Author(s): Rajesh Mangrulkar, M.D., 2011License: Unless otherwise noted, this material is made available under the terms o...
Attribution Key                       for more information see: http://open.umich.edu/wiki/AttributionPolicyUse + Share + ...
Patients and Populations                 Medical Decision-Making            Shifting Probabilities and Questions          ...
Learning Objectives for Today• By the end of this lecture, you will…  – describe Bayesian probabilistic rules, as they    ...
A Clinical Tale• 20 year-old woman presents for genetic  testing• Mother had breast and ovarian cancer,  likely has the BR...
The Tale Continues…FFwd• At age 75 she has not been diagnosed  with breast or ovarian cancer.• Is her probability of havin...
Conditional Probabilities• What is the probability of event B, given  an event A? Written as P(B | A).  – Example: P (BRCA...
Basic Probabilistic Rules        Examples of types of Events• Dependent events: occurrence of 1 depends  to some extent on...
Combining Probabilities of Events• Pr (A   B) = Pr (A) + Pr (B)  – If A and B are mutually exclusive events• Pr (A   B) = ...
Back to our story75 yo woman whose mother very likely  had the BRCA gene, but who has not  herself been diagnosed with bre...
Considering both sides…                Dependent Events              Pr (A   B) = Pr (A) * Pr (B|A)•Step 1:       P (BRCA ...
But that doesn’t tell the full story…• Joint probabilities  – P (BRCA    and no breast ca) = 0.15  – P (NO BRCA and no bre...
WARNINGCONFUSING  MATH  AHEAD
Step 3: Bayes Theorem• Conditional probability is the relative proportion of the  relevant joint probability to the sum of...
Applying Bayes Theorem• P (BRCA | no breast ca) =               0.15        ------------------- = 26%        0.15 + 0.4375...
Why is this important?• Illustration of changing probabilities, and  shifting uncertainty…   …because of test results   …b...
Thread 1: Information Retrieval                    Ask           Acquire                      Apply                       ...
Who is this man?      Image of “The Riddler” is used for illustrative purposes, in an      effort to advance the instructo...
Riddle me this...• How many questions do clinicians ask  while they care for patients? 4 per patient• Why is question-gene...
Patient Care: Questions Decisions• Step 1: Question Search for Answer• Step 2: Assess the strength of answer• Step 3: Weig...
The Well-Structured Clinical Question• Purposes  – Target resources  – Define search terms  – Define what you and the pati...
Anatomy of a Background Question  •   What         •   Disorder  •   How          •   Syndrome  •   Where        •   Findi...
Background Questions -- Examples• Who should get influenza vaccine and when?• Which drugs to treat HIV can cause  pancreat...
Background vs. Foreground          Questions• Background: Designed to improve  general knowledge about a subject• Foregrou...
An Evolution in Question Type100 90 80 70 60                                              Foreground 50                   ...
Background Questions: A CaseUsing the following case, jot down 2 questions with  your partner that may help you care for t...
Background Questions• What kind of medication is glyburide?• In what classes of medication do metformin  and glyburide fal...
Sources for Background Questions   • Course notes, lectures, syllabi   • Textbooks     – MD Consult     – Stat!Ref     – D...
Foreground Questions -- Examples• In patients with chronic       • In patients with acute  atrial fibrillation over the   ...
Foreground Questions -- ExamplesTherapy• In patients with chronic atrial fibrillation over  the age of 70, does warfarin a...
PICO: A Tool to Structure the       Foreground Question      Therapy          DiagnosisP     Patient Pop      DiseaseI    ...
ApplyAppraise
Upcoming SlideShare
Loading in …5
×

The Information Cycle

909 views

Published on

A lecture on the information cycle by Dr. Rajesh Mangrulkar, M.D. This lecture was taught as a part of the University of Michigan Medical School's M1 - Patients and Populations Sequence.

