Enhancing Causal Inference in Observational Studies

Slides:



Advertisements
Similar presentations
Traps and pitfalls in medical statistics Arvid Sjölander.
Advertisements

SLIDE 1 Confounding and Bias Aya Goto Nguyen Quang Vinh.
Chance, bias and confounding
Lesson Designing Samples. Knowledge Objectives Define population and sample. Explain how sampling differs from a census. Explain what is meant by.
11 Populations and Samples.
Knowledge is Power Marketing Information System (MIS) determines what information managers need and then gathers, sorts, analyzes, stores, and distributes.
Association vs. Causation
RESEARCH DESIGN.
Unit 6: Standardization and Methods to Control Confounding.
McGraw-Hill/Irwin McGraw-Hill/Irwin Copyright © 2009 by The McGraw-Hill Companies, Inc. All rights reserved.
Chapter 1: Introduction to Statistics
Chapter 1 - Introduction & Research Methods What is development?
Study Design. Study Designs Descriptive Studies Record events, observations or activities,documentaries No comparison group or intervention Describe.
The Journey Of Adulthood, 5/e Helen L. Bee & Barbara R. Bjorklund Chapter 1 Defining the Journey: Some Assumptions, Definitions, and Methods The Journey.
Lecture 6 Objective 16. Describe the elements of design of observational studies: (current) cohort studies (longitudinal studies). Discuss the advantages.
 Is there a comparison? ◦ Are the groups really comparable?  Are the differences being reported real? ◦ Are they worth reporting? ◦ How much confidence.
Understanding Variability Unraveling the Mystery of the Data’s Message Becoming a “Data Whisperer”
Chapter 1: The Nature of Statistics
Techniques of research control: -Extraneous variables (confounding) are: The variables which could have an unwanted effect on the dependent variable under.
Experimental Design All experiments have independent variables, dependent variables, and experimental units. Independent variable. An independent.
Lecture 7 Objective 18. Describe the elements of design of observational studies: case ‑ control studies (retrospective studies). Discuss the advantages.
Sampling Design and Analysis MTH 494 Ossam Chohan Assistant Professor CIIT Abbottabad.
SAMPLING TECHNIQUES. Definitions Statistical inference: is a conclusion concerning a population of observations (or units) made on the bases of the results.
Design and Analysis of Clinical Study 2. Bias and Confounders Dr. Tuan V. Nguyen Garvan Institute of Medical Research Sydney, Australia.
© 2006 The McGraw-Hill Companies, Inc. All rights reserved. McGraw-Hill Sampling Chapter Six.
Ch. 11 SAMPLING. Sampling Sampling is the process of selecting a sufficient number of elements from the population.
Instructor Resource Chapter 15 Copyright © Scott B. Patten, Permission granted for classroom use with Epidemiology for Canadian Students: Principles,
Matching. Objectives Discuss methods of matching Discuss advantages and disadvantages of matching Discuss applications of matching Confounding residual.
Design of Clinical Research Studies ASAP Session by: Robert McCarter, ScD Dir. Biostatistics and Informatics, CNMC
Sampling technique  It is a procedure where we select a group of subjects (a sample) for study from a larger group (a population)
Chapter Two Methods in the Study of Personality. Gathering Information About Personality Informal Sources of Information: Observations of Self—Introspection,
Chapter 7 Data for Decisions. Population vs Sample A Population in a statistical study is the entire group of individuals about which we want information.
1.3 Experimental Design. What is the goal of every statistical Study?  Collect data  Use data to make a decision If the process to collect data is flawed,
Research Methods & Design Outline
Research and Evaluation Methodology Program College of Education A comparison of methods for imputation of missing covariate data prior to propensity score.
AC 1.2 present the survey methodology and sampling frame used
BIAS AND CONFOUNDING Nigel Paneth.
Unit 1 Section 1.3.
Sampling.
Fukushima Medical University Aya Goto Nguyen Quang Vinh
Principles of Quantitative Research
Chapter 4 Marketing Research
Chapter 4 Marketing Research
Sec 9C – Logistic Regression and Propensity scores
Donald E. Cutlip, MD Beth Israel Deaconess Medical Center
Graduate School of Business Leadership
BIAS AND CONFOUNDING
Data, conclusions and generalizations
Sampling: Design and Procedures
Slides by JOHN LOUCKS St. Edward’s University.
Sampling: Design and Procedures
Chapter Three Research Design.
Sampling: Design and Procedures
Research Methods 3. Experimental Research.
Sampling Design.
Single-Case Designs.
The Aga Khan University
Section 5.1 Designing Samples
Chapter 5: Producing Data
Multivariate Relationships
Critical Appraisal วิจารณญาณ
Where we’ve been and where we’re going…
STATISTICS ELEMENTARY MARIO F. TRIOLA
Relationship Relation: Association: real and spurious Statistical:
Enhancing causal influence (in observational studies)
CHAPTER 4 Marketing Information and Research
Data Collection and Sampling Techniques
Confounders.
Enhancing Causal Inference in Observational Studies
Presentation transcript:

