Cost-effectiveness analysis of out-of-hospital cardiac arrest management strategies

Download 0.81 Mb.
Size0.81 Mb.
  1   2



Haidong Feng

MS, University of Connecticut, 2015
BS, China Pharmaceutical University, China, 2013

Submitted to the Graduate Faculty of

Department of Epidemiology

Graduate School of Public Health in partial fulfillment

of the requirements for the degree of

Master of Public Health

University of Pittsburgh




This essay is submitted


Haidong Feng

April 26, 2017

and approved by

Essay Advisor:

Emma Barinas-Mitchell, PhD ______________________________________

Assistant Professor

Epidemiology, Ultrasound Research Laboratory

Graduate School of Public Health

University of Pittsburgh

Essay Reader:

Hawre Jalal, MD, MSc, PhD ______________________________________

Assistant Professor

Health Policy and Management, Public Health Dynamics Lab

Graduate School of Public Health

University of Pittsburgh

Essay Reader:

Mark Roberts, MD, MPP ______________________________________


Health Policy and Management, Public Health Dynamics Lab

Graduate School of Public Health

University of Pittsburgh

Copyright © by Haidong Feng


Emma Barinas-Mitchell, PhD

Haidong Feng, MPH

University of Pittsburgh, 2017


Background: Out-of-hospital cardiac arrest (OHCA) is the sudden cessation of the heart in an out of hospital setting. In the United States, the incidence of OHCA is estimated at 110 individuals per 100,000 per year with an overall survival rate of 5.2%. The American Heart Association guidelines recommends angiography for patients who have ST elevation in electrocardiogram followed by proper treatments. In patients without ST elevation, other general tests and observations would be conducted before further interventions. Some evidence suggests that immediate angiography and appropriate intervention for OHCA patients could result in better healthcare outcomes regardless of the presence of ST elevation in electrocardiogram. The goal of this study is to investigate whether immediate angiography and PCI are cost-effective compared to the standard of care.

Methods: We built a decision tree in TreeAge Pro Software to compare the cost-effectiveness of immediate angiography followed by proper interventions to standard care. The model calculates the costs and benefits of each strategy over a short-term period. We reviewed the literature to obtain the model parameters, including probabilities for choosing interventions, intervention costs, quality of life and life expectancy estimates. We calculated incremental cost-effectiveness ratio of immediate angiography strategy compared to standard of care. In addition, we tested the robustness of our outcomes using Tornado analysis, and probabilistic sensitivity analysis (PSA) in which we varied all the parameters jointly.

Results: Immediate angiography was less expensive than the standard care ($1,281) per patient treated per year, and more effective [0.03 quality-adjusted life-years (QALYs)]. These findings were robust to all one-way sensitivity analyses. In addition, PSA showed there is more than 91% chance that immediate angiography is more cost effective than the standard care conditional on $100,000/QALY willingness to pay threshold.

Conclusion: Our results suggest that immediate angiography is more cost effective than the standard care for OHCA patients from a societal perspective because the incremental cost-effectiveness ratio (ICER) is well below the threshold that is generally considered to be cost-effective by many health-care agencies. Cost-effectiveness analysis results from our group and others may help inform treatment strategy of OHCA patients and lead to improved allocation of healthcare resources for CVD treatment. Globally, public health resources are limited; the saved resources in the cardiovascular disease area could be used in other areas, like HIV/AIDS, purification water or poverty. From a public health point of view, the research we conducted can provide evidence and support for allocation of limited public health resources.



List of tables vii

List of figures viii

preface ix

preface ix

1.0 Introduction 1

2.0 method 5

3.0 RESULTS 14






bibliography 26

bibliography 26

List of tables

Table 1. Base case estimates and ranges 12

Table 2. Incremental net monetary benefit 14

List of figures

Figure 1. Decision model structure 8

Figure 2. One-way sensitivity analysis (incremental net monetary benefit) 16

Figure 3. Probabilistic sensitivity analysis (PSA) 17


My sincerest thanks go out to Dr. Hawre Jalal, Dr. Emma Barinas-Mitchell and Dr. Mark Roberts. I came in a novice and left with a better understanding of what I am doing, what I should do, and the myriad complications along the way. They gave me the best support in my academic life. My thanks go also to my parents for understanding and support.