View the course materials:
http://open.umich.edu/education/med/m1/patientspop-decisionmaking/2010/materials

Creative Commons Attribution-Non Commercial-Share Alike 3.0 License
http://creativecommons.org/licenses/by-nc-sa/3.0/

Published in: Health & Medicine, Technology
0 Comments
1 Like
Statistics
Notes
  • Be the first to comment

No Downloads
Views
Total views
909
On SlideShare
0
From Embeds
0
Number of Embeds
3
Actions
Shares
0
Downloads
6
Comments
0
Likes
1
Embeds 0
No embeds

No notes for slide
  • This will NOT be the only time you encounter Bayes Theorem. This example is meant to only familiarize you with the principle. You will be spending much more time with this concept in your Human Genetics component and in the small group sessions there.
  • As we practice medicine (and as you begin to take care of patients), you ’ ll realize that there is more to it than just the 4 A ’ s. Step 1 is of course asking the right question and getting answers Step 2 mandates that we assess the strength of the evidence. remember that studies do NOT prove the truth, rather they lend credence to it. They present weak, moderate or strong evidence in support of it. In step 3, this is where EBM stops, and MDM (medical decision-making) begins. It involves weighing the latest knowledge with your own beliefs, the values of the patient, and society ’ s values. Maybe a therapy would not be covered. Maybe a patient ’ s understanding of his or her own disease is such that he would not accept the therapy you are prescribing. This is the most complex step. This is the art.
  • Let ’ s first start with Step 1: Asking questions… First, how does this skill practically help you? Well it does in 3 ways: First, understanding your question can greatly help you figure out where to go to find the answer. For example, a controversial question that you know has been studied extensively (e.g., how to screen for prostate cancer) may be best answered with a practice guideline. A general knowledge question (what is Hashimoto ’ s thyroiditis?), is better found in a textbook, rather than on MEDLINE. Second, defining your question helps determine your search terms. It ’ s very easy to go from the question to the search this way, if you ’ re clear up front what you are interested in. Finally, and I think most importantly, it helps you define what you and your patient are most interested in looking at up-front. This last extremely important benefit will be more clear as we work through some examples.
  • The main categorization that some proponents of EBM make, is that between background and foreground questions. We will define this as follows: <read> Background…We tend to ask these questions very frequently in our careers, especially early on. They may or may not be related to a specific patient, and may not be linked to a specific decision you have to make. They are very important, because they lay the foundation to asking more sophisticated Foreground questions. <read> . Let ’ s step through a case to help illustrate the differences.
  • The Information Cycle

