Disclosure. I have no financial relationships with commercial entities producing healthcare related products and/or services.

Transcription

1 The Challenge of Numeracy: When, Why, and How We Should (or Should Not) Use Numbers to Communicate with Patients Brian J. Zikmund-Fisher, PhD University of Michigan Department of Health Behavior & Health Education Department of Internal Medicine Center for Bioethics & Social Sciences in Medicine Risk Science Center Disclosure I have no financial relationships with commercial entities producing healthcare related products and/or services. Health Numbers Blood test results Bilirubin Hemoglobin A1c CBC Hormone levels Home test results Blood pressure Glucose Disease risk % Cancer Heart disease Diabetes Treatment risk % Success rates Side effects False positives Health Numbers Medication Blood Dosing test Charts results Disease risk % Calorie Counts Bilirubin Cancer Weight Hemoglobin A1c Heart disease CBC Diabetes Fat grams Hormone levels Exercise Monitors Treatment risk % Nutrition labels Home test results Success rates Blood pressure Side effects Serving Sizes Temperature Glucose False positives Why Do We Give People Health Data? Blood Tests Why do clinicians conduct blood tests? Diagnosis Screening Monitoring Blood Tests Why do we want to communicate blood test results to patients? Because! Because it s their data So they know their numbers 1

2 So that they can use them Blood Tests Why do we want to communicate blood test results to patients? Awareness? Concern? Classification? Gist differences? Exact differences? Why Does Need / Purpose Matter? Numbers need to speak to us See also: Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research and Review, 2013 But we can t hear numbers very well Numeracy The ability to understand, transform, and derive meaning from quantitative (health) information Numeracy Measures Which of the following numbers represents the biggest risk of getting a disease? 1 in 100, 1 in 1000, or 1 in 10? Lipkus IM, Samsa G, Rimer BK. General performance on a numeracy scale among highly educated samples. Medical Decision Making in X Formats Subjective Numeracy How good are you at working with fractions? How good are you at figuring out how much a shirt will cost if it is 25% off? How often do you find numerical information to be useful? Fagerlin A, Zikmund-Fisher BJ, Ubel PA, Jankovic A, Derry H, Smith DM. Measuring numeracy without a math test: Development of the subjective numeracy scale (SNS). Medical Decision Making, Numeracy Is Related To Understanding uncertainty yet predictability of health What is a risk factor? Population rates vs. individual propensity Clinical trials Zikmund-Fisher BJ, Mayman G, Fagerlin A. Patient numeracy: what do patients need to recognize, think, or do with health numbers? In B Anderson, J Schulkin (eds.), Numerical Reasoning in Judgments and Decision Making About Health. Cambridge, UK: Cambridge University Press, 2014, pp

3 Numeracy Is Related To More Data Vs. Less Data Understanding uncertainty yet predictability of health Using data Translate between formats Interpret visuals Find relevant data Zikmund-Fisher BJ, Mayman G, Fagerlin A. Patient numeracy: what do patients need to recognize, think, or do with health numbers? In B Anderson, J Schulkin (eds.), Numerical Reasoning in Judgments and Decision Making About Health. Cambridge, UK: Cambridge University Press, 2014, pp Peters E, Dieckmann N, Dixon A, Hibbard JH, Mertz CK. Less is more in presenting quality information to consumers. Medical Care Research & Review, Peters E, Dieckmann N, Dixon A, Hibbard JH, Mertz CK. Less is more in presenting quality information to consumers. Medical Care Research & Review, Less is More Numeracy Is Related To Numeracy Is Related To Understanding uncertainty yet predictability of health Understanding risk information Being able to make better health decisions Adjust decisions based on data Understanding uncertainty yet predictability of health Understand risk information Be able to make better health decisions Completing tasks requiring number skills Medication dose calculation Self-monitoring Peters E, Dieckmann N, Dixon A, Hibbard JH, Mertz CK. Less is more in presenting quality information to consumers. Medical Care Research & Review, Zikmund-Fisher BJ, Mayman G, Fagerlin A. Patient numeracy: what do patients need to recognize, think, or do with health numbers? In B Anderson, J Schulkin (eds.), Numerical Reasoning in Judgments and Decision Making About Health. Cambridge, UK: Cambridge University Press, 2014, pp Zikmund-Fisher BJ, Mayman G, Fagerlin A. Patient numeracy: what do patients need to recognize, think, or do with health numbers? In B Anderson, J Schulkin (eds.), Numerical Reasoning in Judgments and Decision Making About Health. Cambridge, UK: Cambridge University Press, 2014, pp Numeracy, NotLiteracy, Predicts Reading Nutrition Labels Numeracy, NotLiteracy, Relates To Anticoagulation Control Numeracy as Deriving Meaning Rothman RL, Housam R, Weiss H, Davis D, Gregory R, Gebretsadik T, et al. Patient Understanding of Food Labels. The Role of Literacy and Numeracy. Am.J.Prev.Med, Estrada CA, Martin-Hryniewicz M, Peek BT, Collins C, Byrd JC. Literacy and numeracy skills and anticoagulation control. Am. J. Med. Sci,

