12% 3% 8% 5% 0% 6% guidelines Access to other 0% 5% 0% 1% 3% 0% 1% info Time of the day 3% 3% 0% 1% 3% 0% 1%

5.4 Discussion

This study highlights a variation in opinion between emergency physicians concerning management of calcium channel blocker poisoning, and suggests this variation may be influenced by clinician training. Therapies specific to calcium channel blocker poisoning 78

were more often considered by physicians with a five-year FRCPC(EM) certificate. In fact, clinicians without formal toxicology training (family physicians) tended to emphasize guidelines and protocols as influencing their practice. In order to improve recommendations adherence, poison control centres should consider adapting their approach depending on whether the bedside clinicians have toxicology included in their training or not. In fact, Grol et al. (1998) performed an observational study looking at factors influencing guidelines adherence, and determined that recommendations demanding a change in existing practice routines were less likely to be followed. For physicians who never learned about high-dose insulin during their training, it may be more difficult to apply the recommendation even if they called the poison control centre.

Providing a mandatory training, such as ACLS-toxicology for physicians working in acute care setting, having outreach programs or offering bedside consultation services may help solve this problem. Implementation strategies will be detailed in Chapter 9 -

Future directions.

Furthermore, reported resources available varied with practice setting. Having access to extracorporeal life-support and internal pacemakers may be problematic in community/primary centres. In addition, more experienced physicians and FRCPC(EM) physicians articulated that access to critical care might play a role in their patient's management. Therefore, poison control centres should ensure that their recommendations are perceived as easy to implement in the respective physicians' setting. Rogers (2002) mentioned that diffusion of innovation should be perceived as beneficial, compatible with beliefs and easy to implement. In the development of

recommendations for the treatment of calcium channel blocker poisoning, special attention will be paid to values, preferences and the applicability of recommendations.

This study has some limitations. First, despite tremendous effort to recruit centres, only

19 were included. Fortunately, equal representation was maintained as a proportional numbers of community/primary, secondary and academic/tertiary centres declined the invitation to participate in the study. However, because family physicians were absent more often, they were proportionally underrepresented in this study. This may contribute to selection bias, but considering that nearly 100% of physicians who attended meetings participated in the study, it is unlikely that this systematic error was directly related to the survey's subject.

Second, 45 of 140 (32%) participants did not answer the last question concerning the factors influencing their management of calcium channel blocker poisonings. It is possible that the single open-ended question included at the end of the survey discouraged some of the respondents who were less motivated compared to previous questions. In retrospect it is possible we could have collected more information with close-ended questions. Alternatively, the physicians may have had no specific answer for this question.

Finally, this survey reports the opinions of clinicians and may not completely reflect how physicians would actually respond to a calcium channel blocker poisoned patient.

Nevertheless, the previously mentioned retrospective study (St-Onge et al., 2012) identified a 58% non-compliance rate for poison control centre recommendations and

high-dose insulin was not started when indicated. These findings are consistent with the results of the survey outlined in this chapter (only 50% of physicians started high-dose insulin in the fifth scenario), and decontamination also seemed to be erratic, as reflected in the diversity of reported approaches. Therefore, this suggests that the physicians' responses to the hypothetical scenarios may indeed reflect, at least closely, how they would manage a real case.

5.5 Conclusion

In conclusion, physicians' opinion regarding the treatment of calcium channel blocker poisonings can be influenced by their training. Physicians with FRCPC(EM) training were more likely to consider specific therapies used in toxicology, whereas family physicians were the only group who requested toxicology guidelines/protocols and guidance from the poison control centre. Access to resources varied mainly depending on the practice setting. Having access to internal pacemakers or extracorporeal life support was reported as problematic in primary centres, while FRCPC(EM) trained physicians working principally in tertiary centres mentioned access to critical care as influencing the care of calcium channel blocker poisoned patients. The following chapter will assess the evidence for the use each those interventions.

______

Chapter 6 Systematic review:

The evidence

______

This chapter is modified with permission from the following: St-Onge M, Dubé PA, Gosselin S, Guimont C, Godwin J, Archambault PM, Chauny JM, Frenette AJ, Darveau M, LeSage N, Poitras J, Provencher J, Juurlink D, Blais R (2014). Treatment of calcium channel blocker poisoning: A systematic review. Clin Toxicol 52:926-44.

Chapter 5 demonstrated the variation in opinions regarding the management of calcium channel blocker poisoning. The following chapter documents the evidence for each intervention considered for the treatment of calcium channel blocker poisoning and, combined with the cost-effectiveness analysis described in Chapter 7, will provide the required information for decision-making (Step 3) in the process of recommendations development.

6.1 Introduction

In order to facilitate the development of recommendations following the GRADE methodology (Guyatt et al., 2011) and the AGREE II statement (2009), it was necessary to document and characterize the available evidence pertaining to the effects of treatments for calcium channel blocker toxicity. The objective of this systematic review (registry number: CRD42012002823) was to evaluate the reported effects of treatments for calcium channel blocker toxicity where the primary outcomes of interest were mortality and improvement in hemodynamics. This research also evaluated the impact of interventions on secondary outcomes, such as functional outcomes, hospital length of stay, intensive care unit (ICU) length of stay, duration of vasopressor use and serum calcium channel blocker concentrations.

6.2 Methods

Eligible study types: We included controlled trials, observational studies, case series, animal studies, case reports and abstracts from scientific and clinical meetings in any language, with no date restriction. Case reports were defined as articles pertaining to a single case, whereas articles were classified as case series when multiple cases were presented. Cohort studies were differentiated from case series based on an approach proposed by Dekkers et al. (2012).

Types of participants: Studies were eligible if they involved humans or animals poisoned with any calcium channel blocker. Poisoning was defined as an "exposure (...) causing or capable of causing toxicity, regardless of intent" (Lavergne et al., 2012). An adverse effect was defined as an undesirable effect of a drug taken at therapeutic doses for the appropriate indication.

Types of intervention: Studies with defined intervention(s) meant to improve the targeted primary and/or secondary outcomes were eligible.

Types of outcome measures: Studies were required to document at least one of the primary or secondary outcomes. The primary outcomes included mortality (in- hospital, or at the end of experiment for animal studies) and improvement in hemodynamics

(heart rate, blood pressure, stroke volume, cardiac output or peripheral vascular resistance). Secondary outcomes included functional outcomes (defined as return to functional baseline or not), ICU length of stay, hospital length of stay, duration of 84

vasopressor use and serum calcium channel blocker concentrations. Reports of adverse effects of treatments were also documented.

Searches of Medline/OVID, Pubmed, EMBASE, Cochrane Library, Toxline and

International pharmaceutical abstracts up to December 31, 2013 without time restrictions. Two librarians developed the search strategy with the following keywords:

• [calcium channel blockers OR calcium channel antagonist OR calcium channel

blocking agent OR (amlodipine or bencyclane or bepridil or cinnarizine or

felodipine or fendiline or flunarizine or gallopamil or isradipine or lidoflazine or

mibefradil or nicardipine or nifedipine or nimodipine or nisoldipine or nitrendipine

or prenylamine or verapamil or diltiazem)] AND

• [overdose OR medication errors OR poisoning OR intoxication OR toxicity OR

adverse effect].

Conference proceedings and meeting abstracts of the European Association of Poison

Centres and Clinical Toxicologists and the North American Congress of Clinical

Toxicology (2008-2013), trial registries and Google Scholar were also searched. Authors of selected publications (except for case reports) were contacted.

Two independent reviewers blinded to authors and journal names selected the studies based on eligibility criteria. Disagreements were resolved by consensus or, when required, by a third reviewer. The kappa statistic was used to quantify agreement on the articles included. A data abstraction form was used to standardize the data collection process after a pilot version was tested among data abstractors with five articles related to digoxin poisoning. No significant abstraction difference was noted between

abstractors. For each included study, two reviewers independently abstracted study characteristics (year of publication, authors, study design), subjects (number, inclusion/exclusion criteria, age, gender, co-morbidities, co-ingestions, type of animal studied where applicable, sample size calculation, weight for animal studies), treatment and control group characteristics, calcium channel blocker involved (type, dose, route, form), treatment(s) provided, outcomes and results. To ensure uniformity, an independent individual merged the data collection into a single flow sheet.

Two independent reviewers carried out quality analysis for all of studies except case reports. Disagreements were resolved by consensus or by a third party if required. The

Cochrane risk of bias tool (Higgins et al., 2011) was not required because no controlled trials were found. The STROBE checklist (Von Elm et al., 2008) and the Thomas tool

(National Collaborating Centre for Methods and Tools, 2008) were used for observational studies. The Institute of Health Economics tool (Moga et al., 2012) was used for case series, and the ARRIVE guidelines (Kilkenny et al., 2010) and the modified

NRCNA checklist (Institute of Medicine and National Research Council of National

Academies, 2005) for animal studies. The percent agreement was documented for each item, and the kappa statistic was calculated for each type of quality analysis tool.

Qualitative synthesis was used to summarize the evidence for each outcome. A planned meta-analysis was not performed due to heterogeneity of studies, interventions and outcomes.

6.3 Results

The search strategy identified 15,577 citations. Two reviewers selected 216 articles

(kappa on articles inclusion = 0.85, 95% CI 0.73 to 0.89) (Figure 11). Six full-text articles were not found because the foreign language journals were inaccessible (Gadow et al.,

1995; Gutschmidt et al., 1995; Samtleben et al., 1995; Oliver et al., 1992; Pap et al.,

1989; Kandil et al., 1984). Professional translators translated 23 manuscripts. A list of the articles translated and excluded after full-text review is available upon request.

Figure 9 Flow diagram 87

No controlled trial fulfilling eligibility criteria were identified. Human observational studies were published only for high-dose insulin (Musselman et al., 2011; Bryant et al., 2009;

Greene et al., 2007) and extracorporeal life-support (Masson et al., 2012). Comparative studies included observational studies (Bryant et al., 2009; Greene et al., 2007) evaluating two different high-dose insulin regimens and one comparing high-dose insulin to vasopressors (Musselman et al., 2011). One human observational study compared extracorporeal life-support to standard therapy (Masson et al., 2012).

6.3.1 Results of individual studies and risks of bias for medical interventions

Table 11 describes the results for included articles by type of intervention, from highest to lowest level of evidence. A more detailed description of case reports for each intervention is also in appendix 2.

Table 11 Results of individual studies and risks of bias Source Intervention Participants Outcomes Methodological Quality

Observational studies STROBE and Thomas' tool Musselman HDI1 +/- glucagon vs 20 with beta-blocker or CCB - hemodynamics: no significant STROBE: 6/22 et al., 2011 vasopressors only poisonings requiring difference in MAP (no power Thomas' tool: vasopressors (10 in the calculation) • Moderate: Selection intervention group) - hospital or ICU LOS: no signficant bias difference (no power calculation) • Weak: Cohort methods, confounders, blinding, data collection, withdrawals, analysis, intervention integrity Bryant et HDI 0.5-1.0 units/kg 46 poisoned with a CCB - mortality: higher when the HDI is not STROBE: 4/22 al., 2009 bolus followed by hemodynamically unstable provided as per the protocol Thomas' tool: 0.5-1.0 units/kg/h treated with HDI (19 in the • Weak: Selection infusion started intervention group) bias, case-control methods, preceeding or confounders, blinding: weak, shortly after data collection, withdrawals, vasopressors vs analysis, intervention integrity other form of providing HDI Greene et HDI 0.5-2.0 7 poisoned with a CCB - mortality: 1/7 (did not receive a STROBE: 9/22 al., 2007 units/kg/h with 1 hemodynamically unstable bolus) Thomas' tool: unit/kg bolus vs not treated with HDI (3 in the - hemodynamics: increase in more • Moderate: Selection intervention group) than 10 mmHg SBP only in the group bias, confounders, blinding, receiving a bolus withdrawals - mean ICU LOS 2.7 days • Weak: Cohort - adverse effects: 1 non-clinically methods, data collection, significant hypoglycemia (no bolus) analysis, intervention integrity and 2 non-clinically significant hypokalemia (1 with bolus, 1 without) Case series Institute of Health Economics tool for quality of case series Espinoza et HDI 0.5-1 units/kg 46 poisoned with CCB - mortality: 9/46 10/20 (2 unclear) al., 2013 bolus followed by - adverse effects: no hypoglycemia 0.5-1 units/kg/h Holger et HDI bolus with 4 poisoned with amlodipine, - mortality: 0/4 al., 2011 dextrose followed by verapamil or diltiazem - hemodynamics: vasopressors could 13/20 (4 unclear) 1 unit/kg/h, be tapered off with HDI increasing Q15 min - ICU LOS 2, 3, 5 and 30 days by 1-2 units/kg/h up - adverse effects: 2/4 hypoglycemia, to 10 units/kg/h 2/4 hypokalaemia Boyer et al., HDI 0.5-1 units/kg/h 3 poisoned with diltiazem, - mortality: 0/3 2002 amlodipine or verapamil - hemodynamics: improvement in 5/20 (5 unclear) blood pressure within 30 min - adverse effects: hypoglycemia, hypokalaemia Boyer et al., HDI 0.5 units/kg/h 2 poisoned with amlodipine - hemodynamics: rapid reversal of 2001 or diltiazem hemodynamic collapse in both patients 4/20 (4 unclear) Yuan et al., HDI 4-70 units/h 4 poisoned with verapamil - mortality: 0/3 (1 not reported) 1999 SR or amlodipine - hemodynamics: improvement in 4/20 (8 unclear) blood pressure in all cases - hospital LOS 5-14 days - adverse effects: hypoglycemia

1 HDI: high-dose insulin 89

Source Intervention Participants Outcomes Methodological Quality

Case reports (see appendix 2): - Improvement in hemodynamics reported in 20 cases - No improvement in hemodynamics reported in 3 cases - Adverse effects: Hypoglycemia reported in 1 case (Ortiz-Nunoz et al., 2005) Animal studies ARRIVE guidelines and Modified NRCNA Engebretsen Group 1: NS 15 Workshire pigs poisoned - mortality: 4/5 deaths in group 1 et al., 2010 Group 2: HDI 2 with nifedipine 0.0125 compared to 1/5 in group 2 and no ARRIVE: 18/20 units/kg/h mcg/kg/min until 25% of death in group 3 increased Q10 min baseline MAP X CO - hemodynamics: groups 2-3 had NRCNA: 8/16 by 2 units/kg/h ad significant improvement in MAP max 10 units/kg/h compared to group 1, but no Group 3: HDI and significant difference in phenylephrine hemodynamics between group 2 and 3 - adverse effects: hypoglycemia and hypokalaemia Kline et al., Group 1: NS 20 Mongrel dogs - hemodynamics: significant 1997 Group 2: randomized to one of the 4 improvement in heart rate and blood ARRIVE: 17/20 epinephrine groups, poisoned with pressure compared to all other Group 3: glucagon intraportal verapamil infusion groups NRCNA: 9/16 Group 4: HDI 1000 0.04 mg/kg/min X1h then nU/min with 0.08 mg/kg/min X1h then 0.1 dextrose mg/kg/min X1h then 0.2 mg/kg/min until death Kline et al., Group 1: NS 18 Mongrel dogs - hemodynamics: significant 1996 Group 2: HDI randomized to one of the 3 improvement in myocardial ARRIVE: 15/20 Group 3: HDI groups poisoned with contractile function independent of without verapamil intraportal verapamil infusion glucose transport compared to NRCNA: 7/16 toxicity 0.04 mg/kg/min X1h then normal saline 0.08 mg/kg/min X1h then 0.1 mg/kg/min until death Kline et al., Group 1: NS 24 Mongrel dogs poisoned - mortality: 6/6 in group 1, 2/6 in 1993 Group 2: with verapamil infusion 0.1 group 2, 3/6 in group 3, 0/6 in group ARRIVE: 11/20 epinephrine mg/kg/min until 50% 4 (significant) Group 3: glucagon reduction of MAP or - hemodynamics: significant NRCNA: 6/16 Group 4: HDI 4 complete AV block, then improvement in group 2 and 4, units/min +/- vitals kept constant for 30 restoration of sinus rhythm in group dextrose and min, then infusion at 1 3 potassium mg/kg/h until 4h or death - adverse effects: increased lactate and glycemia with epinephrine, hyperglycemia followed by hypoglycemia with glucagon

Source Intervention Participants Outcomes Methodological Quality

Observational studies STROBE and Thomas’ tool Masson et ECLS (mean 62 cardiac arrests or - mortality: 9/11 without ECLS STROBE: 17/22 al., 2012 duration of 6 days severe shock secondary to compare to 23/41 with ECLS Thomas' tool: +/- 2.9 days) vs not poisoning (16 CCB (patients in severe shock) and 0/7 • Strong: Selection bias including verapamil, without ECLS vs 3/3 with ECLS • Moderate: Cohort diltiazem) (patients in cardiac arrest) methods, confounders, blinding, (significant difference) withdrawals, analysis - adverse effects: 4 limb ischemia, 1 • Weak: Data collection IVC thrombus and 2 cases of • Weak to moderate: bleeding requiring surgical revision Intervention integrity

Case series Institute of Health Economics tool for quality of case series and quality of reporting Daubin et ECLS (mean 17 patients with persistent - mortality: 4/17 deaths (2 withdrawal al., 2009 duration of 4.5 days cardiac arrest or severe of life support, 2 from septic shock 13/20 (2 unclear) +/- 2.4 days) shock secondary to and neurological damage) poisoning refractory to - function: 13/17 survived without conventional therapy (4 cardio-vascular or neurologic CCB poisonings with sequelae verapamil) - adverse effects: 10 cannulation- related injuries, 6 limb ischemia with requirement for urgent revascularization in 3/6 (no more cases reported after arterial shunt was added to the cannulation technique), 1 femoral thrombus, 1 IVC thrombus, 2 bleeding at the cannulation site requiring surgical revision Mégarbane ECLS (5-108h, 12 prolonged out-of- - mortality: survival 50% at 24h and et al., 2007 mean duration of hospital cardiac arrests 25% at hospital discharge but none 15/20 (3 unclear) 56h) secondary to poisoning (2 of the CCB poisoning survived CCB, verapamil) - function: survivors (3) were symptom free without deficit at 1 year - ICU LOS: 12-14 days - adverse effects: none noticed Babatasi et ECLS (48-71h, 6 cardiac arrests - mortality: 2/6 al., 2001 mean duration secondary to a cardiotoxic - adverse effects: limb ischemia 9/20 (9 unclear) 59.25+/- 2 hours) drug (2 CCB with (3/6), retroperitoneal hematoma 1/6 verapamil) Case reports (see appendix 2): - Survival without neurologic or cardiac deficit reported in 8 cases - Death reported in 1 case (Schmidt et al., 1995) - Adverse effects: Leg amputation reported in 1 case (Tauro et al., 2009)

Source Intervention Participants Outcomes Methodological Quality

Case series Institute of Health Economics tool for quality of case series and quality of reporting Konca et Calcium 5/7 poisoned with CCB - mortality: 1/5 6/20 (2 unclear) al., 2013 receiving calcium Supradip et Calcium gluconate 2 poisoned with - mortality: 0/2 al., 2008 bolus followed by an amlodipine - hemodynamics: no change in case 9/20 (13 unclear) infusion at 10 mg/h 1 and deterioration in case 2 - hospital LOS: 5 and 10 days Karti et al., Calcium gluconate 2 poisoned with verapamil - mortality: 0/2 2002 10% 10 and 20 ml IV SR - hemodynamics: no improvement 7/20 (5 unclear) Howarth et Calcium chloride or 15 poisoned with - mortality: 4/15 al., 1994 gluconate infusion to verapamil, verapamil SR, - hemodynamics: atropine was 9/20 (3 unclear) maintain a calcemia of diltiazem or nimodipine effective only after calcium 4 mmol/L administration Parikka et Calcium chloride 1-3g 7 poisoned with diltiazem, - mortality: 5/7 al., 1993 IV administered to 4/7 verapamil or nifedipine 3/20 (6 unclear) patients Ramoska Calcium 4.5-95.2 113 poisoned with - hemodynamics: 10/14 reverse of 1993 mmol IV (23/113 verapamil, diltiazem or AVB, 7/11 increased their heart rate, 10/20 (3 unclear) received calcium) nifedipine including 5 16/20 increased their blood pressure sustained-relase Roper et al., Calcium received in 4 fatal poisonings with - mortality: 4/4 1993 1/4 patient diltiazem 3/20 (3 unclear) Bausch et Calcium 3 poisoned with verapamil - mortality: 2/3 al., 1991 (1 had also nefedipine) 4/20 (5 unclear) Horowitz et Calcium chloride 1-5g 2 poisoned with verapamil - mortality: ½ al., 1989 IV - hemodynamics: no change in case 8/20 (4 unclear) 1 and blood pressure improvement in case 2 Henry et al., Calcium chloride 1-2 g 2 poisoned with verapamil - mortality: 0/2 1985 IV - hemodynamics: improved blood 7/20 (3 unclear) pressure Jager et al., Calcium gluconate IV 11 poisoned with - adverse effects: 1 case had more 1984 verapamil arrhythmias after the administration 2/20 (6 unclear) of calcium gluconate Case reports (see appendix 2): - Improvement in hemodynamics reported in 19 cases - No improvement in hemodynamics reported in 3 cases - Adverse effects: Hypercalcemia reported in 1 case (Buylaert et al., 1995)

Source Intervention Participants Outcomes Methodological Quality

Animal studies ARRIVE guidelines and Modified NRCNA Graudins et Group 1: NS 60 Wistar rats (6 groups of - mortality 0-1/10 as opposed to 1- al., 2010 Group 2: 10) poisoned with 2/10 in the other groups ARRIVE: 16/20 levosimendan verapamil 6 mg/kg/h until - hemodynamics: improvement in Group 3: 50% decrease in MAP cardiac output and blood pressure NRCNA: 5/16 levosimendan + 4-AP then 4 mg/kg/h Group 5: calcium chloride loading dose and infusion Group 6: levosimendan and calcium chloride Graudins et Group 1: NS 35 Wistar rats (5 groups of - mortality 1/7 in group 2 and 0/7 in al., 2008 Group 2: calcium 7) poisoned with verapamil group 5 as opposed to 5/7 in group 1 ARRIVE: 16/20 chloride 6 mg/kg/h until 50% - hemodynamics: improvement in Group 3: decrease in MAP then 4 blood pressure and more stability NRCNA: 5/16 levosimendan mg/kg/h Group 4: higher dose of levosimendan Group 5: levosimendan and calcium chloride Strubelt et Group 1: NS 31 Wistar rats (in 6 - mortality: 152% improvement in al., 1990 Group 2: dopamine groups) poisoned with survival with group 6 (138 min vs ARRIVE: 12/20 Group 3: nisoldipine 0.1 mg/kg/min 54.7 min +/-11.1 min) norepinephrine - hemodynamics: normalization of NRCNA: 5/16 Group 4 isoproterenol cardiac output and MAP Group 5: Polygeline Group 6: Calcium chloride Gay et al., Group 1: calcium 23 Mongrel dogs poisoned - hemodynamics: calcium increased 1986 chloride with verapamil bolus 0.72 aortic pressure and left ventricular ARRIVE: 11/20 Group 2: isoproterenol mg/kg followed by an pressure Group 3: atropine infusion of 0.11 mg/kg/min NRCNA: 6/16 Group 4: epinephrine Group 5: norepinephrine Group 6: dopamine Group 7: phenylephrine Group 8: 4-AP

Source Intervention Participants Outcomes Methodological Quality

Strubelt et al., Rats: 95 Wistar rats (in 16 - mortality: 100% improvement with 1986 Group 1: NS groups) poisoned with calcium, isoproterenol or dopamine ARRIVE: 13/20 Group 2: calcium nifedipine 0.5 mg/kg/min in rats, survival also improved with chloride 5 mg/kg/min and 21 bastard rabbits calcium in rabbits, but not with NRCNA: 4/16 Group 3: calcium (in 5 groups) poisoned isoproterenol or dopamine chloride 10 with nifedipine 0.2 - hemodynamics: improvement in mg/kg/min mg/kg/min cardiac output and blood pressure Group 16: calcium with calcium chloride and isoproterenol Rabbits: Group 1: NS Group 2: calcium chloride Group 5: calcium chloride and isoproterenol Strubelt et al., Group 1: calcium 80 Wistar rats (in 11 - mortality: tripled with calcium 1984 Group 2: groups) poisoned with chloride and vasopressors ARRIVE: 11/20 epinephrine verapamil 0.15 - hemodynamics: calcium increased Group 3: dopamine mg/kg/min the blood pressure and LV dp/dt but NRCNA: 5/16 Group 4: not heart rate as opposed to norepinephrine vasopressors Groups 5-6: isoproterenol Group 7: orciprenaline Group 8: prenalterol Group 9: angiotensine amide Group 10: atropine Group 11: haemacel Wesseling et Group 1: NS 12 rabbits (in 3 groups) - mortality: no difference al., 1983 Group 2: calcium- poisoned with verapamil ARRIVE: 10/20 levulate 0.4 ml of 30 mg/kg/h until the SBP 10% drops by one third NRCNA: 7/16 Group 3: 4- aminopyridine 1mg/kg after the drop in SBP

Vick et al., Rabbits calcium 42 rabbits (12 in - hemodynamics: calcium chloride 1983 chloride 300 mg/kg treatment group) restored blood pressure but not ARRIVE: 5/20 after the verapamil poisoned with verapamil heart rate infusion 0.1 mg/kg/min for 15 min NRCNA: 3/16 Pigs 150 mg/kg after preceded by propranolol the verapamil 1 mg/kg and 13 pigs (5 infusion in treatment group) poisoned with verapamil 0.5 mg/kg preceded by propranolol 0.5 mg/kg Case series Institute of Health Economics tool for quality of case series and quality of reporting Levine et al., Vasopressors 48 patients poisoned - mortality: 1/48 14/20 (0 unclear) 2013 with verapamil or - adverse effects: 8 ischemic diltiazem complications in 5 patients, 3 cardiac arrests before the use of vasopressors Konca et al., Dopamine 4/7 poisoned with CCB - mortality: ¼ 6/20 (2 unclear) 2013 on dopamine

Source Intervention Participants Outcomes Methodological Quality Kanagarajan Vasopressin 2.4-4.8 2 poisoned with - mortality: 0/2 et al., 2007 units/h amlodipine and verapamil - hemodynamics: improvement in 9/20 (3 unclear) sustained-release blood pressure for both - hospital LOS 22 and 35 h - vasopressors duration of 36h and 40h Groszek et al, Dopamine 2 poisoned with nifedipine - mortality: 1/2 2003 - hemodynamics: 1/2 improvement in 2/20 (4 unclear) blood pressure Karti et al., Dopamine 2 poisoned with verapamil - mortality: 0/2 2002 SR - hemodynamics: 1/2 improvement in 7/20 (5 unclear) blood pressure Ramoska Dopamine, 113 poisoned with - hemodynamics: 0/10 increase in 1993 isoproterenol verapamil, diltiazem or heart rate or blood pressure with 10/20 (3 unclear) nifedipine (10 received dopamine, 2/3 increase in heart rate vasopressors) with isoproterenol Parikka et al., Dopamine, 7 poisoned with diltiazem, - mortality: 5/7 1993 dobutamine, verapamil or nifedipine - hemodanymics: no improvement 4/20 (5 unclear) epinephrine, isoproterenol, prenalterol Howarth et Dopamine, 15 poisoned with - mortality: 4/15 al., 1994 adrenaline verapamil, diltiazem or - hemodynamics: variable response 9/20 (3 unclear) nifedipine Jaeger et al., Dopamine, 3 poisoned with diltiazem - mortality 0/3 deaths 1990 metaraminol - hemodynamics: 2/3 improvement 2/20 (6 unclear) Sauder et al., Dopamine except 1 6 poisoned with verapamil - mortality: 0/6 1990 patient had - hemodynamics: improvement in the 3/20 (4 unclear) epinephrine and only patient described isoproterenol Case reports (see appendix 2): - Improvement in hemodynamics reported in most reported cases except 2 - Adverse effects: None reported Animal studies ARRIVE guidelines and Modified NRCNA Engebretsen Group 1: NS 15 Workshire pigs - mortality: 4/5 deaths in group 1 et al., 2010 Group 2: HDI 2 poisoned with nifedipine compare to 1/5 in group 2 and no ARRIVE: 18/20 units/kg/h increased 0.0125 mcg/kg/min until death in group 3 Q10 min by 2 25% of baseline MAP X - hemodynamics: significant NRCNA: 8/16 units/kg/h ad max 10 CO improvement in MAP compared to units/kg/h group 1 and no significant difference in Group 3: HDI and hemodynamics between group 2 and phenylephrine 3, the addition of phenylephrine did not make a difference - adverse effects: hypoglycemia and hypokalaemia

Source Intervention Participants Outcomes Methodological Quality

Barry et al., Group 1: NS 18 swines (9 per group) - mortality: 4/8 in the vasopressin 2005 Group 2: vasopressin group vs 2/10 in the control group ARRIVE: 18/20 0.01 units/kg/min (p=0.32) - hemodynamics: no significant NRCNA: 9/16 improvement in MAP, cardiac output or systemic vascular resistance Sztajnkrycer Group 1: NS 12 dogs (6 per group) - mortality: 0/6 in the vasopressin et al., 2004 Group 2: vasopressin group vs 2/6 in the control group ARRIVE: 15/20 0.004 units/kg/min - hemodynamics: no significant increasing to 0.01 difference NRCNA: 6/16 units/kg/min then 0.04 units/kg/min Q 20 min Stone et al., Group 1: NS 18 Mongrel dogs (6 per - mortality: 2/6 deaths in group 1 vs 1996 Group 2: glucagon group) none in the other groups ARRIVE: 15/20 Group 3: glucagon - hemodynamics: significant and phenylephrine increase in MAP in group 3 NRCNA: 6/16 100-180 ug/min compared to group 1, significant increase in cardiac output in group 2-3 compared to group 1 Kline et al., Group 1: NS 24 Mongrel dogs - mortality: 2/6 deaths in group 1 vs 1993 Group 2: epinephrine none in the other groups ARRIVE: 11/20 Group 3: glucagon - hemodynamics: significant Group 4: HDI 4 increase in MAP in group 3 NRCNA: 6/16 units/min compared to group 1, significant increase in cardiac output in group 2-3 compared to group 1 - adverse effects: increased lactate and glycemia with epinephrine Strubelt et Group 1: NS 31 Wistar rats (in 6 - mortality: 135% improvement in al., 1990 Group 2: dopamine groups) poisoned with survival with group 2 (129.2 min vs ARRIVE: 12/20 Group 3: nisoldipine 0.1 mg/kg/min 54.7 min +/-11.1 min), 75.6 min vs norepinephrine 54.7 min in group 3, 187% NRCNA: 5/16 Group 4 isoproterenol improvement in group 4 (157 min vs Group 5: Polygeline 54.7 min) Group 6: Calcium - hemodynamics: groups 2-3-4 chloride showed an increase in blood pressure, cardiac output and stroke volume, group 3 showed an increase in vascular resistances and group 4 a decrease resistance Gay et al., Group 1: calcium 23 Mongrel dogs poisoned - Epinephrine, norepinephrine, 1986 chloride with verapamil bolus 0.72 dopamine and high doses of ARRIVE: 11/20 Group 2: isoproterenol mg/kg followed by an phenylephrine increase cardiac Group 3: atropine infusion of 0.11 mg/kg/min output, AO and LV dp/dt increased NRCNA: 6.5/16 Group 4: epinephrine Group 5: norepinephrine Group 6: dopamine Group 7: phenylephrine Group 8: 4-AP

Source Intervention Participants Outcomes Methodological Quality

Strubelt et al., Group 1: NS 95 Wistar rats (in 16 - mortality: 100% improvement with 1986 Group 2: calcium groups) poisoned with calcium, isoproterenol or dopamine ARRIVE: 13/20 chloride 5 mg/kg/min nifedipine 0.5 mg/kg/min in rats, survival also improved with Group 3: calcium and 21 bastard rabbits (in calcium in rabbits, but not with NRCNA: 4/16 chloride 10 5 groups) poisoned with isoproterenol or dopamine; 50% mg/kg/min nifedipine 0.2 mg/kg/min improvement after epinephrine, Group 16: calcium norepinephrine chloride and - hemodynamics: improvement in isoproterenol cardiac output and blood pressure Rabbits: with most of the vasopressors Group 1: NS Group 2: calcium chloride Group 5: calcium chloride and isoproterenol Strubelt et al., Group 1: calcium 80 Wistar rats (in 11 - mortality: tripled with epinephrine, 1984 Group 2: groups) poisoned with dopamine, norepinephrine, ARRIVE: 11/20 epinephrine verapamil 0.15 mg/kg/min isoproterenol and orciprenaline Group 3: dopamine - hemodynamics: sympathomimetics NRCNA: 5/16 Group 4: increased the blood pressure, heart norepinephrine rate and cardiac output; epinephrine Groups 5-6: and isoproterenol antagonized the isoproterenol decrease in LC dp/dt max Group 7: orciprenaline Group 8: prenalterol Group 9: angiotensine Group 10: atropine Group 11: haemacel Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Lo et al., 2012 Whole bowel 176 children (mean age 2 - mortality: 0/176 irrigation years old) including 57 - adverse effects: 16 patients 8/20 (1 unclear) poisoned with a CCB vomited and 1 had abdominal pain Cumpston et Whole bowel 1 poisoned with diltiazem, - mortality: 1/2 al., 2010 irrigation 1 poisoned with verapamil - adverse effects: 2/2 cardiac arrest 8/20 (4 unclear) SR after vomiting Belson et al., Activated charcoal, 283 poisoned children - mortality: 0/283 and 209/283 never 2000 gastric lavage, (mean age 27 months) developed toxicity 12/20 (1 unclear) whole bowel including 109 cases of irrigation CCB Barrow et al., Gastric lavage, 2 poisoned with verapamil - mortality: 0/2 1994 activated charcoal SR 8/20 (3 unclear) and whole bowel irrigation Buckley et al., Activated charcoal 3 poisoned with verapamil - mortality: 1/3 1993 +/- lavage (2/3) and SR - adverse effects: 1/3 cardiac arrest 8/20 (5 unclear) whole bowel after whole bowel irrigation initiation irrigation (2/3) Bausch et al., Gastric lavage 2 poisoned with verapamil, - mortality: 2/3 1991 1 poisoned with nifedipine - morbidity: 1/3 pulmonary edema, 4/20 (5 unclear) and verapamil 1/3 acute kidney injury Lambert et al., Gastric lavage 134 poisoned with - non-conclusive, hypothesis of 1990 (50/154), activated diltiazem decreased toxicity 6/20 (4 unclear) charcoal (6/134)

Source Intervention Participants Outcomes Methodological Quality

Sauder et al., Gastric lavage (3/6) 6 poisoned with verapamil - mortality: 0/6 1990 3/20 (4 unclear) Case reports (see appendix 2): - Repeated doses of activated charcoal did not increase elimination of diltiazem in 1 case (Roberts et al., 1991) - Adverse effects: Cardiac arrest secondary to verapamil poisoning 2h after gastric lavage (Madera et al., 1977) Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Stachon et al., Endovenous 12 (10 with pacemaker) - hemodynamics: capture and 2011 pacemaker poisoned patients with non- pacing problems 4/20 (2 unclear) dihydropyridine Lee et al., Pacemaker 2 children (2 year old and 14 - mortality: 2/2 2000 months old) poisoned with 5/20 (2 unclear) nifedipine Ramoska et Pacemaker 113 (4 with pacemaker) - hemodynamics: 2/4 al., 1993 poisoned patients with a improvement 10/20 (3 unclear) CCB McGlinchey et Pacemaker 6 poisoned patients, 3 with a - mortality: 1/3 al., 1998 CCB - hemodynamics: 2/3 9/20 (3 unclear) improvement (1 had no capture) Beiträge 1984 Pacemaker 2 patients poisoned with - mortality: 2/2 verapamil - hemodynamics: improvement 4/20 (1 unclear) Immonen et Endovenous 3 patients poisoned with - mortality: 0/3 al., 1981 pacemaker verapamil - hemodynamics: 1 had transient 7/20 (5 unclear) improvement and 1 had sustained improvement

Case reports (see appendix 2): - Improvement in hemodynamics reported in most reported cases - Adverse effects: None reported Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Supradip et Glucagon bolus 10 3 poisoned with amlodipine - mortality: 0/2 al., 2008 mg IV followed by - hemodynamics: no 9/20 (13 unclear) an infusion of 3 improvement in one case and mg/h deterioration in the second one Love et al., Glucagon bolus 3 5/9 patients poisoned with - hemodynamics: improvement 1998 mg followed by an verapamil in 8/9 8/20 (2 unclear) infusion of 3 mg/h Roper et al., Glucagon 4 poisoned patients with - mortality: 4/4 1993 diltiazem (only 1 received 3/20 (3 unclear) the glucagon) Case reports (see appendix 2): - Improvement in hemodynamics reported in 8 cases - No improvement in hemodynamics (Adams et al., 1998) - Adverse effects: Vomiting (Love et al., 1998; Cohen et al., 2011; Mahr et al., 1997) and hyperglycemia (Ashraf et al., 1995; Buylaert et al., 1995; Doyon et al., 1993)

