PRIMARY CARE MANAGEMENT OF THE HEART FAILURE PATIENT
REQUIRING AN IMPLANTED CARDIOVERTER DEFIBRILLATOR
by
Valerie Murphy Moore
BScN, Thompson Rivers University, 2011

PROJECT SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF MASTER OF SCIENCE IN NURSING: FAMILY NURSE
PRACTITIONER

UNIVERSITY OF NORTHERN BRITISH COLUMBIA
August 2019
ã Valerie Murphy Moore, 2019

iii
Abstract
Heart failure is a chronic, progressive disease that remains on the rise partially due to
improvements in preventing cardiac related mortality from other causes. Implanted cardioverter
defibrillators are indicated for patients that are at high risk for sudden cardiac death from
ventricular arrhythmias (primary prevention), or for those who have survived a cardiac death
(secondary prevention). Guidelines suggest simple and easy to follow directions for providers to
determine which patients would be eligible to be considered for a device. However, despite
national guidelines, research supporting cost effectiveness, and a repeatedly demonstrated
reduction in mortality with device therapy, referral and utilization of ICDs remain low, with most
of the non-referral and underutilization rates identified within the primary care provider group. In
addition, for those patients who do receive ICDs, advanced care planning with consideration of
deactivation of the device at end-of-life is rarely completed, leaving many patients at risk of
undignified and painful shocks delivered within the last hours to days of life.
This integrative review seeks to explore: Primary care management of the heart failure
patient requiring an implanted cardioverter defibrillator. A comprehensive search of the
literature was undertaken and 15 articles were selected for inclusion This review establishes that
ICDs are effective and relatively low-risk devices that have significant mortality benefits, and
patients with HF should be screened routinely for eligibility for the device in order to help
prevent SCD. Furthermore, patients who receive an ICD should complete an informed advance
care directive which includes directions for management of the ICD at end-of-life. ICD
management, while often deferred to specialists, requires a collaborative approach and consistent
evaluation from the patient’s primary care provider, as it has the potential to influence all aspects
of patient’s quality of life, as well as quality of end-of-life care. Recommendations for practice,

iv
education, further research, and policy have been made to support the role of primary care
providers, such as Nurse Practitioners, in the management of HF patients requiring an ICD.

v
Table of Contents
Abstract

ii

Table of contents

v

List of tables and figures

viii

Acknowledgements

ix

Chapter I: Introduction and Background

1

Introduction
Background
Epidemiology
Incidence and prevalence
Economic burden
Pathophysiology
Chambers of the heart
Flow of blood through the heart
Conduction system of the heart
Heart Failure
Etiology
Risk factors
Ventricular remodeling
Neurohormonal activation
Arrhythmias
Signs and Symptoms of Heart Failure
Signs
Symptoms
Diagnosis of Heart Failure
BNP
Electrocardiogram
Chest X-ray
Echocardiogram
Stress test
Coronary angiogram
MRI
Management of Heart Failure
Pharmacology
Lifestyle management
Device therapy
Left ventricular assist device
Cardiac transplantation
What is an Implantable Cardioverter Defibrillator?
History

1
3
4
5
5
6
6
7
7
8
8
9
9
9
11
11
12
12
13
13
13
14
14
15
15
15
16
16
18
18
20
20
21
21

vi
Components and function
Guideline Indications for Use
Primary prevention
Secondary Prevention
Implantation
Subcutaneous ICDs
Providers
The NP Role in Primary Care for Patients with Heart Failure

21
22
22
23
24
24
25
25

Chapter II: Methodology
Integrative Literature Review
Literature Search
Inclusion and Exclusion Criteria
Preliminary Search
Focused Search
Grey Literature
Final Results

28
28
28
28
31
32
33
33

Chapter III: Findings
Low Rates and Underutilization of ICDs
Primary Care Provider Knowledge of Indications for Implantable Cardioverter
Defibrillator
Advanced Care Planning and Deactivation of ICDs
Summary

34
34
42

Chapter IV: Discussion
Underutilization of ICDs
Influencing Factors
Provider Perceptions
Guideline Discordant Practice
Provider Knowledge
Provider Education and Knowledge Translation
Advanced Care Planning
Advanced Care Directives
Timing and Prognostication
Deactivation of the Device
Recommendations
Education and Continued Professional Development
Practice
Implantation and referral
Deactivation and end-of-life
Research
Policy
Limitations
Conclusion

59
60
61
62
65
67
69
70
71
72
72
75
76
78
78
79
80
81
82
84

50
58

vii
References
Appendices
Appendix A: Literature Flow Diagram
Appendix B: Literature Review Matrix
Qualitative Research Studies and Systematic Reviews
Primary Research Studies
Appendix C: Nine Essential Elements for Shared Decision-making
(Makoul & Clayman, 2006)
Appendix D: Quality Assessment Tool

86
93
94
94
101
115
116

viii
List of Tables and Figures
Table 1: New York Heart Association Functional Classification and Other Symptom
Descriptors
Table 2: Inclusion and Exclusion Criteria for Integrative Literature Review

16

Table 3: Search Terms for the Integrative Literature Review

31

Table 4: Reasons for Non-Utilization of ICD therapy

36

Table 5: Reasons for Non-Referral for ICD

40

Figure 1: Referral Pathway for Device Therapy in Patients with HF

24

Figure 2: Baseline Characteristics

38

30

ix
Acknowledgements
I would like to extend my sincerest appreciation to my supervisory committee, Dr.
Davina Banner-Lukaris (RN, PhD) and Liz Mulvaney NP (F) for their contributions and support
to this project. Your guidance, encouragement, and passion for cardiac care motivated me to
make this project valuable for us all.
Thank you to my husband, Chris, for taking on everything in our lives while I worked on
this paper. Your dedication to our family and support of my goals is remarkable. To my children
Grayson and Hunter, thank you for being so strong and mature beyond your years. Jolene, your
un-wavering reinforcement of my ambitions were priceless, and I don’t know what I would have
done without you always there to pick me up when I needed help. Mom, Dad, Mike and Sharon,
thank you for keeping life normal in the background of my schooling, and keeping me grounded
in the things that truly matter. Finally, thank you to my exceptional UNBC faculty and peers who
set a standard of excellence to strive for while entering this new chapter of my career.
This project is dedicated to my father-in-law, Gary Moore, who understood the passion
for continuing education, and truly believed in me to pursue this profession. Thank you for your
certainty in my abilities and your undisputed confidence for my future

1

CHAPTER I: INTRODUCTION
“The provider must do everything in his power to save lives, preserve health, or at least
alleviate the suffering”
-Hippocrates
Heart failure (HF) is a chronic, progressive disease with a high rate of morbidity and
mortality. Currently in Canada, at least 600,000 people are affected by this disease, and face
marked impairments in their quality of life along with a poor disease trajectory. While this
population with HF will likely have interactions with cardiologists and other specialists involved
in the variety of needs for this disease, >90% of the care of HF patients is the responsibility of
the primary care provider (Virani et al., 2017). Patients with HF are especially susceptible to
sudden cardiac death (SCD), which in Canada is a leading cause of death. SCD affects both men
and women, and at a rate of approximately 40,000 people per year (Parkash & Tang, 2017).
Implantable cardioverter defibrillators (ICDs) are a common, nationally recommended therapy to
prevent SCD as both primary and secondary prevention in certain patients with HF (Bennett et
al., 2017; Parkash & Tang, 2017). Despite Class I recommendations from national guidelines and
evidence for significant survival benefit, referral for and utilization of this device remains
alarmingly low (Parkash & Tang, 2017).
The Canadian Cardiovascular Society (CCS) updated their ICD guidelines in 2016 to
provide comprehensive updated indications for addressing ICD use in primary and secondary
prevention (Bennett et al., 2017). These guidelines are intended to be utilized by almost all care
providers available in order to provide congruent and collaborative care for these at-risk patients,
and a simple checklist is offered for use as a screening tool to assist clinicians in practice. The

2
ability to increase awareness and understanding of the appropriate indications for device referral
and implantation is an important knowledge and practice gap today in Canada. With only 2530% of patients eligible for the device being referred, there is a large population of HF patients at
risk of SCD that will never have the opportunity to benefit from this therapy. As a primary care
provider, it is important to be aware of opportunities for screening and where to find the
clinically applicable resources to guide these decisions. Guidelines provide current, up-to-date
recommendations for dissemination of research and knowledge and can be simplified to use
quickly and efficiently in busy, primary care practice. This being said, it is also important to
ensure evidence-based knowledge is sufficient to avoid overutilization of ICD therapy in a
population that may not benefit. Barriers for underutilization must be addressed at the individual
provider, hospital, community and regional levels (Ho et al., 2017).
In addition to the concerns discussed above regarding poor referral rates and
underutilization of ICDs, the use of advanced care directives (ACDs) and the need to have clear
instructions for management of the device at end-of-life has been a crucial focus of many
researchers in the past decade. This concern has arisen from a sizable body of literature
documenting clinical experiences of undignified, painful, and anxiety provoking shocks in
patients who were hours to days away from death. These shocks are ultimately avoidable for
patients, when interventions and communication with the primary care provider takes place early
on in the device therapy pathway, and is an identified area for quality improvement of
compassionate, end-of-life care (Javaid, Squirrell, & Farooqi, 2018).
Since its inception, the ICD has offered substantial and constantly evolving options for
the treatment of patients with HF at risk of SCD. With the increasing population experiencing
HF, updated guidelines, and national efforts to improve multidisciplinary care, primary care

3
providers are a prime candidate to offer exceptional and comprehensive care for the patient with
an ICD. As such, this integrative review has been performed in order to explore: Primary care
management of the HF patient requiring an ICD.
Background
Heart failure (HF) is a clinical syndrome that occurs when the heart does not function
adequately to pump blood at an appropriate velocity and volume to meet the metabolic demands
of the body (Lilly, 2016). This results in the classic problems of shortness of breath, fatigue, and
volume overload. Lilly (2016) considers HF to be “the most severe manifestation of nearly every
form of cardiac disease, including coronary atherosclerosis, myocardial infarction, valvular
diseases, hypertension, congenital heart disease, and the cardiomyopathies”, (p. 220). With this
in mind, it can be understood that with longer life spans and improvements in life prolonging
cardiac interventions, HF incidence and prevalence is increasing and is one of the leading causes
of morbidity and mortality worldwide (Metra & Teerlink, 2017). Furthermore, it is the last
clinical syndrome of cardiovascular disease that continues to rise in incidence and is projected to
rise significantly over the coming decades (DiSalvo, 2005). Diagnosis of HF is multifaceted and
includes clinical signs and symptoms, objectively obtained evidence of cardiac dysfunction
through imaging studies, as well as markers of neurohormonal compensation (Mosterd & Hoes,
2007). In addition, a timely diagnosis is essential to facilitate early intervention of therapies, as
HF typically is given a prognosis of a 50-percent five-year mortality rate (Stevenson & Zei,
2005).
The last decade of HF care has shown a great shift towards disease management
interventions that encompass integration of multiple care providers, access to both specialist and
follow-up appointments, and optimization pharmacologic and device therapies. These

4
interventions are essential to assist in improving HF survival rates, outcomes, quality of life, and
decreasing hospital admission rates (Virani et al., 2017). HF was recognized in 2013 by the
Public Health Agency of Canada as a national public health priority, leading to a strategic plan to
reduce the number of hospital related admissions of the disease. This recognition assisted in
furthering the development of the HF care model in Canada to include more primary care
resources; although this action is still strongly supported, the need for a national coordinated
strategy to ensure a more standardized approach to HF care is emphasized by stakeholders of HF
management (Hayes et al., 2015).
Epidemiology
A consensus on a definition of HF has been difficult due to the complex interplay of
alterations in multiple systems that contribute to the development of the syndrome (Dec, 2005).
A broad understanding of the concept however, can be described as a failure of the heart to
produce sufficient pressure-volume work to meet the physical/metabolic demands of organ and
tissue perfusion exclusively from a cardiac cause (Dec, 2005; Lilly, 2016). Two main
considerations in diagnosing HF for management include determining if it is an acute versus
chronic condition, as well as the identification of either systolic or diastolic dysfunction, which
will be described further in the section addressing the pathophysiology and clinical
manifestations.
Chronic HF is the more commonly encountered syndrome, which in itself can have
episodes of acute-on chronic problems as well as decompensation of HF (Mosterd & Hoes,
2007). Acute HF typically presents with some of the classic signs and symptoms of HF such as
dyspnea and pulmonary edema within 24 hours of an insult to the myocardium, as in acute
coronary syndromes (ACS) and cardiogenic shock.

5
HF has high rates of mortality, as well as high rates of readmission to hospital which
subsequently increase the mortality rates. According to the Report from the Canadian Chronic
Disease Surveillance System looking at heart disease in Canada (2018), men over the age of 65
with a diagnosis of HF had a higher all-cause mortality rate than women of the same age
diagnosed with HF. Women and men aged 40-54 years-of-age diagnosed with HF were 27.3 and
16.5 percent more likely to die of any cause compared to those without the disease.
Incidence and prevalence. HF is grouped within the definition of heart disease
nationally, and heart disease reigns as the second leading cause of death in Canada; cancer being
the first. In Canada, during the surveillance years of 2012-2013, the age-standardized incidence
rate of HF for both men and women aged 40 years and older was 3.4%, or approximately
670,000 individuals. 71% of these individuals also suffered from ischemic heart disease. The
incidence rate for HF in the same population in BC was 3.5%, higher than the Canadian average
(Public Health Agency of Canada, 2016).
Economic burden. HF poses a marked financial impact not only to the sufferers, but
extends to caregivers and healthcare systems when it comes to the direct and indirect costs of the
disease, such as a loss of income, employment revenue, and both psychological and
physiological stressors (Viriani et al., 2017). Canada wide, the direct costs of HF were estimated
to be $2.8 billion dollars annually (Viriani et al., 2017). It remains one of the leading causes of
admission to acute care, as well as holding staggering rates of 30-day readmissions (above 20%).
In addition, when a patient with HF is admitted to hospital, their length of stay and overall
complexity of care increase the strain of the costs associated with the disease (Virani et al.,
2017). The pathophysiology of HF will now be discussed as a means of expanding the context
and application of this capstone project.

6
Pathophysiology
Normal cardiac function is essential to sustain cellular metabolism, remove waste
products from the body, and deliver essential requirements such as nutrients, oxygen, hormones
and other substances to tissues (McCance & Huether, 2014). The heart works in collaboration
with other systems of the body including the endocrine, lymphatic, and nervous systems, and
indirectly with the digestive, respiratory and renal systems to maintain normal body functions.
The actual heart can be considered as two pumps; one that serves to deliver blood through the
entire body (the systemic circulation or left heart), and the other that works to pump used blood
through the lungs to obtain oxygen (the pulmonic circulation or right heart) (McCance &
Huether, 2014).
Chambers of the heart. Understanding the anatomical structure of the heart is
valuable when relating to the dysfunctions that occur during heart disease and HF. The heart has
four chamber, two upper chambers called atria (singular is atrium), and two lower chambers
called ventricles. The atria are smaller and have thinner walls, and act as collecting chambers for
return of blood flow from the systemic and pulmonic circulation. The ventricles are much thicker
with a strong myocardial layer responsible for pumping functions, the left larger than the right.
The right atrium is separated from the right ventricle by a one-way valve known as the tricuspid
or right atrioventricular valve. The left atrium is separated from the left ventricle by the mitral or
left atrioventricular valve. The right ventricle must push blood through the bicuspid semilunar
valves into the pulmonic artery and through the low-pressure circulation of the lungs. The left
ventricle must push blood through the typically tricuspid semilunar valves of the aorta and
through the high-pressure systemic circulation (McCance & Huether, 2014).

7
Flow of blood through the heart. Normal blood flow is unchanged after birth and
through the lifespan. Beginning in the right ventricle, unoxygenated blood is pumped through the
semilunar valves into the right and left pulmonary arteries (note this is the only artery in the body
that carries unoxygenated blood). The pulmonary artery is the pathway to the arteries and
arterioles of the lungs, and to the capillaries of the lungs where fresh oxygen is diffused into the
blood. Next, the now oxygenated blood moves through the venules and veins of the lungs, into
the pulmonary veins. The pulmonary veins lead back to the left side of the heart into the left
atrium. The oxygenated blood is then passed through the mitral valve into the left ventricle. The
large demand of the left ventricle requires it to pump against the high-pressure gradients of the
aortic valve, aorta and the systemic circulation. The arteries of the organs systems move blood
into the capillaries of each organ which allows diffusion of oxygen and nutrients into the tissues,
and removal of waste and carbon dioxide. From here, the blood is moved into the venules and
veins of the organs, through the superior and inferior vena cava’s leading into the right side of
the heart; the right atrium, through the tricuspid valve and returning to the right ventricle
(McCance & Huether, 2014).
Conduction system of the heart. While the body is dependent on the heart for its
pumping function, the heart is dependent on the stimulation from its own specialized conduction
system to generate and transmit action potentials through the myocardium to trigger contraction
and relaxation. The cardiac conduction system is unique in that it does not require stimulation
from the nervous system to perform its impulses; however, it is innervated and influenced by
stimuli from the sympathetic and parasympathetic branches of the autonomic nervous system.
The ability of the heart to spontaneously depolarize without external stimuli is known as
automaticity. Normal conduction in the cardiac cycle begins with the pacemaker of the heart,

8
known as the sinoatrial (SA) node, which is located just above the tricuspid valve at the meeting
of the superior vena cava in the right atrium. Its location is very near the visceral pericardium
which makes it very susceptible to influence from physical stimuli such as injury or pericardial
inflammation. When an action potential originates in the SA node, it travels through the left and
right atria and causes them to contract. It then passes to the atrioventricular or AV node, that is
located in the right atria near the tricuspid valve. The conduction that goes through the AV node
then enters the bundle of His, located in the interventricular septum between the two ventricles.
The bundle of His splits into two branches, the right and left bundle branches that serve to help
depolarize the ventricles. Depending on the health and physical structure of the muscle in the
ventricles, these branches can be stimulated or interrupted, causing changes to the conduction
pathway. This is an important concern for the patient with structural changes during HF, and part
of the considerations for risk of malignant arrhythmias and need for ICD therapy. The right and
left bundle branches terminate at the Purkinje fibers (McCance & Huether, 2014).
Heart Failure
Etiology. The etiologies of HF can be classified into two main categories which are
based on the normal or abnormal systolic function of the left ventricular ejection fraction (EF).
HF with reduced EF (HFrEF) is primarily caused by systolic dysfunction. Systolic dysfunction
can result from either impaired contractility or increased afterload, both of which will be
discussed further. HF with preserved EF (HFpEF) is primary caused by diastolic dysfunction that
occurs when there is impairment during the filling process of diastole (Lily, 2016). The most
common cause of HF is ischemic heart disease, followed by hypertension and diabetes; other
causes to consider in conjunction or absence of these causes include the cardiomyopathies, viral

9
or bacterial infection, toxins such as illicit or therapeutic drugs and alcohol, valvular disease, and
arrhythmias (Kemp & Conte, 2012).
Risk factors. Other risk factors include age (there is a rising prevalence of HF with
aging), renal failure, obesity, pregnancy, pulmonary disease, cancers, and other endocrine
disorders (Brashers, 2014). An important consideration of HF is that despite the fact it remains a
heterogeneous condition with varying etiologies, almost all incidences result in the similar
pathway that results in neurohormonal activation as a paramount role in the progression of the
disease, and remains a target of pharmacologic therapy (Metra & Teerlink, 2017).
Ventricular remodeling. Ventricular remodeling is a unique process within the heart
that initially begins as a compensatory mechanism in response to increased hemodynamic
burdens. During the development of HF, the wall stress that accompanies left ventricular
dilatation and chamber radius in order to overcome preload or the need to overcome higher
afterload as in high blood pressure or aortic stenosis causes changes at the cellular level within
the myocardium. These changes, along with the neurohormonal responses which will be
described, stimulate hypertrophic growth of the myocardium along with deposition of
extracellular matrix, contributing to increased mass and thickness. This initial compensation, in a
chronic form, results in increased diastolic ventricular pressures, putting strain on the pulmonic
vasculature and left atrium (Lilly, 2016). If these changes are not slowed or stopped, ventricular
function and compliance will further deteriorate resulting in dilation of the chambers, and the
progression of HF.
Neurohormonal activation. One of the major influences in both early compensation of
reduced cardiac output and ongoing management of HF is the role of neurohormones. The
activity of these neurohormones serve to increase mean arterial pressure (MAP) and are triggered

10
by decreased cardiac output. The adrenergic nervous system, the renin-angiotensin-aldosterone
system (RAAS), and an increase in production of antidiuretic hormone (ADH) serve to increase
the systemic vascular resistance (SVR). When the SVR is increased, vital organs can still be
perfused even when cardiac output is low (Lilly, 2016).
The adrenergic nervous system is activated in HF when baroreceptors in the aortic arch
and carotid sinus sense a fall in cardiac output. When these receptors rate of firing decreases with
the concordant fall in pressure, the 9 th and 10th cranial nerves (glossopharyngeal and vagus)
deliver signals to the cardiovascular control center in the medulla. The medulla stimulates the
sympathetic nervous system (SNS) and acts with a subsequent release of the catecholamines
epinephrine and norepinephrine. These hormones directly affect the heart and peripheral
circulation, resulting in an increased heart rate, increased ventricular contractility, and
vasoconstriction of the peripheral vasculature (Kemp & Conte, 2012; Lilly, 2016). While these
results can be initially beneficial (increased blood return to the heart, increased stroke volume
and therefore increased cardiac output, maintenance of blood pressure and perfusion of vital
organs), chronic adrenergic overstimulation of the SNS will be detrimental to the heart. The
ongoing activation of b1, b2 and a1 may lead to tachycardia, arrhythmias, and decreased EF
within the heart, and subsequently, can lead to further activation of the RAAS (Kemp & Conte,
2012).
The RAAS system plays an important role in HF as management includes ongoing
attempts to reduce its impact. The kidneys are stimulated to release renin from the
juxtaglomerular cells in response to three major incidents: 1) reduced renal blood flow from a
decrease in cardiac output, 2) stimulation of the b-receptors in the juxtaglomerular cells from the
SNS and 3) alterations in intrarenal hemodynamics due to decreased delivery of salt to the

11
macula densa of the kidney (Lilly, 2016). The role of renin is beneficial in the role of
maintenance of systemic blood pressure in the event of massive hemorrhage or dehydration. The
enzymatic role of renin is to cleave the propeptide angiotensinogen in order to form angiotensin
1, which is then immediately cleaved of two amino acids by the endothelial cell-bound
angiotensin converting enzyme (ACE), resulting in the formation of angiotensin II. Angiotensin
II acts primarily on the vascular endothelium, causing potent vasoconstriction in an attempt to
restore a decline in blood pressure. Chronic angiotensin II activity, however, results in in
increased myocardial demand, hypertrophy of the left ventricle, and an overall decrease in
cardiac output. In addition, the effect of angiotensin II on the hypothalamus stimulates thirst and
therefore water intake, as well as increases aldosterone secretion at the adrenal cortex (Lilly,
2016, (McNamara, 2005).
Arrhythmias. The progression of HF leaves the myocardial walls prone to structural
changes and irritability within the electrical conduction system. As the disease becomes more
chronic, examples of arrhythmias such as atrial fibrillation (although this is also a cause of some
tachycardia-induced HF), ventricular tachycardia’s, ventricular fibrillation, and ventricular
arrhythmias that could result in SCD require immediate attention and treatment.
Signs and Symptoms of Heart Failure
Evaluation of signs and symptoms of HF is essential for management; however, they are
insufficient to produce an accurate diagnosis. The range of severity of signs and symptoms
depend not only on extend of HF, but also the chronicity of the condition and other related
comorbid disorders. The initial symptoms experienced by a majority of patients in HF are often
mistaken for those of pulmonary dysfunction, deconditioning, or viral illness (Metra & Teerlink,
2017).

