Michael
EPI 204: Clinical Epidemiology (Fall 2021)
Section outline
-
-
Collaboration with 1 or 2 other classmates on this assignment is permitted and even encouraged. List the people you worked with, but for record-keeping purposes, each member of the team must upload a separate copy of the problem.
If possible, please either include a link to the paper(s) on which your problem is based or upload a pdf of the paper.
1. Please NAME your problem file using this convention:
[Last name of student author closest to the beginning of the alphabet]_ + Ch_ + [Chapter numbers of EBD-2 where material is covered]_ + [a word or two about the topic].
Example: John Brown, Sally Green and Joe Black wrote a problem about brain tumors that has material about ROC curves (Chapter 3) and spectrum bias (Chapter 4). ALL THREE STUDENTS IN THE GROUP would name their problem: "Black_Ch3_4_Brain_tumors.doc". (If you want to add a date at the end, that's fine.)
2. Please name the pdfs of any included articles on which your problem is based starting exactly the same way. If the article was by Gold et al from JAMA in 2019, you could name it
"Black_Ch3_4_Brain_tumors_Gold_JAMA_2019.pdf"
THANK YOU!
-
(You don't need to download the Word document, it has this same information.)
Evidence-Based Diagnosis: an Introduction to Clinical Epidemiology, 2nd Ed.
Thomas B Newman and Michael A Kohn
Illustrated by Martina Steurer
We think the book is well worth buying, but you can download it free from Cambridge University Press if you go through the UCSF Library. Here's a link.
From there, click on
Cambridge University Press online books
You should see a green check for "Access." Then "select all" on the left side and download a zip file.
-
Course introduction; Diagnostic process and measures of interobserver agreement
Introduction to the course and how we will teach it; testing to refine probabilities and guide decisions; concordance, unweighted Kappa, linear and quadratic weighted Kappa
Faculty: Michael Kohn
No Large-Group Discussion (LGD). LGDs start on Tuesday 9/21.
The first small group will be Thursday 9/16 at 1:15 pm. In that small group, we will review Problem Set #0 (not to turn in) and play the Kappa Game. Prior to that small group meeting, please watch Lecture 1, read Chapters 1 and 5 (pages 110 - 121), and do Problem Set #0, which is not to submit.
All subsequent problem sets (#1 - #9) should be uploaded on the CLE prior by 1:15 pm on Thursday, starting with Problem Set #1 on 9/23.
-
Lecture 1 Segment 1 -- Ch 1 material on diagnosis and testing (12:55) Media Resource
-
Lecture 1 Segment 2 -- Ch 5 material on kappa (unweighted) (22:45) Media Resource
-
Lecture 1 Segment 3 -- Ch 5 material on weighted kappa (27:15) Media Resource
-
Lecture 1 Segment 4 Ch 5 What affects kappa? What's a good kappa? (4:41) Media Resource
-
Melanoma pathology: example of agreement among multiple raters (2:35) Media Resource
-
Kohn (2013) Understanding Evidence-Based Diagnosis File
-
Unlike Problem Sets #1-10, Problem Set #0 will not be graded. We will discuss it in the Sept. 16 Small Group Discussion sections.
-
Use the link above to upload Problem Set #1 by 1:15 PM on September 23.
You must be logged in to upload your assignment.
-
Problem Set #1 (M.Laker Section) Assignment
-
PSet #1 Answer Key File
-
Small Group Discussion (see roster below for your section assignment)
Kappa Game
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) File
-
-
Course introduction; Diagnostic process and measures of interobserver agreement
This Large Group Discussion will cover the Chapter 1 material on diagnosis and Chapter 5 material on kappa. The lecture slides and video are all with the 9/14 syllabus entry. We put them there so that you would see them BEFORE the first small group on Thursday 9/16, since that session is spent playing the Kappa Game.
Faculty: Michael Kohn
-
Recording of Interactive Large Group 9/21/2021 (Problem set 0 and Kappa) Media Resource
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and Problem Set #1
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.