Source Intervention Participants Outcomes Methodological quality Animal studies ARRIVE guidelines and Modified NRCNA Stone et al., - Group 1: control 18 Mongrel dogs (6 - mortality: 2 deaths in the control group 1996 - Group 2: glucagon per group) poisoned - hemodynamics: non-significant ARRIVE: 15/20 5 mg IV followed by with verapamil 15 improvement in group 2 and significant infusion of 5 mg /h mg/kg over 30 min improvement in group 3; significant NRCNA: 6/16 - Group 3: glucagon increase in cardiac output in groups 2-3 5 mg followed by infusion of 5 mg + phenylephrine infusion Stone et al., - Group 1: control 15 Mongrel dogs - mortality: 3/7 deaths in the control ARRIVE: 14/20 1995 (n=7) poisoned with group - Group 2: glucagon verapamil 25 mg/kg - hemodynamics: significant increase in NRCNA: 6/16 2.5 mg IV bolus over 30 min cardiac output and heart rate followed by infusion of 2.5 mg/h X 60 min (n=8) Kline et al., - Group 1: normal 24 Mongrel dogs (6 - mortality: 6/6 in group 1, 2/6 in group 2, 1993 saline per group) poisoned 3/6 in group 3 and 0/6 in group 4 ARRIVE: 11/20 - Group 2: with verapamil 0.1 - hemodynamics: significant increase in epinephrine mg/kg/min until 50% heart rate but not in blood pressure NRCNA: 6/16 - Group 3: glucagon reduction in average - adverse effects: hyperglycemia 0.2-0.25 mg/kg, MAP or AV block infusion 150 followed by an infusion mcg/kg/h at 1 mg/kg/h - Group 4: HDI 4 units/min Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Ramoska Atropine 113 patients poisoned - hemodynamics: 2/7 increase heart 1993 with a CCB (7 rate 10/20 (3 unclear) received atropine) Roper et al., Atropine 4 patients poisoned - mortality: 4/4 1993 with diltiazem (only 1 3/20 (3 unclear) received the glucagon) Howarth et Atropine 15 patients poisoned - mortality: 4/15 al., 1994 with a CCB - hemodynamics: effective only with 9/20 (3 unclear) calcium Animal studies ARRIVE guidelines and Modified NRCNA Gay et al., Group 1: calcium 23 Mongrel dogs - hemodynamics: 1986 Group 2: isoproterenol poisoned with significant ARRIVE: 11/20 Group 3: atropine 0.04 mg/kg bolus verapamil bolus of increase in heart Group 4: epinephrine 0.72 mg/kg rate, aortic NRCNA: 6/16 Group 5: norepinephrine followed by a pressure, cardiac Group 6: dopamine continuous output Group 7: phenylephrine infusion of 0.11 Group 8: 4-aminopyridine mg/kg/min Strubelt et al., Group 1: calcium 80 Wistar rats - mortality: no 1984 Group 2: epinephrine poisoned with effect ARRIVE: 11/20 Group 3: dopamine verapamil 0.15 Group 4: norepinephrine mg/kg/min NRCNA: 5/16 Groups 5-6: isoproterenol Group 7: orciprenaline Group 8: prenalterol Group 9: angiotensin amide Group 10: atropine 0.05-0.1 mg/kg/min

Source Intervention Participants Outcomes Methodological quality

Case series (Institute of Health Economics tool for quality of case series and quality of Fiszer et al., 7.2-14.4 mg of 4- 3 patients poisoned with a - mortality: 0/3 reporting) 2007 aminopyridine IV non-dihydropyridine - hemodynamics: improvement 8/20 (4 unclear) Magdalan et 10 mg of 4- 3 patients poisoned with - mortality: 0/3 al., 2003 aminopyridine verapamil - hemodynamics: improvement 9/20 (4 unclear) Animal studies ARRIVE guidelines and Modified NRCNA Graudins et Group 1: NS 60 Wistar rats (6 groups of - mortality 1-2/10 as opposed to 6/10 al., 2010 Group 2: 10) poisoned with verapamil in control group ARRIVE: 16/20 levosimendan 6 mg/kg/h until 50% - hemodynamics: improvement in Group 3: decrease in MAP then 4 cardiac output and blood pressure NRCNA: 5/16 levosimendan + 4- mg/kg/h AP (bolus and infusion) Group 4: levosimendan and 4-AP Group 5: calcium Group 6: levosimendan and calcium Magdalan Group 1: NS 50 Wistar rats (5 groups of - mortality: 3/10 in group 4 as 2003 Group 2: calcium 10) poisoned with verapamil opposed to 10/10 in the control ARRIVE: 15/20 Group 3: 150 mcg/kg/h until 50% group, 8/10 in group 2, 6/10 in group epinephrine reduction of mean arterial 3 or 5/10 in group 5 NRCNA: 5/16 Group 4: 4- pressure and/or heart rate - hemodynamics: improvement aminopyridine Group 5: Bay K8644 Tuncok et al., Group 1: NS 26 Wistar rats poisoned with - mortality: 0 1998 Group 2: 4- verapamil 25 mg/kg/h until - hemodyamics: improvement in ARRIVE: 17/20 aminopyridine 1 45-60% reduction of mean group 3 but not group 2 mg/kg/h arterial pressure and 30% - adverse effects: seizures, NRCNA: 5/16 Group 3: 4- reduction in heart rate bronchorrhea, fasciculations aminopyridine 2 mg/kg/h Group 4: BayK 8644 0.3 mg/kg/h Group 5: BayK 8644 0.6/mg/kg/h Plewa et al., Group 1: NS 18 swines (9 per group) - mortality: no difference 1994 Group 2: 3,4 poisoned with verapamil 10 - hemodynamics: improvement ARRIVE: 16/20 diaminopyridine 1 mg/kg/h for 5 min then 5 - adverse effects: hypertension, mg/kg/min mg/kg/h until a SBP of 55 tachycardia, muscle twitching NRCNA: 8/16 mmHg was achieved Korstanje et Group 1: control 23 New Zealand white - mortality: 2/5 in group 4 as al., 1987 (n=5) rabbits poisoned with opposed to 5/5 in group 1, 1/9 in ARRIVE: 14/20 Group 2: Bay K8644 verapamil 0.1 mg/min for 10 group 2 and 3/6 in group 3 (n=9) min then 0.2 mg/min up to 1 - hemodynamics: no improvement NRCNA: 6//16 Group 3: CGP (n=6) mg/min Goup 4: 4 aminopyridine (n=5) Gay et al., Group 1: control 12 Mongrel dogs (7 in the 4- - mortality: 0/7 deaths 1986 Group 2: 4- aminopyridyne group) - hemodynamics: improvement ARRIVE: 11/20 aminopyrydine 0.5 poisoned with verapamil 90 mg/kg bolus X 3 mcg/kg bolus followed by an NRCNA: 6//16 infusion of 14 mcg/kg/min Agoston et al., Group 1: control 12 cats (6 per group) - mortality: 0/7 as opposed to 4/6 in 1984 Group 2: 4- poisoned with verapamil 8 the control group ARRIVE: 11/20 aminopyridine 0.5 mg/kg/h infusion - hemodynamics: improvement mg/kg IV then 2.5 NRCNA: 6//16 mg

100

Source Intervention Participants Outcomes Methodological Quality

Wesseling Experiment 1: 12 rabbits (4 per group) - mortality: no difference et al., 1983 - group 1: NS poisoned with verapamil - hemodynamics: improvement in ARRIVE: 10/20 - group 2: calcium 30 mg/kg/h until SBP the first experiment - group 3: 4- decreases by 1/3 NRCNA: 7//16 aminopyridine 1 mg/kg Experiment 2: - group 1: NS - group 2: 4- aminopyridine 0.3 mg/kg - group 3: 4- aminopyridine 1 mg/kg Case series (Institute of Health Economics tool for quality of case series and quality of reporting)

Geib et al., Lipid emulsion 20% 5 patients poisoned with a - mortality: 3/5 2009 bolus X 2-3 CCB - adverse effects: non observed 8/20 (4 unclear) Case reports (see appendix 2): - Improvement in hemodynamics reported in 20 cases - No improvement in hemodynamics in 2 cases (West et al., 2010; St-Onge et al. 2010) - Adverse effects: Hypertriglyceridemia (Armenian et al., 2010) and another case without being able to obtain ABG, SatO2, CBC, lytes; extreme lipemia and hyponatremia (N.B.: 2 L given by mistake) Animal studies ARRIVE guidelines and Modified NRCNA Tebbutt et al., Group 1: NS 30 Winstar rats (15 per - mortality: prolonged survival and 2006 Group 2: Lipid group) poisoned with higher median lethal dose in the lipid ARRIVE: 13/20 emulsion 20% 12.4 verapamil 37.5 mg/kg IV emulsion group mg/kg over 5 min - hemodynamics: less decrease in NRCNA: 6//16 heart rate in the lipid emulsion group Perez et al., Group 1: control 30 Sprague-Dawley rats - mortality: dose-response for 2008 group poisoned with verapamil survival up to 18.6 ml/kg ARRIVE: 13/20 Group 2: Lipid infusion 15 mg/kg - hemodynamics: improvement in emulsion 20% 6.2 blood pressure NRCNA: 7//16 ml/kg over 2.5 min Group 3: Lipid emulsion 20% 12.4 ml/kg over 5 min Group 4: Lipid emulsion 20% 18.6 ml/kg over 7.5min Group 5: Lipid emulsion 20% 24.8 ml/kg over 10 min Group 6: Lipid emulsion 37.6 ml/kg over 15 min

101

Source Intervention Participants Outcomes Methodological Quality

Perichon et Group 1: control 30 rodents with oral - mortality: increased mortality al., 2013 group verapamil compared to other groups ARRIVE: 7/20 Group 2: Lipid - hemodynamics: lower heart rate emulsion 20% 4 and blood pressure compared to NRCNA: 6/16 ml/kg over 10 min the calcium group followed by an - sec outcomes: verapamil level infusion at 4 ml/kg/h 50% higher in the lipid emulsion Group 3: Calcium group chloride 0.2 mmol/kg Bania et al., Group 1: NS 14 Mongrel dogs poisoned - mortality: 0/7 as opposed to 6/7 2007 Group 2: Lipid with verapamil 6 mg/kg in the control group ARRIVE: 15/20 emulsion 20% 7 until 50% reduction in - hemodynamics: improvement in ml/kg over 30 min mean arterial pressure, blood pressure, but not in heart NRCNA: 7//16 then changed to 2 mg/kg/h rate, cardiac output or systemic vascular resistance Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Varpula 20091 Levosimendan 2 2 patients poisoned with a - mortality: 0/2 mcg/kg followed by CCB - hemodynamics: improvement in 8/20 (5 unclear) an infusion of 0.1- 2/2 0.2 mcg/kg/min - adverse effects: seizures Case reports (see appendix 2): - Improvement in hemodynamics reported in 3 cases - Adverse effects: none reported Animal studies ARRIVE guidelines and Modified NRCNA Abraham et Group 1: NS 14 Wistar rats (7 per - mortality: decreased survival in ARRIVE: 16/20 al., 2009 Group 2: group) poisoned with the levosimendan group levosimendan 12 verapamil 37.5 mg/kg/h NRCNA: 6//16 mcg/kg bolus then 18 mcg/kg Graudins et Group 1: NS 60 Wistar rats (6 groups of - mortality: 1-2/10 as opposed to al., 2010 Group 2: 10) poisoned with 6/10 in control group ARRIVE: 16/20 levosimendan verapamil 6 mg/kg/h until - hemodynamics: improvement in Group 3: 50% decrease in MAP cardiac output and blood pressure NRCNA: 5/16 levosimendan + 4- then 4 mg/kg/h AP (bolus and infusion) Group 4: levosimendan ant 4- AP Group 5: calcium bolus and infusion Group 6: levosimendan and calcium Kurola et al., Group 1: NS 12 Landrace pigs - mortality: 1/6 as opposed to 5/6 2010 Group 2: poisoned with verapamil in control group ARRIVE: 14/20 levosimendan 1.25 until reduction of 40% in - hemodynamics: improvement in mg bolus cardiac output then 12.5 4/6 compared to 1/6 in the control NRCNA: 8/16 mg/h group Graudins et Group 1: NS 35 Wistar rats (5 groups of - mortality 0/7 in group 3 and 5 as al., 2008 Group 2: calcium 7) poisoned with opposed to 5/7 in group 1 and 1/7 ARRIVE: 16/20 Group 3: verapamil 6 mg/kg/h until in group 2 and 4 levosimendan 50% decrease in MAP - hemodynamics: improvement in NRCNA: 5/16 Group 4: higher then 4 mg/kg/h blood pressure and more stability dose in group receiving calcium but not Group 5: levosimendan levosimendan and calcium Case series (Institute of Health Economics tool for quality of case series and quality of reporting) Kuhlmann et Plasma exchange 2 patients poisoned with - mortality: ½ al., 1999 1000 ml/h for a total verapamil - hemodynamics: improvement 9/20 (3 unclear) of 3000 ml - sec outcomes: decrease in levels Case reports (see appendix 2): - Improvement in hemodynamics and decrease in serum CCB concentrations reported in 3 cases 102

Gastrointestinal decontamination

Five human case series (Lo et al., 2012; Belson et al., 2000; Barrow et al., 1994;

Lambert et al., 1990; Sauder et al., 1990), including two pediatric studies (Lo et al.,

2012; Belson et al., 2000) reported the sequelae-free survival of all patients who underwent gastrointestinal decontamination (including activated charcoal, gastric lavage, or whole-bowel irrigation). Cardiac arrests following initiation of whole-bowel irrigation were documented in a case series of two hemodynamically unstable patients

(Cumpston et al., 2010), and following gastric lavage in one case report (Madera et al.,

1977). In all cases, complications occurred after the patient began vomiting. Given the nature of these reports, neither survival nor cardiac arrest can be attributed with confidence to the decontamination procedures.

High-dose insulin

High-dose insulin (IV bolus of 1.0 unit/kg followed by a 0.5-2.0 unit/kg/h infusion) showed an improvement in hemodynamics in one of two human observational studies

(Bryant et al., 2009; Greene et al., 2007), all five human case series (Espinoza et al.,

2013; Holger et al., 2011; Boyer et al., 2002; Boyer et al. 2001; Yuan et al., 1999), and all four animal studies (Engebretsen et al., 2010; Kline et al., 1997; Luomanmäki et al.,

1997; Kline et al., 1993) assessing that outcome, while a mortality benefit was reported in animal studies (Engebretsen et al., 2010; Kline et al., 1993). Hypoglycemia (2 of 4 subjects in Holger et al., 2011; and 1 of 7 subjects in Greene et al., 2007) and hypokalemia (2 of 7 in Greene et al., 2007; and 2 of 4 subjects in Holger et al., 2011) were reported as adverse effects in human cohort studies and case series, respectively.

103

Calcium

The majority of animal studies (Graudins et al., 2010; Graudins et al., 2008; Strubelt et al., 1990; Gay et al., 1986; Strubelt et al., 1986; Strubelt et al., 1984; Vick et al., 1983) evaluating use of calcium demonstrated a reduced mortality as well as hemodynamic improvement. Human case series (Konca et al., 2013; Supradip et al., 2008; Karti et al.,

2002; Howarth et al., 1994; Parikka et al., 1993; Ramoska et al., 1993; Roper et al.,

1993; Bausch et al., 1991; Jaeger et al., 1990; Horowitz et al., 1989; Henry et al., 1985) and case reports (Devasahayam et al., 2012; Rizvi et al., 2012; Hasson et al., 2011;

Marjanovic et al., 2010; Wills et al., 2010; Nanda et al., 2005; Felgenhauer et al., 2004;

Gerloni et al., 2004; Isbister et al., 2002; Lam et al., 2001; Adams et al., 1998; Li Saw et al., 1996; Buylaert et al., 1995; Ashraf et al., 1995; Lee et al., 1995; Luscher et al.,

1994; Crump et al., 1982; Lipman et al., 1982; Woie et al., 1982; Moroni et al., 1980;

Perkins et al., 1978) demonstrated inconsistent benefits, but adverse effects, such as hypercalcemia were rare. The dose employed was typically an intravenous bolus of calcium chloride (1g to 5g), sometimes followed by an infusion, or the equivalent dose of calcium gluconate.

Vasopressors

An unblinded study using a porcine model of nifedipine-induced cardiogenic shock showed no differences in mortality or hemodynamic parametres (cardiac output, blood pressure, systemic vascular resistance) following the addition of phenylephrine to high- dose insulin (10 units/kg/h) (Engebretsen et al., 2010). Vasopressin was reported as potentially harmful in one blinded randomized controlled trial using a swine model of verapamil poisoning (Barry et al., 2005), although one case series of two patients

104

(Kanagarajan et al., 2007) and one case report (Marraffa et al., 2004) showed blood pressure improvement when added to other vasopressors. Epinephrine was associated with increased cardiac output in animal studies (Kline et al., 1993; Gay et al., 1986;

Strubelt et al., 1984), but hyperglycemia and increasing lactate were noted as adverse effects (Kline et al., 1993). Dopamine and norepinephrine infusions showed improved survival and hemodynamics in animal studies (Strubelt et al., 1990; Strubelt et al.,

1984), but results were inconsistent in case series (Levine et al., 2013; Groszek et al.,

2003; Karti et al., 2002; Howarth et al., 1994; Parikka et al., 1993; Ramoska et al., 1993;

Jaeger et al., 1990; Sauder et al., 1990). No significant ischemic complications were noted with high doses of vasopressors in case series of 48 patients (Levine et al.,

2013). In animal studies, the use of isoproterenol (Parikka et al., 1993; Ramoska et al.,

1993; Gay et al., 1986; Strubelt et al., 1986; Strubelt et al., 1984;) or atropine (Gay et al., 1986; Strubelt et al., 1984) showed occasional improvement in hemodynamics. In one human case series of three patients (Rizvi et al., 2012) and one case report

(Hagège et al., 1990), patients who received isoproterenol improved their heart rate and blood pressure.

Glucagon

Improvement in heart rate and cardiac output were observed with glucagon (IV bolus of

3mg followed by an infusion of 3mg/h) in two of three animal studies (Stone et al., 1996;

Stone et al., 1995), but in only one (Love et al., 1998) of the three human case series

(Supradip et al., 2008; Love et al., 1998; Roper et al., 1993;). Cardiac output was not measured in case series. Hyperglycemia and vomiting were side effects observed in six case reports (Cohen et al., 2011; Love et al., 1997; Mahr et al., 1997; Ashraf et al.,

105

1995; Buylaert et al., 1995; Doyon et al., 1993).

Lipid emulsion

In an animal model (Perez et al., 2008; Bania et al., 2007; Tebbutt et al., 2006) of intravenous verapamil toxicity, the administration of lipid emulsion 20% (IV bolus of

6.2ml/kg up to 18.6ml/kg) was associated with improvement in hemodynamics and survival. However, there was no significant improvement or an increased mortality in two animal studies using an oral verapamil toxicity model (Chu et al., 2013; Perichon et al., 2013). One available human case series (Geib et al., 2009) of five patients demonstrated 60% mortality when using this antidote, compared to a lower mortality reported in retrospective studies of calcium channel blocker poisoning (6% reported by

St-Onge et al. in 2012). Importantly, the mortality reported in observational studies included calcium channel blocker ingestions regardless of severity, whereas the case series published by Geib et al. (2009) only included severe cases. In one case report

(West, 2010), adverse effects, such as hypertriglyceridemia, and hypoxemia were observed with lipid emulsion when used at exceptionally high doses (2 L).

Hyponatremia, extreme lipemia, and inability to obtain a reliable complete blood count, arterial blood gas or electrolyte levels were also noted.

4-aminopyridine

Animal studies (Graudins et al., 2010; Magdalan et al., 2003; Tuncok et al., 1998; Plewa et al., 1994; Korstanje et al., 1987; Gay et al., 1986; Agoston et al., 1984; Wesseling et al., 1983) and human case series (Fiszer et al., 2007; Magdalan et al., 2003) showed survival and hemodynamic benefit with 4-aminopyridine. Seizures were observed in two

106

animal studies (Tuncok et al., 1998; Plewa et al., 1994).

Levosimendan

Animal studies (Graudins et al., 2010; Kurola et al., 2010; Abraham et al., 2009) and a small case series (Varpula et al., 2009) suggested a hemodynamic benefit for levosimendan, although seizures were observed in both patients (Varpula et al., 2009).

One of four animal studies (Abraham et al., 2009) used higher doses of verapamil to induce toxicity, resulting in increased mortality.

6.3.2 Results of individual studies and risks of bias for mechanical interventions

Extracorporeal life-support

The use of extracorporeal life-support was associated with a mortality benefit in patients in severe shock or in cardiac arrest secondary to cardiotoxic poisonings. In the observational study published by Masson et al. (2012), extracorporeal life-support was associated with a lower mortality when initiated in a group of 14 patients compared to conventional therapies provided to a group of 48 patients (48% vs. 86%) after adjustment for SAPS II and beta-blocker intoxication. Most human case series reported positive functional outcomes in the majority of survivors (Daubin et al., 2009;

Mégarbane et al., 2007). However, some patients experienced: limb ischemia (10% in the observational study conducted by Masson et al. in 2012; 0-50% according to the case series published by Daubin et al. in 2009, Mégarbane et al. in 2007 and Babatasi

107

et al. in 2001); thrombosis (2% in the observational study conducted by Masson et al. in

2012; and 0-12% in the case series published by Daubin et al. in 2009, Mégarbane et al. in 2007 and Babatasi et al. in 2001); or hemorrhage (5% in the observational study conducted by Masson et al. in 2012; and 0-12% in the case series published by Daubin et al. in 2009, Mégarbane et al. in 2007 and Babatasi et al. in 2001).

Pacemaker

Results were inconsistent for the success of temporary pacemakers in achieving capture and improving hemodynamics in human case series (Stachon et al., 2011;

McGlinchey et al., 1998; Ramoska et al., 1993; Belträge et al., 1984; Immonen et al.,

1981) and case reports (Gris et al., 1989; Snover et al., 1986). Pacing and capture problems were identified even with transvenous pacemakers (Stachon et al., 2011;

McGlinchey et al., 1998). However, hemodynamic improvement was observed most of the time when capture was successful (McGlinchey et al., 1998; Immonen et al., 1981;

Gris et al., 1989; Snover et al., 1986) and no adverse effect had been reported.

6.3.3 Results of individual studies and risks of bias for interventions for which only small case series, case reports or animal studies are available

The use of amrinone did not show a benefit in animal studies (Koury et al., 1996;

Tuncok et al., 1996) although human case reports using another phosphodiesterase inhibitor (enoximone) observed an increase in inotropy and a decrease in vasopressor

108

requirement (Sandroni et al., 2005; Link et al., 2002). One human case series

(Kuhlmann et al., 1999) and two case reports (Ezidiegwu et al., 2008; Siebenlist et al.,

1990) suggested use of plasma exchange to decrease verapamil concentrations and improve hemodynamics. One human case series of three patients (Pichon et al., 2012) and one case report (Belleflamme et al., 2012) suggested the use of extracorporeal albumin dialysis to improve hemodynamics without a clear impact on the serum calcium channel blocker concentrations. Only human case reports were found for charcoal hemoperfusion (Anthony et al., 1986), continuous venovenous hemodiafiltration

(Rhaman et al., 2009; Pfaender et al., 2008; Roberts et al., 2008; Rosansky et al.,

1991), insertion of an intra-aortic balloon pump (Janion et al., 2010), and methylene blue (Jang et al., 2011). Finally, only animal studies were found for the following interventions: Bay K 8644 and CGP 28932 (Magdalan et al., 2003; Tuncok et al., 1998;

Korstanje et al., 1987), digoxin (Bania et al., 2004; Bania et al., 2000; Ramo et al.,

1992), cyclodextrin (Mottram et al., 2012; Mottram et al., 2011), liposomes (Forster et al., 2012; Bertrand et al., 2010), bicarbonate (Tanen et al., 2000), carnitine (Perez et al.,

2011), fructose 1,6-diphosphate (Kalam et al., 2012), PK11195 (Lheureux et al., 1990) or triiodothyronine (Lynch et al., 2010).

6.3.4 Synthesis of results

Mortality

High-dose insulin (IV bolus of 1 unit/kg followed by an infusion of 0.5-2.0 units/kg/h)

(Bryant et al., 2009; Greene et al., 2007) initiated before or shortly after vasopressors was associated with mortality improvement. In animal studies (rats and rabbits) calcium

109

(Graudins et al., 2010; Graudins et al., 2008; Supradip et al., 2008; Karti et al., 2002;

Howarth et al., 1994; Strubelt et al., 1990; Strubelt et al., 1986; Henry et al., 1985;

Strubelt et al., 1984), epinephrine (Strubelt et al., 1984), dopamine (Groszek et al.,

2003; Karti et al., 2002; Sauder et al., 1990; Strubelt et al., 1990; Strubelt et al., 1986;

Strubelt et al., 1984), norepinephrine (Strubelt et al., 1990; Strubelt et al., 1984) and 4- aminopyridine (Graudins et al., 2010; Magdalan et al., 2003; Plewa et al., 1994;

Korstanje et al., 1987; Gay et al., 1986; Agoston et al., 1984) were associated with reduced mortality. Based on human case series, only calcium (Supradip et al., 2008;

Karti et al., 2002; Howarth et al., 1994; Henry et al., 1985), and dopamine (Groszek et al., 2003; Karti et al., 2002; Sauder et al., 1990;) were associated with reduced mortality. Most human studies did not report a benefit in survival with atropine (Howarth et al., 1994; Roper et al., 1993; Strubelt et al., 1984), glucagon (Kline et al., 1993; Roper et al., 1993), pacemaker (Beträge et al., 1984), levosimendan (Abraham et al., 2009) or plasma exchange (Kuhlmann et al., 1999). Animal studies also did not report any survival benefit with atropine in a rat model (Strubelt et al., 1984), glucagon in a dog model (Kline et al., 1993), or levosimendan in a rat model (Abraham et al., 2009).

Animal studies, two murine models (Perez et al., 2008; Tebbutt et al., 2006;) and one dog model (Bania et al., 2007), each of them of moderate methodological quality, suggested that lipid emulsion improves survival in an intravenous model of verapamil poisoning. In one of these animal studies, this effect was dose-dependent (Perez et al.,

2008), however this was not confirmed in the only human case series available (Geib et al., 2009). Extracorporeal life-support (Masson et al., 2012) for patients in cardiac arrest or severe shock refractory to conventional therapy reported a benefit in survival. Two patients survived with albumin dialysis (Belleflamme et al., 2012; Pichon et al., 2012),

110

and only animal studies supported the use of Bay K8644 (Magdalan et al., 2003;

Tuncok et al., 1998; Korstanje et al., 1987).

Hemodynamics

Positive effects on hemodynamics were well documented with the use of high-dose insulin in human observational studies (Greene et al., 2007), case series (Espinoza et al., 2013; Holger et al., 2011; Boyer et al., 2002; Boyer et al., 2001), and animal studies

(pigs and dogs) (Engebretsen et al., 2010; Yuan et al., 1999; Kline et al., 1992; Kline et al., 1997). As well, extracorporeal life-support in human studies (Masson et al., 2012;

Daubin et al., 2009; Mégarbane et al., 2007; Babatasi et al., 2001), calcium in most animal studies (rodents and dogs) (Graudins et al., 2010; Graudins et al., 2008; Strubelt et al., 1990; Gay et al., 1986; Strubelt et al., 1986; Vick et al., 1983) and some human case series reported improvement in hemodynamics (Ramoska et al., 1993; Henry et al., 1985). Animal studies on the effects of epinephrine, dopamine and norepinephrine in rats (Strubelt et al., 1990; Strubelt et al., 1984) and dogs (Gay et al., 1986) also demonstrated an improvement in hemodynamics. The same effects were documented for 4-aminopyridine in five different types of animal (Plewa et al., 1994; Korstanje et al.,

1987; Gay et al., 1986; Agoston et al., 1984; Wesseling et al., 1983), lipid emulsion in rats (Perez et al., 2008; Tebbutt et al., 2006), and dogs (Bania et al., 2007), and Bay

K8644 in rodents (Magdalan et al., 2003; Tuncok et al., 1998; Korstanje et al., 1987).

The variability in hemodynamic response to calcium observed in human case series

(Supradip et al., 2008; Karti et al., 2002; Howarth et al., 1994; Ramoska et al., 1993;

Horowitz et al., 1989) was also seen with atropine (Howarth et al., 1994; Ramoska et al., 1993), glucagon (Supradip et al., 2008; Kline et al., 1993), and pacemakers

111

(Stachon et al., 2011; McGlinchey et al., 1998; Ramoska et al., 1993). Hemodynamic improvement was reported with levosimendan in two patients (Varpula et al., 2009) and animals (Graudins et al., 2010; Madera et al., 1977). Digoxin was associated with hemodynamic improvement in dogs (Bania et al., 2004; Bania et al., 2000; Ramo et al.,

1992), but had an inconsistent effect on mortality. Animal studies showed an improvement in blood pressure with the use of liposomes (Forster et al., 2012; Bertrand et al., 2010) but this treatment was not tested in humans. The effect of decontamination

(Cumpston et al., 2010; Barrow et al., 1994; Buckley et al., 1993; Bausch et al., 1991;

Lambert et al., 1990; Sauder et al., 1990) on the prevention of toxicity in humans poisoned with calcium channel blocker was limited to small case series.

Impact on functional outcomes

Functional outcomes were only reported in case series involving extracorporeal life- support in humans. Daubin et al. (2009) observed that all survivors in their sample (n=3) were discharged without cardiovascular or neurological sequelae. Mégarbane et al.

(2007) reported that three patients treated with extracorporeal life-support were symptom-free after one year. However, in two isolated case reports, one patient was discharged to a long-term care facility (Rona et al., 2011) and another underwent leg amputation (Tauro et al., 2009).

Impact on other outcomes

Only one observational study concerning high-dose insulin examined the impact on length of stay. No significant differences in patients who received the therapy were observed (Musselman et al., 2011). However, no power calculation was conducted.

112

6.3.5 Risk of bias across studies

Observational studies

The interobserver agreement on the STROBE checklist scoring was excellent for observational studies (kappa: 0.90; 95% CI 0.82 to 0.99). Percent agreement for each element varied from 67-100%, with the exception of the criterion related to the mention of a specific hypothesis, on which observers frequently disagreed. The high-dose insulin studies reported between 4 and 10 of 22 elements in the STROBE checklist, while the extracorporeal life-support study reported 17 of 22. Clear eligibility criteria and reports of data collection methods for high-dose insulin studies, sample size calculation, statistical methods, reported bias and limitations were often missing from observational studies.

The application of the Thomas tool resulted in 67% or higher agreement. However, observers disagreed on selection bias and data quality. Relevant confounders, such as comorbidities were poorly described in all studies. Adherence to the high-dose insulin protocol was often variable. Therefore, the integrity of the intervention (defined as the degree to which the intervention was implemented as planned or intended) was considered weak for high-dose insulin.

Case series

The interobserver agreement with the Institute of Health Economics tool for Quality of

Case Series and Quality of Reporting was substantial (kappa: 0.80; 95% CI 0.76 to

0.84). Percent agreement was higher than 88% when judging the quality of the statistical tests used, which were generally descriptive, and follow-up rates. Case series

113

scored 10/20 except for two of three articles involving extracorporeal life-support (13/20 and 15/20) and one of four high-dose insulin case series scored 13/20. It was often unclear whether several case series were collected in different centres, or if participants were recruited consecutively, or if there was loss to follow-up. A very small number of studies were conducted prospectively with outcomes measured a priori and with adverse events reported.

Animal studies

The interobserver agreement for the use of the ARRIVE guidelines for animal studies was excellent (kappa: 0.90; 95% CI 0.88 to 0.92). Percent agreement for each item varied from 80-100%. All studies obtained a score of 10/20 to 18/20 except for one related to calcium use – 5/20 (Vick et al., 1983). The studies' relevance to human biology was unclear and details concerning the randomization procedure, sample size calculation and husbandry conditions were often missing. When using the modified

NRCNA list, the interobserver agreement was still excellent (kappa: 0.98; 95% CI 0.96 to 0.99) and the percent agreement remained higher than 88% for all items. All studies obtained a score between 4/16 and 9/16. The weaknesses identified by the NRCNA list included use of unanesthetized animals, lack of blood concentration measurements, intervention tested in only one species, oral calcium channel blocker administration, autopsy not conducted, lack of allocation concealment and blinded assessment.

As expected, the risk of bias with case reports was high. Bias risk across studies was high for all interventions and high to moderate for extracorporeal life-support. Table 12 lists bias risk across studies for each intervention. The risk of publication bias was

114

estimated to be high considering inherent risk with case reports.

Table 12 Risks of bias across studies for each intervention Observational studies Case series Animal studies STROBE THOMAS' tool Institute of Health ARRIVE NRCNA checklist checklist Economics (IHE) tool for checklist ( /16) ( /22) quality of case series ( /20) and quality of reporting ( /20) High-dose insulin 6 (4-9) Weak 7 (4-13) 15 (11-18) 8 (6-9) Extracorporeal life 17 Weak to 11 (9-15) support moderate Calcium 6 (2-10) 12 (5-16) 5 (3-7) Vasopressors 7 (2-14) 14 (11-18) 6 (4-9) Decontamination 7 (3-8) Pacemaker 7 (4-10) Glucagon 7 (3-9) 13 (11-15) 6 (6-6) Atropine 7 (3-10) 11 (11-11) 6 (5-6) 4-aminopyridine 9 (8-9) 14 (10-17) 6 (5-8) Lipid emulsion 7 (5-8) 11 (7-15) 6 (6-7) Levosimendan 8 16 (14-16) 6 (5-8) Plasma exchange 9 BayK8644 15 (14-17) 5 (5-6) Digoxin 13 (12-14) 6 (5-7) Phosphodiesterase 16 (15-17) 5 (4-5) inhibitors Cyclodextrin 16 (14-16) 7 (6-7) Liposomes 15 (14-16) 6 (6-7) Albumin dialysis 8 Carnitine 13 (8-17) 6 (5-7) Bicarbonates 15 9 Fructose-1,6- 14 7 diphosphate PK 1195 13 5 Triidothyronine 15 6

6.4 Discussion

The evidence for treatment of calcium channel blocker poisoning derives from a highly biased and heterogeneous literature. Important limitations were identified in the majority of studies. We used different analysis tools to assess the risk of bias with transparency, but to our knowledge this is the first time these tools have been used in toxicology. For many interventions (high-dose insulin, extracorporeal life-support, calcium, dopamine, norepinephrine, epinephrine, 4-aminopyridine), results were consistent across different

115

study types. Inconsistency among studies arose from differences in interventions, populations and outcome measures. Moreover, head-to-head comparisons of treatments were infrequent, making it difficult to evaluate the comparability of treatments. Based upon the published literature, we can draw few valid inferences about the relative merits of one intervention over another.

We designed our search strategy to be as inclusive as possible, including a search of the gray literature. Some articles identified by title could not be retrieved. These were primarily case reports and therefore it is unlikely they would have influenced the overall findings.

6.5 Conclusions

This systematic review found a low level of evidence supporting the use of high-dose insulin and extracorporeal life-support, and a very low level of evidence supporting the use of calcium, dopamine, norepinephrine and epinephrine for the treatment of calcium channel blocker poisoning. This systematic review focused on important outcomes for decision-making in managing patients poisoned with a calcium channel blocker. This chapter, complemented by the cost-effectiveness analysis for the use of extracorporeal life-support in cardiotoxic poisoning described in the following chapter, serves as the starting point for the development of recommendations.

116

______

Chapter 7 Cost-effectiveness analysis

______

This chapter is modified with permission from the following: St-Onge M, Fan E, Mégarbane B, Hancock-Howard R, Coyte P (2014). Venoarterial extracorporeal oxygenation (VA-ECMO) for patients in shock or in cardiac arrest secondary to cardiotoxicant poisoning: a cost-effectiveness analysis. Journal of Critical Care, available online 16 October 2014.

117

Chapter 6 documented the evidence for each intervention considered for the treatment of calcium channel blocker poisoning. Venoarterial extracorporeal membrane oxygenation (VA-ECMO) has been identified as a potential treatment for cardiotoxicant poisonings in severe shock or in cardiac arrest. However, cost may be cited as a reason not to offer this intervention. Consideration of costs is required for appropriate decision-making (Step 2) in the process of recommendations development. To facilitate this process, this chapter details a cost-effectiveness analysis for the use of VA-ECMO for cardiotoxicant poisonings.

7.1 Introduction

Extracorporeal membrane oxygenation (ECMO) is defined as "the use of mechanical devices to temporarily support heart or lung function (partially or totally) during cardiopulmonary failure, leading to organ recovery or replacement (ELSO, 2013). In poisonings, VA-ECMO can be provided in addition to standard therapies, such as high- dose insulin, calcium, glucagon and vasopressors (St-Onge et al., 2014).

As shown in the previous chapter, along with the use of high-dose insulin, VA-ECMO was one of the most strongly supported interventions in the literature, along with the use of high-dose insulin (St-Onge et al., 2014). One observational study (Masson et al.,

2012) showed a significantly lower mortality rate in cardiotoxicant-poisoned patients (n

= 62) treated with VA-ECMO compared with standard care (from 52% to 14%, p =

0.002), and three case series of patients in shock or cardiac arrest (Daubin et al., 2009;

118

Mégarbane et al., 2007; Babatasi et al., 2001) showed survival with good functional outcomes with the use of this invasive strategy. Nonetheless, despite the rapid increase of ECMO usage since 2004 (Paden et al., 2013), only 5% of cardiac arrest patients receiving ECMO between 1992 and 2007 had a diagnosis of poisoning

(Thiagarajan et al., 2009), suggesting that the potential of this intervention is not being realized in cardiotoxicant poisonings.

The cost effectiveness of VA-ECMO for cardiotoxicant poisoning has not been well studied. Vats et al. (1998) previously identified that the use of venovenous ECMO (VV-

ECMO) in paediatric patients with hypoxemic respiratory failure in an American tertiary centre resulted in a cost of $4,190/life year gained (LY). In a UK-based multicenter trial comparing conventional management to transfer for consideration of ECMO treatment in adults with severe respiratory failure, a lifetime model predicted the cost per quality- adjusted life-year (QALY) of ECMO to be £19,252 (95% confidence interval £7,622-

59,200) (Peek et al., 2010). However, no study has been performed concerning the use of VA-ECMO in adults poisoned with cardiotoxicants. Before promoting the therapy in recommendations, it is imperative that the cost-effectiveness be assessed (CIHR,

2013).

The objective of this cost-effectiveness analysis was to assess the incremental cost- effectiveness ratio (ICER) – defined as the difference in costs divided by the difference in effects –- from a societal perspective of using VA-ECMO for adults in shock or cardiac arrest secondary to cardiotoxicant poisoning compared to standard care. The model included the likelihood that a patient would need to be transferred to a different

119

facility to receive the treatment.

7.2 Methods

VA-ECMO cost-effectiveness was analyzed using a decision analysis tree from a societal perspective. The population of adults in shock or cardiac arrest secondary to cardiotoxicant poisoning treated in Canadian hospitals was studied with a lifetime horizon, accounting for future medical and non-medical expenditures when a life is saved (Meltzer, 1997). This study was approved by the Research Ethics Board of the

University Health Network (Toronto).