12
Signs. Objective signs of HF include the following encompass cardiac and systemic
systems. Cardiac signs can include tachycardia from the increased SNS activation, holosystolic
murmurs from mitral or tricuspid valve regurgitation, a displaced apical beat due to left
ventricular dilatation, additional heart sounds (S3 or S4 gallop), elevated jugular venous pressure
(also called jugular venous distention or reflux) as a result of increased right atrial pressure, and
alterations in peripheral pulse pressures known as pulsus alternans. Mitral regurgitation occurs in
left-sided HF when the left ventricle dilates enough to spread the papillary muscles apart from
one another, increasing the size of the valve annulus, preventing the mitral leaflets from proper
closure. Tricuspid regurgitation occurs in right sided HF in a similar manner (Lilly, 2016). The
early diastolic sound S3 is associated with systolic HF, and results from abnormal filling in the
dilated chamber of the left ventricle. S4, the late diastolic sound, results from the atria forcefully
contracting against a stiff ventricle during diastolic dysfunction. Pulmonary signs include those
related to fluid retention, congestion and increased pressures. Pulmonary crackles created by air
moving through the small airways filled with fluid due to increased left atrial pressure, pleural
effusions can develop in left or right-sided HF and are identified by percussing the posterior lung
bases and finding dullness. Systemic signs are a result of combinations of SNS activation, fluid
retention, increased heart chamber pressures and increased metabolic demands. They can include
hepatomegaly, ascites, peripheral edema, diaphoresis, cachexia, peripheral and central cyanosis
(DiSalvo, 2017; Lilly, 2016; Metra & Teerlink, 2017).
Symptoms. While it is important to re-emphasize that no single symptom is pathognomic
for HF, the classic symptoms for HF include dyspnea due to lung congestion and increased
pulmonary/cardiac pressures, orthopnea as a result of lung congestion and venous overload in the
supine position, and paroxysmal nocturnal dyspnea (PND) due to a similar mechanism of

13
orthopnea in addition to depression of the respiratory center. The latter of the three symptoms
have the highest specificity for HF (Dec, 2005). Abdominal bloating, fatigue and a decrease in
functional capacity are also common complaints, along with palpitations and weight gain.
Depression and anxiety are quite common complaints associated with HF, and can occur from
both physiologic and mental health causes (Metra & Teerlink, 2017).
Diagnosis of Heart Failure
As per Arnold et al. (2006), “The diagnosis of clinical HF is made when symptoms and
signs of impaired cardiac output and/or volume overload are documented in the setting of
abnormal systolic and/or diastolic cardiac function.” A relevant clinical history, including the
symptoms and signs discussed above, is the first step towards an accurate, informed diagnosis.
Following this, there are a multitude of tests that can be performed, although not all will be
discussed here
Blood and brain natriuretic peptide (BNP) tests. Serum levels of BNP are an
important value for HF diagnosis as its levels strongly correlate with increased stretch and
pressure of ventricular walls due to volume increases. BNP and N-terminal prohormone BNP
(NT-proBNP) remain plasma markers with distinct sensitivity and specificity for HF in an effort
to rule out and distinguish between cardiac causes of dyspnea or other pulmonary etiologies
(Lilly, 2016; McNamara, 2005). A BNP value of >50 pg/mL or an NT-prBNP value of >125
pg/mL is an indication for consult with a specialist as well as imaging of the heart structures with
echocardiography (Ezekowitz et al., 2017).
Electrocardiogram (ECG). The ECG is a low risk, non-invasive tool that is used in the
diagnosis of HF to assist in determining the rate and rhythm of the electrical conduction of the
heart. It is also useful to identify and measure common conduction abnormalities that occur as a

14
result of the anatomical changes of HF such as ventricular hypertrophy, bundle branch blocks,
pathologic Q waves from previous myocardial infarction, ST-segment changes, and T-wave
abnormalities. ECG provides valuable information that can assist with HF treatment
considerations and drug initiation and titration (Lilly, 2016).
Chest X-Ray. Chest radiography, while specific for some findings of acute HF, is limited
in its sensitivity for the disease but is considered part of the initial assessment (Metra & Teerlink,
2017). The usefulness of the chest x-ray includes assessing the size of cardiac structures, and the
shape and size of the cardiac silhouette in comparison to the surrounding thoracic cage. Specific
patters of chamber and great vessel dilatation can represent specific disease-causing HF, while
manifestations in the pulmonary vasculature can assist to identify changes in pulmonary arterial
and venous blood pressures or flow (Lilly, 2016).
Echocardiogram (transesophageal or transthoracic). Echocardiography is a vital
procedure in the diagnosis as well as the ongoing management of HF. Echocardiography uses
high-frequency ultrasonic waves to produce an image of the heart that allows the reader to
examine structural detail of the heart, including measurements of wall thickness, valve integrity
and size, chamber contents (tumors, thrombi) and size, as well as wall motion, pericardial and
aortic disease, and other anatomical relationships. There are varying methods for
echocardiography including transthoracic, transesophageal, two-dimensional, three-dimensional,
and contrast or bubble study echocardiography. Echocardiography is the most common method
for determining the ejection fraction (EF), which is an essential measurement of left ventricular
ejection fraction, a calculation that is important in classifying HF. Echocardiography is
noninvasive, inexpensive, mobile, relatively fast, and most importantly, safe for the recipient
(Lilly, 2016).

15
Stress test. Stress tests can be performed with either standard exercise or by
pharmacologic means to observe patients suspected of having coronary artery disease. Typically
for standard exercise, the patient is asked to use a treadmill or stationary bike with a progressive
incline in workload to attempt to identify any of the following: angina, evidence of myocardial
ischemia such as ST-segment changes, or achieving a target heartrate predicted for the patient.
The test may be aborted by the patient due to reasons such as fatigue or other discomforts. If the
patient is anticipated to not be able to complete a standard exercise test, a coronary vasodilator
such as dipyridamole or adenosine can be administered. The action of these medications disrupt
flow from areas of the myocardium that may already be experiencing poor perfusion by
enhancing blood flow to other areas, with the result being an increase in angina or evidence of
ischemic ECG changes.
Coronary angiogram. Coronary angiography is a mainstay of imaging to detect,
diagnosis and guide the treatment of ischemic heart disease. Radiopaque contrast dye is injected
after a fluoroscopic guided catheter is inserted; the contrast is used to identify the location and
extent of atherosclerotic lesions that have potential to induce ischemia to the myocardium.
Angiography is an important procedure for diagnosis of HF with suspicion of ischemic disease,
as it is one of the most prevalent causes of HF. If an obstructive lesion is identified, angioplasty,
placement of stent(s), or coronary artery bypass can be used (Lilly, 2016).
Once a diagnosis is completed, the Canadian Cardiovascular Society recommends the use
of the New York Heart Association functional capacity assessment scale (Table 1), as a simple
and well validated tool to document the degree of each patients’ clinical severity of HF.
MRI. Cardiac MRI offers enhanced characterization of the contrasting tissues within the
myocardium, with the ability to differentiate between infarcted, impaired, and viable tissues.

16
This important tool offers the benefit of no radiation, as it relies on radiofrequency and magnetic
fields to develop the valuable detail in the image it produces. MRI offers incredible detail to
diagnose multiple cardiac abnormalities; limitations of cardiac MRI include cost and availability
of the machine (Lilly, 2011).
Table 1: New York Heart Association Functional Classification and Other Symptom Descriptors
Class
I
II
III
IV

Definition
No symptoms
Symptoms with ordinary activity
Symptoms with less than ordinary activity
Symptoms at rest or with any minimal activity

Other descriptor
Asymptomatic
Mild symptoms
Moderate symptoms
Severe symptoms

Ezekowitz, J. A., O’Meara, E., McDonald, M. A., Abrams, H., Chan, M., Ducharme, A., . . . Sussex, B.
(2017). 2017 comprehensive update of the Canadian cardiovascular society guidelines for
the management of heart failure. Canadian Journal of Cardiology, 33(11), 1342-1433.
doi:10.1016/j.cjca.2017.08.022

Management of Heart Failure
Pharmacology. Goal-directed pharmacological therapy for HFrEF has demonstrated
slowing the natural progression of the disease, and improving survival in these patients
(Ezekowtiz et al., 2017). This includes triple therapy consisting of the following: an angiotensin
converting enzyme inhibitor (ACEi) or angiotensin receptor blocker (ARB), a beta blocker (BB),
and a mineralocorticoid receptor antagonist (MRA). Diuretics are also an important part of HF
therapy when volume overload is problematic; however, this is solely for symptoms relief (Metra
& Teerlink, 2017).
ACEis and ARBs play a vital role in the management of HF. Lilly (2016) reinforces the
fact that vasodilators, ACEi’s in particular, have been one of the greatest advances in the
treatment of HF in the past twenty years. Strong evidence supports the introduction of an ACEi
in all patients with HF and an EF of <40% without other contraindications as soon as the
condition allows (Ezekowitz et al., 2017). ACEi’s block the potent vasoconstricting effects of the
end product of the RAAS as previously discussed and enhance the endogenous vasodilator nitric

17
oxide by increasing circulating bradykinin levels. Thus, ACEi’s are a first line therapy for HF
because of its strong effects on blocking the neurohormonal pathways. ACEi’s are initially
prodrugs that require adequate metabolism from the liver and excretion from the kidneys.
Reaching therapeutic goal dosing of these drugs is an important piece of HF management, as the
potential benefits of the drug increase with the dosage, as long as the patient can tolerate side
effects. ARB’s have a similar action and are used when patients cannot tolerate ACEi’s (Dec,
2005; Lilly, 2016).
BB’s provide beta-adrenoreceptor blockade, and have potential to augment cardiac output
in those patients with a reduced EF, decrease deterioration of hemodynamic mechanisms, and
reduce hospitalizations, thus improving survival. BB’s act along with the ACEe’s to block
certain neurohormonal pathways that under chronic stimulation, cause the structural and clinical
changes witnessed in HF. BB initiation should not be started during acute HF, but rather in
patients with more stable HF and uptitrated to goal doses within the limits of the patients’
tolerance (Ezekowitz, 2017; Lilly, 2016).
MRAs block the effects of excess aldosterone that results during HF and contributes to
harmful ventricular remodeling and fibrosis. The two main MRAs have been demonstrated to
improve symptoms of HF as well as improve survival of HF patients from all-cause mortality as
well as SCD post myocardial infarction.
Diuretics provide symptom relief of congestive volume overload in HF, but do not
improve survival. Several different variations of diuretics are available (loop, thiazide, thiazidelike), and should be used in an attempt to achieve euvolemia, and then titrated to each individual
patient to the lowest effective dose, as these medications interfere with sodium retention in the

18
renal tubules and can even potentiate the activation of the RAAS. Careful monitoring is
recommended for evaluation of electrolyte imbalances and hypotension (Dec, 2005).
Digoxin, platelet inhibitors, anticoagulants, and statin medications have important roles
in different types of HF, but must be used on an individual basis. Newer agents, such as
angiotensin receptor-neprilysin inhibitors (ARNi’s) and Ivabradine both also have potential to be
effective in reducing morbidity and mortality in the disease, but are also only indicated for
certain variations of HF and to be used with clinical expertise (Ezekowitz, 2017). The current
CCS guidelines recommend use of ARNi’s for HF patients with NYHA class II-IV with an EF
<40%.
Lifestyle management. Patient education and participation in HF treatment is a
cornerstone to successful management. Sodium and fluid restriction, cardiovascular exercise,
smoking cessation, and decrease of exposure to cardiac toxins such as alcohol and illicit drug use
are all points of focus for every patient. Multidisciplinary HF teams that are available to address
and support these topics and provide interventions to increase patient empowerment, allow for
time-sensitive follow-up, integrate bridged care for patients primary and specialist providers, and
optimize pharmacologic therapy have shown to improve outcomes of quality of life, morbidity
and mortality (Virani, 2017).
Device Therapy. Implantable Cardioverter-Defibrillators (ICDs) continue to be strongly
recommended for prevention of SCD in those patients with HF with an EF <35% and/or a history
of sustained ventricular arrhythmia and/or hemodynamic instability with such arrhythmia after
exclusion of all reversible causes. Patients with additional factors that have demonstrated high
risk of ventricular tachycardia or ventricular fibrillation such as prolonged QRS interval on ECG,
elevated heart rates, higher burdens of premature ventricular complexes, and myocardial scarring

19
or aneurysm may also be considered for ICD therapy (Bennett et al., 2016). ICDs work to
terminate dangerous ventricular arrhythmias (ventricular fibrillation in particular) by monitoring
cardiac electrical activity and administering an electrical current, such as that during external
cardioversion or defibrillation, to eliminate the arrhythmia (Lilly, 2016). In comparison with
antiarrhythmic therapy, ICDs have been shown reduced all-cause mortality than drug therapy
alone, and have a much higher therapeutic benefit with those patients that lower EF’s, higher
NYHA scores, and older age (Ezekowitz et al., 2017). In BC, there were 427 new ICD implants
in 2017 (personal communication, Heather Jackson from Cardiac Services BC, September 21,
2018)
Cardiac Resynchronization Therapy (CRT) is another device method used more often in
HF with a decreased EF and prolonged QRS (>130ms) especially with a left bundle branch block
pattern. A dilated LV may often lead to a LBBB, however this is not part of the criteria for
eligibility. Similar to a pacemaker, but with an additional lead to stimulate both ventricles, the
two lower chambers are paced simultaneously in order to conduct more synchronous contraction
and improve cardiac output. If CRT is implemented correctly, patients usually report improved
exercise tolerance, lessened dyspnea, fewer HF related hospitalizations, and improved quality of
life (Lilly, 2016).
Left Ventricular Assist Device. A left ventricular assist device (LVAD) acts as a bridge
to heart transplant for patients with advanced HF resistant to medical treatment, as destination
therapy, or more recently, a bridge to recovery (Jakovljevic et al, 2017). LVAD therapy is a form
of mechanical circulatory support in which an implanted pump draws blood from the chamber of
the left ventricle and passes it through to the aorta. It is powered by an external system controller
connected to a power source that is attached to the pump by a percutaneous lead (Slaughter et.

20
al., 2009). Two versions of the device are available, one that is capable of delivering a pulsatilelike flow and one that delivers a continuous flow; both have the capacity to deliver up to ten
liters of blood flow per minute (Slaugher et. al., 2009). Use of an LVAD in advanced HF, either
as recovery, destination, or bridge therapy, has demonstrated both improvements in cardiac
biochemistry and myocyte histology, as well as improvements in functional capacity, quality of
live, and survival rates for this group of patients (Jakovljevic et. al., 2017).
Cardiac Transplantation. Heart transplants, or cardiac transplantation, is considered the
gold standard of therapy for patients with advanced HF caused by diffuse atherosclerosis, or
valvular or congenital heart diseases that are not able to be corrected (Jakovljevic et. al., 2017).
Patients will be considered for cardiac transplantation if they have been optimized on medical
and/or surgical therapies, yet continue to experience poor quality of life or have a limited
survival prognosis (Isaac et. al., 2011). However, the availability of suitable hearts is far
exceeded by the number of patients eligible to receive one. Isaac et. al. (2011) report that only
50% of patients eligible for cardiac transplantation will receive one (although this does not
ensure survival). The remainder of patients may either become ineligible due to a deteriorating
health status, suffer multi-organ failure, or die while waiting (Isaac et. al., 2011). Due to the
scarcity of viable organs and success in transplants, contraindications to transplantation include
pulmonary hypertension, age greater than 70 (although this is not absolute), a BMI of <30kg/m2,
malignancy (with consideration for transplant after 5 years of remission), diabetes with endorgan damage, peripheral vascular disease, renal dysfunction, active tobacco or substance use,
some psychosocial issues, immune-incompatibility or active infection, severe hepatic or
pulmonary disease, and some underdeveloped stages of pregnancy (Isaac, 2011).

21
The technologies and science within the pharmacologic and interventional aspects of HF
are continuously advancing which creates exciting and promising futures for both patients and
providers alike for the management of this chronic issue. ICDs, while not new, carry
overwhelming evidence to support their use as a cost-effective and proven intervention to reduce
the risk of SCD in the HF patient (Bennett et al, 2017). The history, composition, and indications
for use of the ICD will now be presented.
What is an Implantable Cardioverter Defibrillator?
History. The first ICD was implanted in a young woman with recurrent ventricular
fibrillation in 1980, after almost a decade of development by Doctors Morton Mower and Michel
Mirowski (DiMarco, 2003). The inspiration came from the death of a colleague, and the doctors
were inspired to create an implantable device to both monitor and analyze the cardiac rhythm. If
the device were to detect ventricular fibrillation, it would be able to produce a shock sufficient to
defibrillate. Fast forward to modern day, ICDs can now detect and treat both ventricular
fibrillation and ventricular tachycardia, and are indicated for both primary and secondary
prevention of SCD (DiMarco, 2003). The efficacy of ICDs in preventing SCD has been well
established in clinical trials such as the Multicenter Unsustained Tachycardia Trial (MUSTT)
(Buxton et al., 1999), the Multicenter Automatic Defibrillator Implantation Trial II (MADIT-II)
(Moss et al., 2002) and the Sudden Cardiac Death in HF Trial (SCD-HeFT) (Bardy et al., 2005).
Components and function. The main body of the ICD is called the pulse generator,
which is a small titanium container that holds a battery made of lithium-silver vanadium oxide.
The container also holds a voltage converter and resister, integrated circuits and microprocessors
that perform the analysis and treatment of dysrhythmia, a memory chip capable of recording
electrophysiological data, as well as a telemetry unit (DiMarco, 2003). A typical size for a

22
common ICD (Medtronic Visia AF model) is 66mmx 51mm x 13mm, and weighs 77 grams
(Medtronic, 2015). An average battery life for each generator can vary from 4-9 years, and the
generator will contain at least one lead, connected to the generator with an epoxy resin header in
order to prevent any damage to the myocardium when a shock needs to be delivered.
Determining whether a patient should receive a single or a dual chamber ICD is based on an
assessment of the likelihood the patient will need permanent pacing in the future due to sinus
node dysfunction; however, multicenter trials strongly suggest that the efficacy of a single
chamber ICD may outweigh the risks of dual-chamber devices, including lead displacement and
infection, and therefore the Canadian Cardiovascular Society and the Canadian Heart Rhythm
Society recommend single chamber ICDs for all patients that require this device for primary
prevention in the absence of sinus node dysfunction or atrioventricular block (Bennett et. al.,
2017). A single chamber ICD has a lead placed in the right ventricle, and a dual chamber ICD
has a lead that is placed into the right atrium and a lead that is placed into the right ventricle
(CorHealth Ontario, 2017).
Guideline Indications for Use
Primary Prevention. Assessment of indication for ICD therapy involves measuring the
risk for the patient of ventricular tachycardia (VT) and/or ventricular fibrillation (VF), and the
treatment or reversibility of conditions that cause VT (Bennett et. al., 2017). Risk factors for
VT/VF include evidence of frequency premature ventricular contractions (PVCs). Patients
without an ischemic cause of HF, who have been on optimal medical therapy for at least three
months and continue to demonstrate a LVEF <35%, who have NYHA II-IV symptoms and an
ECG that shows sinus rhythm with QRS duration <130 msec should be considered for referral
for a primary prevention ICD (Bennett et al., 2017). In addition, those patients with an ischemic

23
etiology of their HF, that are have a LVEF <35%, who have NYHA class II-III symptoms, and a
duration of at least three months post coronary revascularization should be considered. Lastly,
patients with an ischemic etiology of their HF with LVEF < 30% measured at least one month
post myocardial infarction, and three months after post coronary revascularization therapy with
NYHA class I symptoms can be considered (Bennett et al., 2017).
Secondary Prevention. Secondary prevention refers to the indication for patients with
HFrEF that have endured a previous occurrence of sustained ventricular arrhythmia or a
hemodynamically significant arrhythmia to receive ICD therapy. This is a strong
recommendation from the CCS (Bennett et al., 2017) produced from high quality evidence in an
attempt to reduce the risk of SCD in this patient group. Figure 4 demonstrates the CCS 2017
pathway to determine eligibility for device therapy for the HF patient:

24
Figure 1: Referral pathway for device therapy in patients with HF

Figure 4: Referral pathway for device therapy in patients with HF. Adapted from “Implantable
Cardiac Devices,” Ezekowitz, J. et al. 2017 comprehensive update of the CCS guidelines for the
management of HF. Canadian Journal of Cardiology; 33. p.1342-1433
Implantation. The procedure required for implantation of an ICD is relatively simple
and usually performed under local anesthesia. After discussions with the implanting specialist, a
decision is made on whether to place the device on the right or left side of the upper chest area,
just below the clavicle. A small, two to four-inch incision is made creating a pouch beneath the
skin and above the chest wall. One, possible two leads are then guided through the subclavian
vein and then attached to the right atrium and/or right ventricle.
Subcutaneous ICDs. The subcutaneous ICD (S-ICD) is an acceptable option for those
patients that are not eligible for or have circumstances that prevent them from receiving a
transvenous ICD. It may also be preferred for some younger patients, as well as to avoid venous
access problems with leads. The S-ICD allows for detection and delivery of shock for
arrhythmias; it does lack the antitachycardia pacing feature that the ICD has (Kempa, Budrejko

25
& Raczak, 2016). Placement of the S-ICD occurs in the subcutaneous tissue at the midclavicular
line over the 5th and 6th rib. There is in inadequate interventions in approximately 5-16% of
patients these devices are inserted in, and it due to over-sensing.
Providers. Multiple providers may be involved in the assessment to refer a patient for an
ICD. The primary care provider or a cardiologist are typical members of the team for a HF
patient. An electrophysiologist – a cardiologist with additional training to specialize in
abnormalities of the conduction system of the heart – is who most patients will be referred to in
order to make the decision if a patient fits eligibility criterion to have the device implanted. Any
patient with an ICD will require ongoing specialty input, particularly around the time of
insertion. Specialized HF clinic teams can provide exceptional education and ongoing
monitoring for these patients, including cardiac nurses and technicians. The role of the patient’s
primary care provider is critical for the education and support of the HF patient with the ICD,
especially when considering all of the contributing factors into the patient’s quality of life, health
and wellness.
The NP Role in Primary Care for Patients with HF
The NP role in BC was established in 2005, many years behind the extensive utilization
and successes of other areas such as the United Kingdom and United states; even behind
provinces in eastern Canada that have witnessed the success of improved outcomes and patient
satisfaction with a NP provider (BC Nurse Practitioner Association, 2018). NPs are licensed in
BC by the Nurses and Nurse Practitioners of BC (NNPBC,2018), and are advanced practice
nurses that offer autonomous health care services across a variety of primary and speciality
settings, and who have demonstrated expertise in responsive and creative roles to meet the needs
of populations across a continuum (NNPBC, 2018). While the BC Ministry of Health has

26
identified that NPs echo the three main goals of its mandate – enhanced patient experiences,
reduced costs, and improved outcomes – NPs remain vastly underutilized as a provider group
(NNPBC, 2018).
Outpatient HF clinic program development has been on the rise in Canada in order to
address this illness, contain service utilization and cost associated with inpatient admission of
HF, and address the multitude of needs of patients with this condition (BC Ministry of Health,
Services, 2002; Wagner et al., 2001). While most of these programs are physician-led and nursedirected (Thompson & Dykeman, 2007), the US and Australia have multiple recommendations
that advanced practice nurses, specifically NPs could excel in the management of HF through
their specialty education. There is an urgent need to plan ahead to identify strategies to deal with
cost containment in the chronic disease field and implementing the most suitable health care
provider to deal with the complexity of the issue is essential to the future of BC Health Care (BC
Ministry of Health Services, 2002). The benefits related to the management of a HF patient by
an NP include a comprehensive focus of care and ability to move beyond a biophysical model to
promote health through education and capacity building. The approach and lessened time
restrictions of the NP allow the patient to be less reliant on the system by offering decreased
usage of acute care facilities, and improving self-management techniques through education and
lifestyle modification tools, thus reducing overall health care costs through reducing readmission
rates to hospital and decreasing contact hours with outpatient services (Thompson & Dykeman,
2007).
The importance of the nurse’s ability to reframe a solely scientific approach to
assessment and treatment and incorporate viewing the patient with a wider spatial experience can
assist in providing more comprehensive care management (Stockwell, 2008). In considering

27
managing patients with HF and ICDs, the primary care provider, such as a NP, is the optimal
resource to promote the process of shared decision-making with the patient, as they have the
expertise to understand the value of the device, as well as the relationship and aesthetic
knowledge of the patient. NPs value the practice of holistic care and have the ability to adapt
standardized protocol to their patients’ unique needs. The ability for the patient to be informed
and actively participate in decision-making for their health care needs requires an understanding
from the provider of the patient’s mutually agreed upon goals and patient values. The
collaboration of the patient’s health-care team to provide disclosure of appropriate individualized
information related to the implications of the intervention adds to the complexity involved in
participating in patient-centered care (LeMond, Camacho & Goodlin, 2015). The NP is in an
optimal position to contribute to the health care of HF patients who require ICDs.
This project seeks to examine this further by undertaking a critical review of the relevant
literature available. The following chapter will provide an overview of the review methods for
the literature review. This will be followed by a critical review and discussion of the captured
research literature.