-
EPI 204 Section Assignment (includes Zoom Access Information) (copy) File
-
-
LGD: Dichotomous Tests
Sensitivity, specificity, predictive value, prior and posterior probability; 2x2 table method for getting posterior probability; probability and odds; likelihood ratios, the likelihood ratio slide rule, false-positive and false negative confusion; use of X-graphs to understand testing and treatment thresholdsFaculty: Thomas Newman
-
Note: This PPT file is almost identical to the one I used for the recorded lecture. The only differences are:
In the speaker notes I corrected how I misspoke in the recorded lecture on slides 6 and 40.
The substantive difference is that I (TN) made logical error on slide 57 and in my discussion of it in the lecture recording, which is corrected on the slide and in the speaker notes for that slide. If you want a challenge, try watching the lecture to see if you notice the error before finding the answer in slide 57 of this PPT.
-
Lecture 2 -- Dichotomous Tests (1:17:35) URL
Tom Newman explains dichtomous tests: sensitivity, specificity, positive/negative predictive value, and more.
-
Recording of Zoom Large Group Discussion 9/28/2021 Media Resource
-
Cost-Benefit X Graph and calcs MAK File
This is the Excel version of the Regret Graph Calculator available at https://ebd-2.net/regret-calculator/
-
Upload Problem Set #2 (M.Laker section) Assignment
-
PSet #2 Answer Key File
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) File
-
-
LGD: Continuous Tests
ROC Curves, multilevel likelihood ratios and the relationship between them; how dichotmozing tests wastes information; the Walking Man approach to ROC curves; choosing cutoffsFaculty: Michael Kohn
-
This version has one sheet for demo and one with the formulas
-
Lecture 3 Segment 1 -- Multilevel Tests -- Why not make the test dichotomous? (13:34) Media Resource
-
Lecture 3 Segment 2 -- the ROC Curve (17:45) Media Resource
-
Lecture 3 Segment 3 -- Interval Likelihood Ratios (22:34) Media Resource
-
Lecture 3 Segment 4 -- What test result should prompt treatment? (11:29) Media Resource
-
Lecture 3 Segment 5 -- computer-drawn ROC curves (33:30) Media Resource
-
2020 (!) Zoom Recording of Large Group Discussion 10/6/2020 (1:31:00) Media Resource
3:00 – Michael: Review of material covered in the first 4 segments
10:45 – Michael: “Walking Man” approach to understanding computer-drawn ROC curves
46:30 – Student questions
1:03:30 – Tom: problem on using BNP to diagnose peripartum congestive heart failure (CHF)
Includes a detailed discussion of a published ROC table and why NOT to report LR(+) and LR(-) at each of multiple cutoffs.
-
I tried downloading the Zoom recording that includes only the shared screen and audio, not the gallery of faces. It's a bit odd because, when I'm not sharing my screen, there is no video.
-
M.A. Kohn, Key concepts in clinical epidemiology: reporting on the accuracy of continuous tests, Journal of Clinical Epidemiology, https:// doi.org/ 10.1016/ j.jclinepi.2021.07.012 File
-
PSet #3 Answer Key File
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) (copy) (copy) (copy) File
-
-
Bias in Test Accuracy Studies / Systematic Reviews of Test Accuracy Studies
Common biases: incorporation bias, verification bias, double gold-standard bias, spectrum bias; Prevalence, spectrum bias and nonindependence; checklist vs systematic approach to critical appraisal of studies of diagnostic testsFaculty: Tom Newman
-
TN Studies of Diagnostic Test Accuracy Lecture PART 1 (56:24) 2020-1008 URL
-
TN Studies of Diagnostic Test Accuracy Lecture PART 2 (41:32) 2020-1008 URL
-
MAK GRACE Module 1 on Sampling Frame in Diagnostic Test Accuracy Studies Media Resource
GRACE (Guidelines for Reasonable and Appropriate Care in the Emergency Department) is a project of the Society for Academic Emergency Medicine. They asked me (Michael) to record a series of videos on how to evaluate studies of diagnostic test acccuracy (DTA studies). I have recorded two modules so far and am posting them here.
Here is the link for the GRACE project:
https://www.saem.org/publications/academic-emergency-medicine/grace
-
MAK GRACE Module 2 - Incorporation, Partial Verification, and Differential Verification Bias Media Resource
-
Zoom Recording of LGD 10/12/2021 Media Resource
-
Kohn MA Sampling Frame and Partial Verification Bias File
Manuscript under review by the Journal of Clinical Epidemiology
-
Kohn Carpenter Newman -- Understanding the Direction of Bias in Studies of Diagnostic Test Accuracy File
-
Due 10/14/2021 at 1:15 pm.