We generated a decision model comparing the use of VA-ECMO with standard therapies for patients poisoned with a cardiotoxicant in persistent cardiac arrest or severe shock at arrival to the Emergency Department (ED) (Figure 10). As defined by

Masson et al. (2012), persistent cardiac arrest was considered as an absence of return to spontaneous circulation after continuous cardiopulmonary resuscitation, whereas severe shock was described as a persistent arterial hypotension (systolic blood pressure <100mmHg) despite optimal pharmacological treatments as proposed by

Baud et al. (2007) and associated with more than one of the following criteria: left ventricular ejection fraction assessed by echocardiography of less than 30%, severe hypoxemia (PaO2/FiO2 ratio of <150mmHg), and renal failure (urine output of <20ml/h or creatinine of >13mg/dL).

120

Figure 10 Simplified schema of the decision model

121

7.2.1 Effects

The effectiveness measure was documented in LY gained based on survival results obtained from a study yielding the highest level of evidence identified in a recent systematic review (St-Onge et al., 2014), the study published by Masson et al. (2012) with 62 patients, and known life expectancy as reported by Statistics Canada (2013).

The number of LY was calculated by determining the age-matched mean life expectancy stratified by sex for hospital survivors based on the study published by

Masson et al. (2012) (mean age of 48 ± 17 years; 63% women and 37% men).

7.2.2 Costs

Table 13 describes the assumptions, sources and values for costs used in the analysis.

The total cost for the VA-ECMO group included: the costs associated with prehospital transport in a Canadian setting; consultation with the poison control centre; care provided in the emergency department (ED) for a patient in cardiac arrest or severe shock; care provided in the intensive care unit (ICU); the costs associated with the use of VA-ECMO and its complications; the cost of transfer when relevant; the cost of care provided on the ward associated with death or survival (accounting for future medical and non-medical expenditures when a life is saved); and the costs for the patients and their relatives, which were documented with interviews and questionnaires with 14 family members of four patients who were treated with extracorporeal life- support. Direct and indirect costs were considered.

122

The total cost for the group not receiving VA-ECMO also included: the costs associated with prehospital transport; consultation with the poison control centre; care provided in the emergency department for a patient in cardiac arrest or severe shock; and the cost of care provided on the ward or associated with death or survival also accounting for future expenditures when a life is saved. However, the costs of care provided in the

ICU and the costs incurred by the patients and their relatives were different than the

ECMO group as underlined by the different (but not significant) mean (SD) length of stays (20 (±15) days with the use of ECMO vs 13 (±22) days if not) in the study published by Masson et al. (2012).

Table 13 Assumptions, sources and values for costs (Canadian $ in 2013) used in the analysis Type of Costs details Costs Assumptions Sources costs Costs applicable to all patients (i.e. receiving VA-ECMO or not) Poison Nursing: $125/case - Same proportion of - Interview with the control $53/case consultations in both groups. administration of the Ontario centre Poison Control Centre (Annual Referring consultation report), November 2013 physician: - Schedule of Benefits (SOB) $31/case (SOB for Physician Services under K734) the Health Insurance Act Consultant: (October 2013 version) $40/case (SOB K735) Pre-hospital - $240/case - $240/case - No patient's contribution to - Ministry of Health and Long transport the payment. Term Care of Ontario (MOHLTC), Ambulance Services Billing (2013)

123

Stay in the - Physician - $612/ - 2h stay in the ED - SOB (2013) emergency consultation (SOB cardiac - Total ED visits costs divided - Drug costs: Sunnybrook department H055) and life- arrest case by the number of visits (as Health Sciences Center drug (ED) threatening - $558/case per the MOHLTC, 2010) formulary, November 2013 resuscitation with severe represents the baseline cost - MOHLTC, 2010 (SOB G521-522- shock to which physicians fees and 523): $336/case life-threatening interventions - Treatments should be added received for the intubation and stabilization including antidotes (cardiac arrest: $96/case; severe shock $42/case) - ED visit, general cost: $180/case Stay on the - Physician fees $1376/day - The ward length of stay is - Estimated ward length of stay ward (not for a daily visit for 7 days the same for both patients of 7 days based on a ICU) (SOB C122- who received or did not retrospective study: St-Onge et C123-C124- receive ECMO al.,2012 C142-C143) and - Most deaths happen during - Cost of ward stay: Sud et al., a psychiatry the acute phase in the ICU 2011 (Sud et al., 2011) consultation (SOB (intensive care unit) A895): $644 - Cost of ward stay: Sud et - Ward stay: al.,2011 $732/day Death - Physicians fees $6,676/death - One-time cost at time of - Cooke et al., 2009 for death ($5,007- death to approximate the declaration (SOB $8,346) societal cost for care A777) associated with fatal illness - Relative same cost in US and Canada. Survival - Cost related to $3,800/LY - Costs or charges of - Lone et al., 2013 (Canadian healthcare ($1,652- healthcare resource use are studies) resource use after $20,429) the same the first five years hospital discharge after admission and later on. (related and - Healthcare resources after unrelated to discharge in cardiotoxic cardiotoxic poisonings are similar to other poisoning) general ICU admission, severe sepsis, renal failure or acute respiratory failure. Societal Patients $100/day - Annual salaries reported are - Interviews/questionnaires with Costs (mean - Working days representatives of the family members of four patients of 3,5 lost (mean of population ($20,000/year to who were treated with relatives per $100/day lost $36,000/year and 1/4 extracorporeal life-support at patients) each day of unemployed) Toronto General Hospital in hospital stay) - 1 day of work lost each 2014 (n=14) - Missed unpaid week day of hospital stay activities: family - Missed unpaid activities events, reported (no amount exercising. attributed)

124

Relatives $371/day - Annual salaries reported are - Interviews/questionnaires with - Working days ($106/day representatives of the family members of four patients lost (mean of 0.4 for each population ($20,000/year to receiving extracorporeal life- day at $190/day relative) $50,000/year and 4/14 support at Toronto General of work lost, or unemployed as per Hospital (n=14) $76/day lost each interviews; $37,900/year as - Statistics Canada, Median day of hospital per Statistics Canada) total income, 2014. stay) - 0.4 day of work lost each - Transport: week day of hospital stay $15/pers/day - 2/4 patients were transferred - Hostelling: none from other cities (less than 1 - Out-of-pocket hour of transport/day) expenses: - Missed unpaid activities $15/pers/day reported (no amount - Missed unpaid attributed) activities: piano, yoga, soccer, school, community activity Costs applicable only to patients receiving VA-ECMO Stay in the - Physician fees 1 day: $3,535 - Same co-interventions for - Mean ICU length of stay for ICU (ECMO) for first, last day patients receiving ECMO or patients with ECMO: 20 days and daily visits 20 days: not (1-35) in study published by (SOB G557- $65,675 - Costs related to ICU stay Masson et al. (2012) 558-559): $325 documented in the - Schedule of Benefits for for 1 day, 35 days: Canadian study published Physician Services under the $4,315 for 20 $114,127 by Sud et al. (2011) does Health Insurance Act (October days and $7,517 not take into account the 2013 version) for 35 days administration of antidotes - Drug costs obtained from the - Treatments and vasopressors Sunnybrook Health Sciences received such Center drug formulary, Nov as sedation, 2013 analgesia, - Cost of ICU stay: Sud et al., vasopressors, 2011 prophylaxis and antidotes: $432/day - 2D echo (SOB G570-571) and chest X-ray (SOBX90): $164 for day 1 then $15/day for 17 days - ICU stay: $2,614/day ECMO - Physicians $9,035/case - Procedure done by a - Schedule of Benefits for fees (SOB ($2,696- thoracic surgeon assisted Physician Services under the Z743): $15,375) by a perfusionist and an Health Insurance Act (October $517/case intensivist 2013 version) - Mean - Costs/day related to the - Toronto General Hospital, $8,518/case use of ECMO for patients Hospital costs accounting for ($2,179- treated at the Toronto the 2006-2013 ECMO cases $14,858) General Hospital would be (unpublished) the same for poisoned patients

125

ECMO - Circuit Min: $459 - 4-17% of circuit - Minimum and maximum rates complications malfunction: malfunction requiring of complications: DeLange et $2,323/event for Max: $1,215 changing the system, 14- al., 2013 physicians fees 36% of bleeding, 10-21% of - Costs of the ECMO circuit (SOB A935) and infections, 19-21% of limb obtained from an interview with the circuit costs ischemia (amputation), 1- a company selling the products - Bleeding (GI or 22% of thrombosis (PE, (Medtronic) based on the intracranial): DVT or stroke) products used at Toronto $1,056/event for General Hospital physicians fees - Schedule of Benefits for (SOB A935), Physician Services under the medication, CT Health Insurance Act (October scan (SOB 2013 version) X400) - Infections: $1,267/event for antibiotics - Limb ischemia: $25,731/event for physicians fees and surgery (SOB A935) - Thrombosis: $400/event for physicians fees (SOB A615- J198) and medication Transfer for - Physicians $531/transfer - No patient's contribution to - Schedule of Benefits for ECMO fees for phone the payment Physician Services under the conference - A physician is escorting Health Insurance Act (October (CritiCall, K732- the patient 2013 version) 733) and for - A patient not receiving - Ministry of Health and Long escorting the ECMO does not need to be Term Care of Ontario patient transferred (MOHLTC), Ambulance - Pre-hospital Services Billing (2013) transport: 240$/case Costs applicable to patients not receiving VA-ECMO Stay in the - Physician fees 1 day: $3,535 - Same co-interventions for - Mean ICU length of stay for ICU (no for first, last day patients receiving ECMO or patients without ECMO: 13 days ECMO) and daily visits 13 days: not (1-35) in study published by (SOB G557-558- $42,757 - Costs related to ICU stay Masson et al. (2012) 559): $325 for 1 documented in the Canadian - Schedule of Benefits for day, $2,820 for 35 days: study published by Sud et al. Physician Services under the 13 days and $114,127 (2011) does not take into Health Insurance Act (October $7,517 for 35 account the administration of 2013 version) days antidotes and vasopressors - Drug costs obtained from the - Treatments Sunnybrook Health Sciences received such as Center drug formulary, Nov sedation, 2013 analgesia, - Cost of ICU stay: Sud et vasopressors, al.,2011 prophylaxis and antidotes: $432/day - 2D echo (SOB G570-571) and chest X-ray (SOBX90): $164 for day 1 then $15/day for 17 days - ICU stay: $2,614/day 126

7.2.3 Likelihood of events

Table 14 lists the probabilities of clinical events used in the decision model. The probability of survival when receiving VA-ECMO was based on the results of Masson et al. (2012): 100% survival for patients in cardiac arrest and 83% for patients in severe shock. However, for the sensitivity analysis, a more pessimistic probability of survival was used as reported in cohort studies (Daubin et al., 2009; Mégarbane et al., 2007;

Babatasi et al., 2001) and in a population of non-poisoned patients (Paden et al.,

2012): 27% survival for patients in cardiac arrest and 39% for patients in severe shock.

Survival for patients requiring transfer was not modified because no data are available.

Based on a study specifically assessing transport of patients receiving VA-ECMO

(Desebbe et al., 2013), the analysis was predicated on the assumption that 50% of the patients would be transferred. However, for the sensitivity analysis, the impact of 0% vs

50% of patients transferred was assessed. The probabilities of presenting initially in cardiac arrest were also obtained from the study published by Masson et al. (2012).

However, since the initiation of VA-ECMO in adult cardiac arrest patients is not common (Paden et al., 2013), the possibility of treating only patients in severe shock was evaluated in the sensitivity analysis.

127

Table 14 Assumptions, sources and values for probabilities used in the analysis Type of probability Probability Assumptions Sources (Sensitivity Analysis Range) Survival - Survival mimics results - Masson et al. 2012 - Cardiac arrest patients 100% (27-100%) obtained by Masson et al. - Daubin et al., 2009 with ECMO (2012), Daubin et al. (2009), - Mégarbane et al., 2007 Mégarbane et al., (2007) - Babatasi et al., 2001 and Babatasi et al., (2001) - Paden et al., 2013 - Cardiac arrest patients 0% (0%) - Worst case scenario would without ECMO be represented by the

results of the study - Patients in severe shock 83% (39-100%) published by Paden et al. with ECMO (2013) in a non-poisoned

population. - Patients in severe shock 56% (0-100%) without ECMO

Transfer 50% (0-50%) - 50% of the patients would - Desebbe et al., 2013 require to be transferred, but it is still possible that no patient would be transferred. The maximal proportion corresponds to the highest percentage of transport for hemodynamically unstable patients reported in the literature. Clinical presentation - Clinical presentation - Masson et al., 2012 - Cardiac arrest 22% (0-100%) mimics results obtained by Masson et al. (2012) (4), but it is still possible to consider - Severe shock 78% (0-100%) ECMO only for patients in severe shock or only for patients in cardiac arrest

7.2.4 Statistical analysis

The ICER was defined as the difference in costs divided by the difference in effects expressed in life years (LYs) gained (total VA-ECMO costs - total costs without VA-

ECMO)/(effect of VA-ECMO - effect without VA-ECMO). A series of one-way sensitivity analyses was performed to evaluate the uncertainty in the decision model. Model inputs were varied according to their specified ranges (see Table 14) and results were plotted in a tornado diagram (see Figure 10). The impact of uncertainty in all of the parametres was ascertained simultaneously by performing a probabilistic sensitivity

128

analysis (PSA) using Monte Carlo simulation. The PSA was performed using Beta distribution for all clinical probabilities, Gamma distribution for all costs, log-normal distribution for relative risks and normal distribution for all utilities (Briggs et al., 2012).

The PSA notably took into account the large standard deviation for mean hospital length of stay observed by Masson et al. (2012). The PSA results were generated by random sampling model input distributions over 10,000 iterations and forming an ICER after each iteration. Analyses were conducted using TreeAge Pro 2013 (TreeAge

Software; Willamstown, MA). All costs were converted to 2013 Canadian dollars, ($1.00

CDN = $0.9562US in 2013).

7.3 Results

The cost per LY gained in the VA-ECMO group was $145,931/18LY compared to

$88,450/10LY in the non-ECMO group. The ICER of VA-ECMO was in the cost- effective range ($7,185/LY; $8,748/LY if treating only patients in severe shock and

$5,151/LY if treating only patients in cardiac arrest). The one-way sensitivity analyses showed that the ICER was mainly influenced by the probability of survival. The results did not change when other variables were modified. When using a more pessimistic approach (27% survival in cardiac arrest patients and 39% survival in patients in severe shock treated with VA-ECMO), the ICER of VA-ECMO was still within the cost-effective range ($34,311/LY). When only those patients who would not need to be transferred were included in the model, the cost per LY in the VA-ECMO did not change significantly. Figure 11 shows the tornado diagrams illustrating the effect of probability

129

variables on the ICER for VA-ECMO compared to no VA-ECMO.

Figure 11 Tornado diagram - probability variables

The probabilistic sensitivity analysis (Figure 12) indicated variability in both costs and effectiveness. When effective (51% probability), there was a 96% probability that the

ICER was less than $50,000 per LY gained and 98% probability that it was less than

$100,000 per LY gained.

130

Figure 12 Probabilistic sensitivity analysis

7.4 Discussion

We used decision analysis techniques to model the economic outcomes associated with the use of VA-ECMO in cardiotoxicant poisonings. Our results estimate a base case ICER of $7,185/LY with VA-ECMO. This suggests that VA-ECMO is cost- effective, however the probabilistic sensitivity analysis showed that there is some uncertainty in these results. The effectiveness data for VA-ECMO was based on a small sample size, as were the costs from a societal perspective, and requires confirmation in future studies.

131

The willingness-to-pay threshold of US$50,000-$100,000 for a QALY is often mentioned in medical literature without being based on clear scientific evidence

(Shiroiwa et al., 2010). However, an international survey on willingness-to-pay for one additional QALY gained established a threshold of US$62,000 in the United States

(Shiroiwa et al., 2010). In the scenario in which VA-ECMO was effective, the ICER was less then $50,000/LY gained most of the time. However, the acceptable threshold for a

LY gained remains unclear.

The results presented in this chapter are consistent with cost-effectiveness analysis for the use of venovenous-ECMO in other populations (Roncon-Albuquerque et al., 2014;

Peek et al., 2010; Vats et al., 1998; Schumacher et al., 1993), though the outcomes used in this cost-analysis also come from old studies. ECMO technology has evolved substantially over the last 5-10 years and reports suggest that, in more experienced centres, survival may be higher and the complication rate may be lower than the values used in this study. Furthermore, new program models consisting of dedicated multidisciplinary providers using new technology in dedicated space can contribute to lower hospital costs (Cavarocchi et al., 2014).

The cost-effectiveness analysis presented in this chapter has some limitations. First, using VA-ECMO in adult cardiac arrests patients is not common in North America and there is no consensus thus far targeting to which type of cardiac arrest patients VA-

ECMO should be offered. Expanding its use to this group of patients may initially lead to poorer outcomes due to physician inexperience. Furthermore, those cases are less likely to be reported, but may be responsible for high individual costs. Expanding VA-

132

ECMO use beyond its traditional settings (e.g., operating rooms, ICU) may require its implementation in the ED, which involves investments such as the initial cost per machine ($100,000 lasting approximately 20 years, used 3-4 times a year)

(MOHLTCO, 2010) not taken into account in this cost-effectiveness analysis. As additional effective and cost-effective indications for this technology are developed, further implementation issues may have to be explored, such as the appropriate allocation of the devices within hospitals and across hospital systems.

Second, LYs gained and the ICER were calculated, but not quality-adjusted life-years

(QALY). Knowing that 47% of the cardiotoxicant poisoning population had psychiatric disease or drug addiction (St-Onge et al., 2012), there is a possibility that quality of life for those that survive the intervention may be different than that for the general population (Wang et al., 2013). A proportion of survivors may require care associated with ECLS complications, such as wound care, and may also reattempt suicide if no other interventions, such as cognitive therapy, are implemented (Brown et al., 2005).

Future studies should assess the health-related quality of life for cardiotoxicant poisoning patients and long-term survival.

Finally, we used European survival data in a small group of cardiotoxicant poisoned patients and applied it to the Canadian setting where cost might vary across centres. Costs were estimated using a variety of sources, and do not represent the real costs for the patients described in the study published by

Masson et al. (2012). Probabilities were calculated based on a small non- randomized study and limited series including very select patients. It is probable

133

that cases resulting in VA-ECMO failure have been under-published. We were also limited by the fact that this intervention is not yet used in our Canadian population of cardiotoxicant poisonings.

7.5 Conclusion

This cost-effectiveness analysis conducted from a societal perspective may support the use of VA-ECMO in the treatment of cardiotoxicant poisonings if its effectiveness is confirmed, although there are too many open variables for providing any specific number at this point in time. Our results suggest that costs should not be an argument not to offer this treatment to a Canadian population in hospitals where VA-ECMO is available, particularly when used as a salvage manoeuver when standard care is failing.

Cost-effectiveness analysis should also be conducted in other countries to evaluate whether our findings are generalisable to other health care systems. Nonetheless, those results will be considered in future development of recommendations for the treatment of cardiotoxicant overdoses.

134

______

Chapter 8 Recommendations for the treatment of

calcium channel blocker poisoning

______

This chapter is modified from the following: St-Onge M, Anseeuw K, Cantrell L, Gilchrist IC, Hantson P, Baily B, Gosselin S, Kerns W 2nd R, Laliberté M, Mégarbane B, Jang D, Lavonas EJ, Juurlink DN, Muscedere J, Yang CC, Sinuff T, Rieder MJ, Lavergne V. Recommendations for the treatment of calcium channel blocker poisoning. To be submitted.

135

Chapters 6 and 7 gathered the information required for decision making in the recommendations development process detailed in this chapter. To be noted, the word

“guidelines” has been substituted for “recommendations” at the request of an American association that felt that, in the US physicians may still fear the legal implications of the word “guidelines”. Considering that all recommendations are based on very low level of evidence, the committee accepted to use the word “recommendations” instead of

“guidelines”. By providing recommandations in an area where the level of evidence is poor, the committee created consistency in messaging to the clinicians.

8 .1 Introduction

An evidence-based consensus guidelines has been published by Olson et al. for out-of- hospital management of calcium channel blocker ingestion (Olson et al., 2005).

However, current recommendations for in-hospital care have not been not systematically developed based on the evidence and may vary from one poison control centre to another. This observation may also reflect practice variation due to individual, local or regional appraisal of the literature or limited access to resources. Hence, development of recommendations is needed to improve consistency and guide the management of patients with calcium channel blocker poisoning.

8.2 Methods

8.2.1. Objectives, scope, target users and analytical framework

136

The objective for the development of recommendations for the treatment of calcium channel blocker poisoning was to provide an evidence-based document to improve consistency in messaging to the clinicians, with the ultimate goal of improving the management of calcium channel blocker poisoning. The recommendations address what types of in-hospital interventions should be considered for patients who have ingested a potentially toxic amount of calcium channel blocker, according to their clinical status. Targeted users include the following: bedside or telephone-consulting physicians; other healthcare providers (pharmacists and nurses); and poison control centres.

An analytical framework was created, illustrating the links between key questions to be answered during the recommendations development process (Figure 13). This was done, as suggested by the US Preventive Task Force (Harris et al., 2001), to integrate heterogeneous evidence in the development of a treatment approach including multiple interventions.

137

Key questions 1. Is there direct evidence that one (or more than one) intervention reduces mortality, improves functional outcomes, reduces hospital length of stay or reduces intensive care unit length of stay? 2. Does the patient clinical presentation or type of ingestion influence the intervention(s) provided and the outcomes? 3. Does one (or more than one) intervention decrease CCB serum concentration, improve hemodynamics or reduce the duration of vasopressor use? 4. Are the intermediate outcomes reliably associated with reduced mortality or improved functional outcomes? 5. Does one (or more than one) intervention(s) result in adverse effects or demonstrate a lack of cost-effectiveness? Figure 13 Analytical framework for calcium channel blocker poisoning treatment recommendations

8.2.2 Recommendations development working group

A working group representing all participating professional international healthcare organizations (Table 15) was created. The Canadian Association of Poison Control

138

Centres (CAPCC) acted as the leading association and a representative was appointed by each participating organization. The CAPCC named a chair, and a co-chair was named by the working group to represent the pediatric population. Each working group member completed the International Committee of Medical Journal Editors form for disclosure of potential conflicts of interest. No relevant conflict of interest was identified.

The recommendations development process was not externally funded. All working group members were volunteers and only in-kind donations from participating organizations helped support the in-person meetings.

Table 15 Participating organizations • American Academy of Clinical Toxicology • American Association of Poison Control Centres • American College of Medical Toxicology • Asia Pacific Association of Medical Toxicology • Canadian Association of Emergency Physicians • Canadian Association of Poison Control Centres • Canadian Critical Care Society • Canadian Pediatric Society • European Association of Poison Centres and Clinical Toxicologists • European Society of Emergency Medicine • European Society of Intensive Care Medicine • Society of Critical Care Medicine

8.2.3 The evidence

As a starting point, the working group used the systematic review detailed in Chapter 6 that pre-dated the establishment of the consensus group and had been published as a distinct article (St-Onge et al., 2014). The systematic review included all study types involving humans or animals poisoned with a calcium channel blocker, and examined the effects of all interventions on the outcomes targeted by our recommendations, up to

139

December 31, 2013. In order to be thorough and complete, additional literature brought to the attention of the working group by any member was considered.

8.2.4 Definitions and terminology

The stepwise approach used by the working group included the following clinical categories: asymptomatic patients, symptomatic patients requiring first-line treatment, patients refractory to first-line treatment, patients in refractory shock or peri-arrest, patients in cardiac arrest. First-line treatment was defined as treatments initially provided to a symptomatic calcium channel blocker poisoned patient. Patients refractory to first-line treatment were classified as such when desired effects (see Table

17) were not significantly reached with first-line treatments, whereas rescue treatments were those provided to patients in refractory shock or peri-arrest. Refractory shock was defined as persistent cardiovascular failure associated with organ failure despite the administration of supportive care and adequate antidotes.

The members defined signs of calcium channel blocker toxicity as hemodynamic abnormalities, such as low heart rate (<60 per minute in adults), low blood pressure

(systolic <100 mmHg or mean arterial pressure <65 mmHg in adults), myocardial dysfunction or abnormal peripheral vascular resistances. Reference values and definitions supported by the American Heart Association (AHA, 2008) were adopted by the working group, but the members also recognized the importance of clinical judgement. Therefore, as suggested by the American Heart Association (2008), myocardial dysfunction was defined as a decrease in myocardial contractility seen on

140

the echocardiography or a documented cardiac index of less than 2.2 L/min/m while shock was defined as a state characterized by inadequate blood flow and oxygen delivery to organs and tissues.

8.2.5 The recommendation statements

The working group was divided in subgroups. Each subgroup responsible for a specific intervention presented a summary outlining the evidence, the benefits and the risks and costs. The subgroups developed statements focusing on specific interventions to be used in certain clinical presentations (asymptomatic patients, symptomatic patients needing first-line treatments, patients refractory to first-line treatments, patients in refractory shock or peri-arrest requiring rescue treatments or patients in cardiac arrest) in order to achieve a specific outcome. Those statements were used for the subsequent modified Delphi process.

Each statement was associated with a specific level of evidence, which was determined using the GRADE (Guyatt et al., 2011) system (Table 16). In order to determine the strength of recommendation (Table 16), the working group proceeded to a modified Delphi method (four rounds of anonymous online votes using a 9-point

Likert scale followed by telephone meetings and a face-to-face meeting held in

Brussels in May 2014). When a statement needed to be clarified, it was modified and voted on again until the working group members agreed that the results reflected a clear understanding of what the statements meant and implied. For each statement, the strength of recommendation (Table 16) was determined by the results of the votes at

141

the last round using the medians, the lower/upper interquartiles, and the disagreement indexes (RAND/UCLA Appropriateness Method) (Fitch et al., 2001) as described in

Figure 14.

Table 16 Levels of evidence and strength of recommendation Strengths of recommendation: Levels of evidence: • Level 1: Strong recommendation • Grade A: High level of evidence. We (appropriate by the large majority of are confident that the true effect is experts with no major dissension). The close to our estimate of the effect. desirable effects of adherence to the • Grade B: Moderate level of evidence. recommendation outweigh the The true effect is likely to be close to undesirable effects. our estimate of the effect, but there is o In favour: “we recommend” a possibility that it is substantially o Against: “is not recommended” different. • Level 2: Weak recommendation • Grade C: Low level of evidence. The (appropriate by the majority of experts, true effect may be substantially but some degree of dissension exists). different from our estimate of the The desirable effects of adherence to effect. the recommendation probably outweigh • Grade D: Very low level of evidence. the undesirable effects. Our estimate of the effect is just a o In favour: “we suggest” guess, and it is very likely that the o Against: “is not suggested” true effect is substantially different • Level 3: Neutral recommendation. The from our estimate of the effect. course of action could be considered appropriate in the right context. Reproduced with permission: Lavergne et al., 2012

Figure 14 Voting process for recommendations Reproduced with permission: Lavergne et al., 2012

142

8.2.6 Values and preferences

In accordance with AGREE II methodology, the perceived influence of the working group values and preferences on the vote results was documented at each round. In addition, to considering values and preferences of decision makers, users, patients and relatives, the draft recommendations were posted on a blog for two weeks. The public was asked to provide comments and suggestions to which the working group responded. In order to encourage participation, messages were posted on social media websites (Facebook, Twitter, Google +) of relevant organizations of patients and relatives (search key words: patient care, suicide prevention, mental health, substance of abuse, heart disease, hypertension, health ethics, research ethics), poison control centres and professional organizations (search key words: emergency medicine, critical care medicine, toxicology, psychiatry, pharmacy) (De Breejen et al., 2012).

8.2.7 Internal and external review

The recommendations for the treatment of calcium channel blocker poisoning were submitted to all participating associations for internal review, and to anonymous reviewers chosen by the associations for external review. The external reviewers evaluated the recommendations development process with the AGREE tool (Brouwers et al., 2010). The instrument consists of 23 key items organized within six domains: 1) scope and purpose; 2) appropriate stakeholders involvement and representation of the views of its intended users; 3) rigorous methodology to synthesize the evidence and

143

formulate the recommendations; 4) clarity of presentation; 5) applicability; and 6) editorial independence. A second face-to-face meeting was held in New Orleans in

October 2014 to discuss the documented values, preferences, and the results of the internal and external reviews.

8.3 Results

Figure 15 illustrates the progression of care for key recommendations and Table 17 details of the rationale for each recommendation.

144

Figure 15 Progression of care for key recommendations.

145

Table 17: Recommendation table Intervention Clinical presentation(s) Desired Risks Resources Level of Recommendation for which the effects and costs evidence (M = median on Likert scale recommendation applies (1-9),

Asymptomatic line) Symptomatic (first line Refractory to first peri Refractory shock or In cardiac arrest cardiac In L/UIQ = lower/upper interquartile 9as -

arrest appropriate), DI = disagreement index)

IN FAVOUR Decrease None Available D: case series Level 1: strongly in favour Observation +/- X toxicity documented of fair quality (M: 8, LIQ: 7.5, DI 0.1) decontamination Hemodynamic - Extravasation Available D: case series Level 1: strongly in favour Calcium IV X X X X improvement of chloride salt and animal (M: 8, LIQ: 7, DI 0.2) in some cases may lead to studies of (blood severe local poor to fair Typical dose: pressure, tissue injury quality - Calcium chloride 10% IV: contractility) - Hypercalcemia Adults: 10-20 ml Q10-20 min or infusion at 0.2-0.4 ml/kg/h - Calcium gluconate 10% IV: Adults: 30-60 ml Q10-20 min or infusion at 0.6-1.2 ml/kg/h

- Possible - Hypoglycemia - Available D: Level 1: strongly in favour High-dose X X X improvement - Hypokalemia - Requires observational when combined with fluids, insulin in survival - Volume intensive studies of calcium and vasopressors in - overload monitoring poor quality, the presence of myocardial Hemodynamic and case series of dysfunction (M: 8, LIQ: 7, improvement additional poor to fair DI: 0.3) (blood resources quality and pressure, animal studies Level 2: weakly in favour contractility) of good quality when combined with fluids, - Decrease in calcium and vasopressors vasopressors without evidence of requirement myocardial dysfunction (M: 8, LIQ: 6.5, DI: 0.2) or alone in the presence of myocardial dysfunction (M: 7, LIQ: 4.5, DI: 0.5)

Typical dose: 1U/kg bolus followed by an infusion at 1U/kg/h (maintain euglycemia with dextrose if needed) Possible - Anticholinergic Available D: case series Level 2: weakly in favour Atropine X X X X hemodynamic effects and animal (M:7, LIQ: 6.5, DI: 0) improvement studies of (heart rate) poor to fair Typical dose: quality 0.02 mg/kg (min 0.1 mg and max 0.5 mg) Hemodynamic - Organ/tissue Available D: case series Level 1: strongly in favour in Norepinephrine X X X X improvement ischemia and animal presence of undifferentiated (NE), (blood - Increase in studies of shock (M: 8, LIQ: 7-7.5, DI: epinephrine pressure, +/- lactate and poor to fair 0-0.2), but preferentially contractility, glucose with quality norepinephrine in presence +/- heart rate) epinephrine of vasodilatory shock (M: 8, LIQ: 7.5, DI: 0.1) Possible Worsening Available D: case series Level 2: weakly in favour in Dobutamine X X X hemodynamic hypotension of fair quality, patients presenting with improvement but small documented cardiogenic (contractility, number of shock (M: 7, LIQ: 5, DI 0.2) heart rate) patients

146

- - Hypoglycemia - Available D: one small Level 2: when refractory to Incremental X X Hemodynamic - Hypokalemia - Requires case series of first line treatments, weakly doses of high- improvement - Volume intensive fair quality in favour in the presence of dose insulin (blood overload monitoring myocardial dysfunction (M: pressure, and 7, LIQ: 6, DI: 0.4) contractility) additional - Decrease in resources Level 1: for rescue vasopressors treatment, strongly in favour requirement in the presence of myocardial dysfunction (M:7, LIQ: 7 DI 0.2)

Level 2: for rescue treatment, weakly in favour with no documented evidence of myocardial dysfunction (M: 7, LIQ: 4.5, DI 0.4)

Typical dose: Progressive increase of the infusion rate up to 10U/kg/h (maintain euglycemia with dextrose if needed) Possible - Discomfort Transcutan D: case series Level 2: for patients Pacemaker X X hemodynamic - Capture and eous more of poor quality refractory to first-line improvement pacing problems available therapy or as a rescue (heart rate) and faster therapy, weakly in favour in to initiate the presence of unstable than bradycardia or high-grade transvenous AV block and no significant alteration in inotropism (M:8, LIQ: 5.5, DI: 0.2-0.8) Possible - Hyperlipemia Available in D: animal Level 2: for patients Lipid emulsion X X X hemodynamic - Venous most studies of fair refractory to first-line therapy improvement thrombosis centers quality and therapy, weakly in favor (M: (blood - Possible fat case reports 7, LIQ: 5.5, DI: 0.2) pressure +/- embolism heart rate) - Possible Level 1: for rescue decrease treatment, strongly in favour efficacy of other (M: 8, LIQ: 7, DI 0.2) treatments - Laboratory Level 1-2: in cardiac arrest, interference: strongly in favour if not ABG, SatO2, administered previously (M: CBC, lytes 8, LIQ: 7, DI: 0.1), and weakly in favour if a previously administered (M: 7, LIQ: 6, DI: 0.3) - Survival - Limb ischemia - Available D: one Level 2 for rescue VA-ECMO (or X X benefit - Thrombosis only in observational treatments: weakly in favour ECLS) - - Bleeding certain study of good in patients presenting with Hemodynamic centers quality cardiogenic shock (M:7, improvement - Cost including all LIQ: 5.5, DI: 0.2) (mean arterial should not cardiotoxicant pressure) preclude s and case Level 2 in cardiac arrest: consideratio series of fair weakly in favour in patients n of the quality with a low flow of less than therapy 5 min, but weakly against if where it is low flow more than 15 min available (M: 7, LIQ: 5.5, DI 0.4) (St-Onge et al., 2014) AGAINST Unclear - Vomiting Limited D: case series Level 2: weakly against Glucagon X X X X X - Hyperglycemia availability of fair quality (M:3, UIQ: 5.5, DI: 0.4-0.7) - Tachyphylaxis Unclear Ischemic Available D: case series Level 2: weakly against in Dopamine X X complications and animal patients presenting with studies of undifferentiated or poor to fair vasopleglic shock (M: 3, quality UIQ: 5, DI 0.2) Unclear Ischemic Available D: case series Level 2: weakly against in Vasopressin X X complications and animal documented cardiogenic studies of shock (M: 3, UIQ: 4, DI: 0.5) poor to fair quality

147

Possible - Vomiting Available D: Case Level 1: strongly against (M: Methylene blue X X hemodynamic - Blue-green reports 2, UIQ: 3, DI 0.1) improvement discoloration of when used as body fluids a last resort - Serotonin syndrome in patients taking serotonin agents - Large doses: methemoglobine mia, hypoxia

148

8.3.1 Recommendations

8.3.1.1 Therapy in asymptomatic patients

For the treatment of patients who ingested a potentially toxic amount of calcium channel blocker, the working group recommends observation and decontamination following the position statement published by the EACCT/AACCT (2005) (1D).

Rationale

Based on case series (Truitt et al., 2012; Belson et al., 2000) of fair quality, it may be safe to monitor asymptomatic patients who ingested a potentially toxic amount of CCB defined as more than a single therapeutic dose (Olson et al., 2005) and to intervene if they develop signs of toxicity. However, decontamination should also be considered to prevent toxicity. The working group decided to defer to the American Academy of

Clinical Toxicology (AACT) and the European Association of Poison Centres and

Clinical Toxicologists (EAPCCT) position statement (2005) instead of proposing new recommendations for decontamination, in order to avoid confusion and favour consistency.

8.3.1.2 First-line therapy for symptomatic for patients

For the first-line therapy of symptomatic CCB-poisoned patients, the working group recommends the use of:

-Intravenous calcium (1D);

-High-dose insulin therapy (in combination with IV fluids, calcium and

vasopressors) if evidence of myocardial dysfunction is present (1D); and

149

-Norepinephrine and/or epinephrine in the presence of shock (even if the

myocardial function has not yet been assessed), and preferentially

norepinephrine in the presence of vasodilatory shock (1D).

For the first-line therapy of symptomatic CCB-poisoned patients, the working group suggests the use of:

-High-dose insulin therapy as a single therapy in the presence of myocardial

dysfunction, or even in the absence of documented myocardial dysfunction if

used in combination with IV fluids, calcium and vasopressors (2D);

-Dobutamine in the presence of cardiogenic shock (2D); and

-Atropine in the presence of symptomatic bradycardia or conduction

disturbances (2D).

For the first-line therapy of symptomatic CCB-poisoned patients, the working group suggests not to use:

-Dopamine in the presence of shock (2D); and

-Vasopressin in the presence of documented cardiogenic shock (2D).

Rationale

Although fluid resuscitation is commonly used, no formal recommendation was made because there are no fluid repletion studies available specifically for calcium channel blocker poisoning. Nonetheless, the working group considered fluid administration as a first line and continued administration as long as the patient demonstrates signs of fluid responsiveness (e.g. hemodynamic improvement after receiving 500 ml of crystalloid over 10-15 minutes). The working group also recommended intravenous calcium as a first-line treatment based on hemodynamic improvement observed in some case series

150

(Konca et al., 2013; Howarth et al., 1994; Ramoska et al., 1993; Henry et al., 1985) and animal studies (Graudins et al., 2010; Graudins et al., 2008; Strubelt et al., 1990; Gay et al., 1986; Strubelt et al., 1986; Strubelt et al., 1984; Vick et al., 1983). This therapy is readily available and carries little risk provided central venous or secure peripheral venous access is available, particularly when the chloride salt is used. The proposed regimen for the administration of calcium chloride 10% is 10-20 ml q10-20 min or an infusion at 0.2-0.4 ml/kg/h in adults, and 0.1-0.2 ml/kg q10-20 min or an infusion at 0.1-

0.2 ml/kg/h in a paediatric population. If calcium gluconate 10% is given, the dose regimen is 30-60 ml q10-20 min or an infusion at 0.6-1.2 ml/kg/h in adults, and 0.3-0.6 ml/kg q10-20 min or an infusion at 0.3-0.6 ml/kg/h in a pediatric population.