28
CHAPTER II: METHODOLOGY
Integrative Literature Review
To answer the question: What is the role of the primary care provider in supporting HF
patients requiring an ICD? an integrative review has been conducted. The integrative review is
a unique method of collecting existing theoretical or empirical literature pertinent to a particular
topic in order to provide a more comprehensive, evidence-based summary to inform or guide
further theory, practice or policy development (Whittemore & Knaffl, 2005). Searching literature
with a systematic method ensures the process can be reproduced in order to enhance a similar
study in the future, and boosts the rigor of the current study. Whittemore and Knaffl suggest a
modified framework for use when conducting an integrative review, which has been used in this
paper in order to present an orderly sequence of the purpose of the review, the literature search,
eligibility criteria, data analysis of both qualitative and quantitative studies, discussion and
recommendations. The methodology used for this current integrative review will now be
discussed.
Literature Search
Inclusion and exclusion criteria. Prospective studies for this integrative review were
required to meet established inclusion criteria in order to design a high-quality search that
designated a particular clinical and patient characteristic. Exclusion criteria were chosen to
enhance the relevance of the studies to the topic, and eliminate potential factors that may
decrease the validity of the findings. Articles which focused on pacemakers or cardiac
resynchronization therapy devices with ICDs were excluded as the goal of these therapies may
have been introduced for reasons in addition to the risk of SCD. Only articles published between
January 1, 2008 and December 31, 2018 were included to stay focused on current trends and

29
guideline recommendations. Inclusion nor exclusion criteria was used to source articles from any
particular country of origin as this information may assist in development of new policies or
procedures. However, articles were only included if they were available in the English language.
Both qualitative and quantitative studies were included, as the qualitative articled assisted to
augment and further comprehend the statistics found within the quantitative studies. ICDs were
not discriminated against if they were used for primary or secondary prevention, as the purpose
of this review was to identify the role of the primary care provider on the broader management of
the device, and also added to the findings when looking at provider knowledge of the device.
Literature which focused on ICDs within a particular cardiomyopathy were excluded to gain a
more generalizable approach to the provider in a primary care setting. An age of >18yrs was
chosen to reflect only adult patients with HF, as the decision-making piece would be more
independent than a pediatric patient with a guardian. Lastly, studies that focused on ICDs in a
tertiary care or inpatient setting, or on decision-making solely from specialists were excluded in
order to capture literature that concentrated on the role of a care provider similar to a NP in a
primary care setting. Inclusion and exclusion criteria are reviewed in Table 2.

30
Table 2: Inclusion and Exclusion Criteria for Integrative Literature Review
Inclusion
Published between January 1,
2008, through to December
31, 2018

Exclusion
Published prior to January 1,
2008

The research may be
qualitative or quantitative,
and may include primary
studies, systematic reviews,
and meta-analysis

The research may not be nonpeer reviewed articles, case
studies, opinions, analysis of
guidelines, dissertations or
books.

There is an available abstract,
and if not, there is a relevant
title
Addresses at least one of the
following:
1. Decision-making for
ICDs in the heart
failure patient
(primary or secondary
prevention)
2. Referral for ICD
3. Advanced Care
Directives specific to
ICD
4. Primary care
physician’s role in
management of ICD
in HF patients

There is no abstract available
and/or the title does not
appear relevant
Specific focus on
cardiologists, pacemakers,
cardiac-resynchronization
devices, or an individual
cardiac pathophysiology (ex.
non-ischemic dilated
cardiomyopathy), or
management for tertiary care
inpatients exclusively.

Studies in the English
language

Non-English studies

Study population must be >18 Study population must not be
years of age
<18 years of age.

Rationale
These dates were chosen in
order to reflect current
practice, and as well to
include the 2008
ACC/AHA/HRS guidelines
for primary prevention ICD
implantation.
To enhance the value of
information obtained from
quality, peer-reviewed and
evidence-based studies in
order to influence stronger
recommendations.
To assist the search in
obtaining relevant studies
My goal is to address how
ICD management in primary
care needs to be enhanced to
empower the HF patient with
informed-consent and
participation in the decisionmaking process for referral
and deactivation, therefore
limiting to ICD as a lifesustaining device rather than
a cardiac enhancement device
such as PM or CRT was
necessary, and limiting the
providers to be more
generalizable to primary care
was needed to promote the
role of this provider.
To eliminate the need for
translation and possible error
in comprehension of
translation
The focus of this review is
based on adult patients with
HF

31
Preliminary search. The literature search performed began with a preliminary
identification of Medical Education Subject Headings (MeSH) terms in order to identify any
additional words or phrases of relevance that could assist in a more thorough search of current
studies. The terms heart failure, implanted cardioverter defibrillator, and primary care yielded
additional search terms such as defibrillator, implantable, primary health care, and management.
Next, four databases including Medline, Cumulative Index for Nursing and Allied Health
Literature (CINAHL), Ovid, Cochrane Review and CINAHL Complete were searched to obtain
a comprehensive list of relevant articles. These articles included systematic reviews, qualitative
and quantitative primary research studies, dissertations, and some reviews of national and
international guidelines on related topics. The searches were performed by combining keywords
from Table 3 into each individual database, using Boolean search language of AND, OR and * to
truncate certain terms.
Table 3: Search Terms for the Integrative Literature Review
Heart Failure Patients
•
•
•
•
•
•
•
•
•

Heart Failure
Heart Failure,
Systolic
Heart Failure,
Diastolic
Cardiac Failure
Congestive
Heart Failure
Heart Failure,
congestive
Heart
Decompensation
Left-sided heart
failure
Right-sided
heart failure

AND

ICD
•
•
•
•

AND
Defibrillator(s),
Implantable
Defibrillator,
Implant*
Cardioverterdefibrillators,
implantable
Implantable
Defibrillator*

Support in Primary
Care
• Care,
primary
• Health care,
primary
• Care,
primary
health
• Primary
health care
• Family
practice
• Nurse
practitioner
• Practitioner,
nurse
• Management

32
Using the term ‘nurse practitioner’ and ‘practitioner, nurse’ with any of the population or
intervention terms yielded almost no results, and therefore was broadened to the term primary
care in order to encompass what is an evolving area of education for this particular role. Filters
were then applied to limit the timeframe of articles to from January 1, 2008 – December 31,
2018. This filter produces articles that would likely better represent current practice and use of
updated guidelines on the matter of ICD management. Only articles written in the English
language were kept. After these filters were applied, 642 articles were collected and transferred
into Zotero. Zotero is a software manager for collecting, organizing and managing citations,
references and bibliographies. Using Zotero, 185 duplicates were removed, resulting in 475
articles that were then submitted to a title screen for relevance. 66 potentially relevant article
abstracts were screened (9 were removed due to style of the article, for example dissertation,
opinion or case study), and ultimately 15 were chosen for their strength and quality to be
included in this review. Articles which included HF patients requiring ICDs with some regard to
primary care setting or provider were preferentially selected to provide the potential perspective
and needs of a NP. Appendix provides an illustration of the full search strategy
Focused search. Ultimately the preliminary search concluded with 15 relevant articles
with strong quality assessments. This group of articles contained four systematic reviews and
eleven primary research studies. All fifteen articles were read meticulously in their entirety to
ensure they met the inclusion/exclusion criteria and were able to contribute to the purpose of this
integrative review. A scan of the reference lists of these articles was completes to identify further
articles of relevance that fit the inclusion and exclusion criteria; three supplementary primary
research articles were incorporated into the background of this review, however did not meet
inclusion criteria as they were not based within primary care. Using the Health Evidence Quality

33
Assessment Tool from McMaster University, the 15 articles were appraised and consistently
scored highly. The Quality Assessment Tool considers ten different criteria to determine if a
study is transparent, reproducible and supports best-evidence. This assessment tool can be
reviewed in full detail in Appendix D. To provide a detailed and comparative summary of the
information gathered from the studies included in this integrative review, Appendix B contains a
comprehensive literature matrix describing features of each independent article. In this matrix,
dates, methods, study samples, results, strengths, limitations, quality and recommendations are
provided.
Grey literature, A search of the grey literature on management of HF patients with ICDs
was searched, and resulted in further articles and guidelines from the Canadian Cardiovascular
Society, Canadian Heart Rhythm Society, American Heart Association, BC Guidelines, and
Cardiac Services BC. While this search presented valuable information to be included in this
integrative review, most did not meet the eligibility criteria to be included in the findings.
Final results. The 15 articles that met inclusion and exclusion criteria with strong quality
appraisals for the research question in this integrative review are included. Four of these articles
are systematic reviews, and eleven are primary research studies. The following chapter with offer
an investigation of the relevant literature.

34
CHAPTER III: FINDINGS
The purpose of this integrative review was to enhance the understanding of the role
primary care providers play in supporting HF patients who require ICDs. In order to provide
these recommendations, a comprehensive search was conducted to identify relevant literature. A
final cohort of 15 articles were selected for critical analysis. A critical review of the literature
revealed three major themes. 1) Low referral rates and underutilization of ICDs; 2) Primary care
provider knowledge of indications for ICD; and 3) Advanced care planning and deactivation of
ICDs. Ten of the articles included in the final review are original quantitative studies, four of the
final articles are systematic reviews, and one is a qualitative research study. Including these
studies of diverse methodologies allows for a more comprehensive understanding of evidencebased practice and has the potential to generate findings that may contribute to policy and
practice (Whittemore & Knaffl, 20005). An analysis of the themes will now be presented.
Low Referral Rates and Underutilization of ICDs
When examining the amount of low referral rates and underutilization of ICDs in eligible
patients, it is important to consider the many variables that occur during the process. In general,
the national guidelines in both Canada and the US have clear recommendations as to which
patients should be referred for ICD consideration, but as such, do not necessarily reach all
clinical situations equally, and thus the community is faced with added considerations to apply to
each scenario. Three of the primary research studies captured in this review focused their
questions on determining the factors that influence referral, or more specifically, non-referral in
order to better understand what issues need to be addressed to increased rates of utilization
(Bradfield, Warner and Bersohn, 2009; Chae & Koelling, 2010; & Pillarisetti et al., 2015). This
section investigates why a class I therapy for SCD is underutilized in eligible patients.

35
Pillarisetti et al. (2015) conducted a large retrospective cohort study at the University of
Kansas Hospital in Kansas City, Kansas to assess reasons for underutilization and nonimplantation of ICDs in eligible patients. The tertiary care center supported the collection of
patient data from their Cardiovascular Database, which includes all patients that received care
from an entire region with five electrophysiologists and advance electrophysiology laboratories.
The patient population selected consisted of 707 patients with diagnosed LVEF <35%
documented via three consecutive measurements using various acceptable imaging modalities.
As this study used data from a larger, observational study, the patients were followed until May
2009. Patients who had spontaneous improvement of EF, were lost to follow up, or died prior to
2005 were excluded. Rationale for this date range includes the update to ICD recommendations
in the SCD-Heft trail in 2005 (Bardy et al., 2005). Baseline clinical and demographic variables
were obtained, included time to ICD implantation. Two independent investigators independently
reviewed all records including reasons for non-implantation of ICD. Discussion of option for
ICD was considered if there was evidence of mention of ICD implantation or a documented
discussion. As well, if patients refused the option for ICD implantation or carried too many
comorbidities, this was also considered. Statistical analysis was used to compare rates of medical
treatment and ICD implantations received; Chi-square and Exact tests compared the discrete
variables. P = <0.05 was statistically significant. Again, identifying predictors of ICD
implantation was achieved with logistic regression analysis.
After scrutinizing baseline and clinical characteristics including mean EF (26 +7%),
NYHA class, mean QRS duration, sex, race, cardiomyopathy type and duration, percentages of
patients on HF medications, and patients who received ICD implantations, a significant
difference in all-cause mortality was identified between the HF patients with ICDs and those

36
without (25% vs 37%, or p = 0.004). The sample size was then reduced to 510 patients after
taking into consideration the change in guideline recommended practice for ICD therapy in 2005
which included primary prevention ICD, or for patients who died prior to this time and/or were
lost to follow up. With this adjustment, still only 37.6% of eligible patients (192) received an
ICD. The following Table 4 demonstrates the reasons for non-referral or non-utilization in the
62.4% (318) of patients that did not receive an ICD (Pillarisetti et al., 2015).
Table 4: Reasons for non-utilization of ICD therapy

Reasons
Option not discussed (Not
documented)
Patients Refused
Percentage where EF
improved >1yr later
Died
NYHA class I
NYHA class IV
Lost to f/u
Low 1 yr. survival
Low quality of life
Comorbidities
Discussed once but
physician has not followed
up
Discussed but patients not
followed
Miscellaneous causes

Frequency

Percentage

72
52

22.6%
16.3%

1
25
5
18
12
10
6
9

0.3%
7.9%
1.6%
5.7%
3.8%
3.1%
1.9%
2.8%

6

1.9%

28
318

8.9%
100%

74

23.2%

While identifying valid reasons for non-implantation of ICDs in eligible patients, the
findings of Pillarisetti et al. (2015) demonstrate that the most common reason most patients do
not receive this therapy is lack of physician discussion of this important option. The study
revealed that only one third of patients with eligible EF <35% received an ICD, and also
displaying the significantly higher mortality rate in those patients without the device. Patients

37
with more severe NYHA symptoms (II-III), established arrhythmias (example atrial fibrillation),
and on good doses of HF medications were more likely to be referred. Similar to findings from
theme 1, the authors identify targeting educations interventions towards the care provider as a
suggestion to improve the rate of awareness and referral for the population in which the option
for ICD was never discussed. Continuous counselling at subsequent visits despite refusal may
also be indicated, or working in collaboration with a specialized HF team. Limitations of this
study were identified as being only a single center retrospective study and lacking exact causes
of death in the patients who died without ICD. However, these limitations are minor overall and
the quality of the location, adequate power of the sample patient selection for this study
demonstrates important strengths, consistent with findings of other studies observing low-referral
and underutilization rates of ICDs. This study scored a Quality Assessment Tool rating of 10/10,
strong, and emphasizes the importance of enhancing the knowledge of providers and patients of
ICD therapy and eligibility.
Chae and Koelling (2010) had previously conducted a study with a methodology similar
that was adopted by Pillarisetti et al (2015). The study conducted by Chae and Koelling (2010)
was again a retrospective observational cohort study at a single tertiary center at the University
of Michigan Health System (UMHS) HF registry that investigated the rates of ICD utilization in
eligible patients (according to national guideline recommendations), and sought to identify
variables between those patients that received ICDs and those who did not, including those
characteristics of the provider (see Figure 2). The study population consisted of 850 patients with
a documented LVEF <35% of whom met the class I AHA/ACA guideline recommendations for
ICD implantation. In this instance, 70% (594) patients received an ICD, which is a much greater
utilization rate than the population in the Pillarisetti et al., (2015) study. Predictive variables for

38
ICD implantation of statistical significance included those patients that were generally sicker
upon clinical and observational investigation. These were patients that were younger with worse
LVEF, lower systolic and diastolic blood pressures, and wider QRS complexes. In addition, the
patients more likely to receive ICDs were those cared for by a HF specialist (86%) or a general
cardiologist (56%). The patients receiving care by primary care physician had much lower rates
of referral (26%, P = <.0001). Thus, upon analysis, the single greatest predictor of eligible
patients receiving ICD was the provider type who would be referring for care.
Figure 2: Baseline Characteristics of Patients

Pillarisetti, J., Emert, M., Biria, M., Chotia, R., Guda, R., Bommana, S., . . . Lakkireddy, D. (2015).
Under-utilization of implantable cardioverter defibrillators in patients with heart failure - the current
state of sudden cardiac death prophylaxis. Indian Pacing and Electrophysiology Journal, 15(1), 20-29.
doi:10.1016/S0972-6292(16)30838-5

39
There are several limitations to this study, the foremost issue being that the analysis of
rates of ICD implantation did not discriminate between single ICD device vs CRT+ICD devices.
The consequence of this lack of information means that the reader cannot differentiate if the
motivation for the device implantation was for SCD prevention and guideline directed therapy,
or if the goal was symptom reduction and LVEF improvement through CRT with the ICD as an
additional therapy. Secondly, this study was performed at a tertiary center with a specialty group
that focused on HF patients, with enhanced access to HF specialists. This limits the
generalizability of the findings to those areas with limited access or more rural locations. Third,
being a retrospective study with data obtained through charts at a single center leaves room for
possible error from lack of complete data.
Bradfield, Warner and Bersohn (2009) conducted a quantitative retrospective cohort
study to investigate the large number of eligible HF patients that failed to receive an ICD. The
study occurred from April 12, 2002 to June 30, 2006 at the Veterans Affairs Greater Los Angeles
Healthcare System. The purpose of the study was not only to identify patients with low LVEF
who did receive an ICD, but also to further understand the reasons why patients may not have
been referred. Using the echocardiogram database, the authors searched through the timeframe
discussed for values that demonstrated a decreased LVEF and also met guideline criteria for ICD
implantation. A total of 1200 patients met inclusion criteria; 600 patients (120 per study year)
with LVEF <35% and 600 patients (120 per study year) with LVEF < 30%. An analysis of
patients who were not referred for ICD was completed first. Reasons for non-referral are
demonstrated in Table 5.

40
Table 5: Reasons for non-referral for ICD

LVEF 31%-35%

LVEF <30%

(600)

(600)

Terminal illness

38 (6.3)

43 (7.2)

Refractory class IV heart

1 (0.2)

1 (0.2)

Psychiatric illness

6 (1.0)

5 (0.8)

Improved ejection fraction

62 (10.3)

45 (7.5)

Awaiting revascularization

12 (2.0)

12 (2.0)

Work up in progress

28 (4.7)

26 (4.3)

Not on optimal medications

25 (4.2)

32 (5.3)

Lost to follow-up

31 (5.2)

44 (7.3)

Not referred

229 (38)

196 (33)

Patient refused

40 (6.7)

47 (7.8)

Died

5 (0.8)

4 (0.7)

Awaiting ICD implantation

12 (2.0)

16 (2.7)

ICD implanted

111 (18.5)

129 (21.5)

failure

In the first group with LVEF <35%, the ICD implant rate represented only 28% of the
eligible 392 patients. Of the patients that were eligible that did not receive ICD, there was a 34%
mortality rate despite having no contraindications for receiving the device, versus a 15%
mortality rate in the patients who did receive an ICD. Similarly, in the second group with LVEF

41
<30%, the implant rate for eligible recipients was only mildly greater at 33%, with a 51% rate of
non-referral. Again, among patients that were eligible for but did not receive an ICD, there was a
33% mortality rate, versus an 18% mortality rate in the patients who were ICD recipients. This
study demonstrates the hypothesis that Bradfield et al. (2009) discussed, concerned that despite
two large studies (MADIT-II and the SCD-HeFT) that demonstrate benefits of prophylactic ICD
implantation including cost-effectiveness and reduction in mortality rates in specific HF patient
populations, there remains a large number of patients that are not offered this potentially lifesaving therapy. There was only one reviewer for all of the documents of patients included in this
study, and lack of documentation was assumed to be non-discussion. It is possible that these
discussions between providers and patients occurred, therefore overestimation of non-referred
patients may have been a limitation of this study. The results of the SCD-HeFT trial were not
published until 2004, which occurred in the middle of this study, and may be an explanation for
some patients in the LVEF <35% group to have not been referred as they were not yet
considered candidates. The Quality Assessment Rating for this study was 10/10 strong, and its
strengths stemmed from an adequate time frame, population size, detailed inclusion and
exclusion criteria, and data that strongly aligns with results of other studies in this review.
Overall, when critically reviewing the literature that addressed reasons for non-referral and
underutilization of ICDs in HF patients that meet guideline recommended criteria, similar themes
emerged, most prominently lack of discussion, provider type, and patients who were more
obviously ill, identified by either clinical or objective values. While younger age and being of
male sex were independent variables that increased the likelihood of referral for ICD, race,
comorbidities, and medications were not independent predictors of referral. Overwhelmingly, the
rates of ICD implantation in the eligible patients in all of these studies lend to the confirmation

42
that interventions are necessary to increase referral rates and utilize this recommended therapy.
Primary Care Provider Knowledge of Indications for Implantable Cardioverter
Defibrillator
Clinical practice guidelines for ICDs are created by groups of content experts in order to
synthesize pertinent evidence and to provide recommendations for those professionals that care
for patients at risk of SCD. In Canada, the Canadian Cardiovascular Society/Canadian Heart
Rhythm Society provided updated ICD guidelines in 2016. In the U.S., the American College of
Cardiology/American Heart Association/Heart Rhythm Society Guidelines for Device-Based
Therapy of Cardiac Rhythm Abnormalities described very similar recommendations for primary
and secondary SCD prevention with ICDs (Epstein et al., 2008). Despite detailed consensus
upon the selection of patients to be considered for ICD referral, lack of knowledge and/or
comfort of using these guidelines remains one of the primary barriers for non-referral to a
specialist (Bernier et al., 2017; Bradfield et al., 2009; Castellanos et al., 20012; Pillarisetti,
2015). The chronicity of HF suggests that primary care providers will have an increasing role in
the care of patients that may require ICD, and thus it is essential to recognize indications for the
device. Five studies in this review highlighted the consensus of lack of provider knowledge of
ICD indications as barriers to referral, and these limitations must be considered to enhance the
functional capacity of primary care providers in identifying these at-risk patients to provide them
with the opportunity for increased survival. This section will elaborate on the identified lack of
knowledge surrounding ICDs and guideline recommendations from health care providers.
Bernier et al. (2017) undertook a study in Alberta, Canada that explored the knowledge of
indications for ICD’s amongst referring physicians, as the group had identified low rates of
referral for eligible patients. A web-based survey made up of case-scenarios related to ICD

43
indications was sent to 799 physicians. Of these physicians, 14% (109) responded. The sample
comprised 14 cardiology residents, 60 internists, and 35 cardiologists. These participants
represented a group that would typically care for patients that were eligible for device therapy. In
addition to the case scenarios, practice characteristics and baseline demographics of the
responding providers were obtained in order to identify if barriers to referral differed among the
groups. Five case scenarios were presented and the primary outcome of complete guideline
concordance was achieved if all five case scenario questions were answered correctly. Finally,
secondary outcomes of provider perceptions of barriers to referral were assessed. The baseline
distribution of provider types was discussed as above, and furthermore, age, sex, location of
practice, and years of practice were identified. Statistical analysis was performed and reported
descriptively as percentage and counts, and logistic regression examined univariate and
multivariate links between the baseline characteristics of providers and concordance with ICD
guidelines. Interestingly, a total of 34% (37) physicians demonstrated complete guideline
concordance. The specialty of the provider differed significantly when achieving these correct
answers, with cardiologists and cardiology residents both answering 57% correctly, while
internists answered only 15% correctly. Therefore, when predicting guideline-concordant
answers, cardiologists had a p= 0.001 in a multi-variable analysis, while residents had a p =
0.007. When considering the perceived barriers for referral, internists reported a lack of
confidence in their knowledge of guideline recommendations (70%), while cardiologists tended
to have more concern for cost-effectiveness of the intervention (54%), and residents were
concerned about inappropriate shocks (64%). Additionally, younger physicians <40 years of age
were more concerned with their knowledge of guidelines, while those >40 years of age focused
their concern on cost-effectiveness. With this, the only significant reason for barriers to referral

44
between the specialties was lack of knowledge of ICD indications (p<0.001). These results
demonstrate that between different provider groups that should be more exposed to decisions for
patients eligible for device therapy, only one third were completely aware of guideline
concordant indications for ICD, even with access to specialist evaluation or further investigations
available.
A number of limitations exist in relation to this study, including a relatively small group of
respondents and the potential for bias resulting from the engagement of participants who may
have been those that were more comfortable with the guidelines already. However, despite this,
there remained a significant percentage that were able to correctly answer the case scenarios,
demonstrating the evidence of knowledge gaps for those with cardiology as a specialty. These
findings emphasize the need to improve knowledge translation and education for the primary
care provider group, as they will encounter a much larger group of patients with high-risk SCD
indications, and be responsible for the initiation of the referral process. The authors acknowledge
a similar study from Sweden that detected only 2% of family practitioners were aware of updated
ICD guidelines (Hübinette, Lund, Gadler, & Ståhlberg, 2014). Bernier et al.’s study was
conducted within two large university hospitals in Alberta, but only represent the knowledge of
this area of one province, which may reduce its generalizability across Canada or internationally.
The transparency and detail of graphed analysis of this study assisted in its quality, achieving a
Quality Rating Score of 8/10 (strong). This study highlights the opportunity to identify one of the
barriers to referral for ICD, which is provider knowledge. Further studies that may assist this
would include primary care providers such as GP’s and NP’s, and also focus on areas for
strategic education and clinical based pathways to assist in identifying those patients that qualify
for referral.

