In observational studies, investigators have no control over the treatment assignment. The treated and non-treated (that is, control) groups may have large differences on their observed covariates, and these differences can lead to biased estimates of treatment effects. Even traditional covariance analysis adjustments may be inadequate to eliminate this bias. The propensity score, defined as the conditional probability of being treated given the covariates, can be used to balance the covariates in the two groups, and therefore reduce this bias. In order to estimate the propensity score, one must model the distribution of the treatment indicator variable given the observed covariates. Once estimated the propensity score can be used to reduce bias through matching, stratification (subclassification), regression adjustment, or some combination of all three. In this tutorial we discuss the uses of propensity score methods for bias reduction, give references to the literature and illustrate the uses through applied examples.

Propensity score methods for bias reduction in the comparison of a treatment to a non‐randomized control group

99mTc
: technetium-99m
201TI
: thallium 201
ABCB1
: ATP-binding cassette sub-family B member 1
ABI
: ankle-brachial index
ACC
: American College of Cardiology
ACCF
: American College of Cardiology Foundation
ACCOMPLISH
: Avoiding Cardiovascular Events Through Combination Therapy in Patients Living With Systolic Hypertension
ACE
: angiotensin converting enzyme
ACIP
: Asymptomatic Cardiac Ischaemia Pilot
ACS
: acute coronary syndrome
ADA
: American Diabetes Association
ADP
: adenosine diphosphate
AHA
: American Heart Association
ARB
: angiotensin II receptor antagonist
ART
: Arterial Revascularization Trial
ASCOT
: Anglo-Scandinavian Cardiac Outcomes Trial
ASSERT
: Asymptomatic atrial fibrillation and Stroke Evaluation in pacemaker patients and the atrial fibrillation Reduction atrial pacing Trial
AV
: atrioventricular
BARI 2D
: Bypass Angioplasty Revascularization Investigation 2 Diabetes
BEAUTIFUL
: Morbidity-Mortality Evaluation of the If Inhibitor Ivabradine in Patients With Coronary Artery Disease and Left Ventricular Dysfunction
BIMA
: bilateral internal mammary artery
BMI
: body mass index
BMS
: bare metal stent
BNP
: B-type natriuretic peptide
BP
: blood pressure
b.p.m.
: beats per minute
CABG
: coronary artery bypass graft
CAD
: coronary artery disease
CAPRIE
: Clopidogrel vs. Aspirin in Patients at Risk of Ischaemic Events
CASS
: Coronary Artery Surgery Study
CCB
: calcium channel blocker
CCS
: Canadian Cardiovascular Society
CFR
: coronary flow reserve
CHARISMA
: Clopidogrel for High Atherothrombotic Risk and Ischaemic Stabilization, Management and Avoidance
CI
: confidence interval
CKD
: chronic kidney disease
CKD-EPI
: Chronic Kidney Disease Epidemiology Collaboration
CMR
: cardiac magnetic resonance
CORONARY
: The CABG Off or On Pump Revascularization Study
COURAGE
: Clinical Outcomes Utilizing Revascularization and Aggressive Drug Evaluation
COX-1
: cyclooxygenase-1
COX-2
: cyclooxygenase-2
CPG
: Committee for Practice Guidelines
CT
: computed tomography
CTA
: computed tomography angiography
CV
: cardiovascular
CVD
: cardiovascular disease
CXR
: chest X-ray
CYP2C19*2
: cytochrome P450 2C19
CYP3A
: cytochrome P3A
CYP3A4
: cytochrome P450 3A4
CYP450
: cytochrome P450
DANAMI
: Danish trial in Acute Myocardial Infarction
DAPT
: dual antiplatelet therapy
DBP
: diastolic blood pressure
DECOPI
: Desobstruction Coronaire en Post-Infarctus
DES
: drug-eluting stents
DHP
: dihydropyridine
DSE
: dobutamine stress echocardiography
EACTS
: European Association for Cardiothoracic Surgery
EECP
: enhanced external counterpulsation
EMA
: European Medicines Agency
EASD
: European Association for the Study of Diabetes
ECG
: electrocardiogram
Echo
: echocardiogram
ED
: erectile dysfunction