Observational studies (Bryant et al., 2009; Greene et al., 2007), case series (Espinoza et al., 2013; Boyer et al., 2002; Boyer et al., 2001; Yuan et al., 1999) and animal studies (Kline et al., 1997; Kline et al., 1996; Kline et al., 1993), showed hemodynamic improvement and a potential increase in survival with the use of high-dose insulin in patients poisoned with a calcium channel blocker. However, most of the studies used it in combination with fluids, intravenous calcium and vasopressors. Considering that high-dose insulin seems to have a direct positive inotropic effect (Engebretsen et al.,

2011; Kline et al, 1993), the working group made a strong recommendation for its use when there is documented myocardial dysfunction, but still suggested it for patients without a documented myocardial dysfunction. Despite the fact that high-dose insulin requires intensive monitoring, its benefits were thought to outweigh the risks. This intervention can also be considered alone in the presence of myocardial dysfunction.

The proposed dose regimen of high-dose insulin includes a bolus of 1 U/kg, followed by

151

an infusion of 1 U/kg/h with maintenance of euglycemia with a dextrose infusion if needed. Because titrated doses of high dose insulin (up to 10 U/kg/hr) are supported by weaker evidence, the working group suggested this only for patients who do not respond to first-line therapies (see below) (Holger et al., 2011).

The selection of vasopressors should be guided by the type of shock the patient is experiencing and the required doses are likely to be high. In a retrospective study published by Levine et al (2013), the maximal infusion rate reached was 100ug/min for the norepinephrine, 150 ug/min for the epinephrine and 245ug/kg/min for the dobutamine. Based on mechanism of action, the working group recommended the use of norepinephrine in vasoplegic shock, or if myocardial function has not yet been assessed, but had a neutral position when the patient was presenting with cardiogenic shock (Levine et al., 2013; Kline et al., 1996; Gay et al., 1986; Strubelt et al., 1986).

The use of epinephrine is also considered for a CCB-poisoned patient experiencing shock with or without myocardial dysfunction and is supported by a similar evidence base (Levine et al., 2013; Kline et al., 1993; Gay et al., 1986; Strubelt et al., 1984). In the presence of confirmed myocardial dysfunction, clinicians can also use dobutamine.

However, this therapy would not be suggested in other circumstances given the risk of hypotension (Levine et al., 2013). Based on unclear hemodynamic improvement in case series (Konca et al., 2013; Groszek et al., 2003; Karti et al., 2002; Howarth et al.,

1994; Ramoska et al., 1993), the working group did not suggest the use of dopamine.

The use of vasopressin alone was discouraged due to lack of efficacy and worsened survival in animal models (Konca et al., 2013; Barry et al., 2005; Sztajnkrycer et al,

2004). The working group could not make recommendations regarding the use of

152

vasopressin as an adjunct to other vasopressors as there is little documented clinical experience and no preclinical studies. No agreement was reached for the use of phenylephrine in patients poisoned with a calcium channel blocker.

In situations where there is symptomatic bradycardia or conduction disturbances, the working group members suggested using atropine at a dose regimen of 0.5 mg

(paediatric population: 0.02 mg/kg, min 0.1 mg, max 0.5 mg) q3-5 minutes. This suggestion is supported based on considerations that the therapy may temporarily help, is easily accessible, inexpensive and is not associated with important risks

(Howarth et al., 1994; Ramoska et al., 1993; Gay et al, 1986; Strubelt et al., 1986).

8.3.1.3 Therapy for patients refractory to first-line treatments

For the therapy of CCB-poisoned patients refractory to first-line treatments, the working group suggests the use of:

-Incremental doses of high-dose insulin therapy (up to 10 Units/kg/h) if evidence

of myocardial dysfunction is present (2D);

-Pacemaker in the presence of unstable bradycardia or high-grade AV block,

without significant alteration in cardiac inotropism (2D); and

-Intravenous lipid emulsion therapy (2D).

Rationale

In patients refractory to the previously described first-line treatments, the working group members considered therapy supported by weaker evidence, but associated with less risk than rescue treatments. Therefore, in the presence of myocardial dysfunction, the

153

working group suggested to titrate the high-dose insulin infusion rate up to 10 U/kg/h

(Holger et al., 2011). The patient would need to be closely monitored and would likely require a dextrose infusion to maintain euglycemia. Electrical cardiac pacing has been associated with frequent capture and pacing problems, however, there may be potential hemodynamic improvement in patients presenting with unstable bradycardia or high-grade AV block (Stachon et al., 2011; McGlinchey et al., 1998; Ramoska et al.

1993; Beiträge et al., 1984; Immonen et al., 1981). To avoid spending time on a therapy that involves risk and may not be effective, the working group suggested attempting transcutaneous pacing first. Based on possible hemodynamic improvement documented in two animal studies (Perez et al., 2008; Tebbutt et al., 2006) and case reports, the working group members also recommended the use of lipid emulsion therapy. However, this is not recommended for consideration earlier on in treatment given the concern of potentially increasing the absorption of medications still present in the stomach. Although only published as an abstract, his concern was raised by an animal study showing worse outcomes (Perichon et al., 2013) when using oral models calcium channel blocker poisoning instead of intravenous. We believe that lipid emulsion should be used only after incremental doses of high-dose insulin therapy have failed. The working group members felt that there were insufficient data to recommend a specific dose regimen of lipid emulsion therapy. The most commonly recommended dose is 1.5mL/kg of 20% lipid emulsion administered as a bolus, repeated up to two times as needed until clinical stability is achieved, and followed by an infusion of 0.25mL/kg/min for 30-60 minutes (Association of Anaesthetists of Great

Britain and Ireland, 2010).

154

8.3.1.4 Therapy for patients in refractory shock or peri-arrest

For the therapy of CCB-poisoned patients in refractory shock or peri-arrest, the working group recommends, as rescue treatments, the use of:

-Incremental doses of high-dose insulin therapy (up to 10 Units/kg/h) if evidence

of myocardial dysfunction is present if not administered previously (1D); and

-Lipid emulsion therapy (1D).

For the therapy of CCB-poisoned patients in refractory shock or peri-arrest, the working group suggests, as rescue treatments, the use of:

-Incremental doses of high-dose insulin therapy (up to 10 Units/kg/h) even in the

absence of myocardial dysfunction if not administered previously (2D);

-VA-ECMO (or ECLS) in presence of cardiogenic shock in centres where the

treatment is available (2D); and

-Pacemaker in the presence of unstable bradycardia or high-grade AV block,

without significant alteration in cardiac inotropism if not administered previously

(2D).

Rationale

Given the high risk of mortality in patients with severe refractory shock or peri-arrest, the working group members considered therapies with less evidence or higher risks.

Therefore, incremental doses of high-dose insulin therapy would be considered even if no myocardial dysfunction has been documented (Holger et al., 2011) and the use of lipid emulsion therapy is now recommended (Perez et al., 2008; Tebbutt et al., 2006).

Given the risk of mortality in severily poisoned patients and the potential survival

155

benefit demonstrated in an observational study conducted in experienced centres

(Masson et al., 2010), the working group members suggested VA-ECMO as a rescue therapy in patients poisoned with a calcium channel blocker presenting with cardiogenic shock in centres where the treatment is available. In this clinical scenario, the working group evaluated that the benefits outweigh the risks of limb ischemia, bleeding or thrombosis. The members made no recommendation for the use of the

Impella® catheter or other left ventricular- or biventricular-assisted devices as potential alternatives to VA-ECMO as there is simply insufficient clinical or research experience

(Laes et al., 2013).

8.3.1.5 Therapy for patients in cardiac arrest

For therapy of CCB-poisoned patients in cardiac arrest, the working group recommends, in addition to standard advanced cardiac life-support (ACLS), the use of:

-Intravenous calcium, even if previously administered (1D); and

-Lipid emulsion therapy if not administered previously (1D).

For therapy of CCB-poisoned patients in cardiac arrest, the working group suggests the use of:

-Lipid emulsion therapy, even if previously administered (2D); and

-VA-ECMO (or ECLS) in the presence of a low flow, for less than 5 minutes, and

in centres where the treatment is available (2D).

Rationale

Studies looking specifically at patients poisoned with a calcium channel blocker in

156

cardiac arrest are scarce. Most of the recommendations except one for the use of VA-

ECMO are extrapolated from studies conducted in severely-ill patients, but that were not in cardiac arrest. Therefore, the working group reinforced the importance of performing adequate and aggressive resuscitation with previously mentioned modalities. Consequently, the working group members recommended the use of intravenous calcium and lipid emulsion therapy at the same dose regimen previously mentioned. Further, a second dose of lipid emulsion therapy could be considered if the patient already received a bolus before the cardiac arrest.

Concerning the use of VA-ECMO in experienced centres, observational studies have demonstrated a potential survival benefit in cardiac arrest patients (Masson et al.,

2012; Daubin et al., 2009; Mégarbane et al., 2007; Babatasi et al., 2001). The working group members estimated that the benefit of saving a life outweigh the risks of initiating that invasive therapy, if there is reasonable chances of surviving without significant deficit. Therefore, this therapy is suggested if the chest compressions have being ongoing for less than 5 min (low flow period), but would not initiate the therapy in patients who had a low flow period lasting more than 15 min.

8.3.1.6 Rationale for not recommending or not suggesting some treatments

- The working group recommends not to use methylene blue as a first-line treatment given experience is limited to a few case reports (Aggarwal et al., 2013; Kim et al.,

2012; Jang et al., 2011) (1D).

- The working group recommends not to use levosimendan, a calcium channel opening drug based on lack of efficacy in animal studies and unknown risks versus benefits in

157

the clinical setting (Abraham et al., 2009) (1D).

- The working group suggests not to use glucagon because case series reported variable effects (Supradip et al., 2008; Love et al., 1998; Roper et al., 1993). Vomiting

(Cohen et al., 2011; Fant et al., 1997) and hyperglycemia (Fant et al., 1997; Mahr et al.,

1997; Ashraf et al., 1995; Buylaert et al., 1995; Doyon et al., 1993) had been noticed in several case reports, and more effective interventions for the treatment of calcium channel blocker poisoning are available (2D).

- The working group recommends not to or suggests not to use the following treatments based on insufficient experience and scientific scrutiny: plasma exchange, triidothyronine, cyclodextrin, amrinone or other PDE-inhibitors, L-carnitine, intra-aortic balloon pump, CVVHDF, charcoal hemoperfusion, albumin dialysis, MARS, digoxin, liposomes, fructose 1,6 diphosphate, PK11195, BK8644, CPG28932 or potassium antagonists (St-Onge et al., 2014) (1D).

8.3.2 Values and preferences

The working group members reported the following factors as influencing their vote most significantly (≥6/9 on a Likert scale): 1) the evidence (8/9); 2) the balance between risks and benefits (7/9); 3) the feasibility and applicability of the intervention

(7/9); and 4) their experience and training (7/9). Their discussions with the other working group members were perceived to have more influence during the final rounds of voting (first round: 4/9, second round: 5/9, third round: 5/9, fourth round: 6/9). In terms of public involvement, the blog (http://poisoningsguidelines.com; 37 followers on

Twitter, 189 followers on Facebook) documented 796 visitors from 61 countries and

158

1,529 views during the period the draft of recommendations were posted for public feedback (between October 13th and October 27th 2014). Suggestions were made mainly by health care professionals to clarify statements and facilitate its application, but no disagreement was expressed. The working group made the appropriate corrections.

8.3.3 Internal and external reviews

Based on the comments received by the associations during the internal review, the working group agreed to use the word “recommendations” instead of “guidelines” in the final manuscript. An association felt its members would be at higher risk of being exposed to legal implications if not following “guidelines” as opposed to

“recommendations”. The working group also withdrew the recommendations related to decontamination, because those were generating significant disagreement, and a position statement published by the AACT and the EAPCCT has already been produced. Finally, the working group underlined that the recommendations applied mainly to adults, clarified some statements to facilitate its application and added in areas for future research questions raised by the reviewers that could not be answered with the current available evidence.

The external reviewers (2) gave a global score of 6/7 to the recommendations development process. The main suggestions for improvement, which were followed by the working group, concerned the need for more implementation tools and a better defined update process.

159

8.3.4 Implementation and applicability

Considering that effective implementation strategies include multifaceted interventions, interactive education and clinical reminder systems (Prior et al., 2008), we intend to post the algorithm on our blog where a checklist, a video and a quiz will be available.

The blog users will also have the opportunity to provide feedback that will be taken into account when the recommendations are updated. The blog link

(http://poisoningsguidelines.com) will be sent to all relevant professional associations and training programs. Interactive educational meetings and workshops will be encouraged. Interventions that are not widely available, such as VA-ECMO, or those that are not performed on a regular basis by non-toxicologists, such as high-dose insulin, may still be difficult to integrate into practice. Therefore, professional associations must use proactive knowledge translation strategies by using the implementation tools created by the working group.

To monitor the impact of the recommendations implementation, the working group plans to conduct a survey two years post-implementation to assess barriers, facilitators and impact on resources use. Decision makers should consider monitoring adherence to recommendations and outcomes (mortality, functional outcome, ICU and in-hospital length of stay).

160

8.4 Discussion

Guided by the evidence, the balance between risks and benefits, the costs and the use of resources, the working group used a rigorous methodology (including the GRADE system, a modified Delphi technique, and the RAND/UCLA Method) to build recommendations for the in-hospital treatment of patients poisoned with a calcium channel blocker. The recommendations development working group recommended a complementary patient-tailored use of calcium, high-dose insulin and vasopressors as first line treatments; incremental doses of high-dose insulin and pacemaker for some patients refractory to first-line treatments; and considered incremental doses of high- dose insulin, lipid emulsion therapy and the use of VA-ECMO as potential rescue treatments.

The working group encourages clinicians to focus on therapies supported by the highest level of evidence rather than using therapies for which the level of evidence is lower and/or for which there are alternatives with a better safety profile. The target population of our recommendations are adults poisoned with a calcium channel blocker, given that most of the studies available were based on that population.

However, considering the limited literature on the treatment of children poisoned with a calcium channel blocker and, in the absence of evidence that children respond differently than adults to calcium channel blocker poisoning, the working group believe that most of the recommendations can be applied to the pediatric population.

161

The overall evidence available to develop these recommendations was of very low quality. Many interventions had only been studied for surrogate outcomes. With the exception of VA-ECMO for cardiotoxicant poisonings, the use of and costs associated with these resources had not been described (St-Onge et al., 2014). Hence, some of the questions within our proposed analytic framework remain unanswered. Therefore, the working group identified potential areas for future research that may help to improve the care provided to patients poisoned with a calcium channel blocker.

First, observational studies should be conducted to identify which intervention improves the outcome for each specific type of calcium channel blocker (dihydropyridine or nondihydropyridine, sustained-release, etc). Currently, there is not enough evidence available to establish distinctive approaches. Second, observational studies should identify prognostic factors, which are particularly imperative in severe cases that may potentially require VA-ECMO. Third, pharmacokinetic studies, observational studies and clinical trials are required to determine which decontamination methods are more efficient in immediate or extended release formulations, and which therapies for enhanced elimination can be considered in calcium channel blocker poisoning. Fourth, scientists should conduct observational studies to identify which type of patient (with or without myocardial dysfunction) are likely to respond to high-dose insulin therapies, and the appropriate dose of these therapies. Fifth, comparative clinical trials to evaluate the use of antidotes currently recommended or, to assess new antidotes for clinical presentation with limited therapeutic options, such as vasoplegic shock in dihydropyridine poisonings, should also be considered. Finally, clinicians should be encouraged to publish both favourable and unfavourable experiences in order to

162

minimize publication bias.

These recommendation will be updated if there is a significant change in the evidence or every five years. The first author will register to Medline, EMBASE and Google

Scholar search strategies updates and will notify the rest of the working group if a new article may require an update ad hoc. Comments and suggestions will be collected on the working group’s blog in future revisions, and studies will be conducted to monitor adherence to the current recommendations.

8.5 Conclusion

The recommendations for the treatment of patients poisoned with a calcium channel blocker, endorsed by international critical care, emergency medicine and toxicology associations are built to decrease practice variation, improve consistency in messaging to the clinicians, with the ultimate goal of improving the management of this type of poisoning. These recommendations include intravenous calcium, high-dose insulin, vasopressors, incremental doses of high-dose insulin, pacemaker stimulation, lipid emulsion therapy and VA-ECMO in a step-wise treatment approach. The working group also identified potential areas for future research that may help improve the care provided to patients poisoned with a calcium channel blocker.

163

______

Chapter 9 General discussion

______

164

This thesis uses important concepts from the theory of guideline development detailed in Chapter 2 to build recommendations for the treatment of calcium channel blocker poisoning endorsed by international critical care, emergency medicine and toxicology associations. It adapts the very low level evidence outlined by the systematic review

(Chapter 6) to costs (Chapter 7), and values and preferences (Chapter 8) in order to answer key questions outlined by an analytic framework illustrated in Chapter 3.

This approach provides guidance with great transparency to decision makers and recommendations users in an area where the quality of evidence is very low. It also tests innovative ways to involve users, patients and their relatives in the evidence- based decision-making process, which may be challenging in a vulnerable population of suicide attempters treated in an acute care setting.

The following chapter summarizes the work completed, explains how it contributes to the current body of knowledge, and underlines the limitations and remaining issues to be addressed.

9.1 Summary of the research program

The first step of this program of research (Figure 7) was to identify the need for a guideline for the treatment of calcium channel blocker poisoning. A retrospective study

(Chapter 4, St-Onge et al., 2012) demonstrated a non-negligible burden of illness

(morbidity of 50%) for this type of cardiovascular drug. The same study also raised concerns regarding consistency of care and adherence to recommendations. Only 42%

165

of patients received care as recommended by the Quebec Poison Control Centre. Lack of adherence was notably observed with the administration of high-dose insulin, one of the treatments supported by the highest level of evidence (St-Onge et al., 2014).

Moreover, in a survey of emergency physicians, a significant variation in reported opinions and practices associated with background training was noted (Chapter 5, St-

Onge et al., 2014). The burden of illness, coupled with the presence of a lack in adherence to recommendations and evidence of significant practice variation are important criteria underscoring the need for a guideline as underlined by the Institute of

Medicine (Graham et al., 2011).

As a second step, the required information for decision making was collected. The evidence was documented by a systematic review detailed in Chapter 6 (St-Onge et al., 2014). Although based on low quality of evidence, this review illustrated a possible benefit of high-dose insulin on hemodynamics, and on mortality at the risks of hypoglycemia and hypokalemia. Further, a possible role for VA-ECMO in patients in severe shock or in cardiac arrest was shown to improve survival at the cost of bleeding, thrombosis, or limb ischemia. Based on very low quality of evidence, the systematic review underlined that calcium, dopamine, norepinephrine, or epinephrine and lipid emulsion may improve hemodynamics, but found inconsistent results for the use of atropine or pacemaker.

VA-ECMO was identified as one of the interventions for which there was a highest level of supporting evidence; however this therapy is expensive and rarely offered to a poisoned population in North American settings. We suspected cost may be considered

166

as a reason not to use VA-ECMO. As such, a cost-effectivess analysis was conducted from a societal perspective to determine whether this intervention should be considered a viable option in calcium channel blocker poisoning cases. This study, detailed in

Chapter 7 (St-Onge et al., 2014), supported the use of VA-ECMO in the treatment of cardiotoxicant poisonings if its effectiveness is confirmed. Costs should not be a factor in the decision to use VA-ECMO in Canadian hospitals where it is available.

As recommended by Guyatt et al. (2011), the direction (for/against) and strength

(strong/weak) of recommendations in guideline development should be based on the evidence, the balance between risks and benefits, and the use of resources, values and preferences. Therefore, the third step of this program of research used the GRADE methodology (Guyatt et al., 2011) to characterize the level of the evidence. It used a modified Delphi method to characterize the strength of the recommendation and the

RAND/UCLA Appropriateness Method (Fitch et al., 2001) to quantify disagreement within a working group of representatives of international associations in emergency medicine, critical care and toxicology, in order to integrate those important factors into recommendations for the treatment of calcium channel blocker poisoning (Chapter 8).

Using the word « recommendations » instead of « guidelines » was perceived as being more appropriate considering the poor level of evidence. However, the working group felt creating a treatment recommendation would still be beneficial in improving consistency in messaging to the clinicians. The process included face-to-face meetings, the development of an implementation tool, an internal and an external review, including key stakeholders, decision makers, users, patients and relatives.

Rigourous attention was paid to prevent financial or non-financial conflict of interests.

167

9.2 Comparison to other guideline development processes

In toxicology, a consensus guideline for out-of-hospital management of calcium channel blocker poisoning was propsed in 2005 by the American Academy of Clinical

Toxicology, the American Association of Poison Control Centres and the American

College of Medical Toxicology (Olson et al., 2005). The working group clearly identified objectives, the target users and the circumstances in which the recommendations would apply. They also systematically looked at the literature and underlined the level of evidence for each recommendation. However, the impact of values and preferences, and the balance between risks and benefits on the recommendation development was not clearly detailed and only representatives of North-American toxicology associations were part of the working group. An external review following the AGREE instrument

(Brouwers et al., 2009) would have identified those limitations. A multidisciplinary guideline committee involving patients and paying a particular attention to the implementation of those recommendations would have also facilitated adherence to the recommendations.

In 2012, the EXTRIP working group published a guideline methodology manuscript for extracorporeal treatments of poisonings (Lavergne et al., 2012) using the AGREE instrument (Brouwers et al., 2009). The first publications emerging from their work concerned the use of extracorporeal treatment for thallium poisoning (Ghannoum et al.,

2012), tricyclic antidepressant poisoning (Yates et al., 2014), barbiturates poisoining

168

(Mactier et al., 2014), acetaminophen poisoning (Gosselin et al., 2014), and carbamazepine poisoining (Ghannoum et al., 2014). Many issues previously identified were notably solved by a rigourous application of a methodology involving representatives of international associations to build their recommendations. However, given that the working group focused only on the use of one type of therapy, the different options for the management of those poisonings were not clearly presented.

Moreover, the AGREE instrument (Brouwers et al., 2009) domain related to the guideline applicability (description of facilitators and barriers to application, implementation tools and advices, resource implications, monitoring and auditing criteria) was not included in the manuscript. The values and preferences of patients and their relatives were not sought and no external review process was conducted.

Further, the guideline development process funding was obtained from industry via unrestricted educational grants. Notwithstanding these limitations, the EXTRIP working group contributed to raising the standard of guideline development in toxicology.

In critical care, the AGREE instrument was applied for the development of guidelines for the use of noninvasive positive-pressure ventilation and noninvasive continuous positive airway pressure in the acute care setting (Keenan et al., 2011). In these guidelines, the domain of applicability was addressed, the values and preferences of the working group were disclosed and a peer review was performed. However, decision makers, guidelines users, patients and relatives were not involved in the process and the same choice was made for the surviving sepsis campaign (Dellinger et al., 2013).

169

The American College of Cardiology (ACC) and the American Heart Association (AHA) have partnered for nearly 20 years to develop guidelines in emergency medicine and resuscitation sciences. They underline: "Our guidelines development methodologies have evolved over the years as we learned first hand the complexities of combining research findings and clinical judgement to create the best possible recommendations for patient care" (ACC/AHA Task Force on Practice Guidelines, 2014). The ACC/AHA

Task Force follows the recommendations of the Institute of Medicine (Graham et al.,

2011), the AGREE instrument (Brouwers et al., 2009) and addresses the issues previously underlined. They have organized an efficient system for updating their guidelines and accept no industry funding for guideline development (ACC/AHA, 2010).

9.3 Limitations related to these recommendations

The recommendations for the treatment of calcium channel blocker poisoning and their development process presented in the previous chapters have some limitations. First, the data concerning the burden of illness, the lack in recommendations adherence and the practice variation come mainly from North American countries. St-Onge et al.,

(2012), Darracq et al., (2013) and Espinoza et al. (2013) found similar problems in adherence to recommendations, notably with the use of high-dose insulin. However, there are no similar studies conducted in other continents. Therefore, the implementation strategies to consider may vary from one region to another.

170

Second, the quality of evidence available to treat calcium channel blocker poisonings was very low, which means that the true effect may be substantially different from our estimate (Lavergne et al., 2012). The GRADE system (Guyatt et al., 2011) used in the recommendations development process underlines that limitation. The working group members provided guidance for the treatment of calcium channel blocker poisoning based on available evidence, but also identified areas of research to consider in the future. This group stressed the importance of conducting high quality observational studies and comparative clinical trials to improve the quality of evidence.

Third, most of the evidence available measured the effects of interventions on intermediate outcomes, such as lipid emulsion improving hemodynamic, instead of health outcomes, such as mortality, functional outcomes or length of stay. It is likely that some interventions improve intermediate outcomes without improving mortality.

Therefore, a careful interpretation of the evidence is required when the benefit of an intervention favours only intermediate outcomes, but raises doubt concerning its effect on patient-centered health outcomes.

Fourth, although the working group members were representatives of professional organizations from different regions of the world, the RAND/UCLA Appropriateness

Method allows for a transparent consensus method, and the rationale for each recommendation was detailed explicitly. Some associations did not endorse the recommendations related to decontamination. Given that the positions adopted by those associations were leading to opposite opinions, the working group had to defer to an already existing position statement (AACT and EAPCCT, 2005). Decontamination is

171

an area of controversy where practice variation is significant. This issue was highlighted in the study detailed in Chapter 5, and also in a survey of medical directors from 72 poison control centres in the United States and Canada, who suggested 30 different management strategies for the same hypothetical patient acutely poisoned with enteric-coated acetylsalicylic acid (Juurlink et al., 2002). Concordance was poorest for advice on gastric lavage (k=0.093, P=0.445) and multidose charcoal (k=0.039,

P=0.745). The working group responsible for the recommendations pertaining to calcium channel blocker poisoning presented in this document experienced the same type of discordance with those two interventions during the internal review process.

Fifth, the implementability of some recommendations may be challenging depending on the system, practice setting, care providers’ characteristics, recommendations, attributes and the characteristics of the innovation (CIHR, 2014). The system may not have enough resources to provide VA-ECMO or even enough human resources to monitor the administration of high-dose insulin. In certain practice settings, the clinician may perceive that the patients’ coingestants and comorbidities limit the applicability of some recommendations. The care providers’ characteristics may also represent a barrier if the majority of clinicians taking care of poisoned patients never received formal toxicology training as identified in Chapter 5. Further, different people have different motivations for change, attitudes and opinions. Recommendation attributes

(Grol et al., 1998) may constitute a barrier, given that the level of evidence is low; some areas remain controversial and some interventions imply a significant change in usual practice, as in the case of high-dose insulin. Finally, the characteristics of the innovation (Rogers, 2003) may be a barrier notably for the use of VA-ECMO. This is a

172

complex intervention not accessible everywhere, and it is unclear if the use of this invasive strategy in a patient who has attempted suicide is compatible with clinicians’ values and beliefs. Considering the potential barriers to implementation, the working group created multifaceted implementation strategies and proposed to monitor the effect of the implementation of the recommendations on outcomes and resources use.

Sixth, despite tremendous efforts to document preferences and values of patients and relatives, the suggestions provided during the review process came mainly from decision makers and users. Although a few associations in suicide prevention, health ethics and patient advocacy replied to the invitation to join the blog, none of them provided feedback. This may be because they found no area conflicting with their preferences and values, or because they needed more guidance to interact with others on this type of platform, or the manner in which the information was presented did not favour participation from the lay public. Having patients’ representatives joining the working group might have ensured the information was accessible, but some working group members felt that the participation of patients and relatives to the decision- making process in an acute care setting is of limited value.

Finally, there were no representatives from developing countries on the working group.

Therefore, the applicability of the recommendations for the treatment of calcium channel blocker poisoning in these practice settings is unknown. A next step could be to submit the recommendations to representatives of developing countries so as to document their applicability in these countries.

173

9.4 Remaining issues related to these recommendations

9.4.1 Recommendations

These recommendations for the treatment of calcium channel blocker poisoning focuses on the medical management of the acute phase. Strategies for primary, secondary and tertiary prevention could have been discussed, but currently no literature is specific to this poisoning. Interventions, such as limiting access to the number of cardiotoxicant pills to patients at risk of poisoning, may be considered.

Moreover, the support required for families during critical illness of poisoned patients has not been studied. In a general critical care population, the most relevant needs expressed by families are related to obtaining desired results and having sincere communication with healthcare staff (Padilla Fortunatti, 2014). However, those needs may be different in cases of suicide attempts, more specifically for the relative who found the victim.

9.4.2 Patients and public involvement

How to effectively involve patients and relatives in the guideline development process remains unclear. The Guidelines International Network published a toolkit for patients and public involvement (Knaapen et al., 2012). The network stresses that patients and the public should be involved for a variety of reasons, such as well-informed choice,

174

accountability, equality of care, quality of care or improved implementation. Among the possible strategies the network proposes is an official consultation process - including online interface, focus groups, individual interviews, direct participation in the guideline development, and facilitated communication to patients and the public with tools such as decisions aids. However, the efficacy of each strategy has not been measured, and what is considered as successful or significant involvement is unclear. Furthermore, it is difficult to obtain participation from patients and public advocates representing the broad diversity of preferences and values. In a chronic or subacute setting, it may be easier to involve the individual in the decision-making process, but this option is sometimes challenging in an acute care setting.

In the recommendations development process for the treatment of calcium channel blocker poisoning, social media was used to consult decision makers, guidelines users, patients and relatives. However, as previously stated, the public did not actively engage in these opportunities. Prevoius guideline development on infertility used a wiki as a participatory tool for patients (Den Breejen et al., 2012). The authors started with semistructured interviews with infertile couples that were then posted on the wiki. They allowed access for seven months and obtained 265 suggestions. Although it is likely that the general public feels more concern about infertility than calcium channel blocker poisoning, it would be informative to understand why there was limited public participation in the attempts outlined in this thesis.

9.4.3 Impact of the recommendations

175

The impact of the recommendations implementation on consistency, adherence, patient outcome and resources use still needs to be assessed. The working group plans to survey key stakeholders two years after the recommendations implementation to document perceived adherence, barriers, facilitators and impact on resource use.

Ideally, an observational study looking at the effect on health outcomes should also be conducted. However, this may be challenging given that health outcome data collected in poison control centre databases are often incomplete.

In the United States, the need for a valid national database has been recognized. Each year the Association of Poison Control Centres publishes the Annual Report of the

National Poison Data System (Mowry et al., 2013). The American College of Medical

Toxicology has created a Toxicology Investigators Consortium (ToxIC) national registry to complement the poison control centres’ data with cases treated by bedside toxicologists (ACMT, 2014). However, those cases are reported only on a voluntary basis. In Canada, Health Canada is currently working with the Canadian Association of

Poison Control Centres on the development of a national database, however, a lot of work still needs to be done before the data can be used for research purposes. Once the database becomes available, it will be easier to conduct observational studies documenting outcomes pre and post recommendations implementation.

9.4.4 Ethical considerations

Finally, some ethical considerations need to be discussed in more detail. First, it is unclear to what extent a patient who just attempted suicide can actively participate in

176

the decision-making regarding their own care. The assistance of a third party, such as a surrogate, a consultant, or another clinician, to determine capacity may be valuable in discerning the most appropriate action (Larkin et al., 2001). It should be noted that suicidal ingestions neither preclude decision-making capacity nor should be accepted as an “autonomous” expression (Abbott, 2014).

Another important ethical issue revolves around stigma and decisions around care.

When resources are limited, physicians may be tempted not to offer an invasive strategy, such as VA-ECMO, to patients who attempted suicide. However, some research has raised doubts about whether clinicians may be more inclined to stigmatize in cases of attempted suicide than the general population (Scocco et al., 2012). In fact, the rates for later suicide depending on the population can vary between 4-12%

(Abbott, 2014) and a case-control study (Karvellas et al., 2010) of acetaminophen- induced acute liver failure (n = 36, 56% with psychiatric diagnosis and 25% with previous suicide attempts) compared to non-acetaminophen induced acute liver failure

(n = 35, 0% with psychiatric diagnosis or previous suicide attempt) reported only two reattempted suicides post liver-transplant (one died 8 years post liver-transplant).

Consequently, should the decision of offering invasive strategy, such as VA-ECMO, in patients who attempt suicide rely on an isolated physician opinion? More ethical guidance is required.

In conclusion, the need for a guideline was documented, an analytical framework was built, and the information required for decision making was gathered to finally develop recommendations for the treatment of calcium channel blocker poisoning. The

177

approach used was novel notably because it integrated all domains of the AGREE instrument to build a step-wise approach, including different types of interventions for the management of a specific type of poisoning.

Future studies need to not only assess the effect of these recommendations implementation, but also strive to develop a method to maximize public involvement.

This will be particularly important if the updated version of the recommendations includes prevention, support to families and ethical considerations.

178

______

Chapter 10 Conclusion

______

179

Using calcium channel blocker poisoning as an example, the goal of this thesis was to create treatment recommendations in acute care toxicology endorsed by international critical care, emergency medicine and toxicology associations. In doing so, this research may substantially improve consistency in messaging to the clinician based on the poor level of evidence. It may also favors adherence to recommendations by adapting the evidence to costs, resources, values and preferences.

10.1 Process to achieve the aim of this program of research

To achieve the aim of this program of research, once the need for a guideline had been demonstrated, an analytic framework was built as recommended by the US Preventive

Services Task Force (Harris et al., 2001). A thorough systematic review detailed the evidence available for the relevant key questions and a cost-effectiveness analysis was conducted for the intervention (VA-ECMO) for which cost may have a significant impact on its accessibility and use.

Following the AGREE instrument (Brouwers et al., 2009), recommendations were developed as a separate step to clearly distinguish the evidence from the clinical guidance provided by a group of international experts. A rigourous and transparent methodology allowed the group to build recommendations based not only on the evidence, but also on the balance between risks and benefits, costs and resources use, and values and preferences. More importantly, the working group identified knowledge gaps that should be targeted in future research. Modifications made after the internal

180

and external review process were disclosed. The project was not externally funded and none of the working group members had significant conflict of interests.

10.2 Impact of recommendations development for the treatment of calcium channel blocker poisoning

The impact of the recommendations development for the treatment of calcium channel blocker poisoning in acute care toxicology is significant. These are the first recommendations proposed for the in-hospital management of this type of poisoning.

Based on an analytical framework, it integrates multiple treatments in a stepwise fashion instead of considering the indications of a single intervention, which often depend on the other options available. Moreover, as opposed to other recommendations published in acute care toxicology, particular attention was paid to patients and public involvement, implementation and monitoring.

The process used for building recommendations for the treatment of calcium channel blocker poisoning also innovates the field of guideline development methodology by using social media to involve the public, users and decision makers. The use of social media in this manner certainly requires further exploration, but already shows a promising potential to rapidly reach a large numbers of people.

181

______

Chapter 11 Future directions

______

182

This final chapter presents future areas of research in the fields of cardiotoxicants poisonings, and guideline development and implementation in acute care toxicology.

These represent ideas for research programs that may be undertaken in the future.

11.1 Future directions in cardiotoxicants poisonings

Within the field of cardiotoxicants poisonings a variety of programs of research can be considered. These include those related to the prevention of cardiotoxicants poisonings, the use of various decontamination methods, and the use of high-dose insulin or rescue treatments.

Considering that cardiotoxicants poisonings are associated with significant mortality

(Mowry et al., 2013), prevention is obviously an efficient way to decrease mortality. A program of research in this area may answer the question: “In a general population, which preventive intervention(s) decrease(s) the incidence of cardiotoxicants poisonings compared to other interventions or no intervention?” A systematic review of preventive interventions should be completed as a first step in order to identify which interventions or group of interventions should subsequently be compared in an observational study conducted as a second step. A cluster randomized controlled trial is likely to be challenging at this stage given the paucity of observational data available.

However, this could be undertaken as observational data becomes increasingly available.

183

Moreover, the wide discordance in opinions concerning the use of various decontamination methods that was observed during the recommendations development process stresses the importance of conducting comparative studies in that area. A relevant question may be formulated as follows: “In patients who ingested a potentially toxic amount of cardiotoxicants, which decontamination methods decreases toxicity and mortality compared to other intervention or no intervention?” The program of research should first survey clinicians to confirm practice variation, document equipoise and clarify what would be considered a minimal significant clinical effect. Then a pilot randomized controlled trial could be developed to assess the event rate, recruitment rate, protocol adherence and lost to follow up to determine if a formal randomized controlled trial is feasible.

The same type of program of research could also be applied to another relevant clinical question: “In patients poisoned with a beta-blocker or a calcium channel blocker, does the addition of high-dose insulin to vasopressors decrease mortality, length of stay in the critical care unit or in-hospital or improve functional outcomes?” Some scientists may prefer head-to-head comparison between high-dose insulin and vasopressors.

However, no human studies have demonstrated the safety of the use of high-dose insulin alone.

Considering that rescue treatments are reserved for a limited number of patients, case- control studies should be considered for those types of interventions. A program of research investigating the question: “In patients poisoned with a calcium channel blocker, which rescue treatment decreases mortality, length of stay in the critical care

184

unit or in-hospital or improve functional outcomes?” The first step would be to develop a survey to identify which rescue treatment should be studied in priority, an observational study should then look at prognostic factors indicating when to consider rescue treatments, and finally case-control studies should be conducted. The use of incremental doses of high-dose insulin, pacemaker, lipid emulsion therapy or VA-

ECMO are examples of interventions that can be studied.

11.2 Future directions in guideline development in acute care toxicology

Concerning the field of guideline development in acute care toxicology, the same process used in this thesis can be applied to other cardiotoxicants, such as beta- blocker poisonings. Research programs can also be built to further explore patient and public involvement as well as ethical considerations in guidelines development.

To answer the question “Which intervention favours a more active patient and public involvement in guideline development compared to other interventions or no intervention?”, a survey or qualitative study documenting barriers and facilitators can be conducted. As a second step, a review of interventions (including social media) should be completed to compare various interventions or group of interventions in observational studies.

In terms of answering “Should invasive treatments be considered in patients who attempted suicide when they are treated in a practice setting where resources are

185

limited?”, a survey or qualitative study comparing the opinions of physicians, patients and relatives should be developed and administered first before initiating a broader public consultation.