45
A similar study was performed by Castellanos et al. in 2012 in the U.S. This cross-sectional
survey was performed in order to yet again determine physician concordance with guidelines for
ICD referral, using the ACC/AHA/HRS 2008 guidelines. The methodology was similar to that of
Bernier et al. (2017), the authors undertook recruitment at a national level which recruited a total
of 1378 physicians from the American Medical Association Masterfile via a mailed survey
between June 3, 2009 and December 11, 2009. The survey consisted of 34-items as follows:
questions 1-5 evaluated the physicians understanding of current guidelines for primary
prevention ICD implantation, questions 6-13 explored reasons for not referring patients and
potential ICD-related complications, while questions 14-34 established the participating
physicians baseline demographics and habits within their practice. Strategies for participation
were employed including cash incentives and hand-written letters.
The methods of the statistical analysis were rigorous and thorough, and the re-categorization
of the multi-variable groups allowed for further understanding of potential directions for
education. Independent predictors of practice and respondent demographics were quantified with
a p value <0.1 that allowed detection of commonalities for discordant practice. Stata 11 was
utilized for this statistical analysis, and statistical significance was determined to be a two tailed
p < 0.05.
Responding participants represented three clinical specialty areas; general cardiologists
(n=468), internal medicine physicians (n=437), and family medicine physicians (n=473). Family
practice physicians were less likely to refer patients to a subspecialist for ICD consideration, nor
were they likely to provide care concordant with current guidelines. Physicians who were less
likely to refer their patients also reported more concern for peri-procedural infection and worry
that ICD discharges would be more painful. Of note, the only the only statistically significant

46
observation in this study was that primary care physicians that manage systolic HF
independently reported more often that evidence of a ventricular arrhythmia must be required
before primary prevention ICD referral would be considered (p = 0.013). These providers were
most discordant with current ICD guidelines when compared to their cardiologist colleagues, and
in addition, become discordant with Class I indications for ICDs as a primary prevention strategy
within national HF guidelines. In summary, this study found that there is a considerable amount
of family physicians that do not act in concordance with ID guidelines in order to refer patients
for device therapy. This helps to identify where potential barriers exist to appropriate utilization
of ICDs and as well, where to direct knowledge and interventions to increase adherence to
guidelines.
Limitations recognized in this study include the potential lack of representation of the
general referring physician population, however it did exemplify a national sample obtained
from the AMA Masterfile that represents all graduating physicians from the U.S. There was
concern that the survey presented a difficulty in discerning the difference between a lack of
knowledge and an informed opinion; regardless, the lack of ICD guideline adherence was the
ultimate primary outcome of this study and still reflects a critical gap in practice and a need for
enhanced education. Finally, the study did not explore actual device utilization rates, therefore
may have still actually underestimated the number of devices actually implanted after referral.
Strengths of the study include the response rate of the survey population which was high
at 64%, the detailed examination and critique of survey questions performed by researchers wellversed in device therapy, and confirmation of its quality by a group of physicians representing
the same focus as the sample population. The Quality Assessment Rating for this study was 8/10,
strong. The poor referral rates are unfortunate as ICD’s have demonstrated substantial efficacy

47
for improving mortality in eligible patients, and the standard of care for referral of these patients
often originates from the primary care provider. The results clearly identify a barrier of primary
care provider knowledge to ICD guideline adherence, which again is concerning as they are
often the initial step to further management and potentially life-saving therapy. Castellanos et al.
(2012) also noted that these findings reinforced similar results from studies in New Zealand and
the U.S., reiterating that rates of ICD guideline discordant practice from primary care physicians
that is likely attributable to unfamiliarity of indications, a lack of expertise in ICD management,
and a need for improved dissemination of ICD guideline education (McHale, Harding, Lever &
Larsen, 2009; Sherazi et al., 2010). Castellanos et al. stress that standards of care must be
understood by primary care providers to enhance strategies that have been shown to improve
mortality.
Zhang et al. (2015) similarly investigated factors that influenced ICD underutilization,
and amongst the multitude of variables that were investigated, physician-related factors arising
from a lack of awareness of clinical guidelines. Through a case control study within a tertiary
echocardiogram lab in Cedars-Sinai Medical Centre in Los Angeles, 502 patients were recruited.
Eligible patients included those with a LVEF < 35%, and as per the studies inclusion/exclusion
criteria were assessed against guideline-eligibility for the device. Patients whose EF improved
with medical therapy alone were removed. A total of 220 patients were used to compare and
identify differences. The group was split in half, 110 patients who had received ICDs, and 110
who were non-recipients. Charts were reviewed for documented reasons for non-implantation,
evidence of physician discussion of ICD referral or implantation, baseline demographic
characteristics, as well as clinical characteristics. Statistical analysis was performed to express
categorical and continuous variables, and the variables between the recipient and non-recipient

48
groups were compared with either X2 test of Fisher’s exact test. Statistical significance was
considered with a two-tailed P value of <0.05. Key findings that met statistical significance for
the recipient group were factors that were independently associated with ICD utilization and
included male sex (P = 0.01), NYHA symptom class of II or III (P = < 0.001), age <75 years (P =
0.01), and having private health care insurance (p = 0.03). This yields to a suggestion that a
particular group of patients is being more often referred for quaternary care including ICDs,
while there still remains a 43% lower than expected rate of implantations.
Comorbidity burden had little influence on ICD recipients versus non-recipients, not
including those on dialysis. When considering cardiac parameters, ICD recipients were more
likely to have established coronary artery disease (p = 0.03), and be on an ACEi/ARB (p = 0.02).
Zhang and colleagues identified discovered that greater than 50% of the ICD non-recipient group
had no documented discussion on ICD implantation; this corroborates the findings of the
previous two studies described in this theme (Bernier et al., 2017; Castellanos et al., 2012), and
likely reflects that education, awareness, and individual physician perceptions may heavily
influence the decision to refer for ICD, and that targeting this problem should guide future
studies and have a major role in increasing the ICD utilization rates for eligible patients.
Strengths of this study included detailed review of patient charts rather than ICD-9 codes.
Limitations included attendant bias and single center study. Clinical practice patterns may
be generalized to the area and not well representative of other settings. There was also no
evaluation of follow up including mortality in the two groups. The Quality Assessment Tool
rating of this article was 9/10, or strong. Likewise, while it remains somewhat unclear as to what
the true reasons were for non-referral (physician education, opinion, awareness), this study
strongly supports that fact that there are a large number of eligible patients who are never

49
referred for ICD consideration. The roles that are typically responsible for referral may trigger a
pattern bias for providers, who should be more aware of guideline-recommendations in order to
provide optimization of appropriate ICD use in order to decrease patient mortality. Zhang et al.
(2015) elude to the introduction of clinical practice tools that would allow incorporation of
automatic reminders, such as referral to local heart rhythm or function clinics with the result of a
low EF <35%, automated decision support tools, and quality improvement education programs
for the topic in primary care settings.
NPs must be adequately prepared and supported to screen and assess patients eligible for
ICDs. Bradfield et al. (2009), Herman et al., (2018), and Sherazi et al. (2013) all discuss aspects
of provider inexperience, discomfort, or lack of familiarity of device therapy as potential barriers
to referral and management. Despite the complexity of patients and conditions in primary health
care, knowledge of indications for ICD referral in addition to assistant tools to screen for them,
must be heightened to provide HF patients at risk for SCD the opportunity to receive this lifesustaining therapy. Identifying patients who are eligible for ICD referral should be as understood
as colon cancer screening, mammography, and dyslipidemia management. Overall, when
critically examining the studies included in this review, referring provider knowledge and
competence in adhering to national guideline recommendations on eligibility for ICD therapies
was recognized as an area that requires improvement. However, none of these articles examined
these attributes within the Primary Care NP role. This is a critical gap as NPs continue to provide
primary care to a growing proportion of the population and this issue will undoubtedly be an area
of need for ongoing support for NPs also.

50
Advanced Care Planning and Deactivation of ICDs
Complex decision-making in the management of the HF patient is a frequent yet
challenging task. As daily, weekly, monthly choices are required to be made, they are
intertwined with uncertainty, variation in therapies, access, cost, and ultimately, an unknown
prognosis. While the two preceding themes focused on choosing the right patients for ICD
implantation, consideration to those eligible must also be given to how and when the device
should be deactivated, as well as what the impact of shocks delivered may be (Hamel et al.,
2018). Lampert et al. (2010) describes how 20% of ICD patients receive painful, unnecessary
shocks in the last days of life, which contribute to great deals of stress and grief for patients and
members of their families. Similar descriptions by Brady (2016), Herman et al. (2018), Javaid et
al. (2018), Lee et al. (2017), Russo (2011), and Sherazi et al. (2013) demonstrate the urgent
nature to which the professional bodies must investigate ways to enhance the advanced care
planning around options for ICD deactivation in order to avoid this pain and distress, and provide
patients with dignified, respectful days and hours prior to end-of-life. The following articles
focus on the topic of planning for deactivation of ICDs, and explore potential barriers such as
timing of discussions, provider experience and knowledge, patient understanding and the
suggestions for improvements. These will now be discussed.
Hamel et al. (2018) conducted a large meta-synthesis to further understand key points
involved in the decision-making process for deactivation of ICDs. Common barriers during the
decision-making process were found and discussion was provided in order to determine more
optimal shared opportunities between patient and provider. After searching six academic
databases and applying inclusion and exclusion criteria, nine studies published between 2008 –
2016 including 25 132 participating adults (24 770 participants with ICDs and 362 physician

51
providers) were reviewed. Hamel et al. (2018) discovered three focused themes; 1) clinical
practice and management of ICD deactivation, 2) patient perceptions of ICD deactivation, and 3)
provider perceptions of ICD deactivation. The overriding consensus of the three themes was that
patient-provider discussions regarding the deactivation of an ICD are uncommon, and therefore
many patients fail to be informed of this choice during the end-of-life period. When timely
discussions of ICD deactivation are not held, patients and family members may be left with
complex decision-making during a stressful and anxiety provoking time. Moreover, if a patients’
condition deteriorates quickly, the caring provider may be inconsistent with the primary
provider, which can lead to disjointed care. Hamel et al. found that very few HF patients had a
clear written advanced care directive, and of those that did, only a small percentage had mention
of how to manage the ICD at end-of-life, highlighting the need for incorporating this discussion
into earlier advanced care planning – at the time of implantation would be preferable. Timing,
prognostication, involvement of family members, and current health condition at the time of
discussion were also found to be important factors that influenced the ability to concretely
determine an advanced directive.
Limitations of this synthesis include a relatively small group of studies by the same
authors, reflecting the lack of longitudinal data that is available on the topic of decision-making
specific to cardiac care. A team approach to the methodology of the study may have reduced the
risk of bias in this study, instead of a single author. The quality assessment rating of this study
was 10/10, strong. The conclusion of these findings encourages further investigation into the
perspectives of factors involved with complex decision-making in HF, including device therapy
and advanced care directives for these therapies. It would be beneficial to include the
perspectives of patients with more complex disease and their reflection on the decision-making

52
process to accept or decline treatments. These may be able to influence further opportunities for
enhancing care provider’s abilities to provide patient centered care.
Russo (2011) performed a systematic review that continues to seek out factors that delay
discussions around deactivation of ICDs in order to promote timely discussions of the subject
and foster patient centered care. The author acknowledged that despite the necessity of ICDs as
an essential role in the treatment of HF patients at risk for SCD, patients with terminal illnesses
that cannot be protected by the ICDs single role may suffer needlessly at end-of-life due to
painful shocks, and earlier discussions around deactivation are essential to prevent this distress.
Russo (2011) even discussed the practice of not preparing patients early on in treatment for the
possibility of the effects of the ICD shocks on quality of death as unethical, and providers should
be able to concisely describe the risks and benefits of the device before implantation. Commonlaw would support this decision to deactivate a therapy as a right to self-determination, to
withhold or withdraw treatment, which Russo felt providers must be aware of if there was
concern of legal or ethical ramifications. The literature search was undertaken using peerreviewed literature from databases in the time frame between January 1, 1999 – October 31,
2010. English papers with key-word search terms were included. 14 articles were accepted into
the final review. Five of the primary studies has the same principle author, and another two
primary studies had yet another same principle author.
Interestingly, the three main foci that were evident in this study were the same as those
from the Hamel et. al. (2018) study; clinical practice and management of ICD deactivation in
end-of-life care, provider attitudes regarding ICD deactivation, and patient attitudes regarding
ICD deactivation. Six of the studies that examined clinical practice reflected the important issue
that discussions regarding deactivation of ICDs are not common, and that cardiologists were

53
most likely to discuss this issue, while only approximately 25% of family physicians and
internists reported ever having this discussion. Remarkably, one of the studies found that 87% of
the 787 heart rhythm specialists interviewed had experienced a request for deactivation of an
ICD; while some of these requests were identified through a deactivation order, this meant that
the majority of the time there was not consistency in the patients’ primary care provider for the
decision to be made. Additionally, it was acknowledged that most often times ICD deactivation
was performed by an industry representative such as a device manufacturer or device technician,
rather than a direct care provider. If the discussions regarding ICD deactivation were to have
occurred, they were more often at end-of-life rather than proactively, and were even more
unlikely to occur if there was no formal policy or guideline in place.
Interdisciplinary approaches to deactivation discussions led to less inappropriate shocks
for a large number of patients. Physicians believe ICD deactivation should be a part of care
planning, but rarely do so, citing personal discomfort or inadequate knowledge and time
constraints. Most agreed that this discussion should occur prior to ICD therapy being implanted,
however very few felt it was their role to initiate deactivation discussions and assume this
responsibility. One provider even feared having this discussion would be like “shutting off the
hope”, (Russo, 2011, p. 29). Some would rather advocate for further life prolonging therapies, or
felt they required more guidance for end-of-life management of ICDs. Little research in the area
of patient attitudes towards ICD deactivation. 70% of patients identified their primary physician
as the main information source of their device. Unrealistic expectations or insufficient
knowledge of the role of the ICD was also cited as common, lending to the need for further and
comprehensive patient education prior to implantation. In a small group of the studies that
focused on patient attitudes towards ICD deactivation, it was found that in a number of

54
terminally ill patients of whom deactivation had been extensively discussed, they opted to not
deactivate their devices. Russo (2011) did find some contradicting evidence in this focus, from
patients who would not turn their device off despite multiple daily shocks towards end-of-life, to
a group of terminally ill patients who all chose to deactivate their device.
Russo (2011) summarizes his findings by highlighting the need for providers to act as
patient advocates despite discomfort and be able to inform recipients of ICDs with clear
knowledge of its purpose, as well as treatment options including deactivation. Influencing this
movement at a policy or institutional level may help to assist those providers who are less
experienced navigate the topic and support alliances between providers in order to offer
competent and informed care to help reduce stress and anxiety on both patients and families prior
to end-of-life. Limitations of this study are reflected in having only one author writing and
reviewing studies, as this introduces a higher risk of interpretation bias. However, the strength of
the study comes from the rigor of the databases searched, and the extensive timeframe of a
decade. These results would most likely be highly generalizable to other geographical areas, and
thus reinforces the importance of its findings. The study had a quality assessment tool rating of
10/10, strong.
In 2012, Tajouri et al. sought out to establish what percentage of patients with ICDs had
an AD in place, and also of those ADs, the prevalence of any that described management of the
ICD at the end-of-life. The retrospective chart review study was conducted in Minnesota and
consisted of a group of 420 patients that underwent ICD implantation in 2007. This year was
chosen in order to allow enough time after implantation to complete or update an AD (although
only 10% of the patients that were evaluated did so after implantation). No patients were
contacted during this data abstraction. Demographic characteristics of the cohort including age at

55
implantation, race, residence, and clinical characteristics were identified and analyzed for
statistical significance. Finally, the presence of ADs was evaluated, and in further detail explored
for directions for specific life-sustaining treatments. 30% of patients in this study had an AD.
Only 2% (2/126) patients with ICD and an AD mentioned the device; a great difference from the
patients who indicated their preferences for management of other life-sustaining therapies such
as potential tube feeds (46%), mechanical ventilation (17%), hemodialysis (7%), or
cardiopulmonary resuscitation (20%). 46% of patients with ICD and AD discussed pain control
and 30% discussed other comfort measures. When scrutinizing the personal characteristics of the
patients who had ADs, the only statistically significant patient variable was age; patients who
were significantly older at the time of ICD implantation were much more likely to have an AD (p
= <0.001). Clinical characteristics that were more prevalent with having an AD included a
comorbidity of dementia (p = 0.03), chronic obstructive pulmonary disease (p = <0.02), cancer
(p= 0.02) and those that had renal insufficiency (p = 0.005).
As this study spanned only a single year of patient ICD implantations at one institute, it
lacks generalizability to other geographic areas and populations; moreover, a great amount of the
study population were white males. The data is some of the oldest in this integrative review, and
since the data was produced there may have been changes to the uptake of creating advanced
directives. This data was also collected during a time prior to the Heart Rhythm Societies release
of the 2010 expert consensus statement on the management of ICDs in patients requesting
withdrawal of therapy on nearing the end-of-life, which may have resulted in lack of guidance
for some health care providers. Despite some limitation, the study had many strengths and
representing a very specifically defined population that echoes the nature of this particular
theme, and therefore resulted in the studies quality tool rating of 10/10, strong. The results of this

56
study create an incentive to make changes to private and institutional approaches for the care of
the patient with an ICD to create device specific directives for end-of-life management of the
therapy. This change should provide concrete support for providers to advocate for and educate
their patients on the role of ADs for ICDs in particular, in order to support ethically sound and
quality end-of-life care.
Amongst the articles that focused on ADs and deactivation of ICDs, there was only one
Canadian study performed by Habal et al. (2011), which explored the awareness of HF patients
to advanced care directives, and also to examine if they were using them. Based out of a large
Heart Function Clinic in Toronto, Ontario, 41 patients with HF including 19 of which had an
ICD were enrolled in the mixed-methodology, cross-sectional prospective study between July
2007 and December 2007. The study participants were given a 27-item semi-structured interview
by a trained research assistant, and took approximately 30 minutes to complete. Many were
closed ended questions that addressed the awareness of the participant on the use of ADs;
permission was obtained from the participants treating cardiologist to educate the patient on ADs
if they weren’t aware, and education was provided when necessary on the underlying principles
for ICD deactivation. In addition, questions were asked pertaining to the participant’s knowledge
of cardiopulmonary resuscitation processes, and their opinion on the value of discussing ADs
with his/her primary provider.
Habal et al. (2011) discovered findings similar to the other studies captured in the review,
including findings that reflected a lack of familiarity of ADs (Lee et al., 2017; Brady, 2016). For
example, 76% (31) of the study participants in this study did now know what an AD was, and of
those that knew what it was, only three could recall discussing this option with their health care
provider. Two of the 19 participants with ICDs had discussed ADs with their cardiologists had

57
also discussed the preference for deactivation of ICD. When asked by the researcher if they
would like to discuss ADs, 78% of the total participants affirmed they would like to have them
discussed, and preferably early in the diagnosis of HF. Furthermore, 47% of the patients with
ICDs declared they would choose to have their device deactivated if their condition were to
worsen, while 16% were undecided, 26% would not support deactivation, and 11% did not
answer. Similarly, Raphael et al. (2011) found that while the benefit from and need for and ICD
is reiterated often to most eligible patients, the discussions for deactivation are infrequent, yet,
most patients want to be involved in this decision early in the device implantation phase. The
findings resulting from Habal et al.’s study represent a significant demand for increased
communication between patients and their care providers with respect to ICD insertion and
deactivation, as well as the need for information exchange to facilitate shared decision-making
and support patient preferences and values. Both patient and provider related factors exist within
this dilemma, but are an essential area that requires improvement, especially within the context
of HF patients with device therapy. This topic should not only be discussed prior to the
implantation of the device, but also revisited as time passes and patients health status and
lifestyle factors change especially if the use of the device is discordant with the patients end-oflife goals. The smaller and largely younger population included in this study may give rise to
bias and may lack wider generalizability. However, the authors feel that the younger population
allowed for a more comprehensive analysis of the issue as it also may mean that the study
sample contained fewer people with cognitive decline or impairment, allowing more of the study
population that was able to recall important conversations and decision points regarding the ADs.
As the study interviewed only the patients, recall and awareness of previous AD discussions with
their health care providers may have been underestimated, yet, this emphasizes the requirement

58
to evaluate and produce clear and concise information for patients regarding ADs. Finally, as the
interview was done in a small amount of time, and some participants required education on and
AD during the interview, adequate time to process and consider the choices may play a role into
the validity of the answers. The quality assessment tool rating for this study was 9/10, strong.
Summary
This chapter presented the key findings from fifteen articles which discuss the role of the
primary care provider in managing HF patients who require ICDs. The three themes discussed
represent areas of primary care practice that can be generalized to the settings the NP will be
employed, and therefore the evidence is pertinent to the scope of the NP. The next chapter will
critically analyze and consider the findings of these articles and how they relate to the clinical
question posed in this integrative review. Following this, recommendations for education,
practice, and future guideline development will be discussed.

59
CHAPTER IV: DISCUSSION
The benefits of ICDs in eligible adult patients aged >18 years old with heart failure has
consistently proven to be an effective strategy to reduce mortality related to SCD (Pillarisetti et
al., 2015; Zhang et al., 2015). Guidelines in both Canada (2016) and the United States (2008)
have similar recommendations for primary and secondary prevention of SCD with device-based
therapy. Despite having had guideline recommendations in both countries and a multitude of
randomized control trials (MADIT, MADIT-II, MUST and SCD-HeFT) demonstrating reduced
mortality outcomes for the patient population with HF at risk for SCD, there is a significant rate
of underutilization and under-referral for this device. Of those eligible patients, the referral rates
that lead to implantation range from only 28-43% (Pillarisetti et al., 2015, Zhang et al., 2015).
Additionally, lack of advanced care directive and directions regrading deactivation of these
devices at end of life are a major shortcoming in the current care of these patients. Javaid et al.
(2018) discuss that up to 19% of patients with ICDs receive at least one shock in the last minutes
to month of their life. This is incredibly concerning as the threat of ICD therapy delivered within
the last days to hours of life is both futile and undignified, and the focus of comfort and anxiety
reduction could be improved by implementing ACDs during the time the patient is healthy. The
barriers that exist to improve referral and utilization rates, as well as implementing ACDs that
include directions for ICD deactivation need to be addressed and managed to more effectively
focus on patient-centered care.
This integrative literature review was inspired by my own clinical experience, through
which I observed that there was a lack of primary care support when engaging with patients
about this potentially life-saving device. The more robust engagement of primary care providers
has the potential to improve referral and utilization rates, as well as completing of ACDs

60
including deactivation directives, would decrease mortality from SCD and improve the practice
of end of life care for these patients. As presented in the previous chapter, the findings of this
integrative review strongly support the need to enhance ongoing education for primary care
providers, including NPs, on the guideline recommendations for ICD eligibility and referral, as
well as the education and timing on how to make decisions around end-of-life planning for this
particular patient population. This chapter will consider the findings of this review and discuss:
Primary Care Management of the Heart Failure Patient with an ICD. These sections will
specifically explore the core target issues related to ICD use including underutilization of ICDs,
provider knowledge, and ACDs. Further, this chapter will additionally present recommendations
for education, practice, research and policy development.
Underutilization of ICDs
ICD implantation to prevent SCD from sustained ventricular arrhythmias improves
survival in select HF patients (Bennett et al., 2017; Zhang et al., 2015). This therapy has evolved
significantly since its beginnings 40-years-ago into an important and widely accepted treatment
for the heart failure population. Unfortunately, there has been strong documentation that
underutilization of this evidence-based, cost-effective device is a critical issue (Bradfield et al.,
2009; Deyell, Tung & Ignaszewski, 2010; Ho et al., 2017). For example, in Alberta, Canada, the
rates of implantation for ICDs has decreased more than four times in the past decade (Bernier et
al., 2017). Despite this, there are anecdotal reports that rates of ICD use are starting to increase
(personal communications, Liz Mulvaney, July 25, 2019).

61
Influencing factors. Understanding why an evidence-based prevention therapy is
underutilized has led to many researchers investigating the factors that may play a role in the
varied referral rates for an ICD in eligible patients. When exploring key patient demographics,
such as race, sex, age, insurance, and comorbid conditions, there were relatively few
consistencies between studies, suggesting there may be more of a physician or hospital-related
factors that are impacting upon referral for ICD (Chae & Koelling, 2010; Matlock et al., 2011).
As demonstrated in this review of the literature, younger male patients were more likely to be
referred, demonstrating that there may be gender disparities resulting in an under-referral of
eligible females. Other patient characteristics, such as NYHA class and the presence of other comorbid conditions may also be influential. For example, Zhang et al. (2015) in their cohort study
of 110 ICD recipients found that those patients with NYHA classes II-III, composed 75.5% of
the group (P <0.001) and were the most likely to be referred, while those on dialysis were the
least likely. However, when considering who should be screened for ICD eligibility, there needs
to be an understanding of the guideline-recommended risk factors that are actually required for
referral, regardless of these characteristics identifying the most or least likely to be referred.
Furthermore, there is a need to understand which patients would most likely benefit (Zhang et
al., 2015). Thus, optimizing the use of ICDs as primary or secondary prevention strategies
requires consistent and evidence-based evaluation of risk factors for ICD. Based on this, the only
factors to consider other than eligibility criteria for the device, are age (although this should not
be a deterrent), comorbidities, and frailty (Bennett et al., 2017). A key recommendation arising
from this review would be a need to explore best practices for referral for ICD.
Decision support tools, checklists, and automated referrals may assist in increasing
referral and utilization rates in the appropriate and eligible patients, as well as potentially

62
decreasing rates of inappropriate implantation (Zhang et al., 2015). For example, Lewis, Carroll,
Birnie, Stacey and Matlock (2018) discuss the importance of incorporating interventions to
improve high-quality shared decision-making through the duration of referral and implantation
for patients who may benefit from an ICD. The authors of this paper highlight that despite
excellent evidence for mortality benefits in ICD use across the major trials, there is still a large
population of patients who were not well represented in the samples of these trials. They argue
that further research is warranted to explore risk stratification across more diverse populations in
order to fully understand how best to comprehensively determine eligibility for ICD. Lewis et al.
(2018) suggest that the use of written or video-based patient decision aids (PDA’s) may be useful
to both facilitate preparation for consultations on ICDs on to guide the decision-making process.
They suggest utilizing Makoul and Clayman’s (2006) essential elements for shared decisionmaking skills (see Appendix C).
Primary care providers need to be able to detail the criteria for ICD referral to address the
need of this therapy for their patients, and updated training may be required due to recent
changes in guidelines (Pillarisetti et al., 2015). Consistency amongst approaches to selection of
eligible patients would assist in guiding therapeutic choices, as well as assisting providers with
unique patient populations (Kusumoto et al., 2014). Integrating supports for ICD decisionmaking, such as practice support tools and decision-aids may be useful in addressing this clinical
challenge and in turn may improve patient access to evidence-based care.
Provider perceptions. “Let us seek opportunities to coach rather than dictate, to inform
rather than insist, and to listen rather than coerce” (Goetz, Grady & Yancy, 2017). This quote
eloquently describes the transition from relying solely on clinical guidelines to incorporate a
renewed focus on patient-centered care through shared decision-making.