EF
: ejection fraction
ESC
: European Society of Cardiology
EXCEL
: Evaluation of XIENCE PRIME or XIENCE V vs. Coronary Artery Bypass Surgery for Effectiveness of Left Main Revascularization
FAME
: Fractional Flow Reserve vs. Angiography for Multivessel Evaluation
FDA
: Food & Drug Administration (USA)
FFR
: fractional flow reserve
FREEDOM
: Design of the Future Revascularization Evaluation in patients with Diabetes mellitus: Optimal management of Multivessel disease
GFR
: glomerular filtration rate
HbA1c
: glycated haemoglobin
HDL
: high density lipoprotein
HDL-C
: high density lipoprotein cholesterol
HR
: hazard ratio
HRT
: hormone replacement therapy
hs-CRP
: high-sensitivity C-reactive protein
HU
: Hounsfield units
ICA
: invasive coronary angiography
IMA
: internal mammary artery
IONA
: Impact Of Nicorandil in Angina
ISCHEMIA
: International Study of Comparative Health Effectiveness with Medical and Invasive Approaches
IVUS
: intravascular ultrasound
JSAP
: Japanese Stable Angina Pectoris
KATP
: ATP-sensitive potassium channels
LAD
: left anterior descending
LBBB
: left bundle branch block
LIMA
: Left internal mammary artery
LDL
: low density lipoprotein
LDL-C
: low density lipoprotein cholesterol
LM
: left main
LMS
: left main stem
LV
: left ventricular
LVEF
: left ventricular ejection fraction
LVH
: left ventricular hypertrophy
MACE
: major adverse cardiac events
MASS
: Medical, Angioplasty, or Surgery Study
MDRD
: Modification of Diet in Renal Disease
MERLIN
: Metabolic Efficiency with Ranolazine for Less Ischaemia in Non-ST-Elevation Acute Coronary Syndromes
MERLIN-TIMI 36
: Metabolic Efficiency with Ranolazine for Less Ischemia in Non-ST-Elevation Acute Coronary Syndromes: Thrombolysis In Myocardial Infarction
MET
: metabolic equivalents
MI
: myocardial infarction
MICRO-HOPE
: Microalbuminuria, cardiovascular and renal sub-study of the Heart Outcomes Prevention Evaluation study
MPI
: myocardial perfusion imaging
MRI
: magnetic resonance imaging
NO
: nitric oxide
NSAIDs
: non-steroidal anti-inflammatory drugs
NSTE-ACS
: non-ST-elevation acute coronary syndrome
NYHA
: New York Heart Association
OAT
: Occluded Artery Trial
OCT
: optical coherence tomography
OMT
: optimal medical therapy
PAR-1
: protease activated receptor type 1
PCI
: percutaneous coronary intervention
PDE5
: phosphodiesterase type 5
PES
: paclitaxel-eluting stents
PET
: positron emission tomography
PRECOMBAT
: Premier of Randomized Comparison of Bypass Surgery vs. Angioplasty Using Sirolimus-Eluting Stent in Patients with Left Main Coronary Artery Disease
PTP
: pre-test probability
PUFA
: polyunsaturated fatty acid
PVD
: peripheral vascular disease
QoL
: quality of life
RBBB
: right bundle branch block
REACH
: Reduction of Atherothrombosis for Continued Health
RITA-2
: Second Randomized Intervention Treatment of Angina
ROOBY
: Veterans Affairs Randomized On/Off Bypass
SAPT
: single antiplatelet therapy
SBP
: systolic blood pressure
SCAD
: stable coronary artery disease
SCORE
: Systematic Coronary Risk Evaluation
SCS
: spinal cord stimulation
SES
: sirolimus-eluting stents
SIMA
: single internal mammary artery
SPECT
: single photon emission computed tomography
STICH
: Surgical Treatment for Ischaemic Heart Failure
SWISSI II
: Swiss Interventional Study on Silent Ischaemia Type II
SYNTAX
: SYNergy between percutaneous coronary intervention with TAXus and cardiac surgery
TC
: total cholesterol
TENS
: transcutaneous electrical neural stimulation
TERISA
: Type 2 Diabetes Evaluation of Ranolazine in Subjects With Chronic Stable Angina
TIME
: Trial of Invasive vs. Medical therapy
TIMI
: Thrombolysis In Myocardial Infarction
TMR
: transmyocardial laser revascularization
TOAT
: The Open Artery Trial
WOEST
: What is the Optimal antiplatElet and anticoagulant therapy in patients with oral anticoagulation and coronary StenTing