Enhancing Causal Inference in Observational Studies

The five explanations when an association between coffee drinking and myocardial infarction (MI) is observed in a sample

Type of spurious association Strengthening the inference that an association has a cause-effect basis: ruling out spurious associations Type of spurious association Chance (due to random error) Design phase (How to prevent the rival explanation) Increase sample size and other strategies (measurement precision and accuracy, and sample size and power) Analysis phase (How to evaluate the rival explanation) Interpret p value in context of prior evidence

Strengthening the inference that an association has a cause-effect basis: ruling out spurious associations Type of spurious association Bias (due to systematic error) Design phase (How to prevent the rival explanation) Carefully consider the potential consequence between the research question and the study plan: Subject, Predictor, Outcome Analysis phase (How to evaluate the rival explanation) Obtain additional data to see if potential biases have actually occurred. Check consistency with other studies (especially those using different methods).

Strengthening the inference that an association has a cause-effect basis: ruling out other real associations Type of spurious association Effect-Cause (the outcome is actually the cause of the predictor) Design phase (How to prevent the rival explanation) Do a longitudinal study Obtain data on the historic sequence of variables Analysis phase (How to evaluate the rival explanation) Consider biologic plausibility

Strengthening the inference that an association has a cause-effect basis: ruling out other real associations Type of spurious association Effect-Effect (confounding variable is a cause of both the predictor and the outcome)

Design phase strategies for coping with confounders: Specification Advantages Easily understood Focuses the sample of subjects for the research question at hand Disadvantages Limits generalizability May make it difficult to acquire an adequate sample size

Design phase strategies for coping with confounders: Matching Disadvantages May be time consuming and expensive, less efficient than increasing the number of subjects (e.g., the number of controls per case) Decision to match must be made at outset of study and can have irreversible adverse effect on analysis and conclusions Requires early decision about which variables are predictors and which confounders Advantages Can eliminate influence of strong constitutional confounders like age and sex Can eliminate influence of confounders that are difficult to measure Can increase precision (power) by balancing the number of cases and controls in each stratum May be a sampling convenience, making it easier to select the controls in case-control study

Design phase strategies for coping with confounders: Matching Disadvantages Removes option of studying matched variables as predictors or as intervening variables Requires matched analysis Creates the danger of overmatching (i.e., matching on a factor that is not a confounder, thereby reducing power)

Analysis phase strategies for coping with confounders: Stratification Advantages Easily understood Flexible and reversible; can choose which variables to stratify upon after data collection Disadvantages Number of strata limited by sample size needed for each stratum: Few covariables can be considered Few strata per covariable leads to less complete control of confounding Relevant covariables must have been measured

Analysis phase strategies for coping with confounders: Statistical adjustment Advantages Multiple confounders can be controlled simultaneously Information in continuous variables can be fully used As flexible and reversible as stratification Disadvantages Model may not fit Incomplete control of confounding (if model does not fit confounder-outcome relationship) Inaccurate estimates of strength of effect (if model does not fit predictor-outcome relationship) Results are hard to understand Relevant covariables must have been measured