Advanced Quantitative Methods for Health Care Professionals PUBH 742 Spring 2015



Similar documents
Regression Modeling Strategies

Tips for surviving the analysis of survival data. Philip Twumasi-Ankrah, PhD

A Simple Method for Estimating Relative Risk using Logistic Regression. Fredi Alexander Diaz-Quijano

Section Format Day Begin End Building Rm# Instructor. 001 Lecture Tue 6:45 PM 8:40 PM Silver 401 Ballerini

BIOM611 Biological Data Analysis

Unit 12 Logistic Regression Supplementary Chapter 14 in IPS On CD (Chap 16, 5th ed.)

Study Design and Statistical Analysis

Interpretation of Somers D under four simple models

Sun Li Centre for Academic Computing

Service courses for graduate students in degree programs other than the MS or PhD programs in Biostatistics.

11. Analysis of Case-control Studies Logistic Regression

Developing Risk Adjustment Techniques Using the System for Assessing Health Care Quality in the

Department of Epidemiology and Public Health Miller School of Medicine University of Miami

Missing data and net survival analysis Bernard Rachet

Data Analysis, Research Study Design and the IRB

Department/Academic Unit: Public Health Sciences Degree Program: Biostatistics Collaborative Program

MEU. INSTITUTE OF HEALTH SCIENCES COURSE SYLLABUS. Biostatistics

COURSE PLAN BDA: Biomedical Data Analysis Master in Bioinformatics for Health Sciences Academic Year Qualification.

Adequacy of Biomath. Models. Empirical Modeling Tools. Bayesian Modeling. Model Uncertainty / Selection

Organizing Your Approach to a Data Analysis

SOCIOLOGY 311 RESEARCH METHODS

Biostatistics: Types of Data Analysis

Why Taking This Course? Course Introduction, Descriptive Statistics and Data Visualization. Learning Goals. GENOME 560, Spring 2012

BIO 226: APPLIED LONGITUDINAL ANALYSIS COURSE SYLLABUS. Spring 2015

Missing Data in Longitudinal Studies: To Impute or not to Impute? Robert Platt, PhD McGill University

Guide to Biostatistics

Overview Classes Logistic regression (5) 19-3 Building and applying logistic regression (6) 26-3 Generalizations of logistic regression (7)

School of Public Health and Health Services Department of Epidemiology and Biostatistics

Does referral from an emergency department to an. alcohol treatment center reduce subsequent. emergency room visits in patients with alcohol

Name of the module: Multivariate biostatistics and SPSS Number of module:

Multiple logistic regression analysis of cigarette use among high school students

Generalized Linear Models

240ST014 - Data Analysis of Transport and Logistics

Additional sources Compilation of sources:

Multivariate Logistic Regression

The Basics of Regression Analysis. for TIPPS. Lehana Thabane. What does correlation measure? Correlation is a measure of strength, not causation!

MORTGAGE LENDER PROTECTION UNDER INSURANCE ARRANGEMENTS Irina Genriha Latvian University, Tel ,

Sample Size Planning, Calculation, and Justification

STA569 Applied (Nursing) Statistical Methods Instructor/ Facilitator:

Module 223 Major A: Concepts, methods and design in Epidemiology

Students' Opinion about Universities: The Faculty of Economics and Political Science (Case Study)

Multivariate Normal Distribution

Assumptions. Assumptions of linear models. Boxplot. Data exploration. Apply to response variable. Apply to error terms from linear model

Lecture 2: Descriptive Statistics and Exploratory Data Analysis

Statistics Graduate Courses

SPSS TRAINING SESSION 3 ADVANCED TOPICS (PASW STATISTICS 17.0) Sun Li Centre for Academic Computing lsun@smu.edu.sg

Big data size isn t enough! Irene Petersen, PhD Primary Care & Population Health

UNIVERSITY OF KENTUCKY COLLEGE OF PUBLIC HEALTH. Proposal for a Graduate Certificate in Biostatistics. Purpose and Background

Introduction to Statistics and Quantitative Research Methods

ATV - Lifetime Data Analysis

University of Maryland School of Medicine Master of Public Health Program. Evaluation of Public Health Competencies

Fairfield Public Schools

VI. Introduction to Logistic Regression

SUMAN DUVVURU STAT 567 PROJECT REPORT

Multinomial Logistic Regression

Linda K. Muthén Bengt Muthén. Copyright 2008 Muthén & Muthén Table Of Contents

Description. Textbook. Grading. Objective

When to Use a Particular Statistical Test

Statistics for Biology and Health

PARALLEL LINES ASSUMPTION IN ORDINAL LOGISTIC REGRESSION AND ANALYSIS APPROACHES

Master programme in Statistics

ECON643 Empirical Analysis I: Foundations of Empirical Research

SPH CATALOG ADDENDUM

The Harvard MPH in Epidemiology Online/On-Campus/In the Field. Two-Year, Part-Time Program. Sample Curriculum Guide

Course Descriptions. Seminar in Organizational Behavior II

SOCIOLOGY 311 RESEARCH METHODS

Teaching Biostatistics to Postgraduate Students in Public Health

Lecture 25. December 19, Department of Biostatistics Johns Hopkins Bloomberg School of Public Health Johns Hopkins University.