11.3 Future directions in guideline implementation in acute care toxicology

Finally, a program of research can also be developed to answer: “Which implementation strategies facilitate the greatest guidelines adherence compared to other strategies or no strategy?” As a first step, a survey or a qualitative study could be conducted to document barriers and facilitators specific to guideline implementation in acute care toxicology. As a second step, a scoping review or a systematic review would indicate the options already studied and, as a final step, observational studies could be carried out to address this issue.

11.4 Conclusion

In conclusion, this thesis leads to many other possible research programs in the areas of cardiotoxicants poisonings, and guideline development and implementation in acute care toxicology. The research question to prioritize may depend on the resources and opportunities available. However, the US Preventive Task Force (Harris et al., 2001) suggests that priorities should be determined by key organizations on the basis of the burden of suffering to society or individuals and on the potential effectiveness of one or more interventions. As a general rule, research priorities should target subjects having

186

the potential of improving patient-centered outcomes.

187

References

1. Abeysinghe N, Aston J, Polouse S (2010). Diltiazem overdose: a role for high-dose insulin. Emerg Med J 27:802-803.

2. Abbott J (2014). Ethics and the Poisoned Patient. Presentation at the NACCT conference in New Orleans (unpublished document).

3. Aboulsoud S, Huckson S, Wyer P, Lang E (2012). Survey of preferred guideline attributes: what helps to make guidelines more useful for emergency health practitioners ? Int J Emerg Med 5:42.

4. Abraham MK, Scott SB, Meltzer A, Barrueto F Jr (2009). Levosimendan does not improve survival time in a rat model of verapamil toxicity. J Med Toxicol 5:3-7.

5. ACC/AHA Task Force on Practice Guidelines, Methodologies and policies from the ACC/AHA task force on practice guidelines, http://my.americanheart.org/professional/StatementsGuidelines/PoliciesDevelop ment/Development/Methodologies-and-Policies-from-the-ACCAHA-Task-Force- on-Practice-Guidelines_UCM_320470_Article.jsp, consulted June 15th 2014.

6. ACC/AHA Task Force on Practice Guidelines, Methodology manual and policies from the ACCF/AHA Task Force on practice guidelines, ACCF/AHA, June 2010, 88pp.

7. Adams BD, Browne WT (1998). Amlodipine overdose causes prolonged calcium channel blocker toxicity. Am J Emerg Med16:527-528.

8. ADAPTE Collaboration. The ADAPTE process: resource toolkit for guideline adaptation. Version 2.0. Berlin (Germany): Guideline International Network; 2009. Available: www.g-i-n .net /document -store/working-groups- documents/adaptation /adapte-resource -toolkit-guideline-adaptation-2-0.pdf, consulted Jan 25tth 2014.

9. Agarwal A, Yu SW, Rehman A, Henkle JQ (2012). Hyperinsulinemia euglycemia therapy for calcium channel blocker overdose: a case report. Tex Heart Inst J 39:575-578.

10. Aggarwal N, Kupfer Y, Seneviratne C, Tessier S (2013). Methylene blue reverses recalcitrant shock in ß-blocker and calcium channel blocker overdose. BMJ Case Rep 2013. doi:10.1136/bcr-2012-007402.

11. Agoston S, Maestrone E, van Hezik J, Ket JM, Houwertjes, Uges DRA (1984). Effective treatment of verapamil intoxication with 4-aminopyridine in the cat. J Clin Invest 73:1291-6.

188

12. AGREE Next Steps Consortium (2009). Appraisal of Guidelines for Research and Evaluation II, 56pp.

13. Akinci E, Koylu R (2013). Verapamil poisoning, the importance of intravenous lipid therapy: Case report. Akademik Acil Tip Olgu Sunumlari Dergisi 4:130-132.

14. Al-Nabhan M (2011). Lipid emulsion therapy for verapamil overdose. Chest 140:184A.

15. American Academy of Clinical Toxicology and European Association of Poisons Centres and Clinical Toxicologists (2005). Position paper: Single-dose activated charcoal. Clin Toxicol 43:61-87.

16. American College of Medical Toxicology (2014). Toxicology Investigators Consortium (ToxIC), http://www.acmt.net/cgi/page.cgi/ToxIC1.html, consulted November 10th 2014.

17. American Heart Association (2008). ACLS resource text for instructors and experienced providers. American Hearth Association, Texas, 444pp.

18. Anthony T, Jastremski M, Elliott W, Morris G, Prasad H (1986). Charcoal hemoperfusion for the treatment of a combined diltiazem and metoprolol overdose. Ann Emerg Med 15:1344-1348.

19. Armenian P, French D, Smollin C, Olson KR, Wu AHB (2010). Prolonged absorption from a sustained-release verapamil preparation with documentation of serum levels and their response to intralipid. Clin Toxicol 48:646.

20. Ashraf M, Chaudhary K, Nelson J, Thompson W (1995). Massive overdose of sustained-release verapamil: a case report and review of literature. Am J of Med Sci 310:258-263.

21. Assink MAJ, Spronk PE, van Kan HJM, Brader A (2013). Intravenous lipid emulsion in the treatment of verapamil intoxication. Neth J Crit Care 17:18-21.

22. Atkins D, Perez-Padilla R, MacNee W, Buist AS, Cruz AA (2012). Priority setting in guideline development: article 2 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:225-8.

23. Azendour H, Belyamani L, Atmani M, Balkhi H, Haimeur C (2010). Severe amlodipine intoxication treated by hyperinsulinemia euglycemia therapy. J Emerg Med 38:33-35.

24. Babatasi G, Massetti M, Verrier V, Lehoux P, LePage O, Bruno PG, Khayat A (2001). Severe intoxication with cardiotoxic drugs: value of emergency percutaneous cardiocirculatory assistance. Arch Mal Coeur Vaiss 94:1386-92.

189

25. Balshem H, Helfand M, Schünemann HJ, Oxman AD, Kunz R, Brozek J, Vist GE, Falck-Ytter Y, Meerpohl J, Norris S, Guyatt GH (2011). GRADE guidelines: 3. Rating the quality of evidence. J Clin Epi 64:401-6.

26. Bania TC, Chu J, Almond G, Perez E (2004). Dose-dependent hemodynamic effect of digoxin therapy in severe verapamil toxicity. Acad Emerg Med 11:221-7.

27. Bania TC, Chu J, Perez E, Su M, Hahn IH (2007). Hemodynamic effects of intravenous fat emulsion in a an animal model of verapamil toxicity resuscitated with atropine. Acad Emerg Med 14:105-111.

28. Barr J, Kishman CP, Jaeschke R (2013). The methodological approach used to develop the 2013 Pain, Agitation and Delirium clinical practice guidelines for adult ICU patients. Crit Care Med 41:S1-S15.

29. Barrow PM, Houston PL, Wong DT (1994). Overdose of sustained-release verapamil. British J of Anaesthesia 72:361-365.

30. Barry JD, Durkovich D, Cantrell L, Richardson W, Tong T, Offerman S, Clark RF, Tanen DA, Williams S (2005). Vasopressin treatment of verapamil toxicity in the porcine model. J Med Toxicol 1:3-10.

31. Baud FJ, Mégarbane B, Deye N, Leprince P (2007). Clinical review: aggressive management and extracorporeal support for drug-induced cardiotoxicity. Crit Care 11:207.

32. Bausch R, Dirschedl P, Fahrenkrog U, Loellgen H (1991). Clinical observations of calcium-channel-blocker intoxication. Intensive med 28:120-123.

33. Becker M, Neugebauer EAM, Eikermann M (2014). Partial updating of clinical practice guidelines often makes more sense than full updating: a systematic review on methods and the development of an updating procedure. J Clin Epi 67:33-45.

34. Beiträge ZK (1984). Suicide attempt with verapamil, Therapiewoche 34:2483-2491.

35. Beitzke A., Grubbauer HM (1976). Isoptin (verapamil) poisoning, Padiatr Padol 11:570-3.

36. Belleflamme M, Hantson P, Gougnard T, Minon JM, Wittebole X' Laterre PF, Mairesse J, Gugrenier T (2012). Survival despite extremely high plasma diltiazem level in a case of acute poisoning treated by the molecular-adsorbent recirculation system. Eur J Emerg Med 19:59-61.

37. Belson MG, Gorman SE, Sullivan K, Geller R (2000). Calcium channel blocker ingestions in children. Am J Emerg Med 18:581-586.

190

38. Bertrand N, Bouvet C, Moreau P, Leroux JC (2010). Transmembrane pH- gradient liposomes to treat cardiovascular drug intoxication. ACS Nano 4:7552- 8.

39. Bhimani M, Dickie G, McLoed S, Kim D (2007). Emergency medicine training demographics of physicians working in rural and regional southwestern Ontario emergency departments. CJEM 9:449-452.

40. Böhm M, Schwinger RH, Erdmann E (1990). Different cardiodepressant potency of various calcium antagonists in human myocardium. Am J Cardiol 65:1039-41.

41. Bologa C, Lionte C, Coman A, Sorodoc L (2013). Lipid emulsion therapy in cardiodepressive syndrome after diltiazem overdose-case report. Am J Emerg Med 31:1154.e3-4.

42. Bouchard NC, Weinberg RL, Burkart KM, Bacchetta M, Dzierba A, Story D, Livshits Z, Mulcare M, Kuah SS, Farmer BM, Prosser JM, Rao RB, Hoffman RS (2010). Prolonged resuscitation for massive amlodipine overdose with maximal vasopressors, intralipids and veno arterial-extracorporeal membrane oxygenation (VA ECMO). Clin Toxicol 48:613.

43. Boyd EA, Akl EA, Baumann M, Curtis JR, Field MJ, Jaeschke R, Osborne M, Schünemann HJ (2012). Guideline funding and conflicts of interest: article 4 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:234-42.

44. Boyer EW, Duic PA, Evans A (2002). Hyperinsulinemia/euglycemia therapy for calcium channel blocker poisoning. Pediatr Emerg Care 18:36-37.

45. Boyer EW, Shannon M (2001). Treatment of calcium-channel-blocker intoxication with insulin infusion. N Engl J Med 344:1721-2.

46. Briggs AH, Weinstein MD, Fenwick EA, Karnon J, Sculpher MJ, Paltiel AD (2012). Model parameter estimation and uncertainty analysis: A report of the ISPOR-SMDM Modeling Good Research Practices Task Force Working Group– 6. Med Decis Making; 32:722-732.

47. Bronstein AC, Spyker DA, Cantilena LR, Jr, Green JL, Rumack BH, Dart RC (2012). 2011 Annual report of the American Association of Poison Control Centers’ National Poison Data System (NPDS): 29th annual report. Clin Toxicol 50:911-1164.

48. Brouwers MC, Kho ME, Browman GP, Burgers JS, Cluzeau F, Feder G, Fervers B, Chuang R, Heard KJ (2010). Types of Studies Used to Support Treatment Recommendations in Medical Toxicology, J Med Toxicol. 6:345-348.

49. Brown GK, Ten Have T, Henriques GR, Xie SX, Hollander JE, Beck AT (2005). Cognitive therapy for the prevention of suicide attempts: a randomized 191

controlled trial. JAMA 294(5):563-570.

50. Bryant SM, Espinoza TR, Aks SE (2009). Seven years of high dose insulin therapy for calcium channel antagonist poisoning. Clin Toxicol 47:751.

51. Brunton L, Parker K, Blumenthal D, Buxton I (2008). Name of the chapter. In: Goodman & Gilman's: Manual of Pharmacology and Therapeutics. Place of edition, McGraw Hill,1219pp.

52. Buckley N, Dawson AH, Whyte IM (1993). Slow-release verapamil poisoning. Use of polyethylene glycol whole-bowel lavage and high-dose calcium. Med J Aust 158:202-4.

53. Burgers JS, Anzueto A, Black PN, Cruz AA, Fervers B, Graham ID, Metersky M, Woodhead M, Yawn BP (2012). Adaptation, evaluation, and updating of guidelines: article 14 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:304-10.

54. Buylaert WA, Callens B, Decruyenaere JM, Koch AR, Verstraete A, Vogelaers DP (1995). Fatal intoxication with amlodipine. Journal of Toxicol Clin Toxicol 33:253.

55. Canadian Institutes of Health Research, http://www.cihr- irsc.gc.ca/e/39033.html, consulted October 2013.

56. CAPQ (Centre antipoison du Québec) (2009). Annual report 2009. Unpublished document.

57. Cavarocchi N, Wallace S, Hong E, Tropea A, Byrne J, Pitcher H, Hirose H (2014). A cost-reducing extracorporeal membrane oxygenation (ECMO) program model: a single institution experience. Perfusion (Epup ahead of print)

58. Chandrashekar H, Lambert M, George P (2011). Calcium channel blocker poisoning. Acute Med 10:223-224.

59. Chambers D, Rodgers M, Woolacott N (2009). Methods of systematic reviews and meta-analysis: Not only randomized controlled trials, but also case series should be considered in systematic reviews of rapidily developing technologies. J Clin Epi 62:1253-60.

60. Chimienti M, Previtalli M, Medici A, Piccinini M (1982). Acute verapamil poisoning: successful treatment with epinephrine. Clin Cardiol 5:219-222.

61. Choudhry NK, Stelfox HT, Detsky AS (2002). Relationships between authors of clinical practice guidelines and the pharmaceutical industry. JAMA 287:612-7.

62. Chu J, Donovan SP, Lepp D, Rolston D, Patel V, Wall E, Bania T (2013). 192

The effects of increased dosing of L-carnitine in a model of verapamil toxicity. Acad Emerg Med 20:S119.

63. Chu J, Donovan SP, Rolston D, Lepp D, Patel V, Wall E, Bania T (2013). The effect of intravenous lipid emulsions in an oral verapamil toxicity model. Acad Emerg Med 20:S120.

64. Chuang R, Heard KJ (2010). Types of studies used to support treatment recommendations in medical toxicology. J Med Toxicol 6:345-8.

65. CIHR (Canadian Institutes of Health Research), Evidence-Informed Decision Making Process Algorithm. http://www.cihr-irsc.gc.ca/e/45670.html#bc, consulted October 14th 2013.

66. CIHR (Canadian Institutes of Health Research), Moving into action: We know what practices we want to change, now what? An implementation guide for health care practitioners. http://www.cihr-irsc.gc.ca/e/45669.html#a5, consulted November 11th 2014.

67. Clark EG, Nykamp DL, Nguyen VV (2008). High-dose insulin in the treatment of antihypertensive overdose. Hosp Pharmacy 43:206-209.

68. Cluzeau F, Littlejohns P, GrimshawJ, , Feder G, Moran SE (1999). Development and application of a generic methodology to assess the quality of clinical guidelines. Int J Qual Health Care 11:21-28.

69. Cluzeau F, Wedzicha JA, Kelson M, Corn J, Kunz R, Walsh J, Schünemann HJ (2012). Stakeholder involvement: how to do it right: article 9 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:269-73.

70. Cohen V, Jellinek SP, Fancher L, Sangwan G, Wakslak M, Marquart E, Farahani C (2011). Tarka (trandolapril/verapamil hydrochloride extended release) overdose. J Emerg Med 40:291-295.

71. Cooke CR. Kahn JM, Watkins TR, Hudson LD, Rubenfeld GD (2009). Cost-effectiveness of implementing low-tidal volume ventilation in patients with acute lung injury, CHEST 136:79-88.

72. Cooper G, Dyas J, Krishma CV, Thompson JP (2010). Successful use of intravenous fat emulsion in severe poisoning following ingestion of lipid soluble drugs. Clin Toxicol 48:298.

73. Cumpston KL, Rose SR (2009). Titration of hyperinsulinemia euglycemia therapy for the treatment of acute diltiazem toxicity. Clin Toxicol 47:724.

74. Cumpston KL, Aks SE, Sigg T, Pallasch E (2010). Whole bowel irrigation and the hemodynamically unstable calcium channel blocker overdose: primum non nocere. J Emerg Med 38:171-4. 193

75. Crump BJ, Holt DW, Vale JA. Lack of response to intravenous calcium in severe verapamil poisoning. The Lancet 1982; October 23: 939-940.

76. Daubin C, Lehoux P, Ivascau C, Tasle M, Bousta M, Lepage O, Quentin C, Massetti M, Charbonneau P (2009). Extracorporeal life support in severe drug intoxication: a retrospective cohort study of seventeen cases. Crit Care 13:R138.

77. Darracq MA, Thornton SL, Do HM, Clark RF, Cantrell FL (2013). Utilization of hyperinsulinemia euglycemia and intravenous fat emulsion following poison center recommendations. J Med Toxicol 9:226-230.

78. Davis D, Goldman J, Palda VA (2007). Handbook on clinical practice guidelines. Canadian Medical Association, 40pp.

79. De Breejen E ME, Nelen W LDM, Knijnenburg J ML, Burgers JS, Hernens R PMG, Kremer J AM (2012). Feasibility of a Wiki as a participatory tool for patients in clinical guideline development. J Med Internet Rest e138.

80. DeLange DW, Sikma MA, Meulenbelt J (2013). Extracorporeal membrane oxygenation in the treatment of poisoned patients. Clin Toxicol 51:385-393.

81. Dellinger RP, Levy MM, Rhodes A, Annane D, Gerlach H, Opal SM, Sevransky JE, Sprung CL, Douglas IS, Jaeschke R, Osborn TM, Nunnally ME, Townsend SR, Reinhart K, Kleinpell RM, Angus DC, Deutschman CS, Machado FR, Rubenfeld GD, Webb SA, Beale RJ, Vincent JL, Moreno R (2013). Surviving sepsis campaign: International guidelines for management of severe sepsis and septic shock: 2012. Crit Care Med 41:580-637.

82. Dekkers OM, Egger M, Altman DG, Vandenbroucke JP (2012). Distinguishing case series from cohort studies. Ann Intern Med 156:37-40.

83. DeRita F, Barozzi L, Franchi G, Faggian G, Mazzuco A, Luciani GR (2011). Rescue extracorporeal life support for acute verapamil and propranolol toxicity in a neonate. Artif Organs 35:416-420.

84. Deshpande LMK (2004). Diltiazem overdose haemodynamic response to hyperinsulinaemia-euglycaemia therapy: a case report. Crit Care and Resusc 6:28-30.

85. Desebbe O, Rosamel P, Henaine R, et al (2013). Transport interhospitalier sous extracorporeal life support: résultats et perspectives après cinq ans d'expérience. Ann Fr Anesth Reanim; 32:225-230.

86. Deters M, Friesecke S, Hentschel H (2010). Fatal poisoning caused by felodipine. Clin Toxicol 48:281.

194

87. Devasahayam J, Pillai U, Uppaluri C (2012). Acute severe intestinal obstruction secondary to amlodipine toxicity. QJM 105:467-469.

88. De Witt CR, Waksman JC (2004). Pharmacology, Pathophysiology and Management of Calcium Channel Blocker and Beta-Blocker Toxicity. Toxicol Rev 23:223-238.

89. Dogan M, Basaranoglu M, Peker E, Akbayram S, Sahin M, Uner A, Caksen H (2010). Tarka overdose in a young child. Hum Exp Toxicol 30:1392- 1398.

90. Doyon S, Roberts JR (1993). The use of glucagon in a calcium channel blocker overdose. Ann Emerg Med 22:1229-1233.

91. Eccles MP, Grimshaw J, Shekelle P, Schünemann HJ, Woolf S (2012). Developing clinical practice guidelines: target audiences, identifying topics for guidelines, guideline group composition and functioning and conflicts of interest. Implement Sci 7:60.

92. Engebretsen KM, Morgan MW, Stellpflug SJ, Cole JB, Anderson CP, Holger JS (2010). Addition of phenylephrine to high-dose insulin in dihydropyridine overdose does not improve outcomes. Clin Toxicol 48:806-12.

93. El Houari T, Haddiya I, Ouafi NE, Bazid Z (2013). A survival case in a severe amlodipine intoxication. Case Rep Cardiol 2013:842606.

94. Espinoza T, Bryant SM, Aks SE (2013). Hyperinsulin therapy for calcium channel antagonist poisoning: a seven-year retrospective study. Am J Ther 20:29-31.

95. Extracorporeal Life Support Organization (2013). Extracorporeal Life Support Organization (ELSO) General Guidelines for all ECLS Cases, Version 1.1. November 2013, 24pp.

96. Ezidiegwu C, Spektor Z, Nasr MR, Kelly KC, Rosales LG (2008). A case report on the role of plasma exchange in the management of a massive amlodipine besylate intoxication. Ther Apher and Dial 12:180-184.

97. Fabbri LM, Boyd C, Boschetto P, Rabe KF, Yawn B, Leff B, Kent DM, Schünemann HJ (2012). How to integrate multiple comorbidities in guideline development: article 10 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:274- 81.

98. Fant JS, James LP, Fiser RT, Kearns GL (1997). The use of glucagon in nifedipine poisoning complicated by clonidine ingestion. Ped Emerg Care 13:417-419.

99. Felgenhauer EF, Zilker N (2004). Effectiveness of calcium chloride and 195

high dose insulin therapy in a life threatening intentional quinidine and verapamil intoxication. A case report. J Toxicol Clin Toxicol 42:557-558.

100. Field MJ, Lohr KN (1990). Clinical practice guidelines: directions for a new program. Washington, DC: National Academy Press.

101. Fiszer M, Kolacinski Z, Rechcinski T (2007). The application of 4- aminopyridine in calcium channel inhibitors acute poisoning. Przegl Lek 64:293- 7.

102. Fitch K, Bernstein SJ, Aguilar MD, Burnand B, LaCall JR, Lazaro P, Loo MVH, McDonnell J, Vader JP, Kahan JP (2001). The RAND/UCLA Appropriateness Method User's Manual. Library of Congress, United States, 123 pp.

103. Forster V, Luciani P, Leroux JC (2012). Treatment of calcium channel blocker-induced cardiovascular toxicity with drug scavenging liposomes. Biomaterials 33:3578-3585.

104. Franxman TJ, Al-Nabhan M, Cavallazzi RS (2011). Lipid emulsion therapy for verapamil overdose. Ann Int Med 154:292.

105. French D, Armenian P, Ruan W, Wong A, Drasner K, Olson KR, Wu AHB (2011). Serum verapamil concentrations before and after Intralipid therapy during treatment of an overdose. Clin Toxicol 49:340- 344.

106. Frese JH, Rohland L, Schulz M, Schmoldt A (1988). Intoxication with the calcium antagonist Gallopamil: Course and treatment. Dtsch Med Wochenschr 113:770-772.

107. Frierson J, Bailly D, Shultz T, Sund S, Dimas A (1991). Refractory cardiogenic shock and complete heart block after unsuspected verapamil-SR and atemolol overdose. Clin Cardiol 14:933-935.

108. Fung HT, Lai CH, Wong OF, Lam KK, Kam CW (2007). The use of glucagon and other antidotes in a case of beta-blocker and calcium channel blocker overdose. Hong Kong J of Emerg Med 14:113-118.

109. Gadow KA, Sprenger KB (1995). Successful plasmapheresis in severe diltiazem poisoning. Dtsch Med Wochenschr 120:1023-1024.

110. Gagliardi AR, Brouwers MC, Palda VA, Lemieux-Charles L, Grimshaw JM (2011). How can we improve guideline use? A conceptual framework of implementability. Implementation Science 6:26.

111. Gagnier JJ, Kienle G, Altman DG, Moher D, Sox H, Riley D (2014). The CARE guidelines: consensus-based clinical case report guideline development. J Clin Epi 67:46-51. 196

112. Gay R, Algeo S, Lee R, Olajos M, Morkin E, Goldman S (1986). Treatment of verapamil toxicity in intact dogs. J Clin Invest 77:1805-11.

113. Geib AJ, Manini A, Liebelt E (2009). Case series of intravenous lipid emulsion rescue for drug cardiotoxicity. Clin Toxicol 47:704.

114. Gerloni R, Copetti R (2004). Easily reversible hypoxemia and hypotension induced by nimodipine. Eur J Emerg Med 11:295-297.

115. Ghannoum M, Nolin TD, Goldfrank DS, Roberts DM, Mactier R, Mowry JB, Dargan PI, MacLauren R, Hoegberg LC, Laliberté M, Calello D, Kielstein JT, Anseeuw K, Winchester JF, Burdmann EA, Bunchman TE, Juurlink DN, Lavergne V, Mégarbane B, Gosselin S, Liu KD, Hoffman RS (2012). Extracorporeal treatment for thallium poisoning: recommendations from EXTRIP workgroup. Clin J Am Soc Nephrol 7:1682-1690.

116. Ghannoum M, Yates C, Galvao TF, Sowinski KM, Vo TH, Coogan A, Gosselin S, Lavergne V, Nolin TD, Hoffman RS; on behalf of the EXTRIP workgroup (2014). Extracorporeal treatment for carbamazepine poisoning: Systematic review and recommendations from the EXTRIP workgroup. Clin Toxicol 30:1-12.

117. Gonzales J, Meaney C, Sareh H, Hayes B (2012). Intravenous lipid emulsion for amlodipine overdose. Crit Care Med 1:309-310.

118. Goplani K, Gumber M, Kute V, Shah P, Trivedi H, Vanikar A (2011). Successful treatment of refractory hypotension, noncardiogenic pulmonary edema and acute kidney injury after an overdose of amlodipine. Indian J Crit Care Med 15.3:182.

119. Gosselin S, Juurlink DN, Kielstein JT, Ghannoum M, Lavergne V, Nolin TD, Hoffman RS; EXTRIP workgroup (2014). Extracorporeal treatment for acetaminophen poisoning: recommendations from the EXTRIP workgroup. Clin Toxicol 52:856-67.

120. Government of Hong Kong, Virtual AIDS office. Public consultation on draft of recommended guidelines for HIV infection and health care workers, www.info.gov.hk/aids/english/press/2013/130617.htm, consulted in January 2014.

121. Graham ID, Calder LA, Hébert PC, Carter AO, Tetroe JM (2000). A comparison of clinical practice guideline appraisal instruments. Int J Technol Assess Health Care 16:1024-1038.

122. Graham ID, Logan J, Harrison MB, Straus SE, Tetroe J, Caswell W, Robinson N (2006). Lost in knowledge: time for a map? J Contin Educ Health Prof 26:13-24.

197

123. Graham R, Mancher M, Wolman DM, Greenfield S, Steinberg E (2011). Clinical practice guidelines we can trust. The National Academies 391pp.

124. Graham ID, Grimshaw J, Hanna SE, Littlejohns P, Makarski J, Zitzelsberger L (2010). AGREE II: advancing guideline development, reporting and evaluation in health care. CMAJ 182:E839-E842.

125. Graudins A., Wong KKL (2010). Comparative hemodynamic effects of levosimendan alone and in conjunction with 4-aminopyridine or calcium chloride in a rodent model of severe verapamil poisoning. J Med Toxicol 6:85-93.

126. Graudins A, Najafi J, Rur-SC MP (2008). Treatment of experimental verapamil poisoning with levosimendan utilizing a rodent model of drug toxicity. Clin Toxicol 46:50-6.

127. Greene SL, Gawarammana I, Wood DM, Jones AL, Dargan PI (2007). Relative Safety of Hyperinsulinaemia/Euglycaemia Therapy in the Management of Calcium Channel Blocker Overdose: A Prospective Observational Study. Intensive Care Med 33: 2019-2024.

128. Grimshaw J, Eccles M, Thomas R, MacLennan G, Ramsay C, Fraser C, Vale L (2006). Toward Evidence-Based Quality Improvement, Evidence (and its Limitations) of the Effectiveness of Guidelines Dissemination and Implementation Strategies. J Gen Intern Med 21:S14-20.

129. Grimshaw JM, Schünemann HJ, Burgers J, Cruz AA, Heffner J, Metersky M, Cook D (2012). Disseminating and implementing guidelines: article 13 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:298-303.

130. Gris P, Coffernils M, Alame T, Kimbimbi TP (1989). Intoxication volontaire au vérapamil, à propos d'un cas, revue de la littérature. Sem Hôp Paris 65:435- 438.

131. Grol R, Dalhuijsen J, Thomas S, Veld C, Rutten G, Mokkink H (1998). Attributes of clinical guidelines that influence use of guidelines in general practice: observational study. BMJ 317:858-861.

132. Groszek B, Szpak D, Klys M, Scislowski M (2003). Acute poisoning with nifedipine and acebutol, two cases. Przegl Lek 60:262-4.

133. Gutschmidt HJ (1995). Successful plasmapheresis in severe diltiazem poisoning. Dtsch Med Wochenschr 120:81-82.

134. Guyatt G, Akl EA, Hirsh J, Kearon C, Crowther M, Gutterman D, Lewis SZ, Nathanson I, Jaeschke R, Schünemann H (2010). The vexing problem of guidelines and conflict of interest: a potential solution. Ann Intern Med 152:738- 41.

198

135. Guyatt G, Akl EA, Oxman A, Wilson K, Puhan MA, WIlt T, Gutterman D, Woodhead M, Antman EM, Schünemann HJ (2012). Synthesis, grading, and presentation of evidence in guidelines: article 7 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:256-261.

136. Guyatt G, Montori V, Schunemann H, Alonso-Coello P, Dahm P, Campana JPB, Brozek J, Nasser M, Meerpohl J, Rind D, Jaeschke R, Fack- Ytter Y, Norris S (2014). Letter reply to GRADE guidelines articles 14 and 15. J Clin Epi 67:236-240.

137. Guyatt G, Oxman AD, Akl EA, Kunz R, Vist G, Brozek J, Norrie S, Falck- Ytter Y, Glasziou P, deBeer H, Jaeschke R, Rind D, Meerpohl J, Dahm P, Schünemann HJ (2011). GRADE guidelines: 1. Introduction - GRADE evidence profiles and summary of finding tables. J Clin Epi 64:383-94.

138. Guyatt G, Oxman AD, VIst G, Kunz R, Brozek J, Alonso-Coello P, Montori V, Akl EA, Djulbegobiv B, Falck-Ytter Y, Norris SL, Williams JW, Atkins D, Meerpohl J, Schünemann HJ (2011). GRADE guidelines: 4. Rating the quality of evidence - study limitations (risk of bias). J Clin Epi 64:407-15.

139. Hadjipavlou G, Harfeez A, Messer B, Hughes T (2011). Management of lercarnidipine overdose with hyperinsulinaemic euglycaemia therapy: case report. Scand J Trauma Resusc Emerg Med 19.

140. Hagège A, Masquet C, Beaufils P, Slama R (1990). Intoxication massive par le vérapamil à libération prolongée. Arch Mal Coeur 83:1745-7.

141. Harley C, Kannan S, Morrow B (2001). Combined overdose with bisoprolol, nifedipine and diltiazem. Care of the Critically Ill 17.

142. Harris RP, Helfand M, Woolf SH, Lohr KN, Mulrow CD, Teutsch SM, Atkins D (2001). Current methods of the US preventive services task force: a review of the process. Am J Prev Med 20(3S):21-35.

143. Harris NS. Case 24 (2006). A 40-year-old woman with hypotension after an overdose of amlodipine. NEJM 355:602-611.

144. Hartling L, Milne A, Hamm MP, Vandermeer B, Ansari M, Tsertsvadze E, Dryden DM (2013). Testing the Newcastle Ottawa Scale showed low reliability between individual reviewers. J Clin Epidemiol 66:982.93.

145. Hasin T, Leibowitz D, Antopolsky M, Chajek-Shaul T (2006). The use of low-dose insulin in cardiogenic shock due to combined overdose of verapamil, enalapril and metoprolol. Cardiology 106:233-236.

146. Hasson R, Mulcahy V, Tahir H. Amlodipine poisoning complicated with acute non-cardiogenic pulmonary oedema, BMJ Case reports 2011; 199

10.1136/bcr.07.2011.4467.

147. Henderson A, Stevenson N, Hackett LP, Pond SM (1992). Diltiazem overdose in an elderly patient: efficacy of adrenaline. Anaesth Intensive Care 20:507-510.

148. Henry M, Kay MM, Viccellio P (1985). Cardiogenic shock associated with calcium-channel and beta blockers: reversal with intravenous calcium chloride. Am J Em Med 3:334-336.

149. Higgins JPR, Altman DG, Gotzsche PC, Jüni P, Moher D, Oxman AD, Savovic J, Schulz KF, Weeks L, Sterne JAC (2011). The Cochrane Collaboration's tool for assessing risk of bias in randomised trials. BMJ 343:d5928.

150. Hill SR, Olson LG, Falck-Ytter Y, Cruz AA, Atkins D, Baumann M, Jaeschke R, Woitalla T, Schünemann HJ (2012). Incorporating considerations of cost-effectiveness, affordability, and resource implications in guideline development: article 6 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:251- 5.

151. Hobbs DA, Warne MSJ, Markich SJ (2005). Evaluation of criteria used to assess the quality of aquatic toxicity data. Integr Environ Assess Manag 1:174- 180.

152. Holger JS, Stellpflug SJ, Cole JB, Harris CR, Engebretsen KM (2011). High-dose insulin: a consecutive case series in toxin-induced cardioganic shock. Clin Toxicol 49:653-8.

153. Holzer M, Sterz F, Schoerkhuber W, Behringer W, Domanovits H, Weinmar D, Weinstabl C, Stimpfl T (1999). Successful resuscitation of a verapamil-intoxicated patient with percutaneous cardiopulmonary bypass. Crit Care Med 27:2818-2823.

154. Hooijmans CR, Leenaars M, Riskes-Hoitinga M (2010). A gold standard publication checklist to improve the quality of animal studies, to fully integrate the Three Rs, and to make systematic reviews more feasible. Altern Lab Anim 38:167-182.

155. Howarth DM, Dawson AH, Smith AJ, Buckley N, Whyte IM (1994). Calcium channel blocking drug overdose: an Australian series. Hum Exp Toxicol 13:161-6.

156. Horowitz BZ, Rhee KJ (1989). Massive verapamil ingestion: a report of two cases and a review of the literature. Am J Emerg Med 7:624-31.

157. Immonen P, Linkola A, Waris E (1981). Three cases of severe 200

verapamil poisoning. In J Cardiol 1:101-5.

158. Isbister GK (2002). Delayed asystolic cardiac arrest after diltiazem overdose; resuscitation with high dose intravenous calcium. Emerg Med J 19: 335-357.

159. Institute of Medicine and National Research Council of National Academies (2005). Dietary supplements, a framework for evaluating safety. The National Academies. Washington DC, 506 pp.

160. International Committee of Medical Journal Editors (2014), Authorship responsibilities and conflicts of interest. http://www.icmje.org/roles_b.html, consulted Jan 31st 2014.

161. Jacobs AK, Kushner FG, Ettinger SM, Guyton RA, Anderson JL, Ohman EM, Albert NM, Antman EM, Arnett DK, Bertolet M, Bhatt DL, Brindis RG, Creader MA, DeMets DL, Dickersin K, Fonarow GC, Gibbons RJ, Halperin JL, Hochman JS, Koster MA, Normand SL, Ortiz E, Peterson ED, Roach WH, Sacco RL, Smith SC, Stevenson WG, Tomaselli GF, Yancy CW, Zoghbi WA, Harold JG, He Y, Mangu PB, Qaseem A, Sayre MR, Somerfield MR (2013). ACCF/AHA clinical practice guideline methodology summit report: a report of the American College of Cardiology Foundation/American Heart Association Task Force on Practice Guidelines. Circulation 127:268e310.

162. Jaeger A, Haubenstock H, Gossinge K, Pirich A, Donner U, Lruby K (1984). Acute intoxication with calcium channel blocker verapamil, a study of 11 cases. Intensive care 9:74-77.

163. Janion M, Stepien A, Sielski J, Gutkowski W (2010). Is the intra-aortic balloon pump a method of brain protection during cardiogenic shock after drug intoxication? J Emerg Med 38:162-167.

164. Jang DH, Nelson LS, Hoffman RS (2011). Methylene blue in the treatment of refractory shock from amlodipine overdose. Ann Emerg Med 58:565-7.

165. Januszewska K, Kolcz J, Malec E, Mroczek T, Pietrzyk J, Procelewska M (2007). Extracorporeal life support in severe propranolol and verapamil intoxication. J of Intensive Care Med 22:381.

166. Jarry J, Sassoust G (2008). A rare case of nifedipine bezoar. Presse Med 37:428-30.

167. Jaeger A, Sauder P, Bianchetti G, Kopferschmitt J, Dahlet M, Tritsch L, Flesch F (1990). Acute diltiazem poisoning: kinetic and hemodynamic study. J Toxicol Clin Exp 10:243-8.

168. Jones NE, Suurdt J, Ouelette-Kuntz H, Heyland DK (2007). Implementation of the Canadian Clinical Practice Guidelines for Nutrition 201

Support: a multiple case study of barriers and enablers. Nutr Clin Pract 22:449- 57.

169. Juurlink DN, Szalai JP, McGuigan MA (2002). Discrepant advice from poison centres and their medical directors. Can J Clin Pharmacol 9:101-5.

170. Kalam Y, Graudins A (2012). The effects of fructose-1,6-diphosphate on haemodynamic parameters and survival in a rodent model of propranolol and verapamil poisoning. Clin Toxicol 50:546-554.

171. Kalman S, Berg S, Lisander B (1998). Combined overdose with verapamil and atenolol: treatment with high doses of adrenergic agonists. Acta Anaesthesiol Scand 42:379-382.

172. Kanagarajan K, Marraffa JM, Bouchard NC, Krishman P, Hoffman RS, Stork CM (2007). The use of vasopressin in the setting of recalcitrant hypotension due to calcium channel blocker overdose. Clin Toxicol 45:56-9.

173. Kandil A, Karras S, Hamed MR (1984). Effect of calcium on acute toxicity of verapamil. J Drug Res 15:121-125.

174. Karti SS, Ulusoy H, Yandi M, Gunduz A, Kosucu M, Erol K, Ratip S (2002). Non-cardiogenic pulmonary oedema in the course of verapamil intoxication. Emerg Med J 19:458-9.

175. Karvellas CJ, Safinia N, Auzinger G, Heaton N, Muiesan P, O’Grady J, Wendon J, Bernal W (2010). Medical and psychiatric outcomes for patients transplanted for acetaminophen-induced acute liver failure: a case-control study. Liver Int 30:826-33.

176. Katzung BG (2004). Calcium Channel Antagonists In: Basic and Clinical Pharmacology, LANGE 9th edition, USA, p.155-158.