-
PSet 4 Answer Key File
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) File
-
-
LGD: Risk Prediction Models
Prediction models: differences from diagnostic tests; calibration and discrimination; comparing predictions; value of prognostic information.Faculty: Michael Kohn
-
The 2021 Lecture 5 on risk prediction is pre-recorded. The 2019 lecture was recorded live as given on 10/17/2019.
We are making both the 2019 and 2021 lectures available to you.
This is the PPT for the 2021 version. The notation changes that I hand-wrote while recording the lecture have been incorporated into the slides. Also, the rounding problems have been corrected.
-
This is the PPT for the 2019 version of Lecture 5 on Risk Prediction Models as given live on 10/17/2019.
-
2021 Lecture 5 Segment 1 - Calibration (33:16) Media Resource
-
This is the material covered in Lecture 5 Segment 1.
-
2021 Lecture 5 Segment 2 -- Net Benefit and Decision Curves (40:50) Media Resource
-
2019 Lecture 5 Risk Model -- Recorded Live on 10/17/2019 (1:26:25) Media Resource
-
Zoom Recording of Large Group 10/19/2021 Media Resource
-
Duplicate posting under Required Reading. This is the material covered in Lecture 5, Segment 1.
-
Vickers_Net_Benefit_BMJ_2016.pdf File
-
Kerr 2015 Decision Curves File
-
Pset 5 Risk Prediction Key File
-
Stata do-file to create a dataset with meteorologist predictions and (rain) outcomes for HW#6, Problem 1, then run Brier command.
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) File
-
-
LGD: Combining Tests and Multivariable Risk Models
Combining tests/diagnostic models:Importance of test non-independence; recursive partitioning; logistic regression; importance of validation separate from derivationFaculty: Michael Kohn
-
Lecture 6 Segment 1 -- Test Non-independence (29:28) Media Resource
-
Lecture 6 Segment 2 Classification Trees and Logistic Regression (36:51) Media Resource
-
Zoom Recording of Large Group Discussion 10/26/2021 Media Resource
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) (copy) File
-
-
LGD: Quantifying the benefits and harms of treatments using RCTS / Alternatives to RCTs
RCT checklist; importance of baseline incidence; measures of effect size: Relative risk (RR), relative risk reduction (RRR), Odds ratio (OR), Absolute risk reduction (ARR), Number needed to treat (NNT); "Back of the Envelope Cost Effectiveness Analysis"Faculty: Thomas Newman
-
Lecture 7 Part 1, (Introduction and critical appraisal of RCTs; 37:52) 2020-1028 URL
-
Lecture 7, Part 2 (Quantifying treatment effects; 42:33) 2020-1028 URL
-
Zoom Recording of Large Group Discussion 11/2/2021 Media Resource
-
This is an expanded version of Table 8.1 created by Nicole Rodriquez.
-
LeNoury 2015 BMJ Restoring Study 329 File
-
Newman 2004 NEJM Antidepressant Perspective File
-
Stempel 2016 NEJM AUSTRI study Advair v fluticasone alone File
-
Stolker 2014 Circulation rethinking composite endpoints in clinical trials File
-
Ciani 2013 BMJ Surrogate vs patient relevant outcomes File
-
Miner 2017 Am J Gastro Pecanatide for chronic constipation File
-
This recording has some of the same examples as the lecture on Chapter 8; feel free to skip around.
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL. -
Screening Tests
Need for a more critical approach for screening tests, risk factor vs disease screening; biases in observational studies of screening: lead time bias, length time bias, volunteer bias, stage migration bias, pseudodisease; randomized trials of screening
Faculty: Martina Steurer-Muller
-
Lecture 8 Screening (Recorded in 2019) Media Resource
-
This is a problem on length- and lead-time bias that we were going to put on PSet#8, but it's too hard and requires a lot of clinical knowledge. I tried to add some explanation and some additional pieces to make it easier. Martina and Tom are planning to discuss it at large group discussion. They have not reviewed my changes and may think I made things worse rather than better.