Diabetes Prevention in Latinos

Searching for Clinical Prediction Rules in MEDLINE

Example: Credit card default, we may be more interested in predicting the probabilty of a default than classifying individuals as default or not.

The Harvard MPH in Epidemiology Online/On-Campus/In the Field. Two-Year, Part-Time Program. Sample Curriculum Guide

Sampling Error Estimation in Design-Based Analysis of the PSID Data

13. Poisson Regression Analysis

EXPANDING THE EVIDENCE BASE IN OUTCOMES RESEARCH: USING LINKED ELECTRONIC MEDICAL RECORDS (EMR) AND CLAIMS DATA

Calculating the number needed to be exposed with adjustment for confounding variables in epidemiological studies

Credit Risk Models. August 24 26, 2010

Strategies for Identifying Students at Risk for USMLE Step 1 Failure

Certified in Public Health (CPH) Exam CONTENT OUTLINE

It is important to bear in mind that one of the first three subscripts is redundant since k = i -j +3.

Dealing with Missing Data

Introduction to Event History Analysis DUSTIN BROWN POPULATION RESEARCH CENTER

Longitudinal Data Analysis. Wiley Series in Probability and Statistics

Comparison of Survival Curves

Global Infectious Disease Epidemiology PUBH 713 Course Syllabus Fall 2013

STATISTICA Formula Guide: Logistic Regression. Table of Contents

Scholars in HeAlth Research Program (SHARP)

Applied Regression Analysis and Other Multivariable Methods

THE IMPORTANCE OF THE NORMALITY ASSUMPTION IN LARGE PUBLIC HEALTH DATA SETS

Econometrics and Data Analysis I

Silvermine House Steenberg Office Park, Tokai 7945 Cape Town, South Africa Telephone:

Scholars in HeAlth Research Program (SHARP)

Introduction. Survival Analysis. Censoring. Plan of Talk

2015 TUHH Online Summer School: Overview of Statistical and Path Modeling Analyses

Transcription:

1 Advanced Quantitative Methods for Health Care Professionals PUBH 742 Spring 2015 Instructor: Joanne M. Garrett, PhD e-mail: joanne_garrett@med.unc.edu Class Notes: Copies of the class lecture slides are posted on Sakai under Course Documents and PUBH 742 Lectures ; you may want to print them off to bring to class. You will get the most out of the lectures if you read the notes before class the pace of the class is fast. It s difficult to know exactly how much material will be covered in each class. The assigned pages for each lecture are usually correct, but sometimes we will cover more or less material than indicated. There are lecture notes from PUBH 741 that it will help you to review, particularly if you missed any of these lectures. These will be listed in the syllabus and are available on Sakai in Course Documents in a folder called PUBH 741 Lectures. Several articles are posted on Sakai. Many of these are suggested readings, but enhance topics in the lectures and/or make good references. Suggested Texts: (not necessary for class, but you may want to have as references) Hosmer DW and Lemeshow S, Applied Logistic Regression, 3 rd Edition, 2013 Kleinbaum DG and Klein M, Logistic Regression, 3 rd Edition, 2010 Class Location: Class Schedule: Thursday Friday 132 McNider 11:50 1:45 10:10 12:05 Course Work: Problem Sets Assignments 1-5 to be completed and a hardcopy turned in at the end of each problem set session. It s okay to help each other out, particularly if you are stuck on a concept or a Stata command, but you will learn best if you do as much of the work on your own as possible. Final write-ups should be done independently. There will be no final exam. Problem Set Sessions: These sessions will be used to discuss answers to problem sets, as well as any other questions. There is 1 problem set on exploratory data analysis, 3 on logistic regression, and 1 on survival analysis. Additional Stata Resources: www.stata.com (main website for Stata) www.ats.ucla.edu/stat/stata/ (UCLA site with helpful Stata information and tutorials) http://www.cpc.unc.edu/research/tools/data_analysis/statatutorial/index.html (Carolina Pop. Ctr.)