63
While the use of current guidelines can offer simple and clear pathways to determine
which patients should be considered for ICD, the ‘real world’ decision-making process doesn’t
always fit in these neat and organized flow charts. For example, providers must account for
patient preferences but must also be able to perceive risk of SCD in their patients, which reflects
on both their own knowledge, experience, and understanding of current guideline
recommendations. This perceived risk of SCD includes both subjective and objective data, as
well as how the risk of complications from the device and its implantation may affect the patient.
For example, many providers have expressed concerns of peri-procedural infection, pain endured
from delivery of therapies, and psychological risks post-implantation such as depression, anxiety,
and even post-traumatic stress disorder (Matlock et al., 2011).
Communicating this risk to patients in order to obtain informed consent and undertake a
shared decision-making process may also pose its own risk. Providers that feel if they explicitly
describe all of the important risk factors involved in having an ICD, they potentially risk the
patients declining a much-needed therapy because of the way it was conveyed (Matlock et al.,
2011). Alternatively, if the positive aspects of the device are emphasized, such as the common
statement “think of it as an insurance policy for your heart”, or “it is a life-sustaining device”,
patients may understandable cling to these ideas considering their own mortality and not truly
hear the risks involved (Matlock et al., 2011, p. 34). Presenting a balanced discussion of the risks
and benefits of ICDs is complex and provider discomfort in difficulty discussing complications
of HF with patients is well documented (Schallmo, Dudley-Brown & Davdison, 2019);
researchers have found the discomfort may lie in feelings of failure for not finding curative
treatment, and therefore many of the discussions pertaining to death and dying are postponed
until health deterioration is imminent. However, in reviews such as Herman et al. (2018) and

64
Russo (2011), patients explicitly stated that their preference would be to discuss these risks and
complications proactively across the various stages of their illness, including before implantation
and during times without any significant health deterioration. Some health care providers may
also feel hesitant to communicate risk in cardiac patients due to a lack of experience or in
managing ICDs and HF, or as a result of knowledge gaps. This reinforces the need to enhance
primary care providers’ preparedness for management of HF patients with ICDs.
Finally, many providers in the studies described in this review expressed concern that
their ability to engage in a shared decision-making process was limited by the attitude of patients
who always took their providers advice and feared that the patient’s deferral to a provider’s
decision on the subject caused them to implement a more ‘paternalistic’ approach (Bernier et al.,
2017; Matlock et al., 2011). Providers must therefore be able to present information regarding
ICD therapies in a consistent approach that offers both risks and benefits of the device in an
approach that is not necessarily congruent with the current tone of the patient’s emotions. For
example, if a patient is scared or anxious about the risk of SCD, they must also understand the
known risks of the potential therapy instead of simply encouraging the life-sustaining benefits.
All of these considerations should be known to the patient prior to the implantation, and need to
be managed with a shared decision-making approach. HF continues to be a complex disease in
which providers and patients must make ongoing decisions regarding day-to-day and future
treatments.
While the decision-making process for patients with HF and other chronic illnesses is
underrepresented in current literature, it is clear that this process is best shared with the provider
and patient’s family at an early time in the disease course, allowing for longitudinal goal
planning with recurrent reflection on choices to guide future decisions (Hamel et al., 2018).

65
Developing further evidence on shared decision-making would therefore have direct benefits for
both patients with HF and health care providers, but may also resonate to other facets of chronic
disease management. For example, similar decision-making processes occur in other chronic
illnesses, such as patients living with renal failure patients who may be considering dialysis and
transplant. Alike to the care of HF, decision-making for renal care requires well planned
interventions that include careful timing of discussions that transmit information and
prognostication of the illness (Morton et al., 2010). A shared decision-making approach within
the primary care setting that involves collaboration with the patient and the professionals in an
attempt to foster patient-centered care and create an interactive, dynamic decision-making
process is critical to ensure informed consent and patient engagement in complex diseases such
as heart failure, renal failure, cancer, and mental health disorders (Pavlo, O’Connell, Olsen,
Snyder & Davidson, 2019).
Guideline discordant practice. Failure to understand guidelines not only poses a threat
to limiting a patient from a potentially life-threating event such as SCD, but also causes
overutilization of the therapy in patients that do not actually meet the guideline recommended
criteria (Castellanos et al., 2012). While ICD implantation may be considered unique, invasive,
or even risky by some providers, there is little difference when compared to other standard
screening practices. For example, colonoscopy requires referral to a specialist for consideration
of the procedure, and then the patient must undergo sedation, instrumentation, and most certainly
sustains risk of negative health-related outcomes. The whole process however, is considered
quite normal and most patients and providers agree that the screening and procedure are
worthwhile to help prevent devastating outcomes of colon cancer. As discussed earlier by
Castellanos et al. (2012) screening and consideration of HF patient’s risk for SCD should be no

66
different than other primary interventions such as colonoscopy and mammography. These
interventions improve mortality and as such primary care providers need to screen for candidates
that can benefit from this important therapy in order to detect early disease and minimize the
potential of disease-induced negative outcomes. However, Castellanos et al. (2012) provided
some striking evidence that 28% of health care providers never refer patients for consideration of
an ICD, and of that percentage, 7% of these were cardiologists. The authors multivariate analysis
of screening and referral practices highlighted the finding that primary care providers are some
of the most discordant to guideline recommendations when it comes to referral for ICDs, which
also reduces their capacity to practice guideline concordant Class I systolic heart failure
recommendations.
The breadth of knowledge and guideline recommendations are overwhelming in primary
care and most providers have a diverse caseload of patients, meaning that HF and ICD therapies
may occur infrequently. Therefore, it is somewhat difficult to gain an understanding of the term
guideline discordance as it implies not following a rule; however, many providers may simply
not be aware of the most current recommendations due to frequent updates that are available
across a magnitude of clinical guidelines. For example, a self-perceived lack of awareness of
current guidelines accounted for many non-referrals in Bernier et al.’s report (2017) on ICD use.
Clear documentation of subtle distinctions for guideline discordant practice is of the
utmost importance for providers to understand when recording reasons for referral for or refusal
of an ICD (Kusumoto et al., 2014). Multiple studies within this review cited either only “refusal”
or “no documentation” of patients who did not receive ICDs (Bradfield et al., 2009; Pillarisietti
et al., 2015; Zhang et al., 2015). If there had been additional rationale for these nonimplantations, the results of the studies may have highlighted the varied reasons for guideline

67
discordant practice, instead of suggesting a lack of provider knowledge or a difference in opinion
of the therapy. In addition, the multiple studies that are captured in this review focus on a
relatively homogenous group of patients that display a routine group of indications for ICD
therapy (Bardy et al., 2005; Zhang et al., 2015). As many providers and even public can
appreciate, the study populations do not always provide a clear composite of the variations in the
patient population for each provider in different geographical areas. Thus, the understanding of
the unique nature of patient populations, as they relate to those included in many of the clinical
trials, may emphasize the importance of the providers clear understanding of current
recommendations of referral for ICD and how these may be applied in practice (Kusumoto et al.,
2014).
In summary, this review has highlighted that referral and implantation rates of ICDs are
strikingly low, considering ICDs are a class I therapy for prevention of SCD for at risk HF
patients. There appears to be many provider and patient factors that contribute to this guideline
discordant practice, and these gaps in care must be reinforced with education and support in
order optimize appropriate utilization of ICD therapy.
Provider Knowledge
High rates of non-referral to cardiac or device specialists is a continued concern leading
to the above-mentioned underutilization of ICDs. The main issue that emerged in this review was
that provider knowledge is one of the foremost problems causing barriers to ICD use (Bernier et
al., 2017; Castellanos et al., 2012; Matlock et al., 2011; Pillarisetti et al., 2015). The lack of
provider knowledge segregated into three major themes: 1) Lack of understanding of current
guidelines for ICD referral, which will be discussed below, 2) a lack of confidence and
familiarity with the device, and 3) concerns of the ethical and medicolegal nature of the device.

68
As previously discussed, the main reason for non-implantation of ICD in eligible patients,
which exceeds the number of ineligible patients and patients who refused ICD, is non-referral
(Bradfield et al., 2009). For example, Bernier et al. (2017) supported this finding and added that
only one-third of providers in their study were aware of guideline concordant indicators for
eligibility for ICD referral. In Bernier et al.’s study, there was a 32% mortality rate in patients
who did not receive ICDs despite having no contraindications for the implant (2009). Even
despite access to specialists, the differences in knowledge in primary care providers when it
comes to ICD referrals is concerning, as they arguably may play the most essential role in
identifying patients at high-risk of SCD, as they encounter a much larger portion of the
population. Variations in primary care screening are certainly not isolated to cardiac conditions.
For example, Coulton et al. (2017) performed a randomized screening evaluation of 3562
primary care patients who were assessed using a targeted or a universal approach for at-risk
alcohol use. The targeted population, those who presented with a health history of hypertension,
mental health problems, gastrointestinal issues, minor injuries or new patient status had a higher
odds ratio of screening positive for at-risk alcohol use. However, after evaluating the universal
screening approach, it was found that 79.7% of the patients that screened positive for at-risk
alcohol use did not have any of the issues in the targeted population, and would have been
missed for potential interventions if screened only with targeted criteria.
Potential solutions to enhance provider knowledge are presented in the recommendations
section. Enhancing the ease of screening and overcoming barriers to consultation, along with
facilitating opportunities for shared learning and collaboration between primary and specialist
healthcare providers through interdisciplinary workshops or educational sessions may assist in
developing a more concentration on a team approach. Empowering primary care providers with

69
essential knowledge of ICD eligibility and enhanced knowledge translation tools will help to
identify patients who would benefit from referral and potentially receive this life-sustaining
therapy.
Provider education and knowledge translation. In congruence with the findings of this
review, educating patients is a key component of the providers’ role in managing patients with
HF and supporting ICD decision-making. It is extremely important that the primary care provider
be able to address and correctly inform the patient about the role of the ICD, the indications for
it, benefits and risks, and identifying who else should be involved in the decision-making process
(Deyell et al., 2010; Hamel et al., 2018; Lee et al., 2017). While it is easy to assume this role
belongs to the cardiac or device specialist, it is well understood in literature that patients may not
always retain advice or understanding in the first, or even second or third visits. Thus, the
primary care provider is a valuable resource for patients to express their true concerns and ask
additional questions about the device and its therapies.
Ensuring the patient is fully informed about ICD therapy is part of effective patient
centered care as it can be complicated and comes with potential risks (Matlock et al., 2011). As
reported in this integrative review, clarity regarding the role of the device was a frequent
misunderstanding for recipients and their families, and it is essential that patients are able to
understand an ICD will not fix or cure their heart failure, and nor will it prevent the progression
of their heart disease. For example, a study by Lee et al. (2017) that was captured in this review
describes how unrealistic expectations from some patients lead them to believe that their ICD
would improve their heart function through each shock, and therefore should never be turned off.
The sole purpose of an ICD is to protect against fatal arrhythmias and cannot improve other
symptoms of HF. Further information regarding receiving shock therapies, implications for

70
driving, employment (considering the role and duties), and avoidance of certain machinery can
all affect a patient’s quality of life, and deserve an in-depth exploration on to approach this
therapy for individual patients.
Advanced Care Planning
Advanced care planning could arguably be one of the most important decisions in patient
care. While advanced care planning was not a major focus of this review, it emerged as one of
the key approaches to optimizing decision-making around the ICD. The communication between
the provider and patient is essential to ensure that the patients’ decisions regarding end of life
treatment is facilitated, and that the exchange of information is fostered with a trustworthy
interpersonal relationship (Habal et al., 2011). Focusing on the preferences and values of the
patient’s quality of life should be a priority over a sole purpose of following evidence-based
medicine. Multiple studies and media have demonstrated the undignified occurrence of patients
with ICDs who are painfully and repeatedly delivered shock therapy within the last hours to days
of life (Brady, 2016; Herman et al., 2018 & Sherazi et al., 2013). A case report by Nambisan and
Chao (2004) describes a 59-year-old woman dying in hospital of primary lung carcinoma with
metastases, with a completed “Do Not Resuscitate” order, receiving 38 episodes of shocks from
her ICD during the six hours before her death while she was still conscious. The report describes
a lack of direction for the ICD in the advanced care directive and lack of protocol in the clinical
environment for deactivation of an ICD as contributing factors to this unfortunate situation.
While ICDs play an important role in preventing the patient from SCD during years of quality, it
is crucial to consider the purpose and intention of an ICD during the end of life experience
(Svanholm et al., 2015).

71
Advanced care directives. Despite the substantial implications for ICD therapy and its
efficacy, there is an unfortunate trajectory of heart disease that will ultimately lead to an
inevitable end of the patient’s life, and if not heart disease, then the occurrence of other illness
unrelated that would not benefit from ICD therapies (Tajouri et al., 2012). Hamel et al. (2018)
reinforced that primary care providers should engage in discussions early in a HF diagnosis in
order to increase the awareness of options for the patient, as well as promoting clear
documentation of the patient’s health care directives, identification of a substitute decision maker
should their cognitive function decline, and exploring resuscitation preferences. “A surrogate’s
ability to predict a patient’s wishes is only moderately better than chance” (Robinson et al., 2012,
p. 231). While ACDs have become more commonplace for health care providers, especially
within tertiary settings, Tajouri et al. (2012) and Habal (2011) demonstrated the lack of
awareness and discussions between patients and providers on the subject. When patients are
shared between care providers, such a primary care provider and a cardiologist, the
understanding of whom is ultimately responsible to assist the patient in creating an ACD may not
be clear.
While a specialist will have expertise in the particular diagnosis and intervention, ACDs
should ideally be initiated by the primary care provider to facilitate continuity of care.
Furthermore, these providers are most likely to have greater knowledge of the broader medical,
social, and familial factors of importance to the patient. Once an ACD is established, it should
include directions for life-sustaining therapies such as ICDs that will be described below. In
addition, as patient’s preferences change over time with age and experience, it is essential that
this document be reviewed and updated in order to prevent a discordance between the patient’s
current wishes and the physicians’ interpretation of older directives (Brady, 2016). In BC, the

72
ease of access to ACD planning through print, mail, brochures, and online resources should only
encourage providers to start this process early with patients. Having these in printed form, or a
printed link to provide as reference may increase the uptake of patients initiating this document.
Timing and prognostication. Multiple authors captured in this review have identified
that early dialogue and conversations during healthy times are crucial in order to increase
collaboration between patient and provider, and shared decision-making has been a recurrent
theme (Hamel et al. 2018, Herman et al., 2018; Lee et al, 2017; and Russo, 2011). Brady (2016)
discusses eight potential triggers for appropriate timing of discussions related to management of
the ICD: 1) Before implantation or pulse generator change; 2) If the patient has had frequent
hospitalizations (3 admissions in <6 months); 3) When the patient adopts a DNR order; 4) After
an ICD shock, 5) at admission to hospice; 6) If there has been a significant functional decline; 7)
If the patients symptoms are NYHA IV or American Heart Association stage D; and finally, 8)
With intolerance of guideline-directed HF medications or hemodynamic instability. These should
be integrated within current ACD and clinical records to trigger providers to review directives
and amend as needed.
Deactivation of the device. Deactivation of the device is a critical topic and attention is
needed to explore situations of deactivation during the early clinical encounters. In this literature
review, two studies identified that there is a lack of consistency in ICD deactivation discussions
and practices. (Lee et al., 2017; Russo, 2011). In BC, the HF Network provide clear and
comprehensive pre-printed order sets and documents to support referral, consent, and an order
set for deactivation of an ICD (BC HF Network, 2019). These tools are valuable in assisting with
decision-making and both patients and providers need to be aware of them. Most patients in the
studies included were not aware of the option to deactivate their ICD, nor did they understand

73
the impact it may have on their end-of-life experience (Herman et al., 2018). Through the
discussions with patients and families within reports in this review, it is clear that as a patient or
a family member, the best time to discuss potential deactivation of an ICD is when the patient is
well, and more directly, prior to implantation.
The most common timing of deactivation discussions occurs unfortunately once there is a
severe decline in health or diagnosis of another terminal condition. While a provider cannot
determine when end-of-life will occur for most patients with HF, and they may feel very
uncomfortable doing so, informing the patient of their end-of-life options and clearly
documenting them for other health care providers and staff is important dialogue that must occur
to ensure the patient’s wishes are respected, and promote a more peaceful, dignified death (Habal
et al., 2011; Herman et al., 2018, Shallmo et al., 2019). One of the papers not captured in this
review by Shallmo et al. (2019) suggest that increasing communication and skills training for
providers along with a more positive attitude towards palliative approaches can improve overall
quality of life for the patients they care for. Providing a palliative approach is an integral part of
primary care practice, as it encompasses a holistic approach that supports the physical,
emotional, spiritual and social needs of an individual through shared decision-making across the
lifespan. An ACD that includes management of the ICD is a valuable tool for both health care
providers and family members, as it clearly demonstrates the wishes of the patient during a time
they were well enough to reflect on the goals of their care (Lee et al., 2017). Furthermore, while
not all patients will choose to deactivate their device, most want to be informed of the option of
deactivation (Brady, 2016).
There are considerable provider barriers to facilitating discussions regarding deactivation
of ICDs. First and foremost, a lack of experience and knowledge surrounding the legal and

74
ethical aspects of deactivation result in fewer discussions on deactivation in three of the studies
in this review (Herman et al., 2018; Russo, 2011; Sherazi et al., 2013). Kay (2006) well
emphasizes the right to self-determination in current law and reviews that “The individual’s right
to refuse treatment, including her or his right to withdrawal or withholding of treatment, is
considered a constitutional right based on the right to self-determination” (Russo, 2011, p.28).
Reviewing consent to the duration of any treatment should be included in the ongoing dialogue
between patient and provider. As previously mentioned, while there are no direct national
guidelines on the deactivation of ICDs, three panels of experts including the ACC, the AHA, and
the HRS recommend consistent, early and comprehensive discussions pertaining to end of life
issues for HF patients with ICDs that include the consequences and alternatives to deactivation
of the device and a physical documented order for the management of the device if the patient
chooses a do-not-resuscitate (DNR) order (Russo, 2011). A guideline to assist providers
regarding device management at end-of-life within health regions has been developed and is
easily accessible on the BC HF Network website (BC HF Network, 2019). These exemplary
local documents to ICD deactivation through a patient centered care approach are emerging in
the frequency of their use as providers become more aware of them, but need to be integrated
more in practice, particularly in primary care. Through these actions, providers will be able to
adopt an approach to a difficult topic in order to support patient-centered end-of-life care for this
patient population.
In summary, a critical appraisal of the literature captured in this review supports the
involvement of primary care providers undertaking a larger role in the management of the adult
HF patient with an ICD. By increasing referrals to cardiologists or device specialists according to
guideline recommendations, and actively participating in early advanced care planning including

75
the consideration of deactivation of the device, patients at risk of SCD death from HF related
arrhythmias may benefit from this life-sustaining therapy, as well as be supported in a patientcentered approach to experience a dignified death respecting the patients values and wishes.
Appropriate utilization of ICDs requires knowledge and collaboration of all providers, who
should work in a team-based approach to supporting the patient’s needs. NPs in BC are caring
for a larger number of the primary care population, and thus are exposed to a greater number of
the population that will be eligible for this device. The extent of education on HF and device
therapy specifically is minimal in most family practice programs, and thus additional education,
support, and direction is needed to ensure this group of patients receives comprehensive care.
Overall, primary care providers, including Family Nurse Practitioners, are uniquely
situated to care for HF patients with ICDs as they draw from experiences across the lifespan, and
take into consideration not only evidence-based practice, but also the importance of the shareddecision-making process across the continuum of care the HF patient will face with the
complexity of the disease. In this review, key gaps related to referral, utilization and advanced
care planning were identified. Based on these findings, I will present some recommendations for
further areas of research, education, policy and practice development aimed at optimizing the
management of ICD patients in the primary care setting.
Recommendations
The results of this integrative literature review have brought to light recommendations for
areas of improvement for the primary care provider. These areas include practice, educational,
research, and policy recommendations. It is certain that NPs as well as other primary care
providers, will have some of the greatest exposure to patients that may require ICDs and

76
therefore are key stakeholders for implementing these improvements for management of the
patients requiring the device.
Education and Continued Professional Development
Throughout the literature, the emphasis on lack of education recurred frequently for both
provider and patients (Bernier et al., 2017; Herman et al., 2018; Russo, 2011). The main reason
for non-implantation of ICDs was non-referral, which was frequently identified as a result of a
lack of confidence in the referring provider’s knowledge for the indications for ICD (Bradfield et
al., 2009). Additionally, patients and their families must be supplied with adequate and correct
information to improve uptake and understanding of the value of this life-sustaining tool
(Swedberg et al., 2008). While local, national and international guidelines are available at the
click of a button today online, adherence and understanding of the application of these
recommendations should still occur on an interactive basis as a collaboration between
professionals and the patient. Heart failure specialists, electrophysiologists, internal medicine
and primary care providers must be able to cooperate and participate in knowledge translation in
order to improve care, increase the cost effectiveness of treatment, and potentially reduce
mortality and morbidity in their patients (Swedberg et al., 2008).
Interventions to improve provider education on guideline recommendations for
identifying patients eligible for referral for ICDs need to be developed with an understanding of
potential barriers and facilitators facing healthcare providers, along with strategies to optimize
uptake into practice (Grimshaw, Eccles, Lavis, Hill & Squires, 2012). To enable more effective
knowledge translation, a variety of modalities and strategies may be needed that target the
provider, health system and policy levels. Multifaceted interventions that have considered local
barriers to change and enhancing social networking between necessary providers may be more

77
effective than any one single intervention, even though more-costly. For the purpose of this
review, a multifaceted interventional approach to improving ICD referrals could include formal
education from local experts to influence referral behavior, mail outs as two of the primary
research studies did (Bernier et al., 2017; Castellanos et al., 2012), and may also include
reminders of resources such as BC Guidelines.Ca: Chronic Heart Failure – Diagnosis and
Management (2015) and the Canadian Cardiovascular Society/Canadian Heart Rhythm Society
2016 Implantable Cardioverter-Defibrillator Guidelines (Bennett et al., 2017). Providing formal
modes of education to support provider decision-making such as workshops and teaching
sessions, are necessary to address these gaps in knowledge. Mail outs, whether electronic or
physical, have the potential to increase awareness of a topic and the providers own comfort of
that subject in a non-threatening way at a time and place that is comfortable for them.
Reminders, which include reference to guidelines and the use of decision support tools are
helpful and easy prompts to recall important information (Grimshaw et al., 2012).
Clarification of the legal and ethical aspects of device care at the end of life remains an
area that many providers are uncertain of (Sherazi et al., 2013). Workshops, lectures, and email
newsletter style mail-outs that detail the common questions of concern for providers regarding
these issues with well-informed and guidelines recommended answers may help to inform those
that had deliberated over these concerns (Javaid, Squirrell & Farooqi, 2018).
Providers should also be comfortable with local support strategies and how to access
supports when engaging in complex clinical decision-making. An example of this includes the
BC-based Rapid Access to Consultative Expertise (RACE) Line demonstrates an appropriate
strategy for a provider who is unsure or needs reassurance in their decision to reach out to a heart
failure specialist, or cardiovascular risk specialist, to review this decision and gain additional