177. Keenan SP, Sinuff T, Burns KEA, Muscedere J, Kutsogianninis J, Mehta S, Cook DJ, Ayas N, Adhikari NKJ, Hand L, Scales DC, Pagnotta R, Lazosky L, Rocker G, Dial S, Laupland K, Sanders K, Dodek P (2011). CMAJ 183:E195- 214.

178. Kelson M, Akl EA, Bastian H, Cluzeau F, Curtis JR, Guyatt G, Montori VM, Oliver S, Schünemann HJ (2012). Integrating values and consumer involvement in guidelines with the patient at the center: article 8 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:262-8.

179. Kerns II W (2007). Blocker and calcium channel antagonist toxicity. Emerg Med Clin N Am 25:309-31.

202

180. Kilkenny C, Browne WJ, Cuthill IC, Emerson M, Altman DG (2010). Improving Bioscience Research Reporting: The ARRIVE Guidelines for Reporting Animal Research. PLoS Biol 8(6): e1000412. doi:10.1371/journal.pbio.1000412

181. Kim HK, Hoffman RS, Nelson LS, Jan DH (2012). Methylene blue for refractory hypotension in cardiovascular drug overdose. EAPCCT abstract 2012, p.365.

182. King A, Menke NB, Abesamis MG (2013). Volume overload from institution of high dose insulin therapy for calcium channel blocker poisoning. Clin Toxicol 51:575-724.

183. Kline JA, Tomaszewski C, Schroeder JD, Raymon RM (1993). Insulin is a superior antidote for cardiovascular toxicity induced by verapamil in the anesthethetized canine. J Pharmacol Exp Ther 267:744-750.

184. Kline JA, Leonova E, Williams TC, Schroeder JD, Watts JA (1996). Myocardial metabolism during graded intraportal verapamil infusion in awake dogs. J Cardiovasc Pharmacol 27:719-726.

185. Kline JA, Raymon RM, Leonova ED, Williams TC, Watts JA (1997). Insulin improves heart function and metabolism during non-ischemic cardiogenic shock in awake canines. Cardiovasc Res 34:289-298.

186. Knaapen L, Schaefer C, Boivin A, Chalmers S, Davino-Ramaya CM, Gracia J, Rosenfeld R, Rosvik AH, Sakala C, Tyrrell H, Wang M, Weijden Tvd (2012). G-I-N PUBLIC toolkit: Patient and public involvement in guidelines. Guidelines international network, 88pp.

187. Knoch G, Schlepper M, Witzleb E (1993). Isoptin studies on normal subjects and patients with coronary disease. Med Klin 58:1485-9.

188. Kolcz J, Pietrzyk J, Januszewska K, Procelewska M, Mroczek T, Malec E (2007). Extracorporeal life support in severe propranolol and verapamil intoxication. J Intensive Care Med 22:381.

189. Konca C, Yildizdas RD, Sari MY, Yükselmis U, Horoz OO, Yilmaz HL (2013). Evaluation of children poisoned with calcium channel blocker or beta blocker drugs. Turk Arch Ped 48:138-144.

190. Korstanje C, Jonkman FAM, Kemenade JEV, Van Zwieten PA (1987). Bay K 8644, a calcium entry promoter, as an antidote in verapamil intoxication in rabbits. Arch Int Pharmacodyn 287:109-19.

191. Koury SI, Stone CK, Thomas SH (1996). Amrinone as an antidote in experimental verapamil overdose. Acad Emerg Med 3:762-7.

192. Krauth D, Wooruff TJ, Bero L (2013). Instruments for Assessing Risk of 203

Bias and Other Methodological Criteria of Published Animal Studies: A Systematic Review. Environmental Health Perspectives 121:985-992.

193. Kruger P (1984). Suicide attempt with verapamil. Therapiewoche 34:2483-2491.

194. Kuhlmann U, Schoenemann H, Muller TF, Keuchel M, Lange H (1999). Plasmapheresis in fatal overdose with verapamil. Intensive Care Med 25:1473.

195. Kunz R, Fretheim A, Cluzeau F, Wilt TJ, Qaseem A, Lelgemann M, Kelson M, Guyatt G, Schünemann HJ (2012). Guideline group composition and group processes: article 3 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:229-33.

196. Kurola J, Leppikangas H, Magga J, Lindgren L, Kiviniemi V, Rutanen J, Ruokonen E (2010). Effect of levosimendan in experimental verapamil-induced myocardial depression. Scand J of Trauma, Resusc Emerg Med 18:12.

197. Laes J, Orozco BS, Cole JB (2013). Use of percutaneous left ventricular assist device (Impella) in vasodilatory poison-indued shock. Clin Toxicol 51:673.

198. Laine K, Kivistö KT, Neuvonen PJ (1997). Effect of delayed administration of activated charcoal on the absorption of verapamil. J Toxicol Clin Toxicol 35:263-8.

199. Lam YM, Tse HF, Lau CP (2001). Continuous calcium chloride infusion for massive nifedipine overdose. Chest 119:1280-1282.

200. Lambert H, Weber M, Renaud D (1990). Intoxications aiguës par diltiazem. Bilan des centres anti-poisons français. J Toxicol Clin Exp 10:229-42.

201. Lamontagne F, Briel M, Duffett M, Fox-Robichaud A, Cook DJ, Guyatt G, Lesur O, Meade MO (2010). Systematic review of reviews including animal studies addressing therapeutic interventions for sepsis. Crit Care Med 38:2401- 08.

202. Lapatto-Reiniluoto O, Kivistö KT, Neuvoven PJ (2000). Efficacy of activated charcoal versus gastric lavage half an hour after ingestion of moclobemide, temazepam and verapamil. Eur J Clin Pharmacol 56:285-8.

203. Lapatto-Reiniluoto O, Kivistö KT, Neuvoven PJ (2001). Activated charcoal alone and followed by whole-bowel irrigation in preventing the absorption of sustained-release drugs. Clin Pharmacol Ther 70:255-60.

204. Larkin GL, Marco CA, Abbott JT (2001). Emergency determination of decision-making capacity: balancing autonomy and beneficience in the emergency department. Acad Emerg Med 8:282-4. 204

205. Lavergne V, Nolin TD, Hoffman RS, Roberts D, Gosselin S, Goldfarb DS, Kielstein JT, Mactier R, Maclaren R, Mowry JB, Buchman TE, Juurlink D, Mégarbane B, Answeeuw K, Winchester JF, Dargan PI, Liu KD, Hoegberg LC, Li Y, Calello DP, Burdmann EA, Yates C, Laliberté M, Becker BS, Mello-Da- Silva CA, Lavonas E, Ghannoum M (2012). The EXTRIP (EXtracorporeal TReatments In Poisoning workgroup: Guideline methodology. Clin Toxicol 50:403-413.

206. Lee DW, Cohan B (1995). Refractory cardiogenic shock and complete heart block after verapamil SR and metoprolol treatment. Angiology 46:517-519.

207. Lee DC, Greene T, Dougherty T, Pearigen P (2000). Fatal nifedipine ingestions in children. J Emerg Med 19:359-361.

208. Lenzer J, Hoffman J, Furberg C, Ioannidis J (2013). Ensuring the integrity of clinical practice guidelines: a tool for protecting patients. BMJ 347:f5535.

209. Levine M, Curry SC, Padilla-Jones A, Ruha AM (2013). Critical care management of verapamil and diltiazem overdose with a focus on vasopressors: A 25-year experience in a single centre. Ann Emerg Med 63:252-258.

210. Levine M, Boyer EW, Pozner CN, Geib AJ, Thomsen T, Mick N (2007). Assessment of hyperglycemia after calcium channel blocker overdoses involving diltiazem or verapamil. Crit Care Med 35:2071-5.

211. Lheureux P, Vranckx M, Leduc D, Askenasi R (1990). Effect of PK11195 on cardiovascular toxicity due to verapamil: an experimental study in the dog. J Toxicol Clin Exp 10:449-60.

212. Lheureux PER, Zahir S, Gris M, Derrey AS (2006). Penaloza. Bench-to- bedside review: Hyperinsulinaemia/euglycemia therapy in the management of overdose of calcium-channel blockers. Crit Care 10:212.

213. Liang CW, Diamond SJ, Hagg DS (2011). Lipid rescue of massive verapamil overdose: a case report. J Med Case Rep 5:399.

214. Link A, Hammer B, Weigerber K, Bohm M (2002). Verapamil intoxication: therapy with norepinephrine and the phosphodiesterase inhibitor enoximone. Dtsch Med Wochenschr 127:2006-2008.

215. Lipman J, Jardine I, Roos C, Dreosti L (1982). Intravenous Calcium chloride as an antidote to verapamil-induced hypotension. Intensive Care Med 8:55-57.

216. Li Saw F, Lip GYH (1996). Case report: fatal verapamil overdosage despite intensive therapy and use of high dose intravenous calcium. J Human Hypertension 10:495-496.

205

217. Lo JC, Ubaldo C, Cantrell FL (2012). A retrospective review of whole bowel irrigation in pediatric patients. Clin Toxicol 50:414-417.

218. Logan J, Harrison MB, Graham ID, Dunn K, Bissonnette J (1999). Evidence-Based Pressure-Ulcer Practice: The Ottawa Model of Research Use, Can J Nurs Res 31:37-52.

219. Lone NI, Seretny M, Wild SH, et al (2013). Surviving intensive care: A systematic review of healthcare resource use after hospital discharge, Critical care medicine 41:1832-1843

220. Love JN, Howell JM (1997). Glucagon therapy in the treatment of symptomatic bradycardia. Ann Emerg Med 29:181-183.

221. Love JN, Sachdeva DK, Curtis LA, Howell JM, Bessman ES (1998). A potential role for glucagon in the treatment of drug-induced sympatomatic bradycardia. Chest 114:323-326.

222. Luomanmäki K, Tiula E, Kivistö KT, Neuvonen PJ (1997). Pharmacokinetics of diltiazem in massive overdose. Ther Drug Monit 19:240- 242.

223. Luscher TF, Noll G, Sturmer T, Huser B, Wenk M (1994). Calcium gluconate in severe verapamil intoxication. NEJM 330:718-720.

224. Lynch MJ, Katz KD, Callaway CW, Logue ES (2010). Survival of verapamil-poisoned rats treated with triiodothyronine, J Med Toxicol 6:94-9.

225. Maclaren G, Butt W, Cameron P, Preovolos A, McEgan R, Marasco S (2005). Treatment of polypharmacy overdose with multimodality extracorporeal life support. Anaesth Intensive Care 33: 120-123.

226. Mactier R, Laliberté M, Mardini J, Ghannoum M, Lavergne V, Gosselin S, Hoffman RS, Nolin TD; EXTRIP workgroup (2014). Extracorporeal treatment for barbiturate poisoning: recommendations from the EXTRIP workgroup. Am J Kidney Dis 64:347-58.

227. Madera F, Wenger R (1977). About intoxication with Isoptin S (verapamil and pentobarbital). Intensivmedizin 14:373-377.

228. Magdalan J (2003). New treatment methods in verapamil poisoning: experimental studies. Pol J Pharmacol 55:425-32.

229. Mahr NC, Valdes A, Lamas G (1997). Use of glucagon for acute intravenous diltiazem toxicity. The Am J of Cardiology 79:1570-1571.

230. Maldonado I, Megan J, Amber C, Chadwick C (2012). Beta- 206

blockers and calcium-channel blockers overdose: time to abandon a silver-bullet approach. Crit Care Med 40:1182.

231. Marques I, Gomes E, Oliveira J (2003). Treatment of calcium channel blocker intoxication with insulin infusion: case report and literature review. Resuscitation 57:211-213.

232. Marraffa JM, Stork CM, Medicis JJ, Hodgman MH, Cantor R (2004). Massive amlodipine overdose successfully treated using high- dose vasopressin. J Toxicol Clin Toxicol 42:732-733.

233. Masson R, Colas V, Parienti JJ, Lehoux P, Massetti M, Charbonneau P, Saulnier F, Daubin C (2012). A comparison of survival with and without extracorporeal life support treatment for severe poisoning due to drug intoxication. Resuscitation 83:1413-1417.

234. Marjanovic VS, Cokanovic V (2010). A case of severe verapamil intoxication. Serbian Journal of Experimental and Clinical Research 11:33-36.

235. McGlinchey PG, McNeill AJ (1998). Drug overdoses requiring temporary cardiac pacing; a study of six cases treated at Altnagelvin Hospital, Londonderry. Ulster Med J 67:13-8.

236. McLaughlin D, McKenna H, Leslie J, Moore K, Robinson J (2006). Ilicit drug users in Northern Ireland: perceptions and experiences of health and social care professionnals 13:682-6.

237. Meaney C, Sareh H, Hayes B, Gonzales J (2013). Intravenous lipid emulsion in the management of amlodipine overdose. Hosp Pharm 48:848-854.

238. Medical Advisory Secretariat of the Ministry of Health and Long-Term Care of Ontario (2010). Extracorporeal Lung Support Technologies - Bridge to recovery and bridge to lung transplantation in adult patients, an evidence-based analysis. Ontario Health Technology Assessment Series; 10:37-38.

239. Meltzer D (1997). Accounting for future costs in medical cost- effectiveness analysis. J Health Econ 16:33-64.

240. Mégarbane B, Leprince P, Deye N, Résière D, Guerrier F, Rettab S, Théodore J, Karyo S, Gandjbakhch I, Baud FJ (2007). Emergency feasibility in medical intensive care unit of extracorporeal life support for refractory cardiac arrest. Intensive Care Med 33:758-764.

241. Mégarbane B, Karyo S, Abidi K, Delhotal-Landes B, Aout M, Sauder P, Baud FJ (2011). Predictors of mortality in verapamil overdose: usefulness of verapamil concentrations, Basic Clin Pharmacol Toxicol 108:385-9.

242. Merriam-Webster, Dictionary and thesaurus, www.merriam-webster.com, consulted in January 2014. 207

243. Ministry of Health and Long Term Care, Schedule of Benefits for Physician Services under the Health Insurance Act (October 2013 version), 888 pp.

244. Ministry of Health and Long Term Care of Ontario (MOHLTC), Ambulance Services Billing, http://www.health.gov.on.ca/en/public/publications/ohip/amb.aspx, consulted November 2013.

245. Moga C, Guo B, Schopflocher D, Harstall D (2012). Development of a quality appraisal tool for case series studies using a modified Delphi technique. Institute of Health Economics, 79 pp.

246. Moher D, Liberati A, Tetzlaff J, Altman DG (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. BMJ 339:332-6.

247. Montiel V, Gougnard T, Hantson P (2011). Diltiazem poisoning treated with hyperinsulinemic euglycemia therapy and intravenous lipid emulsion. European J Emerg Med 18:121-123.

248. Montori VM, Brito JP, Murad MH (2013). The optimal practice of evidence-based medicine. JAMA 310:2503-2504.

249. Mowry JB, Spyker DA, Cantilena LR, Bailey JE, Ford M (2013). 2012 Annual Report of the American Association of Poison Control Centers' National Poison Data System (NPDS): 30th Annual Report. Clin Toxicol 51:949-1229.

250. Moroni F, Manaioni PF, Dolara A, Ciaccheri M (1980). Calcium gluconate and hypertonic sodium chloride in a case of massive verapmil poisoning. Clin Toxicol 17:395-400.

251. Motokawa K, Takahashi H, Ohashi N, Sato S, Naito H (1993). Poisoning caused by the combined ingestion of nifedipine and metoprolol. J Toxicol Clin Toxicol 31: 631-637.

252. Mottram AR, Bryant SM, Aks SE (2012). Dose dependent response to cyclodextrin infusion in a rat model of verapamil toxicity. West J Emerg Med 13:63-67.

253. Mottram AR, Bryant SM, Aks SE (2011). Effect of cyclodextrin infusion in a rat model of verapamil toxicity. Am J of Ther 18:371-374.

254. Munshi P, Puri N, Haroz R, Gerber D, Rajaram S-S (2010). A fatal case of prolonged hypotension following amlodipine and lisinopril overdose. Critical Care Medicine Conference: 40th Critical Care Congress of the Society of Critical Care Medicine San Diego, CA United States Conference Start: 20110115 208

Conference End: 20110119 Conference Publication: (var pagings) 38:A260.

255. Musselman M, Farber M, Smolinkse S, Aaron C (2011). A Comparison of High Dose Insulin Therapy and Conventional Inotropic Therapy in Calcium Channel Blocker and/or Beta-Blocker Toxic Ingestions. Ann Emerg Med 58:S325.

256. Mycyk MB, Bryant SM (2007). Is simple glucose assessment prognostic in calcium channel blocker overdose? Crit Care Med 35:2216-7.

257. Myles A, Mugford GJ, Zhao J, Krahn M, Wang PP (2011). Physicians' attitudes and practice toward treating injection drug users infected with hepatitis C virus: results from a national specialist survey in Canada. Can J Gastroenterol 25:135-139.

258. Nair R, Aggarwal R, Khanna D (2011). Methods of formal consensus in classification/diagnostic criteria and guideline development. Semin Arthritis Rheum 41:95-105.

259. Nanda U, Ashish A, WHy HJ (2005). Modified release verapamil induced cardiovenic shock. Emerg Med J 22:832-833.

260. National Collaborating Centre for Methods and Tools (2008). Quality Assessment Tool for Quantitative Studies. Hamilton, ON: McMaster University. (Updated 13 April, 2010). Retrieved from http://www.nccmt.ca/registry/view/eng/14.html.

261. National Research Council Institute for Laboratory Animal Research (2011). Guidance for the Description of Animal Research in Scientific Publications. National Academies Press, 246pp.

262. Nickson CP, Little M (2009). Early use of high-dose insulin euglycaemic therapy for verapamil toxicity. Med J Aust 191:350-352.

263. Nimbalkar SM, Patel DV (2013). Near fatal case of amlodipine poisoning in an infant. Indian J Pediatr 80:513-515.

264. Nelson LS, Lewin NA, Howland MA, Hoffman, RS, Goldfrank LR, Flomenbaum NE (2011). Goldfrank's Toxicologic Emergencies. McGraw Hill Medical, 1940pp.

265. Oakes JA, Piquette C, Barthold CL (2009). Successful use of intravenous lipid as adjunctive therapy in a severe calcium channel antagonist poisoning. Clin Toxicol 47:755-756.

266. Ozbilgin S, Ozbilgin M, Kucukoztas B, Kamaci G, Unek T, Yurthu BS, Güneli ME, Hanci V, Gunerli A (2013). Evaluation of the effectiveness of suggamadex for verapamil intoxication. Basic Clin Pharmacol Toxicol 113:280-285. 209

267. Oe H, Taniura T, Ohgitani N (1998). A case of severe verapamil overdose. Jpn Circ J 62:72-76.

268. Oliver TB, Awupor-Reoner C (1992). Isoprenaline infusion and right ventricular pacing in severe diltiazem poisoning. N Z Med J 105:483.

269. Olson KR (2012). Poisoning and drug overdose. Lange, McGraw Hill Medica, 815pp.

270. Olson KR, Erdman AR, Woolf AD, Scharman EL, Christianson G, Caravati M, Wax PM, Booze LL, Manoguerra AS, Keyes DC, Chyka PA, Troutman WG (2005). Calcium channel blocker ingestion: an evidence-based consensus guideline for out-of-hospital management. Clin Tox 43:797-822.

271. Orr K, Bailie R (2010). The use of intralipid in the management of a mixed overdose. JICS 11:268-269.

272. Ortiz-Nunoz L, Rodriguez-Ospina LF, Figueroa-Gonzalez M (2005). Hyperinsulinemic-euglycemic therapy for intoxication with calcium channel blockers. Bol Assoc Med PR 97:182-189.

273. Osthoff M, Bernsmeier C, Marsch SC, Hunziker PR (2010). Levosimendan as treatment option in severe verapamil intoxicaion: a case report and review of the literature. Case Rep Med Article ID 546904.

274. Ovens H, Allen R, Cohen E(1993). Who will practice emergency medicine? Can Fam Physician 39:1356-1365.

275. Paden MK, Conrad SA, Rycus PT, Thiagarajan RR (2013). Extracorporeal Life Support Organization Registry Report 2012. ASAIO Journal; 59:202-210.

276. Padilla Fortunatti CF (2014). Most important needs of family members of critical patients in light of the critical care family needs inventory. Invest Educ Enferm 32:306-16.

277. PADIS (Poison and Drug Information Service) (2013). Annual report 2012. Alberta Health Services, 10 pp.

278. Pap I, Kertesz J, Sagi I (1989). Attempted suicide with verapamil. Orv Hetil 130:735-736.

279. Papadopoulos J, O'Neil M (2000). Utilization of a glucagon infusion in the management of a massive nifedipine overdose. J Emerg Med 18:453-455.

280. Parikka H, Pohjola-Sintonen S, Rapola J, Sipila R, Luomanmaki K (1993). Calcium channel blocker poisoning. Duodecim 109:687-95.

210

281. Patel NP, Pugh ME, Goldberg S, Eiger G (2007). Hyperinsulinemic euglycemia therapy for verapamil poisoning: case report AJCC 16:518-520.

282. Peek GJ, Elbourne D, Mugford M, Tiruvoipati R, Wilson A, Allen E, Clemens F, Firmin R, Hardy P, Hibbert C, Jones N, Killer H, Thalanany M, Truesdale A (2010). Randomised controlled trial and parallel economic evaluation of conventional ventilatory support versus extracoporeal membrane oxygenatino for severe adult respiratory failure (CESAR). Health Technol Assess; 14:1-46.

283. Perez E., Chu J, Bania T, Medlej K (2011). L-carnitine increases survival in a murine model of severe verapamil toxicity. Acad Emerg Med. 118:1135-40.

284. Perez E, Bania TC, Medlej K, Chu J (2008). Determining the optimal dose of intravenous fat emulsion for the treatment of severe verapamil toxicity in a rodent model. Acad Emerg Med 15:1284-1289.

285. Perichon D, Turfus S, Graudins A (2013). Intravenous lipid emulsion does not improve hemodynamics or survival in a rodent model of oral verapamil poisoning. Clin Toxicol 51:252-378.

286. Perkins CM (1978). Serious verapamil poisoining: treatment with intravenous calcium gluconate. BMJ october:1127.

287. Persad E, Raman L, Thompson M, Sheeran P (2012). The use of extracorporeal life support in adolescent amlodipine overdose. Indian J Crit Care Med 16:204-206.

288. Pfaender M, Casetti PG, Azzolini M, Baldi ML, Vailli A (2008). Successful treatment of a massive atenolol and nifedipine overdose with CVVHDF. Minerva Anesthesiol 74:97-100.

289. Pfeffer AN, Williams SR (2014). Treatment of calcium channel blocker poisoning: Should we reprioritize our potpourri of treatment options? Answers to the September 2013 Journal Club Questions. Ann Emerg Med 63:259-65.

290. Pichon N, Dugard A, Clavel M, Amiel JB, François B, Vignon P (2012). Extracorporeal albumin dialysis in three cases of acute calcium channel blocker poisoning with life-threatening refractory cardiogenic shock. Ann Emerg Med 59: 540-544.

291. Plewa MC, Martin TG, Menegazzi JJ, Seaberg DC, Wolfson AB (1994). Hemodynamic effects of 3,4-diaminopyridine in a swine model of verapamil toxicity. Ann Emerg Med 23:499-507.

292. Plumb JAM, Stewart C, Eddleston M, De Beer T (2011). Prolonged refractory hypotension following combined amlodipine and losartan ingestion 211

responsive to metaraminol. Case Rep Med Article ID 283672.

293. Proano L, Chiang WK, Wand RY (1995). Calcium channel blocker overdose. Am J Emerg Med 13:444-450.

294. Pronovost PJ (2013). Enhancing Physicians' Use of Clinical Guidelines. JAMA 310:2501-02.

295. Qaseem A, Forland F, Macbeth F, Ollenschläger G, Pillips S, van der Wees P (2012). Guidelines International Network: toward international standards for clinical practice guidelines. Ann Intern Med 156:525-31.

296. Ramo MP, Grupp I, Pesola MK, Heikkila J, Luomanmaki K, Schroder T, Grupp G (1992). Cardiac glycosides in the treatment of experimental overdose with calcium-blocking agents. Res Exp Med 192:355-343.

297. Ramoska EA, Spiller HA, Winter M, Borys D (1993). A one-year evaluation of calcium channel blocker overdoses: Toxicity and treatment. Ann Emerg Med 22:196-200.

298. Rahman O, Craft M, Mainali R, Hartie J (2009). Massive diltiazem overdose treated with continunous veno-venous hemodiafiltration. Crit Care Med 37:A511.

299. Rizvi I, Ahmad A (2012). Life-threatening calcium channel blocker overdose and its management. BMJ Case reports 10.1136/bcr.01.2012.5643.

300. Roberts D, Pharm NH, Sitar DS, Tenenbein M (1991). Diltiazem overdose: pharmacokinetics of diltiazem and its metabolites and effect of multiple dose charcoal therapy. Clin Toxicol 29:45-52.

301. Roberts DM, Roberts JA, Boots RJ, Mason R, Lipman J (2008). Lessons learnt in the pharmacokinetic analysis of the effect of haemoperfusion for acute overdose with sustained-release diltiazem. Anaesthesia 63:714-718.

302. Roberts DJ, Leigh-Smith S, Faris PD, Ball CG, Robertson HL, Blackmore C, Dixon E, Kirkpatrick AW, Kortbeek JB, Stelfox HT (2014). Clinical manifestations of tension pneumothorax: protocol for a systematic review and meta-analysis. Systematic reviews 3:3.

303. Rogers EM (2002). Diffusion of preventive innovations. Addict Behav 27:989-993.

304. Rogers EM (2003). Diffusion of innovations (5th ed). New York, Free Press.

305. Rona R, Cortinovis B, Marcolin R, Patroniti N, Isgro S, Marelli C, Fumagalli R (2011). Extra-corporeal life support for near-fatal multi-drug intoxication: a case report. J Med Case Rep 5:231. 212

306. Roncon-Albuquerque R, Almeida V, Lopes M, Castro L, Pedrosa A, Paiva JA (2014). Cost analysis of miniaturized ECMO in H1N1-related ARDS managed by a single caregiver. Int Care Med, April (Epub ahead of print)

307. Roper TA, Sykes R, Gray C (1993). Fatal diltiazem overdose: report of four cases and review of the literature. Postgrad Med J 69:474- 6.

308. Rosansky SJ (1991). Verapamil toxicity treatment with hemoperfusion. Ann Int Med 114:340-341.

309. Rosser G, Dubrey W (2012). Massive calcium channel blocker overdose: intravenous insulin and glucose as a therapy. BMJ Case reports 10.1136/bcr.03.2012.6114.

310. Rygnestad T, Moen S, Wahba A, Lien S, Ingul CB, Schrader H, Knapstad SE (2005). Severe poisoning with sotalol and verapamil. Recovery after 4h of normothermic CPR followed by extra corporeal heart lung assist. Acta Anaesthesiol Scand 49:1378-1380.

311. Samtleben W (1995). Successful plasmapheresis in severe diltiazem poisoning. Dtsch Med Wochenschr 120:1023.

312. Sandroni C, Cavallara F, Addario C, Hamilton RJ (2005). Successful treatment with enoximone for severe poisoning with atenolol and verapamil: A case report. Ann Emerg Med 46:99.

313. Sauder P, Kopferschmitt J, Dahlet M, Tritsch L, Flesch F, Siard P, Mantz JM, Jaeger A (1990). Acute verapamil poisoning. 6 cases. Review of the literature. J Toxicol Clin Exp 10:261-270.

314. Schmidt W, Reissig M, Neuhaus KL (1995). Percutaneous cardiopulmonary bypass for cardiogenic shock caused by poisoning with meltidigoxin, nifedipine and indapamide. Dtsch Med Wochenschr 120:996-1002.

315. Schumacher RE, Roloff DW, Chapman R, Snedecor S, Bartlett RH (1993). Extracorporeal membrane oxygenation in term newborns. A prospective cost-benefit analysis. ASAIO J; 3:873-9.

316. Schünemann HJ, Oxman AD, Akl EA, Brozek JL, Montori VM, Heffner J, Hill S, Woodhead M, Campos-Outcalt D, Alderson P, Woitalla T, Puhan MA' Falck-Ytter Y, Bousquet J, Guyatt G (2012). Moving from evidence to developing recommendations in guidelines: article 11 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:282-92.

317. Schünemann HJ, Wiercioch W, Etxeandia I, Falavigna M, Santesso N, Mustafa R, Ventresca M, Brignardello-Petersen R, Laisaar KT, Kowalski S, 213

Baldeh T, Zhang Y, Raid U, Neumann I, Norris SL, Thornton J, Harbour R, Treweek S, Guyatt G, Alonso-Coello P, Reinap M, Brozek J, Oxman A, Akl EA (2013). Guidelines 2.0: systematic development of a comprehensive checklist for a successful guideline entreprise. CMAJ Dec 16 (Epup ahead of print).

318. Schünemann HJ, Woodhead M, Anzueto A, Buist AS, Macnee W, Rabe KF, Heffner J (2012). A guide to guidelines for professional societies and other developers of recommendations: introduction to integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:215-8.

319. Scocco P, Castriotta C, Toffol E, Preti A (2012). Stigma of Suicide Attempt (STOSA) scale and Stigma of Suicide and Suicide Survivor (STOSASS) scale: two new assessment tools. Psychiatry Res 30:872-8.

320. Sena E. van der Worp HB, Howells D, Macleod M (2007). How can we improve the pre-clinical development of drugs for stroke? Trends Neurosci 30:433-439.

321. Sencan A, Adanir T, Terzi G, Atay A, Karahan N (2011). Treatment of calcium channel blocker overdose with levosimendan. Crit Care 15:P95.

322. Siebenlist D (1990). Plasma separation in verapamil poisoning. Dtsch Med Wochenschr 115:797.

323. Sinuff T, Muscedere J, Adhikari NK, Stelfox HT, Dodek P, Heyland DK, Rubenfeld GD, Cook DJ, Pinto R, Manoharan V, Currie J, Cahill N, Friedrich JO, Amaral A, Piquette D, Scales DC, Dhanani S, Garland A (2013). Knowledge translation interventions for critically ill patients: a systematic review. Crit Care Med 41:2627-40.

324. Sinuff T, Muscedere J, Cook DJ, Dodek PM, Anderson W, Keenan SP, Wood G, Tan R, Haupt MT, Miletin M, Bouali R, Jiang X, Day AG, Overvelde J, Heyland DK (2013). Implementation of clinical practice guidelines for ventilator- associated pneumonia: a multicenter prospective study. Crit Care Med 41:15-23.

325. Shaneyfelt T, Mayo-Smith M, Rothwangl J (1999). Are guidelines following guidelines? The methodological quality of clinical practice guidelines in the peer-reviewed medical literature. JAMA 281:1900-1905.

326. Shea BJ, Grimshaw M, Wells GA, Boers M, Andersoon N, Hamel C, Porter AC, Tugwell P, Moher D, Bouter LM (2007). BMC Medical Research Methodology 7:10.

327. Shekelle P, Woolf S, Grimshaw J, Schünemann HJ, Eccles MP (2012). Developing clinical practice guidelines: reviewing, reporting, and publishing guidelines; updating guidelines; and the emerging issues of enhancing guideline implementability and accounting for comorbid conditions in guideline 214

development. Implement Sci 7:62.

328. Shiroiwa T, Sung Y-K, Fukuda T, Lang HC, Bae SC, Tsutani K (2010). International survey on willingness-to-pay (WTP) for one additional QALY gained: What is the threshold of cost effectiveness?. Health Econ 19:422-437.

329. Snover SW, Bocchino V (1986). Massive diltiazem overdose. Ann Emerg Med 15:1221-1224.

330. Socco P, Castriotta C, Toffol E, Preti A (2012). Stigma of suicide attempt (STOSA) scale and stigma of suicide and suicide survivor (STOSASS) scale: two new assessment tools. Psychiatry Research 200: 872-878.

331. Supradip G, Mrinal S (2008). Calcium channel blocker overdose: experience with amlodipine. Indian J of Crit Care Med 12:190-193.

332. Stachon K, Dabek J, Jakubowski D, Rychlik W (2011). Clinical implications in patients after ingestions of excessive doses of calcium channel blockers - our observations. Pol. Merk. Lek. 31:145-149.

333. Statistics Canada, http://www.statcan.gc.ca/tables-tableaux/sum- som/l01/cst01/health26-eng.htm, consulted October 2013.

334. Statistics Canada, Median total income: http://www.statcan.gc.ca/tables- tableaux/sum-som/l01/cst01/famil107c-eng.htm, consulted in April 2014.

335. Stellpflug SJ, Cole JB, Fritzlar SJ, Engebretsen KM, Holger JS (2010). Overdose of diltiazem, metoprolol and amiodarone treated successfully with intravenous fat emulsion and high dose insulin in a awake patient. Clin Toxicol 48:612.

336. Stellpflug SJ, Cole JB, Fritzlar SJ, Engebretsen KM, Holger JS (2011). Cardiotoxic overdose treated with intravenous fat emulsion and high-dose insulin in the setting of hypertrophic cardiomyopathy. J Med Toxicol 7:151-153.

337. Stone CK, May WA, Carroll R (1995). Treatment of verapamil overdose with glucagon in dogs. Ann Emerg Med 25:369-74.

338. Stone CK, Thomas SH, Koury SI, Low RB (1996). Glucagon and phenylephrine combination vs glucagon alone in experimental verapamil overdose. Acad Emerg Med 3:120-125.

339. St-Onge M (2010). Bulletin d'information toxicologique 26:3-9.

340. St-Onge M (2014). Chapter: Calcium channel blocker poisoning in the book "Hantson & Mégarbane. Poisoning with cardiotoxicants" to be published by Springer.

341. St-Onge M, Archambault P, LeSage N, Guimont C, Poitras J, Blais R 215

(2012). Adherence to calcium channel blocker poisoning treatment recommendations in two Canadian cities. Clin Toxicol 50:424-430.

342. St-Onge M, Ajmo I, Poirier D, Laliberté M (2013). L-carnitine for the treatment of a calcium channel blocker and metformin poisoning. J Med Toxicol 9:266-9.

343. St-Onge M, Dubé PA, Gosselin S, Guimont C, Godwin J, Archambault P, Chauny JM, Frenette AJ, Darveau M, LeSage N, Poitras J, Provencher J, Juurlink D, Blais R (2014). Treatment of calcium channel blocker poisoning: A systematic review. Clin Toxicol 52:926-944.

344. St-Onge M, Fan E, Mégarbane B, Hancock-Howard R, Coyte PC (2014). Venoarterial extracorporeal membrane oxygenation for patients in shock or cardiac arrest secondary to cardiotoxicant poisoning: A cost- effectiveness analysis. J Crit Care early online 16 October 2014.

345. Strubelt O (1984). Antidotal treatment of the acute cardiovascular toxicity of verapamil. Acta Pharmacol et Toxicol 55:231-237.

346. Strubelt O, Diederich KW (1986). Experimental investigations on the antidotal treatment of nifedipine overdosage. J Toxicol Clin Toxicol 24:135-49.

347. Strubelt O, Diederich KW (1990). Studies of antidote therapy for nisoldipine intoxication in experimental animals. Arzneimittelforschung 40:747- 51.

348. Sud S, Mittmann N, Cook DJ, Geerts W, Chan B, Gould MK, Guyatt G, Arabi Y, Fowler RA (2011). Screening and prevention of venous thromboembolism in critically ill patients: a decision analysis and economic evaluation. AJCCM; 184:1289-1298.

349. Sztajnkrycer MD, Bond GR, Johnson SB, Weaver AL (2004). Use of vasopressin in a canine model of severe verapamil poisoning: a preliminary descriptive study. Acad Emerg Med 11:1253-61.

350. Tanen DA, Ruha AM, Curry SC, Graeme KA, Reagan CG (2000). Hypertonic sodium bicarbonate is effective in the acute management of verapamil toxicity in a swine model. Ann Emerg Med 36:547-53.

351. Tauro DI, Roy S, Kailuraya M, Mehta RM (2009). Extracorporeal membrane oygenation (ECMO) for amlodipine overdose: A life saving intervention. Chest 136.

352. Taylor R, Giles J (2005). Cash interests taint drug advice. Nature 437:1070–1071.

353. Tebbutt S, Harvey M, Nicholson T, Cave G (2006). Intralipid prolongs 216

survival in a rat model of verapamil toxicity. Acad Emerg Med 13:134-9.

354. Teker MG, Ozdemir H, Saidoglu L, Erkalp K, Basaranoglu G (2010). Levosimendan as a rescue adjunct in amlodipine intoxication. MEJ Anesth 20:869-872.

355. Thiagarajan RR, Brogan TV, Scheurer MA, Laussen PC, Rycus PT, Bratton SL (2009). Extracorporeal Membrane Oxygenation to Support Cardiopulmonary Resuscitation in Adults. Annals of Thoracic Surgery 87:778-785.

356. Truitt CA, Brooks DE, Dommer P, LoVecchio F (2012). Outcomes of unintentional beta-blocker or calcium channel blocker overdose: a retrospective review of poison center data. J Med Toxicol 8:135-9.

357. Tsien RW, Barrett CF (2000). A Brief History of Calcium Channel Discovery. In: Madame Curie Bioscience Database [Internet]. Austin (TX): Landes Bioscience; 2000-. Available from: http://www.ncbi.nlm.nih.gov/books/NBK6465/

358. Tuncok Y, Apaydin S, Kalkan S, Ates M, Guven H (1996). The effects of amrinone and glucagon on verapamil-induced cardiovascular toxicity in anaesthetized rats. In J Exp Path 77:207-212.

359. Tuncok Y, Apaydin S, Gelal A, Ates M, Guven H (1998). The effects of 4- aminopyridine and Bay K 8644 on verapamil-induced cardio-vascular toxicity in anesthetized rats. J Toxicol Clin Toxicol 36:301-7.

360. Wang SM, Chou YC, Yeh MY, Chen CH, Tzeng WC (2013). Factors associated with quality of life after attempted suicide: a cross-sectional study. J Clin Nur 22:2150-2159.

361. Weinberg RL, Bouchard NC, Abrams DC, Bacchetta M, Dzierba AL, Burkart KM, Brodie D (2014). Venoarterial extracorporeal membrane oxygenation for the management of massive amlodipine overdose. Perfusion 29:53-56.