2 Syllabus 2015 Part 1: Sampling and Sample Size (Note: Part 1 will be covered starting on 04/16) Part 2: Exploratory Data Analysis Thurs 1/8 PUBH 742 Course Overview Review of PUBH 741 exam Bring a copy of your final exam and the exam memo, solutions, and grading criteria to class (the memo, solutions, and grading criteria are posted on Sakai for PUBH 742 under Course documents ) Exploratory Class 1 Topics: Continuous variables (univariable) Lecture notes: pp. 2.1.1 2.1.17 [slides: Part 2\Explore1] Fri 1/9 Exploratory Class 2 Topics: Continuous variables (bivariable); Categorical variables (univariable, bivariable) Lecture notes: PUBH 741 Review: Categorizing continuous X variables [ PUBH741-Part5.pdf ; pp. 5.13 5.50]; Transforming X variables [ PUBH741-Part5.pdf ; pp. 5.63 5.98] PUBH 742 notes: pp. 2.1.18 2.1.66 [slides: Part 2\Explore1] Thurs 1/15 Exploratory Class 3 Topics: Influential data points; Collinearity; Missing data; Confounding Lecture notes: pp. 2.2.1 2.2.21; 2.2.26 2.2.36; 2.3.1 2.3.16 [slides: Part 2\Explore2, Explore3] (Note: Skip slides 2.2.22 2.2.25; 2.2.37 2.2.55) Fri 1/16 Exploratory Class 4 Topics: Interaction; Exploratory data analysis summary Lecture notes: pp. 2.3.17 2.3.51 [slides: Part 2\Explore3] Thurs 1/22 Problem set #1 on Exploratory Data Analysis [Exploratory Classes 1 4] [Note: Also, Linear Regression Review lecture on this day see below] Part 3: Logistic Regression Thurs 1/22 Linear Regression Review Topics: Review of multiple linear regression Lecture notes: pp. 3.1.1 3.1.33 [slides: Part 3\Linear1] Note: Read over slides before class not time to cover all of them Fri 1/23 Logistic Class 1 Topics: Overview of logistic regression: model, Odds ratios, Confidence intervals Lecture notes: pp. 3.2.1 3.2.31 [slides: Part 3\Log2]

3 Part 3: Logistic Regression (cont.) Thurs 1/29 Logistic Class 2 Topics: Odds ratios and confidence intervals with interaction; Coding the exposure Lecture notes: pp. 3.2.32 3.2.59 [slides: Part 3\Log2] Fri 1/30 Logistic Class 3 Topics: Coding exposure variable (cont.); Maximum likelihood estimation; Likelihood ratio tests; Modeling strategies Lecture notes: pp. 3.2.60 3.2.85; 3.3.1 3.3.23 [slides: Part 3\Log2, Log3] Thurs 2/5 Logistic Class 4 Topics: Examples traditional Epidemiologic model; Modeling categorical variables with more than two categories (intro) Lecture notes: pp. 3.3.24 3.3.74 [slides: Part 3\Log3] Article: Sun G, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J of Clin Epidemiol, 1996; 49(8):907-916 Fri 2/6 Logistic Class 5 Topics: Modeling categorical variables with more than 2 categories (cont.); Ordinal and nominal outcomes (more than 2 categories) Lecture notes: pp. 3.3.75 3.3.112 [slides: Part 3\Log3] Thur 2/12 Problem set #2 on Logistic Regression [Logistic Classes 1 4] Fri 2/13 Logistic Class 6 Topics: Calculating risk ratios from binomial regression models; Intro to longitudinal models [GEE] Lecture notes: PUBH 741 Review: Cluster sampling pp. 5.3 5.10 [slides: PUBH741-Part5.pdf ] PUBH 742 notes: pp. 3.4.1 3.4.45 [slides: Part 3\Log4] Article: Zou G. A modified Poisson regression approach to prospective studies with binary data. AJE 2004; 159(7):702-6. Mercier RJ, Garrett J, Thorp J,Siega-Riz AM. Pregnancy intention and postpartum depression: secondary data analysis from a prospective cohort. BJOG 2013;120:1116 1122. Thurs 2/19 Logistic Class 7 Topics: Continuation of longitudinal models [GEE] Lecture notes: pp. 3.4.46 3.4.82 [slides: Part 3\Log4] Fri 2/20 Problem set #3 on Logistic Regression [Logistic Classes 5 & 6] Article: Schulman KA, Berlin JA, Harless W, et.al. The effect of race and sex on physicians recommendations for cardiac catheterization. NEJM 1999; 340(8):618-26