78
knowledge to use for future patients. It is essential for NPs as primary care providers to be able
to quickly assess the individual patient on their risk factors for SCD and recognize when further
screening and/or referral is necessary based on this clinical knowledge and judgement. Having
additional education to support practice decision-making, as well as fostering ongoing
competency, is critical and has the potential to improve the delivery of evidence-based patient
care.
Practice
Practice supports are critical when attempting to improve health outcomes in patients
with HF who require ICD. Based on the findings of this review, two areas for practice support
were identified. These are implantation and referral, and deactivation and end-of-life care. These
will now be discussed.
Implantation and referral. The delivery of care to a patient with heart failure require
ICD therapy should ultimately occur through the joint effort of multiple care providers. The
successful practice depends on collaboration between all providers as well as the patients, family
members, and others engaged in the planning and delivery of healthcare (Swedberg et al., 2008).
In the current literature, the main reason for not implanting an ICD in a patient who is eligible
for the therapy is non-referral (Bradfield et al., 2009). Primary prevention screening and
management strategies such as mammography, colon cancer screening, and daily aspirin for
those with certain risk factors are second nature to primary care providers (Castellanos et al.,
2012) ICDs are also included in the application of treatments to improve mortality in select
patients as a primary prevention strategy, and thus should also be considered a standard of care
when approaching screening and identification of health promotion. Risk factors such as EF
<35% and history of documented VF/VT, should all trigger the provider to screen further for

79
ICD eligibility. For example, when reviewing any patient that has completed echocardiography,
any EF other than normal should be documented with a date and value on the face of the
patient’s chart or electronic medical record as a reminder to review risks and eligibility.
Practice recommendations must occur concurrently with education of the provider and
the patient. Providers must be able to approach patients who are eligible for device therapy in a
consistent manner and according to guidelines. Likewise, patients should be provided with
adequate information in order to be able to make informed decisions about ICD uptake
(Swedberg et al., 2008). If education during a single visit is not adequate, another visit should be
scheduled after the patient has had some time to process, discuss with family, and develop
questions specific to their own health. If a patient declines therapy, revisiting this option during
future visits or in response to a changing health status is imperative. In addition to the clinical
interaction, hard copies of educational materials regarding SCD and ICDs should be made
available to the patient. Finally, stronger alliances between primary care providers, palliative
care experts, and cardiology and/or electrophysiology specialists can foster guidance and support
surrounding all issues of ICD implantation, management, and deactivation (Russo et al., 2011).
Deactivation and end-of-life. In a similar fashion to approaching eligibility, deactivation
of ICDs must be approached consistently in practice (Herman et al., 2018). Patients have the
right to be made aware of their “right to refuse or request the withdrawal of a medical
intervention that is ethical and legal” (Sherazi et al., 2013, p.36). The patient and provider should
first have these discussions occurring at the preimplantation stage and should directly address
end-of-life planning within the context of the patient’s goals and quality of life. These
conversations should then again occur at regular intervals, as well as following an ICD shock or
in the event of any life-limiting diagnosis. As patient preferences may change over time with

80
new knowledge, experiences, alterations in health and quality of life, these discussions are
important and can lead to appropriate and timely decisions around ICD deactivation. Likewise,
effective deactivation strategies can reduce suffering and improve quality of life for the patient.
Advanced care planning support and advice should be provided to every patient to bring to light
considerations for their individual end-of-life values, preferences, and goals, clearly documented
in a written advanced care directive (Sherazi et al., 2013). In BC, many health authorities
including Interior Health utilize the Medical Orders for Scope of Treatment (MOST) document
in order to clarify resuscitation goals (Interior Health, 2019). This document could be adapted to
allow space to indicate an addendum to the specific interventions and/or supporting
documentation that includes direction for an active ICD and supporting documentation of any
deactivation requests. Finally, ongoing educational opportunities must exist for providers to
review these deactivation procedures, as these will change and are updated over time.
Research
Most of the data available from the primary studies included in this integrative review
were based on retrospective cohort analysis or chart reviews. As such, opportunities for further
research exist. Future studies exploring patient decision-making processes after being offered an
ICD and their understanding of ICD therapy, as well as understanding reasons for declining
ICDs, are warranted. The role of family members in the influence of these decisions, and patient
barriers to ICD therapy are also valuable areas to investigate (Matlock et al., 2011). In addition
to patient barriers, factors that influence the utilization of ICDs, either in primary care or specific
geographical areas (such as rural and remote areas) would be of interest. The integration of a risk
factor stratification tool for primary care providers to screen for eligibility for ICD, along with a
tools for guidance on deactivation management, would be essential to improve ICD care in

81
practice (Zhang et al., 2015). Finally, longitudinal studies of patients that receive frequent shock
therapy, as well as prospective studies exploring end-of-life management for HF patients with
ICDs have not been performed, and could provide further direction on how to optimize shared
decision-making between the primary care provider and the HF patient (Hamel at al., 2018).
Policy
Recommendations for policy changes at the health system level should be considered in
order to support clear and concise communication around ICDs. This could include the further
integration of provincial or regional protocols and tools such as the BC HF Network End of Life
Tool Kit (2019) to assist providers in their practice and support providers to have the sometimes
difficult and uncomfortable conversations about life-sustaining therapies as end-of-life care. The
ongoing integration of such protocols and policies will also hopefully improve patient care and
reduce any potential moral distress on the part of the care provider as well. (Brady, 2016).
An important step to make progress in the primary care management of HF patients
requiring ICDs is to create and integrate interprofessional device care pathways, that include all
the stages of contemplation, referral, implantation and deactivation. These may assist providers
to work collaboratively to deliver high quality patient care and may can form standards of care
for these patients (Bernier et al., 2017). In addition, national guidelines require updating to
include specific recommendations aimed at improving quality of life, such as specific strategies
to minimize harm due to painful and undignified painful defibrillations during the last hours of
life. Likewise, mandated protocols input into clinic and hospital admission patient records would
clearly identify the patient’s deactivation wishes should they come to a circumstance that they
require this support, as well as non-urgent and urgent deactivation pathways for primary care and
acute care personnel to be trained in. Finally, it is essential to begin incorporating specific

82
management of ICDs for end-of-life care in advance care directives, preferable prior to
implantation, and at the very latest just before the signing of a Do Not Resuscitate order (Tajouri
et al., 2012). All of these require distinct policy-based approaches to the development of more
effective and evidence-based processes of care.
Limitations
This review of the literature on ICD management in primary care has provided some key
recommendations for education, practice, research and policy. Strengths of this review include
the development of a comprehensive literature search strategy and the use of systematic process
for extracting and analyzing data. Despite many strengths, some important limitations exist.
First, the main limitation of this review is related to the varied nature of the available evidence.
Most of the available studies are retrospective in nature and there is a lack of prospective or
randomized control studies evaluating the shared-decision-making process of both primary care
providers referring for and patients eligible for ICDs. Thus, challenges existed when reviewing
studies that completed retrospective interviews or chart analysis because the distinct decisionmaking process, including records of discussions surrounding ICD implantation referral or
refusal, were often not documented. This made it challenging to directly assess gaps in
knowledge and practice. For example, Bradfield et al. (2009) notes the possibility that referral
may have been made prior to a patient’s refusal, and then not documented. Likewise, in the
qualitative studies that were available, there were almost no interviews with patients who had
declined an ICD.
In addition, there were other methodological considerations that may impact upon the
assessment of the evidence and its application in practice. For example, many of the individual
studies were only single center, and therefore this limitation may have led to selection bias. Of

83
the observational studies included, unmeasured confounders may have potentially influenced the
results making it challenging to draw concrete conclusions. In the survey studies, low response
rates may have resulted in selection bias, as the providers that felt more confident in their current
knowledge of the topics might have been more likely to respond. This being said, there were still
some significant knowledge gaps that lead to conclusions similar to this integrative review. It is
difficult to be certain that these knowledge gaps stem from guideline discordant practice due to
lack of education, or if the stem from personal opinion of the provider.
Finally, the most significant limitation in this study may have been that none of the
articles included focused specifically on a NP as the primary care provider. Thus, there are no
available studies that explore the process of managing ICDs in the primary care setting from this
perspective. However, with the increasing scope, autonomy and utilization of NP practice in
Canada, it is difficult to refute that NPs as primary care providers, will ultimately see very
similar populations to primary care physicians and providers that were referred to in the majority
of the studies included in this review. Thus, the concerns and recommendations will likely still
apply to NP practice and will hopefully assist these providers in providing patient-centered and
evidence-based care for patients with HF that require an ICD.

84
Conclusion
Heart failure patients at at an increased risk for SCD may be eligible for device therapy
with an ICD for primary or secondary prevention of fatal arrhythmias. Despite national
guidelines demonstrating repeated reduction in patient mortality with device therapy, the rate of
referral for and utilization of ICDs from primary care providers remains suboptimal. In addition,
there remains a large practice gap in advanced care planning for end-of-life options for ICD
deactivation. This integrative review explored the practice of Primary care management of the
heart failure patient requiring an ICD. A search of key databases and application of eligibility
criteria yielded a final cohort of 15 papers. Primary care providers should be able to screen
patients at risk of SCD for device eligibility, and work in collaboration with the team to offer
patient-centered care and a shared decision-making approach to determine if this is the
appropriate therapy. Furthermore, the primary care provider can influence the quality of ICD
patients end-of-life care by ensuring an informed ACD is completed with specific directions
regarding device deactivation is completed.
While this review has focused solely on ICDs and the role of primary care provider, the
implications of the recommendations should span to other aspects within primary care
management of chronic illnesses. As medical interventions become more advanced, with
multitudes of options and applications, primary care providers will require informed updates on
the patients that are eligible for these products, and guidance on when and who to refer to for a
more specialized assessment. Recommendations for practice, education, research, and policy
have been made to support the role of primary care providers, such as Nurse Practitioners, in the
management of HF patients requiring an ICD.

85
Finally, while all of the advances may be beneficial to the care of patients with significant
health challenges, the focus of informed patient-centered care should be of the utmost
importance in the practice of the primary care NP.

86
References
Arnold, J. M. O., Liu, P., Demers, C., Dorian, P., Giannetti, N., Haddad, H., . . . Canadian
Cardiovascular Society. (2006). Canadian cardiovascular society consensus conference
recommendations on heart failure 2006: Diagnosis and management. Canadian Journal
of Cardiology, 22(1), 23-45. doi:10.1016/S0828-282X(06)70237-9
Bardy, G. H., Lee, K. L., Mark, D. B., Poole, J. E., Packer, D. L., Boineau, R., . . . Sudden
Cardiac Death in Heart Failure Trial (SCD-HeFT) Investigators. (2005). Amiodarone or
an implantable cardioverter–defibrillator for congestive heart failure. The New England
Journal of Medicine, 352(3), 225-237. doi:10.1056/NEJMoa043399
BC Heart Failure Network. (2019). End of life tools. Retrieved from
http://www.bcheartfailure.ca/for-bc-healthcare-providers/end-of-life-tools/
Bennett, M., Parkash, R., Nery, P., Sénéchal, M., Mondesert, B., Birnie, D., . . . Talajic, M.
(2016). Canadian cardiovascular society/Canadian heart rhythm society 2016
implantable cardioverter-defibrillator guidelines. Canadian Journal of Cardiology, 33(2),
174-188. Doi:10.1016/j.cjca.2016.09.009
Bernier, R., Raj, S. R., Tran, D., Reyes, L., Sauve, M., Sumner, G. L., . . . Sandhu, R. K. (2017).
Assessing physician knowledge regarding indications for a primary prevention
implantable defibrillator and potential barriers for referral. Journal of Cardiovascular
Electrophysiology, 28(11), 1334-1341. doi:10.1111/jce.13326
Bradfield, J., Warner, A., & Bersohn, M. M. (2009). Low referral rate for prophylactic
implantation of cardioverter-defibrillators in a tertiary care medical center. Pacing and
Clinical Electrophysiology: PACE, 32 Suppl 1(s1), S194-S197. doi:10.1111/j.15408159.2008.02281.x
Brady, D. (2016). Planning for deactivation of implantable cardioverter defibrillators at the
end of life in patients with heart failure. Critical Care Nurse, 36(6), 24-31.
Brashers, V. (2014). Alterations in cardiovascular function. In K.L McCance & S.E. Huether
(Ed.). Pathophysiology: The biologic basis for disease in adults and children.
(pp.1129-1182). St Louis, Missouri: Elsevier.
Castellanos, J., Smith, L., Varosy, P., Dehlendorf, C. & Marcus, G. (2012). Referring physicians’
discordance with the primary prevention implantable cardioverter-defibrillator
guidelines: a national survey. Heart Rhythm. 9(6), 874-881.
doi:10.1016/j.hrthm.2012.01.029
Chae, S., & Koelling, T. (2010). Patient and physician determinants of implantable
cardioverter defibrillator use in the heart failure population. Congestive Heart Failure
(Greenwich, Conn.), 16(4), 141-146. doi:10.1111/j.1751-7133.2009.00139.x

87
Coulton, S., Dale, V., Deluca, P., Gilvarry, E., Godfrey, C., Kaner, E., . . . Drummond, C.
(2017). Screening for at-risk alcohol consumption in primary care: A randomized
evaluation of screening approaches. Alcohol and Alcoholism (Oxford,
Oxfordshire), 52(3), 312-317. doi:10.1093/alcalc/agx017
CorHealth Ontario. (2017). Implantable cardioverter defibrillator (ICD) deactivation: a guide
for health care professionals. Retrieved from
https://www.corhealthontario.ca/Implantable-Cardioverter-Defibrillator-Deactivation-AGuide-for-Health-Care-Professionals.pdf
Dec, G. (2005). Conventional therapy of chronic heart failure: Diuretics, angiotensin-converting
enzyme inhibitors, and digoxin. In Dec, G. (Ed.), Heart failure: A comprehensive guide
to diagnosis and treatment (pp. 181-214). New York, NY: Marcel Dekker.
Deyell, M., Tung, S., & Ignaszewski, A. (2010). The implantable cardioverter-defibrillator: from
Mirowski to its current use. BC Medical Journal 52(5). 248-253
DiMarco, J. P. (2003). Implantable cardioverter-defibrillators. New England Journal of
Medicine, 349(19). 1836-1847. doi:10.1056/NEJMra035432
DiSalvo, T. (2005). The clinical syndrome of heart failure. In Dec, G. (Ed.), Heart failure: A
comprehensive guide to diagnosis and treatment (pp. 181-214). New York, NY: Marcel
Dekker.
Epstein, A., DiMarco, J., Ellenbogen, K et al. (2008). ACC/AHA/HRS 2008 guidelines for
device-based therapy of cardiac rhythm abnormalities: A report of the American college
of cardiology/American heart association task force on practice guidelines (writing
committee to revise the ACC/AHA/NASPE 2002 guideline update for implantation of
cardiac pacemakers and antiarrythmia devices). Circulation, 117(21): e 350-6408.
Ezekowitz, J. A., O’Meara, E., McDonald, M. A., Abrams, H., Chan, M., Ducharme, A., . . .
Sussex, B. (2017). 2017 comprehensive update of the Canadian cardiovascular society
guidelines for the management of heart failure. Canadian Journal of Cardiology, 33(11),
1342-1433. doi:10.1016/j.cjca.2017.08.022
Goetz, P. W., PhD, Grady, K. L., PhD, & Yancy, Clyde W., MD, MSc. (2017). The ICD and
shared decision making: Nothing is ever easy. Journal of Cardiac Failure, 23(11), 800801. doi:10.1016/j.cardfail.2017.05.012
Grimshaw, J. M., Eccles, M. P., Lavis, J. N., Hill, S. J., & Squires, J. E. (2012). Knowledge
translation of research findings. Implementation Science, 7(1), 50-57. doi:10.1186/17485908-7-50
Habal, M. V., Micevski, V., Greenwood, S., Delgado, D. H., & Ross, H. J. (2011). How aware of
advanced care directives are heart failure patients, and are they using them? Canadian
Journal of Cardiology, 27(3), 376-38. doi:10.1016/j.cjca.2010.12.067

88

Hamel, A. V., Gaugler, J. E., Porta, C. M., & Hadidi, N. N. (2018). Complex decision-making in
heart failure: A systematic review and thematic analysis. The Journal of Cardiovascular
Nursing, 33(3), 225-231. doi:10.1097/JCN.0000000000000453
Hayes, S. M., Peloquin, S., Howlett, J. G., Harkness, K., Giannetti, N., Rancourt, C., & Ricard,
N. (2015). A qualitative study of the current state of heart failure community care in
Canada: What can we learn for the future? BioMed Central Health Services Research,
15(1), 290. doi:10.1186/s12913-015-0955-4
Herman, M., Horner, K., Ly, J., & Vayl, Y. (2018). Deactivation of implantable cardioverterdefibrillators in heart failure: A systematic review. Journal of Hospice and Palliative
Nursing: JHPN: The Official Journal of the Hospice and Palliative Nurses
Association, 20(1), 63-71. doi:10.1097/NJH.0000000000000399
Ho, E. C., Cheskes, S., Angaran, P., Morrison, L. J., Aves, T., Zhan, C., . . . Dorian, P. (2017).
Implantable cardioverter defibrillator implantation rates after out of hospital cardiac
arrest: are the rates guideline-concordant? Canadian Journal of Cardiology, 33(10),
1266-1273. doi:10.1016/j.cjca.2017.05.013
Hübinette, C., Lund, L. H., Gadler, F., & Ståhlberg, M. (2014). Awareness of indications for
device therapy among a broad range of physicians: A survey study. Europace: European
Pacing, Arrhythmias, and Cardiac Electrophysiology: Journal of the Working Groups
on Cardiac Pacing, Arrhythmias, and Cardiac Cellular Electrophysiology of the
European Society of Cardiology, 16(11), 1580-1586. doi:10.1093/europace/eut416
Interior Health. (2019). Medical orders for scope of treatment. Retrieved from
https://www.interiorhealth.ca/YourCare/PalliativeCare/ToughDecisions/Documents/MOS
T-Orders%20for%20scope%20of%20treatment.pdf
Kay, R. S. (2006). Causing death for compassionate reasons in American law. The American
Journal of Comparative Law, 54(suppl_1), 693-716
Kusumoto, F.M., Calkins, H., Boehmer, J., Buxton, A. E., Chung, M. K, Gold, M. R., . . .
Welikovitch, L. (2014). HRS/ACC/AHA expert consensus statement on the use of
implantable cardioverter-defibrillator therapy in patients who are not included or not well
represented in clinical trials. JACC (Journal of the American College of
Cardiology), 64(11), 1143-1177. doi:10.1016/j.jacc.2014.04.008
Isaac, D., Chan, M., Haddad, H., Cheung, A., West, L., Rao, V., & Dipchand, A. (2011).
Cardiac transplantation: Eligibility and listing criteria in Canada. Retrieved from
https://www.ccs.ca/images/Affiliates/CCTN/FINAL_Cardiac_Transplant_Eligibility_and
_Listing_Criteria_20121.pdf
Jakovljevic, D. G, Yacoub, M. H, Schueler, S., MacGowan, G. A., Velicki, L., Seferovic, P.M., .
. . Tan, L., DPhil. (2017). Left ventricular assist device as a bridge to recovery for

89
patients with advanced heart failure. JACC (Journal of the American College of
Cardiology), 69(15), 1924-1933. doi:10.1016/j.jacc.2017.02.018
Javaid, M., Squirrell, S., & Farooqi, F. (2018). Improving rates of implantable cardioverter
defibrillator deactivation in end-of-life care. British Medical Journal Open Quality, 7(2),
e000254. doi:10.1136/bmjoq-2017-000254
Kemp, C., & Conte, J. (2012). The pathophysiology of heart failure. Cardiovascular
Pathology, 21(5), 365-371. doi:10.1016/j.carpath.2011.11.007
Kempa, M., Budrejko, S., & Raczak, G. (2016). Subcutaneous implantable cardioverterdefibrillator (S-ICD) for secondary prevention of sudden cardiac death. Archives of
Medical Science: AMS, 12(5), 1179-1180. doi:10.5114/aoms.2016.61921
Lee, M. C., Sulmasy, D. P., Gallo, J., Kub, J., Hughes, M. T., Russell, S., . . . Nolan, M. T.
(2017). Decision-making of patients with implantable cardioverter-defibrillators at end of
life: family members’ experiences. American Journal of Hospice and Palliative
Medicine, 34(6), 518-523. doi:10.1177/1049909116641622
LeMond, L., Camacho, S. A., & Goodlin, S. J. (2015). Palliative care and decision making in
advanced heart failure. Current Treatment Options in Cardiovascular Medicine, 17(2), 111. doi:10.1007/s11936-014-0359-5
Lewis, K., Carroll, S., Birnie, D., Stacey, D., & Matlock, D. (2018). Incorporating patients’
preference diagnosis in implantable cardioverter defibrillator decision-making: a review
of recent literature. Current Opinion in Cardiology, 33(1), 42-49
Lilly, L. S., & Harvard Medical School. (2016). Pathophysiology of heart disease: A
collaborative project of medical students and faculty (6.th ed.). Philadelphia: Wolters
Kluwer.
Makoul G., & Clayman M. (2006). An integrative model of shared decision making in
medical encounters. Patient Education Counsel, 60: 301 – 312.
Matlock, D. D, Nowels, C. T., Masoudi, F. A., Sauer, W. H., Bekelman, D. B., Main, D. S., &
Kutner, J. S. (2011). Patient and cardiologist perceptions on decision making for
implantable Cardioverter-Defibrillators: A qualitative study. Pacing and Clinical
Electrophysiology, 34(12), 1634-1644. doi:10.1111/j.1540-8159.2011.03237.x
McCance, K. L., & In Huether, S. E. (2014). Pathophysiology: The biologic basis for disease in
adults and children (Ed.7). St Louis, Missouri: Elsevier
McHale, B., Harding, S., Lever, N., & Larsen, P. (2009). A national survey of clinician's
knowledge of and attitudes towards implantable cardioverter defibrillators.
Europace, 11(10), 1313-1316. doi:10.1093/europace/eup236

90
McNamara, D. (2005). Neurohormonal and cytokine activation in heart failure. In Dec, G. (Ed.),
Heart failure: a comprehensive guide to diagnosis and treatment (pp. 181-214). New
York, NY: Marcel Dekker.
Medtronic. (2015). Visia AFTM VR DVAB1D1. Retrieved from
http://manuals.medtronic.com/content/dam/emanuals/crdm/CONTRIB_235956.pdf
Metra, M., & Teerlink, J. (2017). Heart failure. The Lancet; 390(10106), 1981-1995.
doi:10.1016/S0140-6736(17)31071-1
Morton, R. L., Tong, A., Howard, K., Snelling, P, Webster, A.C. (2010). The views of patients
and carers in treatment decision making for chronic kidney disease: systematic review
and thematic synthesis of qualitative studies. British Medical Journal, 340(c112), 1-10.
doi.10.1136/bmj.c112
Mosterd, A., & Hoes, A. W. (2007). Clinical epidemiology of heart failure. Heart, 93(9), 11371146. doi:10.1136/hrt.2003.025270
Nambisan, V., & Chao, D. (2004). Dying and defibrillation: A shocking experience. Palliative
Medicine, 18(5), 482-483. doi:10.1191/0269216304pm907cr
Parkash, R., & Tang, A. S. L. (2017). Implantable cardioverter-defibrillators in sudden cardiac
death survivors: Are we doing all we can? Canadian Journal of Cardiology, 33(10),
1215-1216. doi:10.1016/j.cjca.2017.07.016
Pavlo, A. J., O'Connell, M., Olsen, S., Snyder, M. K., & Davidson, L. (2019). Missing
ingredients in shared decision-making? The Psychiatric Quarterly, 90(2), 333-338.
doi:10.1007/s11126-019-9624-9
Pillarisetti, J., Emert, M., Biria, M., Chotia, R., Guda, R., Bommana, S., . . . Lakkireddy, D.
(2015). Under-utilization of implantable cardioverter defibrillators in patients with heart
failure - the current state of sudden cardiac death prophylaxis. Indian Pacing and
Electrophysiology Journal, 15(1), 20-29. doi:10.1016/S0972-6292(16)30838-5
Public Health Agency of Canada. (2018). Report from the Canadian chronic disease surveillance
system: Heart disease in Canada, 2018. Retrieved from
https://www.canada.ca/content/dam/phac-aspc/documents/services/publications/diseasesconditions/report-heart-disease-canada-2018/pub1-eng.pdf
Raphael, C. E., Koa-Wing, M., Stain, N., Wright, I., Francis, D. P., & Kanagaratnam, P. (2011).
Implantable cardioverter-defibrillator recipient attitudes towards device deactivation: how
much do patients want to know? Pacing and Clinical Electrophysiology, 34(12), 16281633. doi:10.1111/j.1540-8159.2011.03223.x