362. Wells CD, Luckritz TC, Rady MY, Zornik JM, Leighton JA, Patel BM (2006). Bezoar formation requiring endoscopic removal after intentional overdose of extended-release nifedipine. Pharmacotherapy 26:1802-5.

363. Wilson KC, Irwin RS, File TM, Schünemann HJ, Guyatt GH, Rabe KF (2012). Reporting and publishing guidelines: article 12 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:293-7.

364. Wilt TJ, Guyatt G, Kunz R, Macnee W, Puhan MA, Viegi G, Woodhead M, Akl EA, Schünemann HJ (2012). Deciding what type of evidence and outcomes to include in guidelines: article 5 in integrating and coordinating efforts in COPD 217

guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:243-50.

365. Woie L, Storstein L (1981). Successful treatment of suicidal verapamil poisoning with calcium gluconate. Eur Heart J 2:239-242.

366. Uhlig K, Leff B, Kent D, Dy S, Brunnhuber K, Burgers JS, Greenfield S, Guyatt G, High K, Leipzig R, Mulrow C, Schmader K, Schunemann H, Walter LC, Woodcock J, Boyd CM (2014). A framework for crafting clinical practice guidelines that are relevant to the care and management of people with multimorbidity. J Gen Inter Med Jan 18 (epub ahead of print).

367. Varpula T, Rapola J, Sallisalmi M, Kurola J (2009). Treatment of serious calcium channel blocker overdose with levosimendan, a calcium sensitizer. Anesth Analg 108:790-2.

368. Vats A, Pettignano R, Culler S, Wright J (1998). Cost of extracorporeal life support in pediatric patients with acute respiratory failure. Crit Care Med; 26:1587-92.

369. Vick JA, Kandil A, Herman EH, Balazs T (1983). Reversal of proprabolol and verapamil toxicity by calcium. Vet Hum Toxicol 25:8-10.

370. Von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) Statement: guidelines for reporting observational studies. Ann Intern Med. 147:573-577.

371. Weinberg RL, Bouchard NC, Abrams DC, Bacchetta M, Dzierba AL, Burkart KM, Brodie D (2014). Venoarterial extracorporeal membrane oxygenation for the management of massive amlodipine overdose. Perfusion 29:53-56.

372. Wesseling H, Houwertjes MC, DeLangen CDJ, Kingma JH (1983). Hemodynamic Effects of High Dosages of Verapamil and the Lack of Protection by 4-Aminopyridine in the Rabbit. Arch int. Pharmacodyn 266:106-112.

373. West PL, McKeownNJ, Hendrickson RG (2010). Iatrogenic lipid emulsion overdose in a case of amlodipine poisoning. Clin Toxicol 48:393-396.

374. Wills B, Liu M, Wahl M (2010). Third-degree AV block from extended- release diltiazem ingestion in a nine-month old. J Emerg Med 38:328-331.

375. Wilson BJ, Cruikshank JS, Wiebe KL, Dias VC, Yarema MC (2012). Intravenous lipid emulsion therapy for sustained release diltiazem poisoning: a case report. J Popul Ther Clin Pharmacol 19:e218-e222. 218

376. Wolf LR, Spadafora MP, Otten EJ (1993). Use of Amrinone and glucagon in a case of calcium channel blocker overdose. Ann Emerg Med 22:150-152.

377. Woolf SH, Grol R, Hutchinson A, Eccles M, Grimshaw J (1999). Potential benefits, limitations, and harms of clinical guidelines. BMJ 318:517-30.

378. Woolf S, Schünemann H, Eccles M, Grimshaw JM, Shekelle P (2012). Developing clinical practice guidelines: types of evidence and outcomes; values and economics, synthesis, grading, and presentation and deriving recommendations. Implement Sci 7:61.

379. Yates C, Galvao T, Sowinski KM, Mardini K, Botnaru T, Gosselin S, Hoffman RS, Nolin TD, Lavergne V (2014). Extracorporeal treatment for tricyclic antidepressant poisoning: recommendations from the EXTRIP workgroup. Semin Dial 27:381-9.

380. Yawn BP, Akl EA, Qaseem A, Black P, Campos-Outcalt D (2012). Identifying target audiences: Who are the guidelines for?: article 1 in integrating and coordinating efforts in COPD guideline development. An official ATS/ERS workshop report. Proc Am Thorac Soc 9:219-24.

381. Young AC, Velez Ll, Kleinschmidt KC. Intravenous fat emulsion therapy for intentional sustained-release verapamil overdose. Resuscitation 2009; 80:591-593.

382. Yuan TH, Kerns WP 2nd, Tomaszewski CA, Ford MD, Kline JA (1999). Insulin-glucose as adjunctive therapy for severe calcium channel antagonist poisoning. J Toxicol Clin Toxicol 37:463-74.

383. Zhou H, Liu Y, Li GQ, Wei LQ (2013). The use of a novel dosing regimen for calcium infusion in a patient of massive overdose of sustained-release nifedipine. Am J Med Sci. 2013;345:248-51.http://www.merriam- webster.com/dictionary/, consulted January 23rd 2014.

219

Appendices

Appendix 1

INTERVENTIONS USED FOR CALCIUM CHANNEL BLOCKER POISONING

1) Age: _____years 2) Sex: ! male ! female

3) Practice setting: ! primary center ! secondary center ! tertiary center

4) Training: ! Family physician ! CCFP (EM) ! FRCP(EM)

5) Number of years of practice: _____years

6) Did you get clinical experience, training or recently went to a conference related to the subject of calcium channel blocker poisoning over the past 2 years? ! yes ! no

7) CASE #1: A 35 year-old male with no prior history is brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release calcium channel blocker) approximately 4 hours ago. He has no sign or symptom and his vital signs on arrival are:

HR 65 (sinusal) BP 110/60 RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 15/15

7a) Do you monitor this patient? ! yes ! no

7b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

7c) How often do you measure blood glucose level? ! prn ! every _____ minutes ! every _____ hours

7d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

7e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

7f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

7g) Do you call the Poison Control Center concerning this patient? ! yes ! no

YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

220

8) CASE #2: A 35 year-old male with no prior history is brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release calcium channel blocker) approximately 30 minutes ago. He has no sign or symptom and his vital signs on arrival are:

HR 65 (sinusal) BP 110/60 RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 15/15

8a) Do you monitor this patient? ! yes ! no

8b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

8c) How often do you measure blood glucose levels? ! prn ! every _____ minutes ! every _____ hours

8d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

8e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

8f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

8g) Do you call the Poison Control Center concerning this patient? ! yes ! no

YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

221

9) CASE #3: A 35 year-old male with no prior history is brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release calcium channel blocker) approximately 4 hours ago. He is dizzy and his vital signs on arrival are:

HR 36 (1° AV block) BP 80mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

9a) Do you monitor this patient? ! yes ! no

9b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

9c) How often do you measure blood glucose level? ! prn ! every _____ minutes ! every _____ hours

9d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

9e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

9f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

9g) Do you call the Poison Control Center concerning this patient? ! yes ! no

9h) If you do not have access to an colleague or an authority such as the Poison Control Center to give you advice, what intervention(s) would you prescribe for the next 2-3 minutes for this patient?

! Fluids ! Phosphodiesterase III inhibitors ! Atropine ! Mechanical interventions (specify) ! Vasopressors ! external pacemaker ! IV calcium ! intravenous pacemaker ! Glucagon ! extracorporeal life-support ! High-dose insulin ! Lipid emulsion

! Other: ______

YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

222

10) CASE #4: A 35 year-old male with no prior history was brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release calcium channel blocker) approximately 4 hours ago. He was dizzy and his vital signs on arrival were:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

Your colleague gave him a bolus of 500 cc of NS, administered a dose of atropine without response and started a dopamine perfusion (doses recommended by the ACLS). The patient is still dizzy and his vital signs are now:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

You have confirmed that there was no diagnostic error. He is now your patient.

10a) Do you monitor this patient? ! yes ! no

10b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

10c) How often do you measure blood glucose level? ! prn ! every _____ minutes ! every _____ hours

10d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

10e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

10f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

10g) Do you call the Poison Control Center concerning this patient? ! yes ! no

10h) If you do not have access to a colleague or an authority such as the Poison Control Center to give you advice, what intervention(s) would you prescribe for the next 2-3 minutes for this patient? ! Fluids ! Phosphodiesterase III inhibitors ! Atropine ! Mechanical interventions (specify) ! Vasopressors ! external pacemaker ! IV calcium ! intravenous pacemaker ! Glucagon ! extrecorporeal life-support ! High-dose insulin ! Lipid emulsion

! Other: ______

YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

223

11) CASE #5: A 35 year-old male with no prior history was brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release calcium channel blocker) approximately 4 hours ago. He was dizzy and his vital signs on arrival were:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

Your colleague gave him a bolus of 500 cc of normal saline, administered a dose of atropine without response, started a dopamine perfusion (doses recommended by the ACLS) and then administered two boluses of IV calcium. The patient is still dizzy and his vital signs are now:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

You have confirmed that there was no diagnostic error. He is now your patient.

11a) Do you monitor this patient? ! yes ! no

11b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

11c) How often do you measure blood glucose level? ! prn ! every _____ minutes ! every _____ hours

11d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

11e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

11f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

11g) Do you call the Poison Control Center concerning this patient? ! yes ! no

11g) If you do not have access to a colleague or an authority such as the Poison Control Center to give you advice, what intervention(s) would you prescribe for the next 2-3 minutes for this patient? ! Fluids ! Phosphodiesterase III inhibitors ! Atropine ! Mechanical interventions (specify) ! Vasopressors ! external pacemaker ! IV calcium ! intravenous pacemaker ! Glucagon ! extrecorporeal life-support ! High-dose insulin ! Lipid emulsion

! Other: ______YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

224

12) CASE #6: A 35 year-old male with no prior history was brought in by ambulance after a suicide attempt. He ingested an undetermined amount of diltiazem tablets (immediate release CCB) approximately 4 hours ago. He was dizzy and his vital signs on arrival were:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

Your colleague gave him a bolus of 500 cc of normal saline, administered a dose of atropine without response, started a dopamine perfusion (doses recommended by the ACLS) and then administered two boluses of IV calcium. The insulin-glucose protocol was also started over 30 minutes ago. The patient is still dizzy and his vital signs are now:

HR 36 (1° AV block) BP 80 mmHg systolic RR 20 Sat 99% T° 37.5 (rectal) Gluco 11 GCS 14/15

You have confirmed that there was no diagnostic error. He is now your patient.

12a) Do you monitor this patient? ! yes ! no

12b) How often do you measure HR and BP? ! prn ! every _____ minutes ! every _____ hours

12c) How often do you measure blood glucose level? ! prn ! every _____ minutes ! every _____ hours

12d) Do you do an ECG? ! yes ! no If yes, how often do you repeat the ECG? ! prn ! every _____ minutes ! every _____ hours

12e) How often do you measure electrolytes? ! prn ! every _____ minutes ! every _____ hours

12f) Do you decontaminate the patient? ! yes ! no If yes, how? ! ipecac syrup !gastric lavage !dialysis ! activated charcoal !activated charcoal repeated ! whole bowel irrigation ! other (specify): ______

12g) Do you call the Poison Control Center concerning this patient? ! yes ! no

12h) If you do not have access to a colleague or an authority such as the Poison Control Center to give you advice, what intervention(s) would you prescribe for the next 2-3 minutes for this patient? ! Fluids ! Phosphodiesterase III inhibitors ! Atropine ! Mechanical interventions (specify) ! Vasopressors ! external pacemaker ! IV calcium ! intravenous pacemaker ! Glucagon ! extracorporeal life-support ! High-dose insulin ! Lipid emulsion

! Other: ______YOU WILL NOT BE ABLE TO MAKE CHANGES ONCE THIS PAGE IS HANDED IN

225

13) If you were to prescribe the following treatments, specify how you would do so:

13a) IV calcium (check only one answer) ! a single bolus ! repeated boluses ! bolus + infusion ! bolus + infusion only if bolus is effective ! infusion alone

13b) Glucagon (check only one answer) ! a single bolus ! repeated boluses ! bolus + infusion ! bolus + infusion only if bolus is effective ! infusion alone

13c) Insulin-glucose: ! insulin ! bolus, if yes, how much?__ U/kg or ! do not know the amount ! infusion, if yes, how much? __U/kg/h or ! do not know the rate ! glucose ! bolus, if yes, how much? ___ g/kg or ! do not know the amount ! infusion, if yes, how much? __g/kg/h or ! do not know the rate !potassium ! infusion or bolus immediately ! perfusion or bolus prn Indications for discontinuing the protocol: ! hemodynamic improvement, discontinuation of vasopressor ! hemodynamic improvement, non-discontinuation of vasopressor ! deterioration of the hemodynamic status ! hypokalemia ! hypoglycemia ! other (specify): ______! do not know

14) Check off the interventions which are not available in your hospital:

! glucagon in a sufficient quantity to start an infusion ! IV calcium ! high-dose insulin ! external pacemaker ! intravenous pacemaker ! extracorporeal life-support ! phosphodiesterase III inhibitors ! lipid emulsion ! adequate material or human resources for continuous monitoring and to take capillary or venous blood samples every 15-30 min

15) In your opinion, what factors would influence the treatments considered for your patient with calcium channel blocker poisoning?

Thanks for your participation!

Maude St-Onge

226

Appendix 2 Case reports focusing on an intervention or reporting a side effect Source Subject and patient Type of CCB Intervention and co- Outcome characteristics poisoning interventions

High-dose insulin (HDI)

Abeysinghe - Subject: 62 yo male 2.5g diltiazem SR - Intervention: HDI 70 units - Mortality: N/R et al., 2010 - Comorbidities: hypertension, bolus followed by an infusion - Morbidity: metabolic coronary artery disease at 0.5 units/kg/h acidosis, acute kidney - Coingestions: atorvastatin, - Co-interventions: fluids, injury aspirin, isosorbide mononitrate calcium chloride, glucagon, - Hemodynamic - State at arrival: unstable vasopressors improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Agarawal et - Subject: 60 yo male - 5.4g diltiazem - Intervention: HDI 0.1 to 0.3 - Mortality: no al, 2012 - Comorbidities: ER (1st units/kg/h - Morbidity: bowel schizoaffective disorder, admission) - Co-interventions: fluids, ischemia, acute kidney coronary artery disease, heart - 1.8g nifedipine calcium, dopamine, glucagon injury failure, hypertension, long-acting (2nd (2nd and 3rd admission), - Hemodynamic peripheral vascular disease admission) activated charcoal (3rd improvement: yes - Coingestions: none - 5.4g diltiazem admission) - Functional status at - State at arrival: unstable ER (3rd discharge: N/R admission) - Secondary outcomes: N/R - Adverse effects: N/R Azendour et - Subject: 20 yo female 240mg of - Intervention: HDI 1 unit/kg - Mortality: no al., 2010 - Comorbidities: none amlodipine bolus followed by an infusion - Morbidity: N/R - Coingestions: none at 0.5 unit/kg/h - Hemodynamic - State at arrival: unstable - Co-interventions: fluids, improvement: yes vasopressors, calcium - Functional status at gluconate infusion discharge: N/R - Secondary outcomes: 3 days in ICU - Adverse effects: N/R Clark et al., - Subject: 52 yo female 960mg of - Intervention: HDI 0.5-1 - Mortality: no 2008 - Comorbidities: hypertension, verapamil SR unit/kg/h - Morbidity: N/R peptic ulcer disease - Co-interventions: - Hemodynamic - Coingestions: trandolapril, vasopressors, glucagon improvement: N/R olmesartan, escitalopram, - Functional status at benzodiazepines, ethanol discharge: 3 days in - State at arrival: unstable hospital - Secondary outcomes: N/R - Adverse effects: N/R Cumpston - Subject: 49 yo male 6.48g of diltiazem - Intervention: HDI 1 unit/kg - Mortality: yes et al., 2009 - Comorbidities: atrial SR bolus followed by an infusion - Morbidity: N/R fibrillation, hypertension, at 1 unit/kg/h titrated up to 360 - Hemodynamic dilated cardiomyopathy, units/h improvement: yes hyperlipidemia, coronary - Co-interventions: - Functional status at artery disease vasopressors, calcium discharge: N/A - Coingestions: other gluconate, glucagon - Secondary outcomes: antihypertensives, statin, N/R warfarin - Adverse effects: N/R - State at arrival: unstable Deshpande - Subject: 59 yo female 5.76g diltiazem - Intervention: Insulin started - Mortality: N/R et al., 2004 - Comorbidities: hypertension, at 5 units/h increased to 25 - Morbidity: N/R depression, suicide attempts units/h after 15 hours - Hemodynamic - Coingestions: zolpidem and - Co-interventions: activated improvement: yes lorazepam charcoal, fluids, calcium - Functional status at - State at arrival: unstable chloride, glucagon, discharge: N/R vasopressors - Secondary outcomes: N/R - Adverse effects: N/R

227

Devasahay - Subject: 60 yo female Amlodipine 10 mg - Intervention: HDI - Mortality: no am et al., - Comorbidities: hypertension, PO BID X 10 days - Co-interventions: - Morbidity: small bowel 2012 hypothyroidism, depression vasopressors, calcium obstruction and metabolic - Coingestions: none acidosis - State at arrival: unstable - Hemodynamic improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 28 days in hospital - Adverse effects: N/R El Houari et - Subject: 27 yo female 150 mg - Intervention: HDI 0.5 U/kg/h - Mortality: no al., 2013 - Comorbidities: none amlodipine - Co-interventions: gastric - Morbidity: N/R - Coingestions: none lavage, fluids, calcium - Hemodynamic - State at arrival: unstable gluconate, dobutamine, improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 7 days in hospital - Adverse effects: N/R Felgenhaue - Subject: 65 yo female 1.4g of verapamil - Intervention: HDI 35 units/h - Mortality: no r et al., - Comorbidities: N/R - Co-interventions: calcium, - Morbidity: pneumonia 2007 - Coingestions: quinidine glucagon, vasopressors, - Hemodynamic - State at arrival: N/R bicarbonates, pacing improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Goplani et - Subject: 28 yo female 250mg of - Intervention: HDI 10 units IV - Mortality: no al., 2011 - Comorbidities: none amlodipine bolus followed by a continuous - Morbidity: N/R - Coingestions: N/R infusion at 10 units over 2-4h - Hemodynamic - State at arrival: unstable - Co-interventions: gastric improvement: yes lavage, fluids, vasopressors, - Functional status at calcium, glucagon discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Hadjipavlou - Subject: 49 yo female 560mg of - Intervention: HDI 1 unit/kg - Mortality: no et al., 2011 - Comorbidities: hypertension lecarnidipine bolus followed by an infusion - Morbidity: seizures - Coingestions: N/R at 0.5 unit/kg/h for 8 hours - Hemodynamic - State at arrival: unstable - Co-interventions: improvement: yes vasopressors - Functional status at discharge: N/R - Secondary outcomes: 2 days in hospital - Adverse effects: N/R Harris et al, - Subject: 40 yo female 1g of amlodipine - Intervention: HDI 80 units - Mortality: no 2006 - Comorbidities: mental illness bolus followed by an infusion - Morbidity: ARDS, - Coingestions: N/R at 40 units/h ventilator-associated - State at arrival: stable - Co-interventions: activated pneumonia, ileus charcoal, whole bowel Hemodynamic irrigation, calcium, glucagon, improvement: yes vasopressors Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Hasin et al., - Subject: 59 yo female 3g of verapamil - Intervention: Insulin 1-2 - Mortality: no 2006 - Comorbidities: hypertension units/h - Morbidity: pulmonary - Coingestion: enalapril - Co-interventions: fluid, edema - State at arrival: unstable calcium, glucagon, - Hemodynamic vasopressors improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 3 days in hospital - Adverse effects: N/R

228

King et al., - Subject: 58 yo female Diltiazem - Intervention: HDI 1 U/kg/h - Mortality: no 2013 - Comorbidities: N/R - Co-interventions: fluids - Morbidity: pulmonary - Coingestions: N/R edema, anasarca - State at arrival: unstable - Hemodynamic improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 15 days in hospital - Adverse effects: volume overload Li Saw et - Subject: N/R Verapamil - Intervention: HDI - Mortality: yes al., 1996 - Comorbidities: N/R - Co-interventions: calcium, - Morbidity: N/R - Coingestions: N/R vasopressors - Hemodynamic - State at arrival: N/R improvement: N/R - Functional status at discharge: N/A - Secondary outcomes: N/R - Adverse effects: N/R Maldonado - Subject: 63 yo male 450 mg - Intervention: HDI 0.5- - Mortality: yes et al., 2012 - Comorbidities: N/R amlodipine 3.5U/kg/h - Morbidity: N/R - Coingestions: metoprolol - Co-interventions: calcium, - Hemodynamic - State at arrival: unstable glucagon, vasopressors, improvement: N/R dobutamine, milrinone, lipid - Functional status at emulsion, plasmapheresis, discharge: N/R pacemaker, intra-aortic - Secondary outcomes: balloon pump N/R - Adverse effects: N/R Marques et - Subject: 75 yo female Diltiazem - Intervention: HDI 0.5 - Mortality: no al., 2003 - Comorbidities: hypertension, units/kg/h - Morbidity: N/R atrial fibrillation and psychiatric - Co-interventions: fluids, - Hemodynamic disorder vasopressors, calcium improvement: yes - Coingestions: indapamide, - Functional status at lorazepam, alprazolam, discharge: N/R cetirizine, diclofenac, - Secondary outcomes: 4 celecoxib, digoxin, diosmin days in ICU and trimetazide - Adverse effects: N/R - State at arrival: unstable Montiel et - Subject: 18 yo female 3.6g of diltiazem - Intervention: HDI 40 units - Mortality: no al., 2011 - Comorbidities: none SR bolus followed by an infusion - Morbidity: N/R - Coingestions: none at 25 units/h (0.5 units/kg/h) - Hemodynamic - State at arrival: unstable - Co-interventions: fluids, improvement: yes calcium chloride, lipid - Functional status at emulsion discharge: N/R - Secondary outcomes: 9 days in the ICU - Adverse effects: N/R Munshi et - Subject: 19 yo male 700mg of - Intervention: HDI - Mortality: yes al, 2010 - Comorbidities: N/R amlodipine - Co-interventions: activated - Morbidity: N/R - Coingestion: lisinopril charcoal, fluids, vasopressors, - Hemodynamic - State at arrival: unstable calcium chloride, glucagon, improvement: no lipid emulsion, bicarbonate - Functional status at discharge: N/A - Secondary outcomes: N/R - Adverse effects: N/R Nickson et - Subject: 49 yo male Verapamil - Intervention: HDI 30 units - Mortality: no al, 2009 - Comorbidities: depression, bolus followed by an infusion - Morbidity: N/R ethanol abuse, chronic back at 30 units/h - Hemodynamic pain, hypertension, previous - Co-interventions: activated improvement: yes suicide attempts charcoal, fluids, vasopressors - Functional status at - Coingestions: morphine, discharge: N/R diazepam, tramadol - Secondary outcomes: - State at arrival: stable N/R - Adverse effects: no

229

Nimbalkar - Subject: 11 month-old infant 90 mg amlodipine - Intervention: HDI 0.5U/kg/h - Mortality: no et al., 2013 - Comorbidities: lower - Co-interventions: calcium - Morbidity: renal failure, respiratory tract infection gluconate, glucagon, pulmonary edema - Coingestions: none vasopressors, peritoneal - Hemodynamic - State at arrival: unstable dialysis improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 10 days in hospital - Adverse effects: N/R Ortiz-Nunoz - Subject: 77 yo male Nifedipine - Intervention: HDI 0.5 - Mortality: no et al., 2005 - Comorbidities: hypertension, units/kg/h - Morbidity: pulmonary dyslipidemia, hypothyroidism, - Co-interventions: gastric edema stroke lavage, activated charcoal, - Hemodynamic - Coingestion: atenolol fluids, calcium gluconate, improvement: N/R - State at arrival: unstable glucagon - Functional status at discharge: N/R - Secondary outcomes: 1 day in ICU - Adverse effects: hypoglycemias Patel et al., - Subject: 49 yo female 5.8g of verapamil - Intervention: HDI 0.1 unit/kg - Mortality: no 2007 - Comorbidities: N/R SR bolus followed by an infusion - Morbidity: acute kidney - Coingestions: captopril, at 0.25 units/kg/h injury, pancreatitis glyburide - Co-interventions: activated - Hemodynamic - State at arrival: unstable charcoal, atropine, calcium improvement: yes chloride, vasopressors, - Functional status at bicarbonate, pacemaker discharge: N/R - Secondary outcomes: N/R - Adverse effects: no hypoglycemia Rizvi et al., - Subject: young female 280mg of - Intervention: HDI - Mortality: no 2012 - Comorbidities: N/R amlodipine - Co-interventions: calcium - Morbidity: N/R - Coingestions: N/R gluconate, glucagon - Hemodynamic - State at arrival: unstable improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Rosser et - Subject: 43 yo female 3,36g of diltiazem - Intervention: HDI 2 units/kg/h - Mortality: no al., 2012144 - Comorbidities: hypertension, in 2 units increments up to 10 - Morbidity: N/R chest pain units/kg/h - Hemodynamic - Coingestions: gabapentin, - Co-interventions: fluids, improvement: N/R citalopram, diazepam, pacemaker, hemofiltration, - Functional status at irbesartan levosimendan discharge: left ventricular - State at arrival: unstable dysfunction - Secondary outcomes: 30 days in the ICU - Adverse effects: N/R Stellpflug et - Subject: 30 yo female Diltiazem - Intervention: HDI 0.5 units/kg - Mortality: no al., 2011 - Comorbidities: hypertension, bolus followed by an infusion - Morbidity: N/R congestive heart failure at incremental doses up to 10 - Hemodynamic - Coingestions: metoprolol, units/kg/h improvement: yes amiodarone - Co-interventions: fluids, - Functional status at - State at arrival: unstable calcium, lipid emulsion discharge: N/R - Secondary outcomes: 5 days in the ICU - Adverse effects: N/R St-Onge - Subject: 47 yo female 4.5g of verapamil - Intervention: HDI 10 units/kg - Mortality: no 2010 - Comorbidities: depression, bolus followed by an infusion - Morbidity: kidney injury, hypothyroidism, chronic pain, up to 500 units/h liver dysfunction, arrhythmias - Co-interventions: atropine, metabolic acidosis - Coingestions: none vasopressors, calcium, - Hemodynamic - State at arrival: unstable glucagon, bicarbonates, lipid improvement: yes emulsion, pacemaker - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R 230

ECLS Bouchard et - Subject: 50 yo male 500 mg of - Intervention: ECLS for 9 days - Mortality: no al., 2010 - Comorbidities: depression, amlodipine - Co-interventions: - Morbidity: acute kidney hypertension, ethanol vasopressors, glucagon, injury, sepsis - Coingestions: calcium, HDI (1.5 units/kg/h), - Hemodynamic hydrochlorothiazide, lisinopril methylene blue, lipid emulsion improvement: yes - State at arrival: unstable - Functional status at discharge: back to baseline - Secondary outcomes: N/R - Adverse effects: no DeRita et - Subject: 33 days old (5600 g) Verapamil IV - Intervention: ECLS - Mortality: no al., 2010 - Comorbidities: paroxysmal - Co-interventions: fluid, - Morbidity: N/R SVT atropine, vasopressors - Hemodynamic - Coingestions: N/R improvement: yes - State at arrival: unstable - Functional status at discharge: back to normal neurologic and cardiac functions - Secondary outcomes: N/R - Adverse effects: N/R Holzer et - Subject: 41 yo male 4.8g-6.4g of - Intervention: ECLS - Mortality: no al., 1999 - Comorbidities: N/R verapamil - Co-interventions: - Morbidity: N/R - Coingestions: N/R vasopressors, theophylline - Hemodynamic - State at arrival: cardiac arrest improvement: yes - Functional status at discharge: back to baseline - Secondary outcomes: 18 days in ICU - Adverse effects: N/R Kolcz et al., - Subject: 15 yo female 960mg of - Intervention: ECLS - Mortality: yes 2007 - Comorbidities: N/R verapamil - Co-interventions: activated - Morbidity: dysrhythmias - Coingestions: ethanol, charcoal, atropine, - Hemodynamic propranolol vasopressors, bicarbonate, improvement: yes - State at arrival: unstable glucagon, pacemaker, - Functional status at milrinone, frozen plasma discharge: back to baseline - Secondary outcomes: 32 days in hospital - Adverse effects: bleeding Maclaren et - Subject: 45 yo female 7.2g verapamil SR - Intervention: ECLS - Mortality: no al., 2005 - Comorbidities: N/R - Co-interventions: activated - Morbidity: N/R - Coingestions: doxepin, charcoal, fluids, vasopressors, - Hemodynamic quetiapine, diazepam, glucagon, calcium, HDI, improvement: yes temazepam, clonazepam atropine, dobutamine, - Functional status at - State at arrival: stable CVVHD, plasmapheresis discharge: N/R - Secondary outcomes: 8 days of ICU - Adverse effects: N/R Persad et - Subject: 13 yo female 300 mg - Intervention: ECLS - Mortality: no al., 2012 - Comorbidities: N/R amlodipine - Co-interventions: fluids, - Morbidity: N/R - Coingestions: ethanol, calcium chloride, vasopressors - Hemodynamic barbiturates improvement: yes - State at arrival: unstable - Functional status at discharge: neurologically intact - Secondary outcomes: 2 days in ICU - Adverse effects: N/R Rona et al., - Subject: 36 yo male Nifedipine - Intervention: ECLS - Mortality: no 2011164 - Comorbidities: hypertension, - Co-interventions: activated - Morbidity: N/R depression, bipolar, suicide charcoal, whole bowel - Hemodynamic attempts irrigation, fluid, bicarbonate, improvement: yes - Coingestions: atenolol, vasopressors, calcium, - Functional status at sertraline, lacidipine, fluoxetine glucagon, HDI, plasma discharge: long-term care - State at arrival: unstable exchange, CVVH facility one year post event - Secondary outcomes: 28 days in ICU 231

- Adverse effects: N/R Rygnestad - Subject: 29 yo female 3.6g verapamil - Intervention: ECLS - Mortality: no et al., 2005 - Comorbidities: N/R - Co-interventions: fluids, - Morbidity: compartment - Coingestions: sotalol vasopressors, glucagon, syndrome of left forearm, - State at arrival: unstable calcium, pacemaker, CVVHF, acute kidney injury, milrinone rhabdomyolysis, gastrointestinal bleed, transient nerve palsy - Hemodynamic improvement: N/R - Functional status at discharge: back to work 6 months after poisoning - Secondary outcomes: N/R - Adverse effects: N/R Schmidt et - Subject: 38 yo male Nifedipine - Intervention: ECLS - Mortality: yes al., 1995 - Comorbidities: depression - Co-interventions: Digibind - Morbidity: severe - Coingestions: methyldigoxin, neurologic damage indapamide - Hemodynamic - State at arrival: unstable improvement: yes - Functional status at discharge: N/A - Secondary outcomes: N/R - Adverse effects: N/R Tauro et al., - Subject: 18 yo male 275mg of - Intervention: ECLS - Mortality: no 2009 - Comorbidities: N/R amlodipine - Co-interventions: gastric - Morbidity: acute lung - Coingestions: N/R lavage, fluid, vasopressors, injury - State at arrival: stable glucagon, calcium, HDI 1 - Hemodynamic unit/kg/h improvement: yes - Functional status at discharge: back to baseline with an amputated leg - Secondary outcomes: N/R - Adverse effects: leg compartment syndrome with requirement for an amputation Weinberg et - Subject: 50 yo male 500 mg - Intervention: ECLS - Mortality: no al., 2013 - Comorbidities: depression, amlodipine - Co-interventions: calcium - Morbidity: cardiac failure ethanol abuse gluconate, glucagon, - Hemodynamic - Coingestions: lisinopril, vasopressors, HDI, methylene improvement: yes hydrochlorothiazide blue, lipid emulsion - Functional status at - State at arrival: unstable discharge: N/R - Secondary outcomes: 56 days in hospital - Adverse effects: N/R Calcium Adams et - Subject: 76 yo male 100mg of - Intervention: calcium - Mortality: yes al., 1998 - Comorbidities: coronary amlodipine over gluconate and chloride (7g - Morbidity: acute kidney artery disease, congestive 24h total) injury, pulmonary edema, heart failure, hypertension, - Co-interventions: retroperitoneal hematoma, moderate mitral regurgitation, vasopressors, glucagon, ARDS, sepsis aortic valve replacement furosemide, salbutamol - Hemodynamic - Coingestions: N/R improvement: yes - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Ashraf et - Subject: 45 yo female 24g of verapamil - Intervention: calcium - Mortality: no al., 1995 - Comorbidities: N/R - Co-interventions: - Morbidity: N/R - Coingestions: N/R vasopressors, glucagon, - Hemodynamic - State at arrival: unstable pacemaker improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 15 days in ICU and 28 days 232

in hospital - Adverse effects: N/R Buylaert et - Subject: 63 yo female 70 mg of - Intervention: calcium - Mortality: yes al., 1995 - Comorbidities: depression amlodipine gluconate 2g followed by an - Morbidity: N/R - Coingestion: oxazepam infusion, then calcium chloride - Hemodynamic - State at arrival: unstable 1g/h improvement: yes - Co-interventions: atropine, - Functional status at vasopressors, glucagon, discharge: N/R insulin - Secondary outcomes: N/R - Adverse effects: hypercalcemia Crump et - Subject: 39 yo female 1.2g of verapamil - Intervention: calcium - Mortality: no al., 1982 - Comorbidities: N/R gluconate 10%, 20 ml - Morbidity: anoxic brain - Coingestions: N/R - Co-interventions: glucagon, injury - State at arrival: unstable prenalterol, atropine, - Hemodynamic vasopressors, isoprenaline improvement: no - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Devasahay - Subject: 60 yo female Amlodipine 10 mg - Intervention: Calcium - Mortality: no am et al., - Comorbidities: hypertension, PO X 10 days chloride (total 3.5g) - Morbidity: small bowel 2012 hypothyroidism, depression - Co-interventions: obstruction, metabolic - Coingestions: none vasopressors, HDI acidosis - State at arrival: unstable - Hemodynamic improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 28 days in hospital - Adverse effects: N/R Fergenhaue - Subject: 65 yo female 1.4g verapamil - Intervention: calcium chloride - Mortality: no r et al., - Comorbidities: atrial - Co-interventions: - Morbidity: pneumonia 2007 fibrillation vasopressors, urine - Hemodynamic - Coingestion: quinidine alkalinisation, pacemaker, HDI improvement: yes - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Gerloni et - Subject: 81 yo female Nimodipine - Intervention: calcium - Mortality: no al., 2004 - Comorbidities: N/R gluconate 10% 20ml IV - Morbidity: N/R - Coingestions: N/R - Co-interventions: none - Hemodynamic - State at arrival: unstable improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Hasson et - Subject: 22 yo female 280mg of - Intervention: calcium - Mortality: no al., 2011 - Comorbidities: N/R amlodipine gluconate - Morbidity: pulmonary - Coingestions: N/R - Co-interventions: fluids, edema - State at arrival: N/R furosemide - Hemodynamic improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Isbister - Subject: 51 yo male 1.8-3.6g of - Intervention: 13.5g of calcium - Mortality: no 2002 - Comorbidities: N/R diltiazem SR over 12 min - Morbidity: seizures - Coingestions: - Co-interventions: activated - Hemodynamic acetaminophen, aspirin, charcoal, fluids, vasopressors, improvement: isosorbide nitrate and alcohol pacemaker, dobutamine improvement followed by - State at arrival: stable cardiac arrest - Functional status at discharge: back to baseline - Secondary outcomes: 233