    1. 1. Author(s): Rajesh Mangrulkar, M.D., 2011License: Unless otherwise noted, this material is made available under the terms ofthe Creative Commons Attribution–Non-commercial–Share Alike 3.0 License:http://creativecommons.org/licenses/by-nc-sa/3.0/We have reviewed this material in accordance with U.S. Copyright Law and have tried to maximize your ability to use, share, andadapt it. The citation key on the following slide provides information about how you may share and adapt this material.Copyright holders of content included in this material should contact open.michigan@umich.edu with any questions, corrections, orclarification regarding the use of content.For more information about how to cite these materials visit http://open.umich.edu/education/about/terms-of-use.Any medical information in this material is intended to inform and educate and is not a tool for self-diagnosis or a replacement formedical evaluation, advice, diagnosis or treatment by a healthcare professional. Please speak to your physician if you have questionsabout your medical condition.Viewer discretion is advised: Some medical content is graphic and may not be suitable for all viewers.
    2. 2. Attribution Key for more information see: http://open.umich.edu/wiki/AttributionPolicyUse + Share + Adapt { Content the copyright holder, author, or law permits you to use, share and adapt. } Public Domain – Government: Works that are produced by the U.S. Government. (17 USC § 105) Public Domain – Expired: Works that are no longer protected due to an expired copyright term. Public Domain – Self Dedicated: Works that a copyright holder has dedicated to the public domain. Creative Commons – Zero Waiver Creative Commons – Attribution License Creative Commons – Attribution Share Alike License Creative Commons – Attribution Noncommercial License Creative Commons – Attribution Noncommercial Share Alike License GNU – Free Documentation LicenseMake Your Own Assessment { Content Open.Michigan believes can be used, shared, and adapted because it is ineligible for copyright. } Public Domain – Ineligible: Works that are ineligible for copyright protection in the U.S. (17 USC § 102(b)) *laws in your jurisdiction may differ { Content Open.Michigan has used under a Fair Use determination. } Fair Use: Use of works that is determined to be Fair consistent with the U.S. Copyright Act. (17 USC § 107) *laws in your jurisdiction may differ Our determination DOES NOT mean that all uses of this 3rd-party content are Fair Uses and we DO NOT guarantee that your use of the content is Fair. To use this content you should do your own independent analysis to determine whether or not your use will be Fair.
    3. 3. Patients and Populations Medical Decision-Making Shifting Probabilities and Questions Rajesh S. Mangrulkar, M.D. University of Michigan Department of Internal Medicine Division of General MedicineFall 2011
    4. 4. Learning Objectives for Today• By the end of this lecture, you will… – describe Bayesian probabilistic rules, as they apply to a basic diagnostic question – summarize how uncertainty in diagnostic reasoning interacts with trust of the practitioner. – explain the difference between background and foreground clinical questions – recognize how individual targeted searches for the answers to clinical questions drive self-directed learning that is crucial for all practitioners – be able to craft foreground questions for both diagnosis and treatment, using the PICO format
    5. 5. A Clinical Tale• 20 year-old woman presents for genetic testing• Mother had breast and ovarian cancer, likely has the BRCA gene (autosomal dominant)• With this assumption, the patient’s likelihood of having the gene is…50%• She decides not to get tested.
    6. 6. The Tale Continues…FFwd• At age 75 she has not been diagnosed with breast or ovarian cancer.• Is her probability of having the BRCA gene different at age 75 than it was at age 20? – Yes: it is lower – How much lower?
    7. 7. Conditional Probabilities• What is the probability of event B, given an event A? Written as P(B | A). – Example: P (BRCA | no breast cancer)• Key concept: – Conditional probabilities can be combined with prior probabilities to create joint probabilities
    8. 8. Basic Probabilistic Rules Examples of types of Events• Dependent events: occurrence of 1 depends to some extent on the other – Example: The same person passing step 1 of the boards and then passing step 2 of the boards 2 years later.• Independent events: both can occur – Example: 2 different people passing step 1 of the boards (abiding by the honor code)• Mutually exclusive events: cannot both occur – Example: A person getting >250 on step 1 of the boards, or the same person getting 220-250 on step 1.
    9. 9. Combining Probabilities of Events• Pr (A B) = Pr (A) + Pr (B) – If A and B are mutually exclusive events• Pr (A B) = Pr (A) * Pr (B) – If A and B are independent events• Pr (A B) = Pr (A) * Pr (B|A) – If A and B are dependent events (Joint probability) = OR = AND
    10. 10. Back to our story75 yo woman whose mother very likely had the BRCA gene, but who has not herself been diagnosed with breast cancer.• Our patient wants to know: – What is P (BRCA | no breast ca)?