Guidelines summarize and evaluate all evidence available, at the time of the writing process, on a particular issue with the aim of assisting physicians in selecting the best management strategies for an individual patient with a given condition, taking into account the impact on outcome, as well …

/pdf/2013-esc-guidelines-on-the-management-of-stable-coronary-2178xukyyy.pdf

2013 ESC guidelines on the management of stable coronary artery disease: the Task Force on the management of stable coronary artery disease of the European Society of Cardiology.

The aim of many analyses of large databases is to draw causal inferences about the effects of actions, treatments, or interventions. Examples include the effects of various options available to a p...

Estimating Causal Effects from Large Data Sets Using Propensity Scores

Dynamic Bayesian Networks: Representation, Inference and Learning by Kevin Patrick Murphy Doctor of Philosophy in Computer Science University of California, Berkeley Professor Stuart Russell, Chair Modelling sequential data is important in many areas of science and engineering. Hidden Markov models (HMMs) and Kalman filter models (KFMs) are popular for this because they are simple and flexible. For example, HMMs have been used for speech recognition and bio-sequence analysis, and KFMs have been used for problems ranging from tracking planes and missiles to predicting the economy. However, HMMs and KFMs are limited in their “expressive power”. Dynamic Bayesian Networks (DBNs) generalize HMMs by allowing the state space to be represented in factored form, instead of as a single discrete random variable. DBNs generalize KFMs by allowing arbitrary probability distributions, not just (unimodal) linear-Gaussian. In this thesis, I will discuss how to represent many different kinds of models as DBNs, how to perform exact and approximate inference in DBNs, and how to learn DBN models from sequential data. In particular, the main novel technical contributions of this thesis are as follows: a way of representing Hierarchical HMMs as DBNs, which enables inference to be done in O(T ) time instead of O(T ), where T is the length of the sequence; an exact smoothing algorithm that takes O(log T ) space instead of O(T ); a simple way of using the junction tree algorithm for online inference in DBNs; new complexity bounds on exact online inference in DBNs; a new deterministic approximate inference algorithm called factored frontier; an analysis of the relationship between the BK algorithm and loopy belief propagation; a way of applying Rao-Blackwellised particle filtering to DBNs in general, and the SLAM (simultaneous localization and mapping) problem in particular; a way of extending the structural EM algorithm to DBNs; and a variety of different applications of DBNs. However, perhaps the main value of the thesis is its catholic presentation of the field of sequential data modelling.

/pdf/dynamic-bayesian-networks-representation-inference-and-yy4fhfxdcj.pdf

Dynamic bayesian networks: representation, inference and learning

Preface. I. BASICS. 1. Introduction to Bayesian Networks. 2. More DAG/Probability Relationships. II. INFERENCE. 3. Inference: Discrete Variables. 4. More Inference Algorithms. 5. Influence Diagrams. III. LEARNING. 6. Parameter Learning: Binary Variables. 7. More Parameter Learning. 8. Bayesian Structure Learning. 9. Approximate Bayesian Structure Learning. 10. Constraint-Based Learning. 11. More Structure Learning. IV. APPICATIONS. 12. Applications. Bibliography. Index.

Learning Bayesian networks

We are developing a software simulation system for emission tomography (SimSET). All the software is being designed and documented using modern software engineering techniques. As modules in the system are completed they will be placed in the public domain. The first module of SimSET to be released, the photon history generator (PHG), tracks photons through the tomograph field-of-view. The PHG incorporates importance sampling [I] and allows source activity and attenuation to be modeled as voxelized objects in three dimensions. The input format for the source object has been chosen to match the digitized anthropomorphic phantom of Zubal et a1 [2], allowing physiologically realistic phantoms to be easily simulated.

Preliminary Experience With The Photon History Generator Module Of A Public-domain Simulation System For Emission Tomography

Graphical Markov models determined by acyclic digraphs (ADGs), also called directed acyclic graphs (DAGs), are widely studied in statistics, computer science (as Bayesian networks), operations research (as influence diagrams), and many related fields. Because different ADGs may determine the same Markov equivalence class, it long has been of interest to determine the efficiency gained in model specification and search by working directly with Markov equivalence classes of ADGs rather than with ADGs themselves. A computer program was written to enumerate the equivalence classes of ADG models as specified by Pearl & Verma's equivalence criterion. The program counted equivalence classes for models up to and including 10 vertices. The ratio of number of classes to ADGs appears to approach an asymptote of about 0.267. Classes were analyzed according to number of edges and class size. By edges, the distribution of number of classes approaches a Gaussian shape. By class size, classes of size 1 are most common, with the proportions for larger sizes initially decreasing but then following a more irregular pattern. The maximum number of classes generated by any undirected graph was found to increase approximately factorially. The program also includes a new variation of orderly algorithm for generating undirected graphs.

Enumerating Markov Equivalence Classes of Acyclic Digraph Models

Graphical Markov models determined by acyclic digraphs (ADGs), also called directed acyclic graphs (DAGs), are widely studied in statistics, computer science (as Bayesian networks), operations research (as influence diagrams), and many related fields. Because different ADGs may determine the same Markov equivalence class, it long has been of interest to determine the efficiency gained in model specification and search by working directly with Markov equivalence classes of ADGs rather than with ADGs themselves. A computer program was written to enumerate the equivalence classes of ADG models as specified by Pearl & Verma's equivalence criterion. The program counted equivalence classes for models up to and including 10 vertices. The ratio of numbers of classes to ADGs appears to approach an asymptote of about 0.267. Classes were analyzed according to number of edges and class size. By edges, the distribution of number of classes approaches a Gaussian shape. By class size, classes of size 1 are most common, with the proportions for larger sizes initially decreasing but then following a more irregular pattern. The maximum number of classes generated by any undirected graph was found to increase approximately factorially. The program also includes a new variation of orderly algorithm for generating undirected graphs.

/pdf/enumerating-markov-equivalence-classes-of-acyclic-digraph-48vk5czcgj.pdf

Steven Gillispie

Papers

Preliminary Experience With The Photon History Generator Module Of A Public-domain Simulation System For Emission Tomography

Enumerating Markov Equivalence Classes of Acyclic Digraph Models

Enumerating Markov Equivalence Classes of Acyclic Digraph Models

Medical versus Early Surgical Therapy in Patients with Triple-Vessel Disease and Mild Angina Pectoris: A CASS Registry Study of Survival

The size distribution for Markov equivalence classes of acyclic digraph models