  1. Introduction

Out-of-hospital cardiac arrest (OHCA) is defined as the sudden cessation of the heart in out-of-hospital setting. OHCA is a leading global cause of death, in the US. Each year around 400,000 people experience non-traumatic OHCA assessed by emergency medical services (EMS) with a 5.2 % survival overall and this number has been increasing from 2011 to 2014.1 In the United States, the incidence rate per emergency medical service (EMS) has been estimated around 110.8 individuals per 100,000 population.2 The overall survival to hospital discharge is around 10.8% and the median age for OHCA is around 65 years.3 Improving the survival and health care outcomes of these patients is important when considering management strategies.

According to the OHCA surveillance, in about 78.7% of OHCA patients the cause is due to no obvious extra-cardiac reasons, for example, ventricular fibrillation (VF), pulseless ventricular tachycardia, asystole or pulseless electric activity. Among them, acute myocardial infarction (MI) and VF have been major causes.4,5 The high mortality of out-of-hospital cardiac arrest highlights the importance of identifying the risk factors. However, There are also factors that are not know well. Since coronary disease is the main cause for OHCA, cardiovascular risk factors could increase the risk of having OHCA. For example, smoking and diabetes can significantly increase the risk of suffering OHCA.6,7 Brain-related issues are other causes of OHCA. Arnaout et al.8 performed a retrospective review (1999-2012) of adults with a primary neurologic cause of OHCA and compared two randomly selected groups of patients with OHCA of non-neurologic etiology. Among the patients they studied, 2.3% suffered OHCA from brain-related etiology and most of the patients had OHCA caused by subarachnoid hemorrhage (SAH). All patients died within three days. Patients with brain-related causes are not covered by our research but it is very important to exclude these patients from the model because they should receive different treatment from patients with cardiac causes.

Acute myocardial infarction (MI) is one of the major causes of OHCA and there are two major types of MI, ST-segment elevation myocardial infarction (STEMI) and Non-ST-segment elevation myocardial infarction (NSTEMI). The formation of blood clots in the major coronary arteries previously affected by atherosclerosis can result in STEMI, which main mechanism of atherosclerosis is cholesterol deposition within the artery wall.9 The deposited cholesterol eventually forms a plaque called atherosclerosis plaque. In a long-term process, the atherosclerosis plaque may develop into blood clots and finally block the coronary artery and interrupt the blood supply, ultimately, leading to a ST-segment elevation myocardial infarction.

The pathophysiology of NSTEMI is different from STEMI. In STEMI, a complete occlusion could develop in a major artery that is previously affected by atherosclerosis and lead to necrosis or death of the entire thickness of the myocardium, known as transmural infarction, downstream from the blockage. While, NSTEMI results from a complete occlusion developed in a minor coronary artery or a partial occlusion developed in a major coronary artery that is previously affected by atherosclerosis. In this case transmural infarction is not evident per ECG findings and the myocardium may have partial necrosis.10 Usually, the ECG findings of NSTEMI are T-wave inversion or ST-depression without showing ST-segment elevation. On the other hand, STEMI shows ST-segment elevation plus pathological Q-wave formation and T-wave inversion in ECG.11

The 2010 American Heart Association (AHA) guideline has introduced general strategies for the diagnoses and treatments of cardiac-causes of OHCA, categorized by the presence of ST-segment elevation.12,13 These guidelines suggest immediate angiography for patients who have ST-segment elevation on electrocardiogram (ECG) followed by appropriate treatments, such as percutaneous coronary intervention (PCI), stenting or coronary artery bypass graft surgery (CABG). For patients without ST-segment elevation, these guidelines suggest other tests first, such as an echocardiogram and in-hospital observation. In the United States, health care practitioners are usually following these guidelines. Although the guidelines recommend the strategy of employing coronary angiography based on the results of the electrocardiogram (ECG), it is not confirmed whether the ECG are credible and reliable initial test. A recent review found that electrocardiogram after successful resuscitation is not quite helpful for defining coronary lesions.14

Recently, health care practitioners in emergency room and cardiologists are tending to use immediate angiography without considering ECG results. Dumas and his colleauges15 found that an immediate angiography could lead to better outcomes in survival and neurological functions regardless of the presence of ST-segment elevation in ECG for OHCA patients. Because of the high probability of occurrence of false negative ECG results, interventions could be delayed and the best treatment window could be missed for OHCA patients who do have culprit occlusions in vessels and can be more beneficial than ECG. Under these circumstances, the immediate angiography should be considered as a diagnosis tool.16 Although there is still some controversy, research supports that there is improved outcomes following immediate angiography in OHCA patients.14,17-19