91
Robinson, C., Kolesar, S., Boyko, M., Berkowitz, J., Calam, B., & Collins, M. (2012).
Awareness of do-not-resuscitate orders: What do patients know and want? Canadian
Family Physician, 58(4), e229-233
Ross, H., Ross, D., Howlett, J., Arnold, J., Liu, P., O’Neill, B., . . . Canadian
Cardiovascular Society Access to Care Working Group. (2006). Treating the right patient
at the right time: Access to heart failure care. Canadian Journal of Cardiology, 22(9),
749-754. doi:10.1016/S0828-282X(06)70290-2
Russo, J. E. (2011). Original research: deactivation of ICDs at the end of life: a systematic
review of clinical practices and provider and patient attitudes. The American Journal of
Nursing, 111(10), 26-37
Schallmo, M. K., Dudley-Brown, S., & Davidson, P. M. (2019). Healthcare providersʼ perceived
communication barriers to offering palliative care to patients with heart failure: An
integrative review. The Journal of Cardiovascular Nursing, 34(2), E9-E18.
doi:10.1097/JCN.0000000000000556
Sherazi, S., Mcnitt, S., Aktas, M. K., Polonsky, B., Shah, A. H., Moss, A. J., . . . Zareba, W.
(2013). End-of-life care in patients with implantable cardioverter defibrillators: A
MADIT-II substudy. Pacing and Clinical Electrophysiology, 36(10), 12731279. doi:10.1111/pace.12188
Slaughter, M. S, Rogers, J. G., Milano, C. A., Russell, S. D., Conte, J. V., Feldman, D., . . .
Frazier, O. H. (2009). Advanced heart failure treated with continuous-flow left
ventricular assist device. The New England Journal of Medicine, 361(23), 2241-2251.
doi:10.1056/NEJMoa0909938
Stevenson, L.W. & Zei, P. (2005). Clinical profiles and prognosis in heart failure. In Dec, G.
(Ed.), Heart failure: A comprehensive guide to diagnosis and treatment (pp. 181-214).
New York, NY: Marcel Dekker.
Stockwell, F. (2008). An unpopular view. Interview by Adele Waters. Nursing Standard (Royal
College of Nursing (Great Britain): 1987), 23(12), 22-23
Svanholm, J., Nielsen, J., Mortensen, P., Christensen, C., & Birkelund, R. (2015). Refusing
implantable cardioverter defibrillator (ICD) replacement in elderly persons—the same as
giving up life: A qualitative study. Pacing and Clinical Electrophysiology, 38(11), 12751286. doi:10.1111/pace.12702
Swedberg, K., Cleland, J., Cowie, M. R., Nieminen, M., Priori, S. G., Tavazzi, L., . . . & Zannad,
F. (2008). Successful treatment of heart failure with devices requires
collaboration. European Journal of Heart Failure, 10(12), 1229-1235

92
Tajouri, T., Ottenberg, A., Hayes, D., & Mueller, P. (2012). The use of advance directives
among patients with implantable cardioverter defibrillators. Pacing and Clinical
Electrophysiology, 35(5), 567-573
Thompson, T. L., & Dykeman, M. (2007). Nurse practitioners in Canadian heart failure clinics:
Evidence to support their presence on healthcare teams. Canadian Journal of Nursing
Leadership, 20(2), 81-94. doi:10.12927/cjnl.2007.18904
Virani, S. A., Bains, M., Code, J., Ducharme, A., Harkness, K., Howlett, J.G, . . .Zieroth, S.
(2017). The need for heart failure advocacy in Canada. Canadian Journal of Cardiology,
33(11), 1450-1454. doi:10.1016/j.cjca.2017.08.024
Whittemore, R., & Knafl, K. (2005). The integrative review: Updated methodology. Journal of
Advanced Nursing, 52(5), 546-553. doi:10.1111/j.1365-2648.2005.03621.x
Zhang, L., Narayanan, K., Chugh, H., Shiota, T., Zheng, Z., & Chugh, S. S. (2015). Factors
influencing utilization of the primary prevention implantable defibrillator. PloS
One, 10(3), e0121515. oi:10.1371/journal.pone.0121515

93
Appendix A: Literature Flow Diagram

Overview of Search Process
(October 29 – December 5, 2018)

Medline OVID
(October 29,
2018) (143)

CINAHL
COMPLETE
(December 4,
2018) (306)

CINAHL
(December 5,
2018) (122)
Total identified articles (642)
Removal of duplicates

Duplicates (185)
Primary relevance assessment
Non-relevant (based on title
and abstract screening) (391)
Potentially relevant articles (66)
Relevance assessment of full document versions (66)
Non-relevant articles (49)
Article Style
(Dissertation, Review, Opinion, Case Study)
(9)
Background Information
(40)Total relevant articles (15)

Quality assessment of relevant articles (17)
Weak articles (0)

Strong articles (15)

Moderate articles (0)

OVID/Cochrane
(December 5)
(71)

94
Appendix B: Literature Matrix
Qualitative Research Studies and Systematic Reviews
Author, Date,
Title

Review,
Purpose

Databases, Dates

Methods

Hamel, A.,
Gaugler, J.,
Porta, C., and
Hadidi, N.

To understand
key points in
the decisionmaking
process of
initiating
complex
treatment
interventions

Search of Ovid
MEDLINE,
CINAHL,
PsycInfo and Web
of Science were
searched for
qualitative data.

Qualitative
literature review
and synthesis

May/June
2018
Minnesota,
United States
of America.
Complex
DecisionMaking in
Heart
Failure: A
Systematic
Review and
Thematic
Analysis

The goal is to
identify
common
issues within
the decisionmaking
process for
treatments
such as ICD
implantation
so that health
care providers
can identify
opportunities
to introduce
these
discussion
points and
encourage
shareddecisionmaking

The author Hamel,
A. conducted the
searches and
applied the
inclusion criteria
for relevant
articles.
English studies
were included, no
limits on
publication date,
and no limits on
perspective
(clinician, patient
or caregiver).
Articles with
treatments defined
as ICD,
ventricular assist
device and/or
cardiac transplant
were included.

The COREQ
checklist was
utilized by the
author as a final
assessment for
inclusion (Tong,
Sainsbury &
Craig, 2007)
12 studies were
finally included
in the review
Followed
example of
treatment
decisions for
chronic renal
failure patients
by Morton et al.
(2010).
Framework was
based on a
trajectory model
for qualitative
research
throughout.
“This model
conceptualizes
human beings as
interactive with
others and
objects in their
environments
and within a
temporal
context” (p.
226).
The computer
program NVivo
was used for
data extraction,

Data Analysis, Results

Three major themes,
each with two
subthemes were
identified within the 12
studies:
1) Processing the
decision, which
included
transmission of
information and
changing
decisions
2) Timing and
prognostication,
which included
both of the
factors above
3)Considering the
future, which
represented
considerations towards
life or death and looking
forwards.

Key Findings,
Further
Research
Recognizing
these common
themes can help
providers
optimize shared
decision-making
opportunities
and support
patient-centered
care
Further studies
to examine the
decision-making
process on
complex
treatment
interventions
throughout the
trajectory of HF
(advanced care
planning, when
to deactivate
devices or
transition to
more intense
options) could
also enhance the
opportunities for
health care
providers to
identify areas for
communication
and timing of
conversations.

Strengths,
Limitations,
Qualities
Due to the
study design
raw data from
the original
studies can be
misinterpreted
by another
researchers’
aims
The data
collection,
inclusion,
analysis and
synthesis
were
performed by
one sole
researcher;
therefore, a
team
approach
would have
helped to
eliminate
some
potential bias
and increase
the rigor.
Quality
Assessment
Tool:
10/10
(Strong)

95
coding and
analysis
performed again
by Hamel, A.
A total of 313
participants were
involved in the
identified
studies, and were
considered
decision-making
stakeholders
including
patients,
clinicians, and
caregivers.

96
Herman, M.,
Horner, K.,
Ly, J., and
Vayl, Y.
February
2018
Deactivation
of
Implantable
CardioverterDefibrillators
in Heart
Failure

The purpose
of this
systematic
review was to
identify
problems that
may delay the
deactivation
of ICD and
address
possible
considerations
for ICD
management
to improve
end-of-life
care in adult
patients with
HF
Many patients
with ICD’s
will
experience
one or
multiple
shocks within
the last 24 hrs
of life, and it
is essential to
consider if
deactivation
of ICD
therapy is
appropriate
with endstage HF

Six electronic
databases were
used for this
literature review
including
Cumulative Index
to Nursing and
Allied Health
Literature
(CINAHL),
Cochrane, Ovid,
PubMed,
EMBASE and
Scopus.
Key words used
for search
included
“deactivation”,
“end-of-life”,
“heart failure” and
“Implantable
cardioverterdefibrillator”.
Limits placed
included clinical
studies from 20072017, adults ages
>18 years of age,
and English
language.
Inclusion criteria
were quantitative
or qualitative
studies, patients
with HF, patients
with ICD, full-text
articles, outcomes
that identify
problems that may
delay ICD
deactivation, and
considerations for
end-of-life in
patients with an
ICD.
A total of (9)
studies met
inclusion criteria
for this systematic
review, and were
between the years
of 2008-2016

Three categories
of results were
compiled:
1.

Clinical
practice and
management
for
deactivation
of ICD
49% of
deactivation
requests were
made by
patients, while
51% were made
by substitute
decision makers.
57% had ACD’s,
while only (1)
patient had
directly
described the
management of
the ICD
55% of ICD
deactivations
were carried out
by nurses, 31%
by physicians,
and 15% by
other allied
health
professionals.
More frequent
palliative care
consults, female
sex, white race
and higher
socioeconomic
status, or longer
hospital stays
were associated
with higher
incidence of
having an ACD
2. Patient
perceptions
of ICD
deactivation
Most patients do
not accurately
understand the
indication for or
of the role of
their ICD. Most

Many physicians do not
feel prepared or
comfortable to initiate
discussions surrounding
deactivation around
ICD.
As well, many patients
are not well informed
regarding the option for
deactivation at end of
life, and many don’t
comprehensively
understand the reasons
for ICD. However,
patients deserve a
thorough explanation of
risks and benefits of the
therapy in order to avoid
reduced quality of life.
ACD’s are valuable
tools that patients with
HF and ICD’s should
include in their medical
record to assist
physicians and alleviate
strain from family
members.
Many physicians feel
discussions around ICD
deactivation are situation
dependent, and also feel
they lacked knowledge
regarding ICDs and their
management.
Limitations to this study
include having only
historical data available
within the previous
articles.
The common themes
emerging from the
different articles
strengthen the quality of
this review.
Future recommendations
include development of
protocols to guide
providers in initiating
and addressing end-oflife issues for patients
with ICD.

Limitations
include the
observational
nature of this
study, as a
randomized
control trial at
end-of-life in HF
patients with
ICD has not yet
been performed.
The sample sizes
of these studies
were relatively
small and from
single centers,
which may limit
the
generalizability
to other
geographic
populations.
Qualities of the
study include the
relevance and
clarity of the
description of
the context and
data collection.
Quality
Assessment
Tool: 9/10
(Strong)

97
patients never
recalled having a
discussion with
their provider
regarding
deactivation, and
most did not
know this option
existed, yet
could express
when they felt
this should have
occurred (before
implantation,
without
significant
deterioration in
health, and at
end of life)
3. Provider
perceptions
of ICD
deactivation.
Few physicians
reported wanting
to accept
responsibility for
these
discussions, and
felt that patients
and/or families
should initiate
these
discussions.

98
Lee, M.,
Sulmasy, D.,
Gallo, J.,
Kub, J.,
Hughes, M.,
Russell, S.,
Kellogg, A.,
Owens, S.,
Terry, P., and
Nolan, M.
2017
DecisionMaking of
Patients with
Implantable
CardioverterDefibrillators
at End of
Life: Family
Members’
Experiences

This was a
qualitative
pilot study
that used a
mixed
methods
approach to
try to
understand
how health
care providers
can further
comprehend
the decisionmaking
process of
patients with
ICD at end of
life, with a
goal of
enhancing
informed
decisionmaking that
may improve
the quality of
care during
this difficult
time.

The data for this
study was
extracted from a
larger mixed
methods pilot
study called the
Trial of
Ascertaining
Individual
preferences for
Loved Ones’ Role
in End-of-lifedecisions, (the
“TAILORED”
ICD study).
The timeframe
this study took
place was between
2012 - 2012
The study was
approved by the
Institutional
Review Board of
Johns Hopkins
University.

Inclusion criteria
were patients
with ICD’s who
had either
NYHA III or IV
symptoms. They
were put in the
study via
convenience
dyad sampling,
and were either
inpatients or
outpatients. The
location was a
mid-Atlantic
tertiary medical
center in the
United States.
The
interviews were
completed with
the family
members
(spouses or
children) of 6
different patients
using an
interview guide
to help assist
with
consistency. The
questions were
implemented by
a single nurse
who had
experience in
research and
similar themes.
The interviews
were 60-90
minutes long.

The interview
transcriptions were input
into the Nvivo 9
program, and then were
analyzed line-by-line
with a thematic
approach by three of the
authors independently.
Lincoln and Guba
guidelines of truthvalue, applicability,
consistency, and
neutrality were utilized
to assess the
trustworthiness of the
study.

Three prevalent
themes were
emerged and
were coded.
Communication
methods,
perception on
ICD
deactivation, and
decision-making
preferences.
The concept of
deactivation was
unheard of to the
majority of
family members,
and only 1
family member
recalls having a
discussion about
the ICD with a
physician.
Family members
would have
preferred direct
verbal
communication
regarding the
ICD at EOL
rather than
written as an
AD. Even with
knowledge of
deactivation,
most patients
requested to
keep ICD active
during EOL.
Informed
decision-making
and direct
communication
with families
prior to end of
life can assist to
ease the
decision-making
process and
allow patients
and families to
be actively
engaged in this
process.

Limitations of
this pilot
study
included the
relatively
small sample
size of (6)
patient
families. The
majority of
families were
Caucasian as
well (5/6)
which means
that other
ethnic groups
may have
preferences
not discussed
here.
Qualities
included
recurrent
themes which
should be
strongly taken
into
consideration,
and recall bias
was
minimized by
performing
the interviews
very soon
after the
patient’s
death.
Further
research may
look into
enhanced
education for
providers to
promote this
style of
decisionmaking.
Quality
Assessment
Tool: 9/10
(Strong)

99
Russo, J.
October,
2011
Deactivation
of ICD’s at
the End of
Life: A
Systematic
Review of
Clinical
Practices and
Provider and
Patient
Attitudes

Many
providers are
not discussing
the
possibilities
of ICD
deactivation
when illness
or heart
disease
changes to
palliative
care.
The purpose
of this
systematic
review was to
1) identify
factors that
delay
discussions
around ICD
deactivation,
and
2)Promote
ways to
encourage
timely
discussions
about
deactivation
in an effort to
foster
improved
patient
centered endof-life care

8 databases
including
PubMed,
PsycARTICLES,
MEDLINE,
EBSCO Health
Source:
Nursing/Academic
Edition, the
Cochrane
Database of
Systematic
Reviews, the
Cohrane Library
and CINAHL
were searched.
The date ranges
applied for results
were limited to
January 1, 1999
through to
October 31, 2010.

The terms used
for searching
within these
databases
included
“death”,
“dying”,
“palliative”,
“terminal”,
“end-of-life”,
“end of life”,
“deactivation”,
deactivate”, plus
“ICD” or
“defibrillator” or
“implantable
cardioverterdefibrillator”.
Bibliographies
and reference
lists of relevant
articles were
also hand
searched.
Only peerreviewed,
primary research
studies were
included.
Articles were
excluded if they
focused on a
population <
18years old.
14 articles were
ultimately
included in this
review.

Three themes were
identified and
information was grouped
within these domains:
Patients attitudes about
ICD deactivation,
providers’ attitudes
towards ICD
deactivation, or
providers’ practice and
management of ICD
deactivation.

Deactivation
This systematic review
discussions
is of
were
high quality and
not
included several
commonplace,
important primary
and only research studies on these
occurred in
approximately
Strengths also
27% of cases,
included a
and even in
large time
patients with
frame and
DNR orders,
excellent
ICD deactivation databases
was discussed <
examined.
45%.
Only 25% of
Limitations
family
include only
practitioners
one author
were reported to
examining
have had
data, which
discussions
may lead to
around ICD
interpretation
deactivation
bias.
Continuity of
care during
These results
deactivation was would most
often lost, and
likely be
was carried out
reproducible
by another
and
professional.
applicable to
These
many other
discussions were areas and
also less likely
similar patient
when there was a populations
lack of
institution
Quality
policies or
Assessment
guidelines.
Tool: 10/10
Interdisciplinary (Strong).
approaches to
deactivation
discussions led
to less
inappropriate
shocks for a
large number of
patients.
Physicians
believe ICD
deactivation
should be a part
of care planning,
but rarely do so,
citing personal
discomfort or
inadequate

100
knowledge and
time constraints.
Some would
rather advocate
for further life
prolonging
therapies.
Little research in
the area of
patient attitudes
towards ICD
deactivation.70%
of patients
identified their
primary
physician as the
main
information
source of their
device.
Unrealistic
expectations or
insufficient
knowledge of the
role of the ICD
was also cited.
Patients must be
informed of
risks, rationale
and benefits of
the devise.
Institutional
policies should
be examined.

101
Primary Research Studies
Author,
Date, Title
Bernier, R.,
Raj, S., Tran
D., Reyes, L.,
Sauve, M.,
Sumner, G.,
Exner, D.,
Sandhu, R.
December,
2017
Assessing
physician
knowledge
regarding
indications
for a primary
prevention
implantable
defibrillator
and potential
barriers for
referral

Study Type,
Purpose,
Background
The purpose of
this quantitative
cross-sectional
study was
threefold, and
hoped to
identify
potential barriers
for referral for
ICD therapy, to
determine
whether or not
these barriers
differed amongst
physician
groups, and to
understand the
knowledge of
indications for
primary
prevention ICD.
Despite wellestablished
evidence that
primary
prevention ICD
improves
survival rates for
select
populations,
referral rates and
awareness of
guidelines
remains poor.

Methods

Data Analysis
and Results

The study
population was
selected from a
database of
physicians with
active 2016
Alberta Medical
Association
members, and
these individuals
were sent an
electronic
survey
containing five
case scenarios,
and questions
regarding
potential ICD
candidates.
Between
September 1 st,
2015 and
December 31,
2015, the
physicians were
given 3 months
to respond, as
well as an
incentive of a
$10 coffee gift
card if
completed.
109/799
physicians
completed the
survey.

Count and
percentage
were the
method in
which statistics
were described
and reported.
The values of
correct case
responses were
compared
across the age
groups,
location of
practice, and
physician
specialties with
chi-square
tests.
Analysis

Key Findings,
Further
Research
Baseline
characteristics
were compared,
and knowledge
of ICD guideline
concordance was
predicted in a
multivariate
analysis. Correct
answers for the
case scenarios
ranged from
72%-86% for
residents, 45%88% for
internists, and
77%-97% for
cardiologists.
Common barriers
for referral or
underutilization
of ICD involved
primarily a lack
of confidence in
personal
knowledge of the
recommended
guidelines,
concern for
inappropriate
shocks, and costeffectiveness,
however,
knowledge of the
indications for
and ICD was the
only significant
difference
(P<0.001)
Ultimately,
approximately
only one-third of
referring
physicians had
complete
awareness for
current
guidelines of
ICD implantation

Strengths,
Limitations,
Quality
Limitations of this
study include very
low response rate
on survey
participation, and
due to this method
of distribution and
voluntary
participation,
physicians who
felt more
confident in their
knowledge of this
topic may have
participated.
This was a
primary study of
this topic in
Canada, and may
not reflect other
areas of the
country. The
timeframe to be
allowed to
respond was
limited, only 3
months. Lastly,
patient
knowledge,
reflection, barriers
or decisionmaking processes
were not
addressed.
Qualities of this
study include an
appropriate
description of
methodological
quality and
transparency of
results including
graphs and
comparisons of
results.
Quality
Assessment

102

Bradfield, J.,
Warner, A.,
and Bersohn,
M.
March, 2009
Low Referral
rate for
Prophylactic
Implantation
of
CardioverterDefibrillators
in a Tertiary
Care Medical
Center

In this
retrospective
observational
cohort study, the
researchers
aimed to
quantify the
number of
appropriate
referrals for ICD
implantation, as
well as
understanding
the reasons for
which referral
did not occur.
While ICD’s
have become a
standard of care,
there continues
to be a large
number of
patients that are
not offered this
potentially lifesaving device,
and reasons for
this may include
the need for
patient and
provider
education.

The VA Greater
Los Angeles
Healthcare
System was
screened for
patient charts
that qualified
with specified
values was
searched from
April 12, 2002
until June 30,
2006.
Inclusion
criteria were
patients with
LVEF <35%
that met
guidelines for
ICD
implantation and
received a
device during
the timeframe.
The records
were also
classified of
those patients
who met
guideline
criteria, but did
not receive ICD.
This second
group was
further classified
and evaluation
for reasons for
nonimplantation,
including

120 patients
that met
inclusion
criteria were
selected from
each year of
the study, and
were randomly
allocated to the
following
groups: those
with LVEF
<35% (600)
and those that
had LVEF
<30% (600).
Percentages
were calculated
within each
group for those
who received
ICD and those
who did not, as
well as death
rates in each
group.

or referral.
Overcoming
these barriers is a
strong area of
need for future
research, and the
ability for
primary care
providers to
enhance their
knowledge of
referral processes
may have a
bigger impact on
the population
that qualifies.
Of the <35%
group, 111
patients received
ICD (19%),
which
represented only
28% of the
eligible patients.
58% were not
referred.
Amongst the
patients not
referred only
28% were valid
reasons (terminal
illness, etc.)
There was a 34%
mortality rate in
the group of
patients not
referred for ICD
despite no
contraindications
Of the <30%
group, 129
patients received
ICD (22%),
which
represented only
33% of the
eligible patients.
51% were not
referred.
Amongst patients
not referred, only
24% were valid
reasons, and
mortality was
33% for patients
not referred for

Rating: 8/10
(Strong)

Limitations of this
study include only
a single reviewer
of the medical
records. Some
patients may have
been offered the
referral for device
and turned this
down but was not
documented.
Patients who
experienced an
increase in LVEF
>35% but were
not referred may
have
overestimated
non-referral for
legitimate reasons.
The dates of this
study fall during
the results of the
SCD-HeFT trial,
and therefore
patients with
LVEF <35% may
not have been
referred prior to
2004.
Qualities include
an appropriate
timeframe of chart
retrieval, a good
sample size, and
thorough analysis
of the reasons for
non-implantation.
Inclusion/exclusio
n criteria were

103
limited life
expectancy,
those who
required further
medical
optimization, or
those awaiting
coronary
revascularizatio
n.

Castellanos,
J., Smith, L.,
Varosy, P.,
Dehlendorf,
C., and
Marcus, G.
June, 2012
Referring
Physicians’
Discordance
with the
Primary
Prevention
Implantable
CardioverterDefibrillator
Guidelines: A
National
Survey

This crosssectional survey
study was
performed to
further
understand
concordance
rates of
physicians
referring
patients for ICD
with the primary
prevention ICD
guidelines.
Primary care
providers are
ultimately the
initial referee for
patients who
may be eligible
for ICDs, which
have been
shown to
improve
mortality, yet
many studies
demonstrate that
only a small
minority of
patients eligible
for the device

A national
sample of 3000
physicians
(family
medicine,
general
cardiology, and
internal
medicine) were
randomly
selected from
the American
Medical
Association
Masterfile.
Each received a
hand-mailed
survey that
consisted of 34items testing
physician
knowledge of
current ICD
guidelines,
complications,
reasons for not
referring
patients, practice
habits and
patient
demographics.
The surveys
were sent

ICD despite no
contraindications
.
A substantial
number of
patients who
qualify for ICD
are not referred,
and of those a
large number
experience a
high mortality
rate despite lack
of
contraindications
for the device.
An increase in
physician and
patient
knowledge of
indication for
ICD implantation
is essential
1459 responded,
yet 713 had
incorrect
information.
Family practice
physicians were
less likely to
refer patients to a
subspecialist for
ICD
consideration,
nor were they
likely to provide
care concordant
with current
guidelines.
Physicians who
were less likely
to refer their
patients also
reported more
concern for periprocedural
infection and
worry that ICD
discharges would
be more painful
Patient
preference was
the most
important factor
obtained in

well described,
and the center
chosen for the
study was well
experienced in
implantation of
ICDs.
Quality
Assessment
Rating: 10/10
(Strong)

Limitations of this
study include the
method of
surveys, leaving
only those that
volunteer to
complete as the
representatives,
yet does not
encompass all
providers that
refer for ICDs.
The survey was
not able to
differentiate if
guidelines
discordant
answers were due
to lack of
knowledge or
professional
opinion. This
study is also
limited as it does
not include the
actual rates of
device utilization,
more so just the
first step in the
process.
Qualities of this
papers include a
detailed

104
are actually
receiving them.

between June 3,
2009 and
December 11,
2009. Incentives
were provided in
the form of $10
cash.

understanding
referral rates.
1/3 of physicians
provided
discordant
answers to
current
guidelines for
ICD
implantation.
Guideline
adherence may
require further
education and
attention in order
to ensure eligible
HF patients are
referred for ICD.