N/R - Adverse effects: N/R Lam et al., - Subject: 37 yo female 4.2g of nifedipine - Intervention: calcium chloride - Mortality: N/R 2001 - Comorbidities: diabetes, 20ug/min titrated to maintain a - Morbidity: N/R hypertension, schizophrenia calcemia of 2 mmol/L - Hemodynamic - Coingestions: N/R - Co-interventions: gastric improvement: yes - State at arrival: unstable lavage, activated charcoal, - Functional status at fluids discharge: N/R - Secondary outcomes: N/R - Adverse effects: necrosis at the IV site secondary to extravasation Lee et al., - Subject: 78 yo female Verapamil - Intervention: calcium chloride - Mortality: no 1995 - Comorbidities: - Co-interventions: atropine, - Morbidity: N/R hyperlipidemia, hypertension, vasopressors - Hemodynamic diabetes, coronary bypass improvement: yes - Coingestion: metoprolol - Functional status at - State at arrival: unstable discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Lipman et - Subject: 71 yo male Verapamil 5 mg IV - Intervention: calcium - Mortality: no al., 1982 - Comorbidities: hypertension bolus then 5 mg/h gluconate 10%, 20 ml - Morbidity: N/R - Coingestions: none infusion - Co-interventions: N/R - Hemodynamic - State at arrival: stable improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Li Saw et - Subject: N/R Verapamil - Intervention: calcium - Mortality: yes al., 1996 - Comorbidities: N/R - Co-interventions: HDI, - Morbidity: N/R - Coingestions: N/R vasopressors - Hemodynamic - State at arrival: N/R improvement: no - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Luscher et - Subject: 65 yo female Verapamil - Intervention: calcium - Mortality: N/R al., 1984 - Comorbidities: hypertension gluconate 10%, 20 ml - Morbidity: N/R - Coingestions: N/R repeated to a total of 18g - Hemodynamic - State at arrival: unstable - Co-interventions: gastric improvement: yes lavage, pacemaker, - Functional status at epinephrine discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Marjanovic - Subject: 33 yo female 4g verapamil - Intervention: calcium - Mortality: no et al., 2010 - Comorbidities: N/R gluconate - Morbidity: N/R - Coingestions: N/R - Co-interventions: gastric - Hemodynamic - State at arrival: unstable lavage, activated charcoal, improvement: yes atropine, vasopressors - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Moroni et - Subject: 31 yo male 3.2g of verapamil - Intervention: calcium - Mortality: no al., 1980 - Comorbidities: ethanol gluconate 10%, 20 ml every - Morbidity: N/R - Coingestions: N/R 8h - Hemodynamic - State at arrival: unstable - Co-interventions: improvement: yes dexamethasone - Functional status at discharge: N/R - Secondary outcomes: 14 days in hospital - Adverse effects: N/R Nanda et - Subject: 58 yo male 240mg verapamil - Intervention: calcium chloride - Mortality: no al., 2005 - Comorbidities: atrial 10%, 10 ml IV - Morbidity: N/R fibrillation - Co-interventions: atropine, - Hemodynamic - Coingestions: digoxin, fluids, pacemaker improvement: N/R 234

warfarin - Functional status at - State at arrival: unstable discharge: N/R - Secondary outcomes: 2 days in hospital - Adverse effects: N/R Perkins et - Subject: 19 yo female 3.2g of verapamil - Intervention: calcium - Mortality: N/R al., 1978 - Comorbidities: ventricular gluconate 10%, 10 ml over 5 - Morbidity: N/R extrasystoles min - Hemodynamic - Coingestions: none - Co-interventions: none improvement: yes - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Rizvi et al., - Subject: young female 280mg of - Intervention: calcium - Mortality: no 2012 - Comorbidities: N/R amlodipine gluconate - Morbidity: N/R - Coingestions: N/R - Co-interventions: HDI, - Hemodynamic - State at arrival: unstable glucagon improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Wills et al., - Subject: 9 month old (9.8kg) 120 mg of - Intervention: calcium - Mortality: no 2010 - Comorbidities: N/R diltiazem ER gluconate 500mg - Morbidity: no - Coingestions: N/R - Co-interventions: activated - Hemodynamic - State at arrival: unstable charcoal improvement: yes - Functional status at discharge: back to baseline - Secondary outcomes: 42h in ED - Adverse effects: no Woie et al. - Subject: 33 yo male 3g of verapamil - Intervention: calcium - Mortality: no 1981 - Comorbidities: N/R gluconate 10%, 10 ml - Morbidity: N/R - Coingestions: N/R - Co-interventions: gastric - Hemodynamic - State at arrival: unstable lavage, metaraminol improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Vasopresseurs Adams et - Subject: 76 yo male 100mg of - Intervention: dopamine, - Mortality: yes al., 1998 - Comorbidities: coronary amlodipine dobutamine, norepinephrine - Morbidity: acute kidney artery disease, congestive - Co-interventions: injury, pulmonary edema, heart failure, hypertension, gastrointestinal ARDS, sepsis mitral regurgitation decontamination, fluids, - Hemodynamic - Coingestions: N/R glucagon, calcium improvement: yes - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: 10 days of vasopressors use, 112 days in hospital - Adverse effects: N/R Chimenti et - Subject: 38 yo female 2.4 g of verapamil - Intervention: orciprenalin, - Mortality: no al., 1982 - Comorbidities: cardio- dopamine, epinephrine - Morbidity: no vascular disease, depression - Co-interventions: gastric - Hemodynamic - Coingestions: none lavage, calcium gluconate improvement: no - State at arrival: unstable improvement with dopamine, yes with epinephrine - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Gris et al., - Subject: 62 yo female Verapamil - Intervention: dobutamine, - Mortality: no 1989 - Comorbidities: none dopamine - Morbidity: N/R - Coingestions: none - Co-interventions: pacemaker - Hemodynamic - State at arrival: unstable improvement: yes 235

- Functional status at discharge: N/R - Secondary outcomes: 1 day in hospital - Adverse effects: N/R Hagège et - Subject: 18 yo female 16.8g of verapamil - Intervention: isoproterenol - Mortality: no al., 1990 - Comorbidities: N/R SR 0.5 ug/kg/min - Morbidity: no - Coingestions: N/R - Co-interventions: gastric - Hemodynamic - State at arrival: unstable lavage, pacemaker, inotropres improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Henderson - Subject: 73 yo male 4.8g of diltiazem - Intervention: adrenaline 1mg - Mortality: no et al. 1992 - Comorbidities: angina, bolus followed by an infusion - Morbidity: pulmonary depression at 0.2 mcg/kg/min edema - Coingestions: N/R - Co-interventions: atropine, - Hemodynamic - State at arrival: unstable calcium gluconate, glucagon, improvement: yes isoprenaline, pacemaker - Functional status at discharge: N/R - Secondary outcomes: 6 weeks in hospital - Adverse effects: N/R Kalman et - Subject: 55 yo female 2.88g of verapamil - Intervention: adrenaline, - Mortality: N/R al., 1998 - Comorbidities: SVT SR noradrenaline, dobutamine, - Morbidity: N/R - Coingestion: atenolol isoprenaline - Hemodynamic - State at arrival: unstable - Co-interventions: activated improvement: yes charcoal, fluids, atropine, - Functional status at calcium gluconate, prenalterol, discharge: N/R - Secondary outcomes: 5 days in ICU - Adverse effects: none Marraffa et - Subject: 22 yo female 800mg of - Intervention: vasopressin 4.8 - Mortality: N/R al., 2004 - Comorbidities: N/R amlodipine units/h - Morbidity: N/R - Coingestions: N/R - Co-interventions: fluids, - Hemodynamic - State at arrival: unstable vasopressors (dopamine and improvement: yes epinephrine), calcium, insulin - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Oe et al., - Subject: 73 yo female 3.6g of verapamil - Intervention: norepinephrine - Mortality: no 1998 - Comorbidities: depression, - Co-interventions: activated - Morbidity: N/R suicide attempts, SVT charcoal - Hemodynamic - Coingestions: N/R improvement: N/R - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Plumb et - Subject: 50 yo female 770mg of - Intervention: Noradrenaline, - Mortality: no al., 2011 - Comorbidities: diabetes, amlodipine norepinephrine, vasopressin - Morbidity: acute kidney hypertension, obesity then metaraminol injury - Coingestion: losartan - Co-interventions: fluids, - Hemodynamic - State at arrival: stable glucagon, calcium, HDI improvement: with metaraminol - Functional status at discharge: N/R - Secondary outcomes: 6 days in hospital - Adverse effects: N/R Proano et - Subject: 42 yo female Diltiazem CD - Intervention: dopamine, - Mortality: no al., 1995 - Comorbidities: diabetes, dobutamine - Morbidity: aspiration depression, hypertension, - Co-interventions: activated pneumonia, ARDS, sepsis suicidal attempts, ethanol, charcoal in sorbitol, fluids, - Hemodynamic benzodiazepine abuse atropine, calcium, insulin 5 improvement: no - Coingestions: N/R units IV, pacemaker - Functional status at - State at arrival: unstable discharge: N/R - Secondary outcomes: 20 days in hospital 236

- Adverse effects: N/R Decontamination Madera et - Subject: 33 yo male 2.8g of verapamil - Intervention: gastric lavage - Mortality: yes al., 1977 - Comorbidities: N/R - Co-interventions: - Morbidity: pulmonary and - Coingestion: pentobarbital bronchodilator, bicarbonate, cerebral edema - State at arrival: unstable cortisone, calcium, - Hemodynamic vasopressors, doxycycline improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: cardiac arrest 2h after gastric lavage Roberts et - Subject: 38 yo female 900mg of diltiazem - Intervention: repeated doses - Mortality: no al., 1991 - Comorbidities: N/R of activated charcoal 50g Q6h - Morbidity: N/R - Coingestion: ethanol - Co-interventions: ipecac, - Hemodynamic - State at arrival: stable fluids, calcium, vasopressors improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: decrease serum CCB concentrations - Adverse effects: N/R Pacemaker Dogan et - Subject: 3.5 yo female (14 240mg of - Intervention: pacemaker - Mortality: no al., 2010 kg) verapamil - Co-interventions: activated - Morbidity: N/R - Comorbidities: N/R charcoal, dopamine, - Hemodynamic - Coingestion: trandolapril epinephrine improvement: yes - State at arrival: unstable - Functional status at discharge: normal at discharge and at 6 months follow-up - Secondary outcomes: 3 days in hospital - Adverse effects: N/R Gris et al., - Subject: 62 yo female Verapamil - Intervention: pacemaker - Mortality: no 1989 - Comorbidities: N/R - Co-interventions: dopamine - Morbidity: N/R - Coingestions: N/R - Hemodynamic - State at arrival: unstable improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Snover et - Subject: 74 yo male 1.5g of verapamil - Intervention: pacemaker - Mortality: no al., 1986 - Comorbidities: glaucoma, - Co-interventions: atropine, - Morbidity: pneumonia peripheral vascular disease vasopressors, isoproterenol - Hemodynamic - Coingestions: N/R improvement: yes - State at arrival: unstable - Functional status at discharge: N/R - Secondary outcomes: 32 days in hospital - Adverse effects: N/R Glucagon Adams et - Subject: 76 yo male 100mg of - Intervention: glucagon - Mortality: yes al., 1998 - Comorbidities: coronary amlodipine infusion 5-10 mg/h - Morbidity: acute kidney artery disease, congestive - Co-interventions: calcium, injury, pulmonary edema, heart failure, hypertension, vasopressors, dobutamine, ARDS, sepsis cardiac bypass, aortic valve albuterol, furosemide, aspirin - Hemodynamic replacement improvement: N/R - Coingestions: N/R - Functional status at - State at arrival: unstable discharge: N/R - Secondary outcomes: 112 days in hospital - Adverse effects: N/R Ashraf et - Subject: 45 yo female 24g of verapamil - Intervention: glucagon - Mortality: N/R al., 1995 - Comorbidities: N/R SR infusion at 5 mg/h - Morbidity: aspiration - Coingestions: N/R - Co-interventions: naloxone, - Hemodynamic - State at arrival: unstable glucose, atropine, thiamine, improvement: N/R dopamine, calcium, - Functional status at 237

norepinephrine, pacemaker, discharge: N/R insulin drip - Secondary outcomes: 28 days in hospital - Adverse effects: hyperglycemia Buylaert et - Subject: 63 yo female 70mg of - Intervention: glucagon 10 - Mortality: yes al., 1995 - Comorbidities: depression amlodipine mg/h - Morbidity: N/R - Coingestion: oxazepam - Co-interventions: activated - Hemodynamic - State at arrival: unstable charcoal, fluids, vasopressors, improvement: N/R calcium - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: hyperglycemia Cohen et - Subject: 60 yo male Verapamil - Intervention: glucagon 3 mg, - Mortality: no al., 2011 - Comorbidities: hypertension, 4 mg followed by an infusion - Morbidity: N/R arthritis, asthma, diabetes, at 4 mg/h for 30 min - Hemodynamic hyperlipidemia, cardiac - Co-interventions: activated improvement: N/R disease, right total knee charcoal, HDI, calcium, fluids, - Functional status at replacement vasopressors discharge: N/R - Coingestion: trandolapril - Secondary outcomes: 2 - State at arrival: unstable days in hospital - Adverse effects: vomiting Doyon et - Subject: 73 yo female 2.88g of verapamil - Intervention: glucagon 5 mg - Mortality: no al., 1993 - Comorbidities: coronary followed by an infusion at 2-5 - Morbidity: N/R artery disease mg/h - Hemodynamic - Coingestion: temazepam - Co-interventions: activated improvement: yes - State at arrival: unstable charcoal, whole bowel - Functional status at irrigation, pacemaker, discharge: N/R vasopressors - Secondary outcomes: N/R - Adverse effects: hyperglycemia Fant et al., - Subject: 13 yo female (60kg) 120mg of - Intervention: glucagon - Mortality: no 2010 - Comorbidities: depression nifedipine - Co-interventions: gastric - Morbidity: N/R - Coingestions: clonidine lavage, activated charcoal, - Hemodynamic - State at arrival: stable atropine, calcium chloride, improvement: yes vasopressors, glucagon - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: vomiting, hyperglycemia Fung et al., - Subject: 45 yo female Nifedipine SR - Intervention: glucagon 3 mg, - Mortality: no 2007 - Comorbidities: hypertension, then 6 mg followed by an - Morbidity: N/R diabetes, hyperlipidemia, infusion at 9 mg/h - Hemodynamic chronic kidney disease - Co-interventions: activated improvement: yes - Coingestions: metoprolol, charcoal, fluids, calcium - Functional status at prazosin discharge: N/R - State at arrival: unstable - Secondary outcomes: N/R - Adverse effects: N/R Love et al., - Subject: 79 yo female Verapamil - Intervention: glucagon 5 mg - Mortality: no 1997 - Comorbidities: N/R followed by an infusion at 5 - Morbidity: N/R - Coingestion: timolol mg/h - Hemodynamic - State at arrival: unstable - Co-interventions: atropine improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: vomiting Mahr et al, - Subject: 71 yo female Diltiazem 20 mg - Intervention: glucagon 5 mg - Mortality: no 1997 - Comorbidities: cardiac bolus followed by followed by an infusion at 2 - Morbidity: N/R disease, congestive heart an infusion at 10 mg/h - Hemodynamic failure, peripheral vascular mg/h - Co-interventions: fluids, improvement: yes disease, diabetes, flutter calcium chloride, dopamine - Functional status at - Coingestions: none discharge: N/R - State at arrival: unstable - Secondary outcomes: N/R - Adverse effects: vomiting 238

Motokawa - Subject: 42 yo male 600mg of - Intervention: glucagon 3 mg - Mortality: no et al., 1993 - Comorbidities: hypertension, nifedipine followed by an infusion at 0.6 - Morbidity: N/R suicide attempt mg/h - Hemodynamic - Coingestions: metoprolol, - Co-interventions: gastric improvement: yes etizolam, ethanol lavage, calcium, vasopressors, - Functional status at - State at arrival: unstable methylprednisolone discharge: N/R - Secondary outcomes: 13 days in hospital - Adverse effects: N/R Papadopoul - Subject: 59 yo male 2.7g of nifedipine - Intervention: glucagon 1 mg - Mortality: no os et al., - Comorbidities: N/R every 3 min X 4 followed by an - Morbidity: N/R 2000 - Coingestions: N/R infusion at 4 mg/h - Hemodynamic - State at arrival: unstable - Co-interventions: gastric improvement: N/R lavage, activated charcoal, - Functional status at whole bowel irrigation, fluids, discharge: N/R calcium, vasopressors - Secondary outcomes: N/R - Adverse effects: N/R Wolf et al., - Subject: 30 yo female 3.6g of verapamil - Intervention: glucagon 14mg - Mortality: no 1993 - Comorbidities: N/R followed by an infusion at - Morbidity: pulmonary - Coingestions: N/R 5mg/h edema - State at arrival: unstable - Co-interventions: amrinone, - Hemodynamic atropine, calcium improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Lipid emulsion Akinci et al., - Subject: N/R 240mg verapamil - Intervention: lipid emulsion - Mortality: no 2013 - Comorbidities: N/R - Co-interventions: fluids, - Morbidity: N/R - Coingestions: trandolapril calcium, glucagon, - Hemodynamic - State at arrival: stable vasopressors improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Al-Nabran, - Subject: 39 yo female 4.08g of verapamil - Intervention: lipid emulsion - Mortality: no 2011 - Comorbidities: hypertension ER 20% 100ml over 20 min - Morbidity: N/R - Coingestions: N/R followed by an infusion at - Hemodynamic - State at arrival: unstable 0.5ml/kg/h X 8h improvement: yes - Co-interventions: activated - Functional status at charcoal, fluids, calcium, discharge: N/R glucagon, vasopressors - Secondary outcomes: N/R - Adverse effects: N/R Armenian et - Subject: 47 yo male 6g of verapamil - Intervention: lipid emulsion - Mortality: no al., 2010 - Comorbidities: N/R SR 20% 100 ml X 2 followed by - Morbidity: N/R - Coingestions: N/R an infusion of 500 ml over 30 - Hemodynamic - State at arrival: unstable min than another bolus of 100 improvement: yes ml followed by 150 ml over 15 - Functional status at min discharge: N/R - Co-interventions: pacemaker - Secondary outcomes: N/R - Adverse effects: hypertriglyceridemia Assink et - Subject: N/R Verapamil - Intervention: lipid emulsion - Mortality: N/R al., 2013 - Comorbidities: N/R - Co-interventions: CRRT and - Morbidity: N/R - Coingestions: N/R others - Hemodynamic - State at arrival: N/R improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Bologna et - Subject: 81 yo female 5,8g diltiazem - Intervention: lipid emulsion - Mortality: no al., 2013 - Comorbidities: hypertension, 20% 100 ml followed by an - Morbidity: N/R ischemic heart disease infusion at 0.5ml/kg/h X 12h - Hemodynamic - Coingestions: indapamid - Co-interventions: gastric improvement: yes - State at arrival: unstable lavage, activated charcoal, - Functional status at 239

fluids, calcium gluconate, discharge: N/R vasopressors - Secondary outcomes: N/R - Adverse effects: N/R Bouchard et - Subject: 50 yo male 500mg of - Intervention: lipid emulsion - Mortality: N/R al., 2010 - Comorbidities: depression, amlodipine 20% 100ml X 2 followed by an - Morbidity: hypoxia, hypertension, ethanol abuse infusion of 0.25ml/kg/min X 1h congestive heart failure, - Coingestions: lisinopril, - Co-interventions: fluids, acute kidney injury, sepsis hydrochlorothiazide vasopressors, calcium, - Hemodynamic - State at arrival: unstable glucagon, HDI, methylene improvement: yes blue, ECLS - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Chandrashe - Subject: 56 yo female 3.36g of verapamil - Intervention: lipid emulsion - Mortality: no kar et al., - Comorbidities: hypertension, ER 20% 100ml over 20 min - Morbidity: N/R 2011 atrial flutter followed by an infusion of - Hemodynamic - Coingestions: N/R 0.5ml/kg/h X 8h improvement: yes - State at arrival: unstable - Co-interventions: fluids, - Functional status at calcium, glucagon, discharge: N/R vasopressors - Secondary outcomes: N/R - Adverse effects: N/R Cooper et - Subject: 52 yo female 1.68g of diltiazem - Intervention: lipid emulsion - Mortality: no al., 2010 - Comorbidities: N/R SR 20% 500 ml over 30 min - Morbidity: aspiration - Coingestions: amitriptyline, - Co-interventions: fluids, pneumonia dihydrocodeine, temazepam, naloxone, atropine, calcium - Hemodynamic diazepam, citalopram gluconate, vasopressors, improvement: yes - State at arrival: unstable bicarbonate - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R French et - Subject: 47 yo male 6.3g of verapamil - Intervention: lipid emulsion - Mortality: no al., 2011 - Comorbidities: hypertension SR 20% 100 ml X 2 followed by - Morbidity: hypoxia, - Coingestions: N/R an infusion of 500 ml over 30 pleural effusion - State at arrival: unstable min and another bolus of 150 - Hemodynamic ml over 15 min improvement: yes - Co-interventions: atropine, - Functional status at calcium, glucagon, HDI, discharge: N/R vasopressors, pacemaker - Secondary outcomes: decrease serum CCB concentrations once lipid removed from the sample - Adverse effects: N/R Gonzales et - Subject: 47 yo female 350 mg - Intervention: lipid emulsion - Mortality: no al., 2012 - Comorbidities: amlodipine 20% 100 ml - Morbidity: N/R - Coingestions: ethanol - Co-interventions: activated - Hemodynamic - State at arrival: stable charcoal, fluids, calcium, improvement: yes vasopressors - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Liang et al., - Subject: 41 yo female 19.2g of verapamil - Intervention: lipid emulsion - Mortality: no 2011 - Comorbidities: depression SR 20% 100 ml over 20 min - Morbidity: pulmonary - Coingestions: N/R followed by an infusion at edema, ARDS, radial - State at arrival: unstable 0.5ml/kg/h for 3 days (total artery thrombosis, urinary 4.2L) tract infection, ischemic - Co-interventions: repeated colitis doses of activated charcoal, - Hemodynamic fluids (40.7L), vasopressors, improvement: yes calcium, HDI, glucagon, - Functional status at pacemaker, dialysis discharge: help with daily livings activity at 55 days post ingestion - Secondary outcomes: N/R - Adverse effects: N/R Meaney et - Subject: 47 yo female 350 mg - Intervention: lipid emulsion - Mortality: no 240

al., 2013 - Comorbidities: ethanol amlodipine 20% 100 ml followed by an - Morbidity: N/R abuse, psychiatric disorder, infusion at 100 ml/h increased - Hemodynamic hypertension, delirium to 500 ml/h (total 2.3L) improvement: yes tremens, post-traumatic - Co-interventions: activated - Functional status at disorder, mitral valve prolapse charcoal, fluids, calcium discharge: N/R - Coingestions: ethanol, gluconate, glucagon, - Secondary outcomes: 4 ziprasidone, pregabalin vasopressors, bicarbonates days in ICU and 8 days in - State at arrival: unstable hospital - Adverse effects: lipemia Montiel et - Subject: 18 yo female 3.6g of diltiazem - Intervention: lipid emulsion - Mortality: no al., 2011 - Comorbidities: none SR 20% 1.5ml/kg followed by an - Morbidity: MSSA - Coingestions: none infusion at 0.25mg/kg/h over bacteremia - State at arrival: unstable 1h - Hemodynamic - Co-interventions: fluids, improvement: yes calcium, HDI - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Oakes et - Subject: 47 yo female 3g of diltiazem ER - Intervention: lipid emulsion - Mortality: no al., 2009 - Comorbidities: bipolar - Co-interventions: lorazepam, - Morbidity: pancreatitis, disorder, hypertension calcium, atropine, liver dysfunction, acute - Coingestions: lamotrigine, vasopressors, HDI, glucagon, kidney failure citalopram milrinone, intra-aortic balloon - Hemodynamic - State at arrival: unstable pump, pacemaker, digoxin improvement: yes - Functional status at discharge: neurologically intact - Secondary outcomes: N/R - Adverse effects: unable to draw electrolytes for 8h, lipemia Orr et al., - Subject: 34 yo male Amlodipine - Intervention: lipid emulsion - Mortality: no 2010 - Comorbidities: N/R 20% 1.5 ml/kg X 3 followed by - Morbidity: acute kidney - Coingestions: diazepam, an infusion of 15 ml/kg/h for a injury temazepam, citalopram, total of 1L or 12ml/kg - Hemodynamic antifreeze, perindopril, - Co-interventions: fluids, improvement: yes doxazosin, codeine, vasopressors, ethanol - Functional status at paracetamol infusion, bicarbonate, dialysis discharge: neurologically - State at arrival: unstable intact on day 12 - Secondary outcomes: N/R - Adverse effects: N/R Stellpflug et - Subject: 30 yo female Diltiazem - Intervention: lipid emulsion - Mortality: no al, 2011 - Comorbidities: congestive 20% 1.5ml/kg followed by an - Morbidity: N/R heart failure, hypertrophic infusion of 1.5L over 1h - Hemodynamic cardiomyopathy - Co-interventions: fluids, improvement: yes - Coingestions: metoprolol, calcium, HDI - Functional status at amiodarone discharge: N/R - State at arrival: unstable - Secondary outcomes: N/R - Adverse effects: N/R St-Onge - Subject: 47 yo female 32.4g verapamil - Intervention: lipid emulsion - Mortality: no 2010 - Comorbidities: depression, 20% 70 ml followed by an - Morbidity: acute kidney hypothyroidism, chronic pain, infusion for 1 h injury, liver dysfunction, arrhythmias - Co-interventions: fluids, increased INR - Coingestions: none glucagon, calcium, atropine, - Hemodynamic - State at arrival: unstable vasopressors, bicarbonate, improvement: no HDI, pacemaker - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R West et al., - Subject: 71 yo female 135mg of - Intervention: lipid emulsion - Mortality: yes 2010 - Comorbidities: hypertension, amlodipine 20% 400ml bolus but total of - Morbidity: acute kidney emphysema, depression 2L given over 12.5h injury, pulmonary edema - Coingestions: none - Co-interventions: fluids, - Hemodynamic - State at arrival: unstable vasopressors, calcium, HDI, improvement: no blood transfusion - Functional status at discharge: N/R 241

- Secondary outcomes: N/R - Adverse effects: 3h without being able to obtain an ABG, SatO2, CBC and electrolytes. Extreme lipemia and hyponatremia. Wilson et - Subject: 57 yo female 360mg of diltiazem - Intervention: lipid emulsion - Mortality: no al., 2012 - Comorbidities: hypertension, SR 20% 1.5 ml/kg followed by an - Morbidity: N/R peripheral vascular disease, infusion at 0.25 mg/kg/min for - Hemodynamic smoking, ethanol abuse, a total of 8 ml/kg improvement: yes depression - Co-interventions: whole - Functional status at - Coingestions: bisoprolol, bowel irrigation, fluids, calcium discharge: N/R candesartan/hydrochlorothiazi chloride, glucagon, - Secondary outcomes: de, acetaminophen/caffeine, vasopressors N/R codeine, dimenhydrinate, - Adverse effects: N/R ethanol - State at arrival: unstable Young et al, - Subject: 32 yo male 13.4g of verapamil - Intervention: lipid emulsion - Mortality: no 2009 - Comorbidities: N/R SR 20% 100 ml over 20 min - Morbidity: N/R - Coingestions: levothyroxine, followed by an infusion of 0.5 - Hemodynamic bupropion, zolpidem, ml/kg/h X 24h improvement: yes quetiapine, clonazepam, - Co-interventions: gastric - Functional status at benazepril lavage, activated charcoal, discharge: no neurological - State at arrival: unstable fluids, vasopressors, calcium, deficit glucagon - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Levosimendan Osthoff et - Subject: 60 yo male 480mg of - Intervention: levosimendan - Mortality: no al., 2010 - Comorbidities: N/R verapamil 24mcg/kg followed by an - Morbidity: N/R - Coingestions: N/R infusion at 0.1 mcg/kg/h - Hemodynamic - State at arrival: unstable - Co-interventions: atropine, improvement: yes fluids, calcium, HDI, - Functional status at vasopressors discharge: N/R - Secondary outcomes: 5 days in hospital - Adverse effects: N/R Sencan et - Subject: 20 yo male 250mg of - Intervention: levosimendan - Mortality: no al., 2011 - Comorbidities: N/R amlodipine 0.2 mcg/kg/min - Morbidity: N/R - Coingestions: N/R - Co-interventions: - Hemodynamic - State at arrival: unstable dobutamine, vasopressors improvement: yes - Functional status at discharge: N/R - Secondary outcomes: 3 days in hospital - Adverse effects: N/R Teker et al., - Subject: 16 yo female 500mg of - Intervention: levosimendan - Mortality: no 2010 - Comorbidities: N/R amlodipine 12 mcg/kg/min for 10 min then - Morbidity: N/R - Coingestions: N/R 0.1 mcg/kg/min for 12h then - Hemodynamic - State at arrival: unstable 0.2 mcg/kg/min for 12h improvement: yes - Co-interventions: fluids, - Functional status at calcium chloride, discharge: N/R vasopressors, HDI - Secondary outcomes: 6 days in hospital - Adverse effects: N/R Phosphodiesterase inhibitor Goenen et - Subject: 30 yo female 3.6g of verapamil - Intervention: amrinone - Mortality: N/R al. - Comorbidities: N/R SR 1mg/kg bolus followed by an - Morbidity: pulmonary 1986 - Coingestions: N/R infusion at 6 mcg/kg/min edema - State at arrival: unstable - Co-interventions: activated - Hemodynamic charcoal, atropine, fluids, improvement: yes calcium, glucagon - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Link et al. - Subject: 40 yo patient 2.4g of verapamil - Intervention: enoximone - Mortality: N/R 2002 - Comorbidities: N/R - Co-interventions: - Morbidity: N/R - Coingestions: N/R decontamination, calcium, - Hemodynamic 242

- State at arrival: unstable vasopressors improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Sandroni et - Subject: 57 yo male 1.6g of verapamil - Intervention: enoximone 1 - Mortality: no al. - Comorbidities: depression, mg/kg followed by an infusion - Morbidity: N/R 2005 coronary artery disease - Co-interventions: gastric - Hemodynamic - Coingestion: atenolol lavage, activated charcoal, improvement: yes - State at arrival: unstable vasopressors - Functional status at discharge: N/R - Secondary outcomes: 15 days in ICU - Adverse effects: N/R Wolf et al. - Subject: 30 yo female 3.6g of verapamil - Intervention: amrinone - Mortality: N/R 1993 - Comorbidities: N/R SR 1mg/kg bolus followed by an - Morbidity: pulmonary - Coingestions: N/R infusion at 6 ug/kg/min edema - State at arrival: unstable - Co-interventions: activated - Hemodynamic charcoal, atropine, fluids, improvement: yes calcium, glucagon - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Methylene blue Agarawal et - Subject: 69 yo woman Amlodipine - Intervention: methylene blue - Mortality: no al, 2013 - Comorbidities: hypertension, 1 mg/kg over 10 min followed - Morbidity: no hypothyroidism, depression by an infusion at 1 mg/kg/h for - Hemodynamic - Coingestions: atenolol 10h improvement: yes - State at arrival: unstable - Co-interventions: fluids, - Functional status at calcium chloride, glucagon, discharge: N/R vasopressors, HDI, - Secondary outcomes: 4 defibrillation, pacemaker days in hospital - Adverse effects: N/R Jang et al., - Subject: 25 yo female 400mg of - Intervention: methylene blue - Mortality: no 2011 - Comorbidities: none amlodipine 2mg/kg over 20 min followed - Morbidity: no - Coingestions: (no other by an infusion at 1 mg/kg/h - Hemodynamic cardioactive medication) - Co-interventions: activated improvement: yes - State at arrival: unstable charcoal, fluids, calcium - Functional status at gluconate, glucagon, discharge: no sequelae vasopressors, HDI - Secondary outcomes: 6 days in hospital - Adverse effects: N/R Kim et al., - Subject: 69 yo female Amlodipine - Intervention: methylene blue - Mortality: no 2012 - Comorbidities: hypertension, 1 mg/kg bolus over 10 min - Morbidity: N/R depression followed by 2 5-hour infusions - Hemodynamic - Coingestions: atenolol, of 2 mg/kg/h 8h apart improvement: yes thioridazine - Co-interventions: fluids, - Functional status at - State at arrival: N/R glucagon, calcium chloride, discharge: N/A vasopressors, HDI, - Secondary outcomes: pacemaker N/R - Adverse effects: N/R Autres St-Onge et - Subject: 68 yo male 300 mg - Intervention: L-carnitine 1g IV - Mortality: no al., 2013 - Comorbidities: hypertension, amlodipine Q4h - Morbidity: no type II diabetes, benign - Co-interventions: fluids, - Hemodynamic prostate hypertrophy, chronic calcium gluconate, HDI, improvement: yes anemia vasopressors, bicarbonates - Functional status at - Coingestions: metformin, discharge: N/R ethanol - Secondary outcomes: - State at arrival: unstable N/R - Adverse effects: N/R Ezidiegwu - Subject: 22 yo female 425mg of - Intervention: plasma - Mortality: no et al., 2008 - Comorbidities: none amlodipine exchange 1 plasma volume - Morbidity: acute kidney - Coingestions: N/R daily for 2 days injury, right foot ischemia - State at arrival: unstable - Co-interventions: gastric and amputation lavage, whole bowel irrigation, - Hemodynamic calcium, glucagon, improvement: yes vasopressors, HDI, - Functional status at hemodiafiltration discharge: N/R 243

- Secondary outcomes: decrease in serum CCB concentrations, 30 days in hospital - Adverse effects: N/R Maclaren et - Subject: 45 yo female 7.2g of verapamil - Intervention: plasmapheresis - Mortality: no al., 2005 - Comorbidities: N/R SR - Co-interventions: activated - Morbidity: N/R - Coingestions: doxepin, charcoal, fluids, dobutamine, - Hemodynamic quetiapine, diazepam, vasopressors, glucagon, improvement: yes temazepam, clonazepam calcium, HDI, - Functional status at - State at arrival: stable hemodiafiltration, ECLS discharge: N/R - Secondary outcomes: 8 days in ICU - Adverse effects: N/R Anthony T - Subject: 49 yo female 1.2g diltiazem - Intervention: ipecac - Mortality: no et al., - Comorbidities: coronary - Co-interventions: atropine, - Morbidity: N/R 1986123 bypass glucagon, calcium gluconate, - Hemodynamic - Coingestions: metoprolol, calcium chloride, pacemaker, improvement: fall in blood ethanol vasopressors pressure at initiation then - State at arrival: stable improvement - Functional status at discharge: no sequelae - Secondary outcomes: N/R - Adverse effects: N/R Luomanmäk - Subject: 22 yo male 3.6g of diltiazem - Intervention: charcoal - Mortality: no i et al., 1997 - Comorbidities: none SR hemoperfusion - Morbidity: aspiration - Coingestions: N/R - Co-interventions: gastric pneumonia - State at arrival: unstable lavage, fluids, calcium, - Hemodynamic vasopressors, dobutamine, improvement: yes pacemaker - Functional status at discharge: no neurological or cardiac deficit - Secondary outcomes: no decrease in serum CCB concentrations - Adverse effects: N/R Roberts et - Subject: 19 yo male 3.6g of diltiazem - Intervention: charcoal - Mortality: N/R al., 2008 - Comorbidities: schizophrenia SR hemoperfusion - Morbidity: N/R - Coingestions: atenolol, - Co-interventions: activated - Hemodynamic venlafaxine, irbesartan- charcoal, atropine, calcium, improvement: yes hydrochlorothiazide, glucagon, vasopressors, - Functional status at clozapine, sertraline bicarbonate, metaraminol discharge: N/R - State at arrival: unstable - Secondary outcomes: increased clearance - Adverse effects: N/R Rosansky et - Subject: 48 yo male 480mg of - Intervention: charcoal - Mortality: yes al., 1991 - Comorbidities: hypertension verapamil daily X hemoperfusion - Morbidity: N/R - Coingestion: cimetidine 4 days - Co-interventions: fluids, - Hemodynamic - State at arrival: stable atropine, calcium, improvement: no vasopressors, furosemide, - Functional status at pacemaker discharge: N/R - Secondary outcomes: decrease in serum CCB concentrations - Adverse effects: N/R Goplani et - Subject: 28 yo female 250mg of - Intervention: SLED - Mortality: no al., 2011 - Comorbidities: none amlodipine - Co-interventions: gastric - Morbidity: N/R - Coingestions: N/R lavage, fluids, vasopressors, - Hemodynamic - State at arrival: unstable calcium, glucagon, improvement: N/R dexamethasone - Functional status at discharge: back to baseline - Secondary outcomes: N/R - Adverse effects: N/R Harvey et - Subject: 57 yo male 960mg diltiazem - Intervention: CVVHDF - Mortality: no al., 2001 - Comorbidities: hypertension, SR and 120mg - Co-interventions: activated - Morbidity: N/R angina, hyperlipidemia, nifedipine charcoal, calcium, glucagon, - Hemodynamic coronary artery disease vasopressors improvement: yes - Coingestions: bisoprolol, - Functional status at 244

aspirin discharge: N/R - State at arrival: unstable - Secondary outcomes: N/R - Adverse effects: N/R Pfaender et - Subject: 45 yo female Nifedipine - Intervention: CVVHDF - Mortality: no al., 2008 - Comorbidities: hypertension, - Co-interventions: activated - Morbidity: liver liver cirrhosis, stroke, right charcoal, whole bowel dysfunction, acute kidney ophthalmoplegia, psychiatric irrigation, fluids, atropine, injury, rhabdomyolysis disorder vasopressors, calcium - Hemodynamic - Coingestions: gluconate, HDI, glucagon improvement: N/R benzodiazepines, warfarin, - Functional status at atenolol discharge: back to - State at arrival: unstable baseline - Secondary outcomes: 21 days in ICU, 32 days in hospital, more extraction of atenolol from the blood - Adverse effects: N/R Raham et - Subject: 38 yo female 7.4g of diltiazem - Intervention: CVVHDF - Mortality: no al., 2009 - Comorbidities: N/R ER - Co-interventions: activated - Morbidity: ARDS, acute - Coingestions: N/R charcoal, atropine, calcium kidney injury - State at arrival: unstable gluconate, glucagon, HDI, - Hemodynamic vasopressors, HFO improvement: N/R - Functional status at discharge: N/R - Secondary outcomes: 11 days in ICU, 39 days in hospital - Adverse effects: N/R Belleflamme - Subject: 70 yo female Diltiazem SR - Intervention: MARS - Mortality: no et al., - Comorbidities: N/R - Co-interventions: calcium, - Morbidity: N/R 2012122 - Coingestions: N/R vasopressors, HDI, intra-aortic - Hemodynamic - State at arrival: unstable balloon pump improvement: yes - Functional status at discharge: N/R - Secondary outcomes: decrease un serum CCB concentrations - Adverse effects: hypoglycemia Deters et - Subject: 33 yo female 600mg felodipine - Intervention: MARS - Mortality: yes al., 2010 - Comorbidities: N/R - Co-interventions: calcium, - Morbidity: compartment - Coingestions: N/R glucagon, HDI, lipid emulsion syndrome - State at arrival: unstable - Hemodynamic improvement: yes - Functional status at discharge: N/R - Secondary outcomes: N/R - Adverse effects: N/R Janion et - Subject: 17 yo female Verapamil and - Intervention: IABP - Mortality: no al., 2010 - Comorbidities: N/R amlodipine - Co-interventions: - Morbidity: N/R - Coingestion: metoprolol vasopressors, dobutamine - Hemodynamic - State at arrival: stable improvement: mild - Functional status at discharge: back to baseline - Secondary outcomes: 56 days in hospital - Adverse effects: N/R Frierson et - Subject: 57 yo female Verapamil SR - Intervention: IABP - Mortality: no al., 1991 - Comorbidities: - Co-interventions: atropine, - Morbidity: N/R hyperlipidemia, hypertension fluid, vasopressors, - Hemodynamic - Coingestions: N/R dobutamine, pacemaker, improvement: back to - State at arrival: unstable baseline - Functional status at discharge: no neurologic deficit - Secondary outcomes: 7 days in hospital - Adverse effects: N/R 245

Laes et al., - Subject: 58 yo female Amlodipine - Intervention: Impella X 10 h - Mortality: no 2013 - Comorbidities: N/R - Co-interventions: HDI 10 - Morbidity: N/R - Coingestions: zolpidem U/kg/h, calcium gluconate, - Hemodynamic - State at arrival: stable vasopressors improvement: yes - Functional status at discharge: neurologically intact - Secondary outcomes: N/R - Adverse effects: arterial tear requiring surgical repair

246