4 Part 3: Logistic Regression (cont.) Thurs 2/26 Predictive Class 1 Topics: Predicted risk; Strategy: variable selection, descriptive statistics, specifying starting model, variable reduction, quantifying predictive ability, validation Lecture notes: pp. 3.5.1 3.5.34 [slides: Part 3\Predict5] Article: Concato J, Feinstein AR, Holford TR. The risk of determining risk and multivariable models. Ann Intern Med 1993; 118:201-210 Sun G, Shook TL, Kay GL. Inappropriate use of bivariable analysis to screen risk factors for use in multivariable analysis. J of Clin Epidemiol, 1996;49(8):907-916 Fri 2/27 Predictive Class 2 Topics: Example study; Quantifying predictive ability Lecture notes: pp. 3.5.35 3.5.80 [slides: Part 3\Predict5] Article: Wasson JH, Sox HC, Neff RK, Goldman L. Clinical prediction rules. N Engl J Med 1985; 313:793-799. Pryor DB, Harrell FE, Lee KY, Califf RM, Rosati RA. Estimating the likelihood of significant coronary artery disease. Amer J Med 1983; 75:771-9 Thurs 3/5 Predictive Class 3 Topics: Comparing ROC curves; Reliability; Model validation Lecture notes: pp. 3.5.81 3.5.118 [slides: Part 3\Predict5] Fri 3/6 Predictive Class 4 Topics: Validation (cont.); Strategy for a predictive linear regression model Lecture notes: pp. 3.5.119 3.5.156 [slides: Part 3\Predict5] Article: Carson SS, Garrett J, Hanson LC, et.al. A prognostic model for one-year mortality in patients requiring prolonged mechanical ventilation. Crit Care Med, 2008;36(7):2061-9. Thurs 3/12 Spring Break Fri 3/13 Spring Break Thurs 3/19 Problem set #4 on Predictive Models [Predictive Classes 1 4; Carson article] [Note: Depending on Bill Miller s schedule, his Clinical applications of predictive models class may be today and Problem Set 4 may be on Fri 3/20] Fri 3/20 Predictive Class 5 Topics: Clinical applications of predictive models (Bill Miller)

5 Part 4: Survival Analysis Thurs 3/26 Survival Class 1 Topics: Review of survival analysis Lecture notes: pp. 4.1.1 4.1.51 [slides: Part 4\Surv1] Articles: Tibshirani R. A plain man s guide to the proportional hazards model. Clinical and Investigative Medicine 1982;5(1):63-8. Fri 3/27 Survival Class 2 Topics: Cox proportional hazards model; Proportional hazards (PH) assumption Lecture notes: pp. 4.2.1 4.2.56 [slides: Part 4\Surv2] Thurs 4/2 Survival Class 3 Topics: PH assumption (cont.); Modeling strategy; Exploratory data analysis; Example (one exposure) Lecture notes: pp. 4.2.57 4.2.72; pp. 4.3.1 4.3.45 [slides: Part 4\Surv2 & Surv3] Fri 4/3 Good Friday (No Class) Thurs 4/9 Survival class 4 Topics: Example (predictive model); Time dependent covariates; Repeated events Lecture notes: pp. 4.3.46 4.3.60; pp. 4.4.1 4.4.26 [slides: Part 4\Surv3 & Surv4] Fri 4/10 Survival Class 5 Topics: Competing risks Lecture notes: pp. 4.4.27 4.4.73 [slides: Part 4\Survival] Article: Gourlay ML, Fine JP, Preisser JS, May RC, Li C, Lui L, Ransohoff DF, Cauley JA, Ensrud KE. Bone-Density Testing Interval and Transition to Osteoporosis in Older Women. NEJM 2012;366(3):225-233. Thurs 4/16 Problem set #5 on Survival Analysis [Survival classes 1 4] [Note: Also, first sampling lecture on this day see below] Part 1: Sampling and Sample Size Thurs 4/16 Sampling 1 Topics: Sample selection; Sample size background Lecture notes: pp. 1.1.1 1.1.36 [slides: Part 1\Sample1] Fri 4/17 Sampling Class 2 Topics: Calculations (2-sample t-test, paired t-test, two proportions, risk ratio, odds ratio, ANOVA) Lecture notes: pp. 1.2.1 1.2.47 [slides: Part 1\Sample2]

6 Part 1: Sampling and Sample Size (cont.) Thus 4/23 Sampling Class 3 Topics: Calculations (cluster design, equivalency trials, non-inferiority trials) Lecture notes: pp. 1.2.48 1.2.xx [slides: Part 1\Sample2] Fri 4/24 Sampling Class 4 [Note: Last day of classes] Topics: Correcting for survey design (sample weights, etc.) Lecture notes: PUBH 741 Review: Cluster sampling pp. 5.3 5.10 [slides: PUBH741-Part5.pdf ] PUBH 742 notes: pp. 1.3.1 1.3.36 [slides: Part 1\Sample3] Thurs 4/30 Extra class [make-up] Fri 5/1 Extra class [make-up]