-
Merenstein D. Winners & Losers. JAMA 2004;291:242-5 File
-
Heath 2014 BMJ Role of fear in overdiagnosis File
-
Heath 2013 BMJ Overdiagnosis When good intentions meet vested interests File
-
Newman 2016 JAMA IM Lipid screening in children-low value care File
-
Annas 1995 NEJM Reframing the Debate on Healthcare by replacing Metaphors File
-
Tilman 2014 Nature Global diets link environmental sustainability and human health File
-
Problem Set #8 (M.Laker Section) Assignment
-
PSet #8 Answer Key File
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Friday NOT Thursday, November 12, 2021; 1:15 PM - 2:45 PM (Additional session: Thursday 11/11, 6:00 - 7:30 PM)
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL.-
EPI 204 Section Assignment (includes Zoom Access Information) File
Adam Zakaria's section has been re-distributed to the other 3 sections. He has emailed you to assign you to one of Nicole, Anita (Michael), or Vidya (Tom).
You may also attend Nasim's section on Thursday evening at 6 pm.
-
-
P-values and confidence intervals, Alternatives to randomized trials
Introduction and justification: why this Bayesian stuff is important and why we think you can handle it; what P-values and confidence intervals don' t mean; what they do mean; confidence intervals for negative studies and for proportions with small numerators. Alternatives to RCTs: When randomized trials are and may not be needed; instrumental variables and natural experiments; measuring additonal predictor and outcome variables to estimate bias; propensity scores.
Faculty: Thomas Newman
-
Tom's 2019 Lecture on Alternatives to RCTS and P-values and CI (WATCH FOR 2020) URLUpdate for Slide 28: Vinay Prasad is in our department at UCSF now!
-
Zoom Recording of Large Group 11/16/2021 Media Resource
-
Goodman S 2008 Seminars Hematol Dirty Dozen P Value fallacies File
-
Laptook 2017 JAMA Effect of therapeutic hypothermia after 6 hours Bayesian example File
-
Quintana 2017 JAMA Bayesian analysis. Using prior information to interpret results of clinical trials File
-
Browner WS, Newman TB. Are all significant P-values created equal? The analogy between diagnostic tests and clinical research. JAMA 1987;257:2459-63 File
-
Goodman, S. Toward evidence-based medical statistics. 1: The P value fallacy. Annals of Internal Medicine 1999; 130: 995-1004 File
-
Goodman, S. Toward evidence-based medical statistics. 2: The Bayes factor. Annals of Internal Medicine 1999; 130: 1005-1013 File
-
Newman, T. If almost nothing goes wrong, is almost everything all right? JAMA 1995; 274: 1013 File
-
Goodman S 2013 Circulation Are we there yet File
-
Ray 2006 American Journal of Epidemiology New User DEsigns File
-
Problem Set #9 (M.Laker Section) Assignment
-
Drop-in Help
Course faculty are available to address questions
Location (Access restricted to registered students): Zoom URL
-
Small Group Discussion (see roster for your section assignment)
Review of prior lecture and problem set
Faculty: Vidya Eswaran, Anita Hargrave, Nicole Rodriguez, Nasim Sobhani, Adam Zakaria, Martina Steurer, Tom Newman, Alexis Beatty, and Michael Kohn
Location (Access restricted to registered students): Zoom URL. -
Cognitive Biases / Course Review
Criticisms of evidence-based medicine; stories and statistics; psychology of medical decision making: heuristics used to estimate probability. Course review.Faculty: Michael Kohn
-
2021 Students: Note that this is last year's course review.
-
2018 Lecture Cognitive Biases and Course Review (2021 students should watch this) URL
-
Newman, TB. The Power of Stories over Statistics.BMJ. 2003;327(7429):1424-7 File
-
If possible, please upload as a Word .docx. We will grade using "Track Changes".
Due 12/6/21. 11:59 PM.
Do NOT collaborate with classmates.
-
Final Exam Review
Faculty: Michael Kohn-
Zoom link (click on the word Zoom) or +1 669 900 6833 (Meeting ID: 945 0978 8609) (Password: 169414)
-
Review of the Final Exam 12/8/2021 Media Resource
-
Final Exam with Answers File
-