Milosevic et al20 compared immediate and delayed invasive interventions for non-STEMI patients and concluded that immediate angiography could lead to a better survival and neurological outcomes. However, up to now, no published literatures have reported cost-effectiveness analysis of immediate angiography for OHCA patients with no obvious extra-cardiac causes globally. Patients, healthcare practitioners, payers and policy makers would benefit from knowing whether the improvements in outcomes are worthy for the additional cost of immediate angiography. Thus, our aim is to construct a decision analytic model to compare the cost-effectiveness of applying the new immediate angiography strategy versus the standard care management strategy initiated by ECG to treat patients with OHCA due to cardiac causes. We report our findings to provide important data that may assist in the decision of implementing immediate angiography for every patient resuscitated from OHCA.
  1. method

    1. Decision analytical model

We built a decision-analytical model to compare the cost-effectiveness of immediate angiography plus appropriate interventions versus the standard care management strategy initiated by ECG for the patients who experienced out-of-hospital cardiac arrest (OHCA) in 1-year time horizon. We chose the short-term time horizon because 1-year survival rate is from 5% to 10% and most of the differences in cost and benefit occur in this period after hospitalization.14 We used the societal perspective when considering the cost and healthcare outcome of treatments. Patients with no obvious extra-cardiac causes were included in the model. However, patients with obvious extra-cardiac causes, such as respiratory failure, brain stroke, metabolic disorder, hemorrhage, or any other non-cardiac reasons, were excluded from the model as these patients would be treated following other procedures. We built the model in TreeAge 2017 software (Williamstown, Massachusetts) A schematic representation of the decision tree is shown in Figure 1.
    1. Treatment strategies

Patients who were resuscitated from OHCA either receive 1) standard care, where patients would receive the electrocardiogram (ECG) followed by the regular angiography and proper interventions based on the results of ECG, or 2) immediate angiography, where every patient would receive an immediate angiography and proper interventions. In the standard care, patients will undergo emergency electrocardiogram (ECG) followed by angiography when it shows a ST-segment elevation. If the angiography result indicates that there are culprit lesions then invasive interventions like coronary artery bypass graft (CAGB) and percutaneous coronary intervention (PCI) will be implemented for patients with severe occlusions based on the angiographic results13. Patients with diabetes, hypertension or other conditions who are not eligible for PCI will undergo CABG. If ECG shows negative results or there isn’t any culprit lesion based on angiography, patients will receive conservative or medical treatment (MEDS). The procedure of medical treatment was defined through the research of Cannon and his colleagues.21 In the medical or conservative treatment, the catheterization would be performed only for those who experienced recurrent ischemia or had an abnormal stress test. Otherwise, patients would just receive conservative or medical treatments.

In the strategy of immediate angiography, however, without implementing emergency ECG, an immediate angiography will be conducted for every patient who was resuscitated from OHCA. Based on the angiographic results, which shows the existence, locations and severity of occlusions, patients will be assigned into different intervention groups (PCI vs CABG vs MEDS) based on conditions described above. All patients would receive regular testing plus medicines and followed by routine physician visits during the treatment period.

Patients with less than 50% occlusion were considered to be at moderate risk and were assigned into MEDS group, while patients with equal to or more than 50% occlusion in vessels were considered to have a severe condition and were assigned into PCI or CABG group. In addition to the procedures indicated by these two strategies, patients would receive appropriate care according to the 2010 American Heart Association guidelines for CPR and ECC.12,13
    1. Key assumptions

Several assumptions were made in our model. First, we assumed that the patients in standard care with non-ST-segment elevation were in similar condition as those with non-ST-segment elevation myocardial infarction (NSTEMI). Similarly, we assumed that patients in standard care with ST-segment elevation were similar as those with ST-segment elevation myocardial infarction (STEMI). Second, we assumed that crossover among CABG, PCI and MEDs or revascularizations would have little influence on quality adjusted life-years for patients. Third, because of the lack of extant information for out-of-hospital cardiac arrest, for example, life expectancy, quality of life and cost of OHCA patients with negative ECG results, we chose to use data from acute myocardial infarction (AMI) or coronary artery disease (CAD), in which patients share common conditions with those of OHCA.

Figure 1. Decision model structure

    1. Share with your friends:
  1   2

The database is protected by copyright © 2019
send message

    Main page