    11. 11. Considering both sides… Dependent Events Pr (A B) = Pr (A) * Pr (B|A)•Step 1: P (BRCA and no breast cancer) = P(BRCA) * P(no breast ca | BRCA) = 0.5 * 0.3 (from studies) = 0.15•Step 2: P (NO BRCA and no breast ca) = P(NO BRCA)*P(no breast ca|NO BRCA) = 0.5 * 0.875(from studies) = 0.4375
    12. 12. But that doesn’t tell the full story…• Joint probabilities – P (BRCA and no breast ca) = 0.15 – P (NO BRCA and no breast ca) = 0.4375• The assumption is that these are NOT independent events.• Again, our patient wants to know: – What is P (BRCA | no breast ca)?
    13. 13. WARNINGCONFUSING MATH AHEAD
    14. 14. Step 3: Bayes Theorem• Conditional probability is the relative proportion of the relevant joint probability to the sum of all the joint probabilities.• P(BRCA | no breast ca) = P (BRCA & no breast ca) P (no breast ca) = P(BRCA) * P (no breast ca | BRCA) P (no breast ca)• P (no breast ca) = sum of all the joint probabilities • P (no breast ca & BRCA) • P (no breast ca & NOT BRCA)
    15. 15. Applying Bayes Theorem• P (BRCA | no breast ca) = 0.15 ------------------- = 26% 0.15 + 0.4375• 26% is significantly lower than 50% (our prior probability)
    16. 16. Why is this important?• Illustration of changing probabilities, and shifting uncertainty… …because of test results …because of events …because of time• Fundamentally, clinicians deal with shifting probabilities and uncertainty with each patient they encounter – Many tools to help (Bayes, 2x2, Likelihood ratios)
    17. 17. Thread 1: Information Retrieval Ask Acquire Apply Appraise
    18. 18. Who is this man? Image of “The Riddler” is used for illustrative purposes, in an effort to advance the instructor’s teaching goals. This use is Fair and consistent with the U.S. Copyright Act. (USC 17 § 107)Frank Gorshin - a.k.a. “The Riddler”
    19. 19. Riddle me this...• How many questions do clinicians ask while they care for patients? 4 per patient• Why is question-generation a critical skill? (1) Targeting Information Resources (2) Generating Search Terms
    20. 20. Patient Care: Questions Decisions• Step 1: Question Search for Answer• Step 2: Assess the strength of answer• Step 3: Weigh against: – Patient Values – Physician Values – Society’s Values
    21. 21. The Well-Structured Clinical Question• Purposes – Target resources – Define search terms – Define what you and the patient care about• Two Types – Background – Foreground
    22. 22. Anatomy of a Background Question • What • Disorder • How • Syndrome • Where • Finding • When • Health state • Who • Concern • Why
    23. 23. Background Questions -- Examples• Who should get influenza vaccine and when?• Which drugs to treat HIV can cause pancreatitis?• What is the metabolic pathway for cholesterol synthesis?• Why do patients with sleep apnea have high blood pressure?
    24. 24. Background vs. Foreground Questions• Background: Designed to improve general knowledge about a subject• Foreground: Patient-specific questions, strong implications for decisions, often with comparisons
    25. 25. An Evolution in Question Type100 90 80 70 60 Foreground 50 Background 40 30 20 10 0 M1-M2 M3-M4 Residency Practice
    26. 26. Background Questions: A CaseUsing the following case, jot down 2 questions with your partner that may help you care for this patient:A 42 year old woman comes to her primary care practitioner’s office for follow up of her diabetes. She is currently on glyburide 10 mg twice daily. However, her morning and evening blood sugars still stay elevated. You are the medical student who sees this patient with your attending. Afterwards, your attending asks whether you think she should add metformin to her regimen. You say that you don’t know because your knowledge of diabetes medications are sketchy.
    27. 27. Background Questions• What kind of medication is glyburide?• In what classes of medication do metformin and glyburide fall?• What is the initial dosage of metformin?• What are the adverse effects of metformin that I must be cautious about?• Is it safe to be on glyburide and metformin at the same time?
    28. 28. Sources for Background Questions • Course notes, lectures, syllabi • Textbooks – MD Consult – Stat!Ref – DynaMed • Review articles • Practice Guidelines
    29. 29. Foreground Questions -- Examples• In patients with chronic • In patients with acute atrial fibrillation over the chest pain of less than age of 70, does warfarin 6 hours’ duration, what anticoagulation reduce is the diagnostic the rate of stroke and accuracy of a single death when compared troponin level when with aspirin? compared with serial EKG’s and enzymes?
    30. 30. Foreground Questions -- ExamplesTherapy• In patients with chronic atrial fibrillation over the age of 70, does warfarin anticoagulation reduce the rate of stroke and death when compared with aspirin?Diagnosis• In patients with acute chest pain of less than 6 hours’ duration, what is the diagnostic accuracy of a single troponin level when compared with serial EKG’s and enzymes?
    31. 31. PICO: A Tool to Structure the Foreground Question Therapy DiagnosisP Patient Pop DiseaseI Intervention TestC Comparison Gold StandardO Outcome Accuracy
    32. 32. ApplyAppraise

    ×