Chae, S., and
Koelling, T.
August, 2010
Patient and
Physician
Determinants
of
Implantable
Cardioverter
Defibrillator
Use in the
Heart Failure
Population

The purpose of
this
retrospective
cohort study was
to understand
the practice of
ICD utilization
in a single
university based
tertiary care
center in order
to examine the
low utilization
rates, and
explore
characteristics of
physicians and
patients that
may be
influencing the
rates of ICD
utilization.
Guidelines
recommend ICD
implantation as
a Class 1
recommendation
for HF patients
(ischemic and
non-ischemic)

Chart reviews
were completed
from ambulatory
patients
registered in the
University of
Michigan Health
System (UMHS)
Heart failure
registry between
December 1,
2004 and
December 31,
2007.
Patients with an
LVEF of <35%
at any time were
included in the
study, and
detailed
demographic
data was
compiled.
Patients who
refused ICD
were included
but recorded.
Excluded
patients

850 patients
eligible for
ICD
implantation
were reviewed.
594 (70%)
received an
ICD.
47 patients
(18%) refused.
Statistical
analysis using
SPSS statistical
software was
used.
Multivariate
stepwise
logistic
regression was
used to analyze
independent
variables
surrounding
non-utilization
of ICD.
P values <0.05
were
considered
statistically
significant.

Patients who
received ICD
were more likely
to be younger,
with poorer
LVEF, more
severe
hypotension, and
were more often
men. Those with
other terminal
illnesses such as
dementia or
cancer were also
less likely to
receive ICD.
This study found
no other
comorbidities
influenced
utilization, and
that race nor
health insurance
status influenced
rates of ICD
implantation.
Patients with
cardiologists
who specialized
in heart failure or
general

examination of the
survey by both
researchers,
family
practitioners and
cardiologists prior
to its mail out.
The methods of
the statistical
analysis were
rigorous and
thorough, and the
re-categorization
of the multivariable groups
allowed for
further
understanding of
potential
directions for
education.
Quality
Assessment
Rating: 8/10
(Strong)
Strengths of this
study include
large sample size,
and high rates of
ICD utilization,
demonstrating the
use of decisionmaking tools for
device therapy
such as The
Guidelines
Applied in
Practice for Heart
Failure (GAP-HF)
can assist to
enhance treatment
therapies for
qualifying
patients.
Limitations of this
study include
methodological
limitations of
retrospective
analysis.
Limitation to
information
provided on
documentation

105
with NYHA
class II-III
symptoms and
LVEF <35% or
Ischemic
patients with
NYGA class I
symptoms and
LVEF <30%
(All at least 40
days post MI).
Despite this,
major
retrospective
observational
studies
demonstrate
very low
utilization of
ICDs.

Habal, M.,
Micevski, V.,
Greenwood,
S., Delgado,
D., Ross, H.
(2011)
How aware of
advanced
care
directives are
heart failure
patients, and
are they using
them?

Prospective,
single-center
study
The purposes of
this study were
to (1) determine
patients’
awareness,
comprehension,
and utilization
of ACDs and (2)
determine their
knowledge of
the process of
cardiopulmonar
y resuscitation
and their current
resuscitation
preference.
The increasing
prevalence of
heart failure and

included those
who had an
improvement of
LVEF >35% or
those who had a
nonischemic
cardiomyopathy
and NYHA class
1 symptoms.

cardiology were
3x and 2x as
likely to receive
an ICD as those
patients who had
care from a
primary care
practitioner
(86%, 56% vs
26%, P <0.0001)
This study found
that the type of
physician
coordinating the
care of the
patient was most
predictive of
ICD utilization.

A mixed
methodology,
prospective,
cross-sectional
single centered
study at a Heart
Function Clinic
at academic
hospital in
Ontario, Canada

Descriptive
statistics used
to analyze data.

41 subjects who
were attendees
of HFC with age
>18, and a
diagnosis of left
ventricular
systolic
dysfunction
were
interviewed

3 participants
who were
aware of
ACD’s had had
a discussion
with a health
care provider;
1 had had a
discussion with
cardiologist
regarding
deactivation
and

A trained
research

Awareness of
ACD’s:
76% of patients
were not aware
or had heard of
ACD’s.

Patients’ end-oflife wishes are
not being
adequately
addressed,
discussed, or
documented
Less than one
third of patients
have had end-oflife discussions
Most patients
would prefer to
have ACD
discussions with
their PCP’s early
on in the disease
trajectory instead
of when
complications
arise

demonstrates
another weakness.
This study also
did not
differentiate
whether the
device was only
ICD or also CRT,
leading to a
possible
difference in
patient
preferences and
indication.
This may not be
generalizable to
some areas as this
institute was a
tertiary center
with multiple
divisions of
specialists that
focused on CHF
and EP, which
may not be
practical in more
rural or remote
populations.
Quality
Assessment Tool:
10/10 (Strong)
Relatively young
group (56.5 +/10.6 years), may
limit
generalizability to
an older group
A large majority
of the group did
not recollect any
discussions
surrounding
ACD’s, possibly
due to cognitive
difficulties which
further identifies
the need to
enhance these
conversations
Due to the
unawareness of
ACD/CPR
options, the

106
its unpredictable
trajectory
highlight the
need for patients
to make their
end- of-life care
wishes known
using advanced
care directives
(ACDs)

assistant
conducted a 27point semi
structured
interview from
July – December
2007

resuscitation
preferences,
the other two
had discussed
with social
workers.
78% wished
they had had
this discussion.
Participants
with ICD:
Only 2/19
participants
had discussed
deactivation of
ICD, yet after
discussion 9
would choose
the option of
deactivation if
their condition
deteriorated. 2
were unknown,
5 would
decline.

Increased control
in end-of-life
care is essential
to increasing
patient autonomy
Further research
should focus on
approaches to
increase patients
awareness and
uptake of ICDs

brevity of the
interview may not
have been
sufficient time to
determine true
preferences.
Quality
Assessment Tool:
9/10 (Strong)

Resuscitation:
61% of patients
knew what
CPR was, the
rest were
unsure or
haven’t heard
of it.

Pillarisetti, J.,
Emert., M.,
Biria, M.,
Chotia, R.,
Guda, R.,
Bommana, S.,
Pimentel, R.,
Vacek, J.,
Dendi, R.,
Berenbom, L.,
Dawn, B., and
Lakkireddy,
D.
2015
UnderUtilization of

This was a large,
retrospective
observation
study.
The purpose of
this study was to
identify reasons
for
underutilization
of ICD’s from a
clinical
standpoint.
Despite strong
evidence (a
Class I
indication) that

It was produced
from a
prospectively
collected
cardiovascular
database in
Kansas City
Kansas
University
Hospital.
Date range for
accessed
information
were from 1998
and 2006,
completed by

SAS institute
software was
used to analyze
rates of ICD
implantation,
the medical
treatment that
was received.
These were
expressed as
mean +/standard
deviations.
Kaplan-Meier
survival curves
were acquired
for those

ACC/AHA 2002
guidelines did
not indicate or
recommend ICD
for non-ischemic
cardiomyopathy
or for primary
prevention until
2005, and
therefore another
group of patients
who died prior to
this time
(December,
2005) were
removed from
the study,

Limitations
included a large
number of patients
lost to follow up
(125), variances
between
institutional
differences. It is
also representative
of only a single
institution.
Strengths included
the center being a
tertiary referral
with 30
physiologists as

107
Implantable
Cardioverter
Defibrillators
in Patients
with Heart
Failure – The
Current State
of Sudden
Cardiac
Death
Prophylaxis

ICD’s are
effective in
preventing
sudden cardiac
death in HF
patients,
utilization rates
are low around
37.6%

two
investigators
after approval
from that
institutions
review board.
Inclusion was all
patients
diagnosed with
low LVEF
<35%
(documented by
various
modalities)
Excluded
patients were
those who had
spontaneous
improvement of
EF within 1
year, those who
were lost to
follow up, and
those who died
Baseline
demographics,
clinical
variables, and
medication
treatment were
included.

patients with
and those
without ICDs,
and ICD-nonutilization in
patients were
evaluated for
their reasoning.
707patients
were reviews
after meeting
inclusion
criteria. Mean
RF = 26%, and
the majority of
patients were
white males.
59% had
ischemic CM,
while 41% had
non-ischemic
etiology. ICD’s
were implanted
in only 200
patients (28%)
The differences
in all-cause
mortality
between the
ICD group and
the non-ICD
group was
statistically
significant
p=004.

reducing the
sample size to
510 patients. Of
this group,
37.6% (192) had
an ICD.
Mean duration
between
diagnosis of
HFrEF and
implantation was
2.9+/-2.1 years.
ICD implantation
could be
predicted by the
following
universal
predictors: use of
ACEi/ARB, BB,
diuretic, statins
and higher QRS.
The single
largest reason for
non-implantation
was lack of
discussion
between
cardiologist and
patient
suggesting the
option of ICD.
Other reasons
included refusal,
improvement of
EF, expected
death within 1
year, or death.
Rates of ICD’s
(and discussion
of option for) in
patients with
LVEF remain
low, although
they are shown
to be cost
effective and
significantly
reduce mortality.
Targeting
educational
intervention on
ICD’s to
physicians may
help improve
this.

well as an EP
program.
This study is of
excellent quality
and essential to
the approach for
patients who are
considering ICD
implantation.
Quality
Assessment Tool:
10/10 (Strong)

108
Continued
counselling on
ICD’s is
recommended
despite refusal
“Another
alternative is to
refer to the heart
failure clinic
where a
physician and/or
a nurse
practitioner
would explain to
the patient in
detail about the
benefits of ICD”
Raphael, C.,
Koa-Wing,
M., Stain, N.,
Wright, I.,
Francis, D.,
Kanagaratna
m, P.
December,
2011
Implantable
CardioverterDefibrillator
Recipient
Attitudes
towards
Device
Deactivation:
How Much
Do Patients
Want to
Know?

This study
explored the
unclear aspects
of if and when
end of life issues
and device
deactivation
should be
discussed with
patients with
ICDs
It looked at the
recollection of
patients with
ICD’s of the
consent
procedure prior
to receiving the
device, and
followed by
discussions of
when device
deactivation
may be
considered or
device
programming
would need to
occur
It also explored
how much
patients want to
know about the

200 patients
with ICDs who
were regularly
followed at St.
Mary’s Hospital
in the UK were
recruited for
interviews, and
were completed
before or after
their routine
ICD check
Inclusion
criteria were a
diagnosis of
heart failure
with an ICD and
the ability to
give informed
consent
Exclusion
criteria included
patients with
ICD’s for
arrhythmias
only, and/or
inability to
communicate
without an
interpreter
A pre-specified
questionnaire

Analysis was
both
quantitative
and qualitative
Two-tailed
Students t-test
was used to
compare
baseline
characteristics
Patients were
numbered and
expressed as a
percentage and
compared
using x2 test;
P-value <0.05
significant.
68 patients
were
approached,
80% enrolled.
Reasons for
non-enrollment
included the
need to return
to the clinic
and limited
English.

Most ICD
deactivation
discussions occur
in the last few
hours or days of
life; however,
patients would
prefer to discuss
these issues in
advance
The two main
barriers to
potential end of
life discussions:
The patients
understanding of
the device and
The willingness
of the provider to
broach the
subject.
Only
approximately
1/3 of patient
were aware the
device could be
deactivated
without being
removed, and
most patients are
known to

Strengths of this
study stem from
the inclusion of
patients with
ICD’s that had
either had shocks
as well as those
that hadn’t, and a
common finding
of no association
between shocks
and quality of life
(all regarding
continued life
expectancy higher
than a reduction in
pain from shocks)
Limitations of this
retrospective
study include
recall bias, and a
likely poor
recollection of the
consent process.
The small sample
size limits the
power of this
study, as well as
the fact all of the
participants were
recruited from the
same center which
may be reflective
of that particular

109
device and the
decisions that
will be required
regarding its
management

Sherazi, S.,
McNitt, S.,
Aktas, M.,
Polonsky, B.,
Shah, A.,
Moss, A.,
Daubert, J.,
and Zareba,
W.
October, 2013
End-of-Life
Care in
Patients with

This was a casecontrol study
performed to
identify the
pattern of ICD
management and
frequency of
shocks delivered
before death.
ICD shocks
cause substantial
distress, pain,
and anxiety
especially
towards end-of-

was used and
conducted by
one doctor with
ICD experience
Two groups
were formed:
Group 1 had
recently
received an ICD
but not
experienced any
shocks, and
Group 2 had
ICD’s at least 6
months and had
at least (1)
shock
Group 1 was
asked questions
regarding
consent towards
the procedure,
awareness of
options for
deactivation,
feelings towards
shocks, and
quality of care.
Group 2 was
asked addition
questions
regarding end of
life, if
deactivation was
an option, and
what influenced
their decisions
to either keep
the device active
or deactivate it
The cases were
recruited in an
acceptable way;
98 deceased
patients’ charts
were analyzed
retrospectively
and chosen from
a group that
receive ICD’s in
the MADIT II
study.
Three groups
were identified:

54 patients
completed
questionnaires,
Group 1 = 29
Group 2 = 25
High
prevalence of
older males
>70 yrs with
borderline EFs
and NYHA
class between
1.5-2.5

overestimate the
expected benefit
of their ICD.
93% of
cardiologists
believe their
patients
understood what
the ICD did and
assumed they
were aware of
deactivation
End of life
discussions did
not cause distress
to the patients
Further research
may include the
option of a
variable switchoff device that
could potentially
have periods of
inactivity such as
overnight, to
avoid shocks
during sleep and
support a
peaceful, quick
death.

Data was
analyzed using
the KruskalWallis one-way
analysis of
variance,
which allows
for comparing
two or more
sample sizes
and does not
assume a
normal
distribution of
variables.

In the 30 days
preceding, 19
patients received
59 appropriate
shocks, and 7
patients received
17 inappropriate
shocks.
Characteristics of
Group 1 were
men with NYHA
II-IV. Most
deactivations
were requested

services as
opposed to
representing other
institutions
The quality of this
study is good as
there is no bias of
the participants or
observers.
However, the
MADIT II trial in
which these charts
were retrieved
from began in
1997, therefore
over 20 years
have passed and
practices may
have changed
significantly.
Quality
Assessment Tool:
9/10 (Strong)

End-of-life
discussions with
patients and
family members
should occur
while the patient
remains coherent
and has decisionmaking capacity
in order to help
clarify goals and
guide healthcare
provider’s
management.

110
Implantable
Cardioverter
Defibrillators
: A MADIT-II
Sub study

life, which
requires further
consideration of
ICD
deactivation and
management
during this time
of patient care,
rather than in
reaction to
declining health

Group 1 were 15
patients who had
their ICD
deactivated the
week before
their death
Group 2 were 36
patients without
ICD
deactivation
who were in
hospice care or
had DNR orders
Group 3 were 47
patients without
ICD
deactivation,
were not in
hospital, and did
not have DNR
orders.
Patient
demographics,
time of days
from ICD
deactivation to
death, timing
and number of
shocks before
death, initiation
of discussions
and requests for
ICD
deactivation or
refusal for ICD
deactivation
were collected.
Data on all
groups heart
rate, blood
pressure, NYHA
class and CHF
symptoms
before terminal
event were
identified.

Statistical
analysis was
completed with
SAS software
and a P-value
of <0.05 was
considered
significant.
52% of all the
deaths occurred
within the first
year after ICD
implantation,
and CHF was
the leading
cause of death.
Groups 2 and 3
had no
documented
discussion
regarding ICD
deactivation.

by family, and
never a PCP
There is low
frequency of
ICD deactivation
at end of life in
terminally ill
patients. Higher
frequency of
ICD shocks
delivered due to
VF/VT in the
week prior to
death spurred the
initiation of ICD
deactivation.
There is a lack of
comprehensive
planning
regarding end of
life care in
patients with
ICD, and one
explanation is
lack of
experience with
implanted
devices
preventing
discussions for
deactivation.
Time is also
another
consideration,
where rapid
deterioration of
patients’ status
makes it difficult
to find time to
have these
discussions
End-of-life
discussions with
patients and
family members
should occur
while the patient
remains coherent
and has decisionmaking capacity
in order to help
clarify goals and
guide healthcare

Health care
providers must
learn to clearly
communicate
risks associated
with active ICDs
at end of life.
There needs to be
a clear
differentiation
between death
after withdrawal
of support versus
physician-assisted
suicide/euthanasia
Limitations of this
study include the
retrospective
nature and
information solely
retrieved from
chart reviews,
therefore may
have incomplete
information.
Many of the
ICD’s were not
interrogated after
the patient’s death
which may have
misrepresented
the number of
shocks in the
hours surrounding
death.
Quality
Assessment Tool:
10/10 (Strong)

111
providers
management.
Health care
providers must
learn to clearly
communicate
risks associated
with active ICDs
at end of life.
There needs to
be a clear
differentiation
between death
after withdrawal
of support versus
physicianassisted
suicide/euthanasi
a
Tajouri, T.,
Ottenberg,
M., Hayes,
D., and
Mueller, P.
March, 2012
The Use of
Advance
Directives
among
Patients with
Implantable
Cardioverter
Defibrillators

This was a
retrospective
cohort study via
chart review.
The goal of this
study was to
ascertain the
number of
advance
directives
(AD’s) amongst
patients with
ICD’s.
Additionally,
establishing the
AD’s that
addressed ICD
management
during end-oflife care.

This study took
place at the
Mayo Clinic in
Rochester,
Minnesota.
Patients who
received an ICD
at this institution
in the year 2007
were chosen, as
this gave an
appropriate
amount of time
for patients to
develop an AD.
Exclusion
criteria involved
eliminating
patients who
were
incarcerated, or
who lacked
research
authorization.
Particular
demographics
were obtained
from each
patients’ chart,
including those
who had an AD
and furthermore
characteristics

420 patients
were identified
with ICD in
2007 after
exclusion
criteria was
implemented.
Patients were
grouped and
summarized
into two
characteristics,
presence or
absence of an
AD (127/293).
The Fisher
exact test or x2
test was used to
compare
categorical
variables.
SAS JMP
statistical
software was
utilized, and a
P-value <0.05
was considered
significant.

A higher number
of patients with
an AD were
white, lived in
the nearby area.
The two groups
did not differ
significantly in
demographics
such as sex,
education,
religion, marital
status, smoking
or alcohol use.
Those with AD
had a higher
incidence of
renal
insufficiency,
COPD, cancer,
or dementia.
Primary
prevention was
the indication for
77% of patients
with ICD in the
AD group, which
was similar and
not statistically
significant to
with the non-AD
group (P=0.11)

The qualities of
this study include
a large,
specifically
defined
population. It
exemplifies the
beliefs and values
of AD’s in
patients with
ICD’s who
request end-of-life
deactivation of the
device.
Limitations of this
study included the
majority of
patients being
white males,
which may not be
generalizable to
other populations.
The retrospective
nature of the chart
review did not
involve any
contact with
actual patients.
Quality
Assessment Tool:
10/10 (Strong)

112
involved in the
AD.

ICD deactivation
prior to death
was measured
but a large
percentage of
this statistic was
unknown.
126 patients had
an AD, and data
is available for
the time in which
this was
completed; of
these 126 with
AD’s, only 2
discussed ICD
management.
Patients may
have terminal
illness for which
ICD is no longer
effective, and
may present as a
barrier to a
natural death.
It is ethical and
legal for
clinicians to
carry out
requests for
device
deactivations in
informed
patients.

Zhang, L.,
Narayanan,
K., Chugh,
H., Shiota, T.,
Zheng, ZhiJie., & Chugh,
S.
(2015)
Factors
Influencing

This was a case
control study
performed to
assess the
factors that
influence
recommendation
s for ICD
utilization and
deployment in

Adult patients
with LVEF
<35% (severe)
were identified
at an
echocardiogram
lab in CedarsSinai Medical
Centre in LA,
from February –
October, 2013.

Primary
prevention ICD
recipients and
non-recipients
were compared
with respect to
several clinical
and
demographic
variables
including

Decisions and
goals for health
care need to be
documented, and
a device-specific
AD tool should
be developed.
63 patients
(57.2%) of ICD
non-recipients
had no
documentation of
discussion with
physician
regarding
implantation, and
of those with
implantation

Results show
there is a need to
improve
optimizing
appropriate
therapy for ICD
use in primary
prevention.
Targeting health
care provider
education

113
Utilization of
the Primary
Prevention
Implantable
Defibrillator

patients eligible
for the device.
Literature
suggests
underutilization
of primary
prevention ICD

These patients
were then
screened for
details on if they
had received a
primary
prevention ICD.
If patient met
criteria for ICD
but did not
receive one, the
records were
searched for
evidence of
physician
discussion of the
ICD or evidence
of patient refusal
to undergo
implantation.
Inclusion
criteria accepted
patients with
single or dual
chamber
devices, as well
as CRT-D.
Exclusion
criteria included
patients with
ICD for
secondary
prevention,
those with EF
assessment
within 40 days
of ACS, newly
diagnosed HF in
<3months, those
who lacked
follow-up
information and
those with other
incomplete data.
As well, those
with prior heart
transplant, or
mechanical
circulatory
support were
also excluded.

public health
care insurance,
private health
care insurance
or no
insurance. The
Charlson
Comorbidity
Index (CCI)
was used to
measure each
patients burden
of comorbidity,
categorized as
high (>3) or
low (<3).
A statistical
analysis was
performed
using the SPSS
(an IBM based
software
package for
analyzing
social sciences)
Ultimately, 110
patients with
primary
prevention ICD
and 110 ICD
non-recipients
were
compared.
Demographic
characteristics
included age,
gender,
obesity, race,
health
insurance,
alcoholism,
smoking or
drug abuse.
Cardiac
parameters
included
LVEF, QRS,
CAD, AF,
CABG, PCI,
NYHA, and
use of BB and
ACEI/ARB.
Comorbid data
included PVD,

advised, 24
(51.1%) refused
the procedures.
Unfortunately, a
clear reason for
refusal was not
available. 5 were
identified with
reasons,
including
accepting risk of
ICD, insurance
issues, while
another was
waiting
reassessment of
EF.
Compared to
non-ICD
recipients, ICD
recipients were
found to be
younger (<75),
male, less likely
to use alcohol,
tobacco or have a
hx of drug abuse.
More were also
found to have
public or
supplemental
private
insurance. No
significant
difference
between race.

regarding primary
prevention
guidelines may
improve rates and
requires further
study.
Future studies
should include
patients with
significant
comorbidities to
evaluate the
benefit in this
population.
Ensuring that the
candidates most
likely to benefit
from ICD is most
essential, and this
may be improved
through enhanced
risk stratification.
Suggestions to
improve ICD
implantation rates
included
computerized
decision support
tools, and
automated
referrals for
patents with low
EF <35%.
Strengths of this
study included
detailed review of
patient charts
rather than ICD-9
codes.
Limitations
included attendant
bias and single
center study.
Clinical practice
patterns may be
generalized to the
area and not well
representative of
other settings.
There was also no
evaluation of
follow up
including
mortality in the
two groups.

114
DM, CKD,
COPD,
cerebrovascula
r disease,
dialysis,
tumor/leukemi
a, liver disease,
and HIV
infection.

Quality
Assessment Tool:
9/10 (Strong)

115
Appendix C: Nine Essential Elements for Shared Decision-Making

Nine Essential Elements for Shared Decision-Making (Makoul & Clayman, 2006)
1. Define/explain problem
2. Present options
3. Discuss benefits/risks/costs
4. Clarify patient’s values/preferences
5. Discuss patient ability/self-efficacy
6. Discuss doctor knowledge/recommendations
7. Check/clarify patient understanding
8. Make or defer a decision
9. Arrange follow-up

116
Appendix D: Quality Assessment Tool – Review Articles
Instructions for completion:
Please refer to the attached dictionary for definition of terms and instructions for completing
each section. For each criteria, score by placing a check mark in the appropriate box.
First Author: Year: Journal: Reviewer:
CRITERIA
Q1. Did the authors have a clearly focused question [population, intervention
(strategy), and outcome(s)]?
Q2. Were appropriate inclusion criteria used to select primary studies?
Q3. Did the authors describe a search strategy that was comprehensive?
Circle all strategies used: § health databases
§ psychological databases
§ social science databases § educational databases
§ other
§ handsearching § key informants § reference lists § unpublished
Q4. Did search strategy cover an adequate number of years?
Q5. Did the authors describe the level of evidence in the primary studies included in
the review?
§ Level I § Level II § Level III
® RCTs only
® non-randomized, cohort, case-control ® uncontrolled studies

YESNO

117
Q6. Did the review assess the methodological quality of the primary studies, including:
(Minimum requirement: 4/7 of the following)
§ Research design
§ Study sample
§ Participation rates
§ Sources of bias (confounders, respondent bias)
§ Data collection (measurement of independent/dependent variables) § Followup/attrition rates
§ Data analysis
Q7. Are the results of the review transparent?
Q8. Was it appropriate to combine the findings of results across studies?
Q9. Were appropriate methods used for combining or comparing results across studies?
Q10. Do the data support the author’s interpretation?
TOTAL SCORE:
Quality Assessment Strong

Moderate

Weak Rating:

(total score 8 – 10) (total score 5 – 7) (total score 4 or less)