4 Can Patients Use This? Patient Portals What Is Out of Range? Increasing direct access to test results BUT, patients must be able to recognize out-of-range values before they can derive any meaningful use Scenario Test Results Type 2 diabetes scenario Hemoglobin A1c 3 months ago = 6.8% Explicit goal of A1c<7% Being tested in-between appointments Participants received tables of: CBC counts CBC differential % s Hemoglobin A1c Renal panel Tables included standard range but did not include high/low flags Participants 1817 adults age Recruited from a demographically diverse Internet panel Measured both health literacy and numeracy Experimental Design A1c level 7.1% 8.4% Number of out-of-range values A1c only A1c + viral infection WBC, platelet, MCH, MCHC, neutrophil %, lymphocyte %, monocyte %, ANC, and serum glucose Effects of Numeracy and Literacy Zikmund-Fisher BJ, Exe NL, Witteman HO. Numeracy and literacy independently predict patients ability to identify out-of-range test results. Journal of Medical Internet Research 2014;16(8):e187. 4

5 Estimated Likelihood of Calling a Doctor Houston, we have a problem! Tables of test results are NOT the right tool for less numerate/literate Problem: We can know what numbers are without knowing what they mean. Information Evaluability Zikmund-Fisher BJ, Exe NL, Witteman HO. Numeracy and literacy independently predict patients ability to identify out-of-range test results. Journal of Medical Internet Research 2014;16(8):e187. Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research & Review, 2013 Hsee, 1996, Organizational Behavior & Human Decision Processes. Hsee, et al., 1999, Psychological Bulletin. Zikmund-Fisher, et al., 2004, Medical Decision Making. Good or Bad? Choosing a Fertility Clinic Preference Reversal in Ratings of Fertility Clinics Fertility Partners Reproductive Associates Separate Versions F (n=58) R (n=47) Joint Version (n=51) Amend B. Welcome to Jasorassic Park, 1998, p.36. Distance 15 min. 45 min. Success rate 33% 40% IVF procedures / yr Fertility Partners 7.1 (33% success, 15 min.) Reproductive Assoc. 6.2 (40% success, 45 min.) F>R p= R>F p=.051 Note: Rating scale Zikmund-Fisher BJ, Fagerlin A, Ubel PA. Is 28% good or bad? : Evaluability and preference reversals in health care decisions. Medical Decision Making, 2004 Zikmund-Fisher BJ, Fagerlin A, Ubel PA. Is 28% good or bad? : Evaluability and preference reversals in health care decisions. Medical Decision Making, 2004 Good or Bad? 5-year Breast Cancer Risk: 2.6% Decision Making Hard-to-evaluate data requires reference standards to be meaningful Such data are generally ignored unless comparative data are provided Hsee, 1996, Organizational Behavior & Human Decision Processes. Hsee, et al., 1999, Psychological Bulletin. Zikmund-Fisher, et al., 2004, Medical Decision Making. 5

6 The Curse of Knowledge Good or Bad? Information Evaluability Functional Numeracy Once we know something, we find it hard to imagine what it was like to not know it. Chip Heath & Dan Heath, Made to Stick, 2007, p. 20. Dioxin Blood Concentration: 33 parts per trillion TEQ Good or Bad? What Is Dangerous? Platelets: 145 x 10 9 What is normal? Vs. What is dangerous? NOTHING! Context? Context! Ongoing Design Research Your Result 145 Your Result 145 Standard (Normal) Range Bleeding Risk Standard (Normal) Range No Surgeries DRAFT Concept Image AHRQ R01 HS (B.J. Zikmund-Fisher, PI) 6

7 Ongoing Design Research Why Do We Give People Health Risk Estimates? Imagine Robert Your 10-year risk of cardiovascular disease is: 11.22% DRAFT Concept Images AHRQ R01 HS (B.J. Zikmund-Fisher, PI) Robert s Tale Problems Problems Am I at high risk, or not? Excess precision Do we really think we know Robert s risk to a hundredth of a percent? Integers are more believable and easier to recall than decimals Witteman, Zikmund-Fisher, et al., JMIR, 2011 Excess precision Number-only formats Robert s Risk Problems Iconarray.com Excess precision Number-only formats Unmet information needs Created at iconarray.com 7

8 Risk Estimates Risk Estimates Question Why do clinicians estimate health risks? Evaluation Decision making Surveillance Inform? Why do we want to Change beliefs? communicate Change behavior? risk estimates to patients? Is Robert informed about his cardiovascular disease risk? Brewer NT. Ch. 2 Goals in Fischhoff, Brewer, Downs, eds. Communicating Risks and Benefits: An Evidence-Based User s Guide. FDA, A Taxonomy of Risk Concepts Risk Concept Illustrative Risk Messages It could happen to me. Relative / Comparative It is more likely to happen to me. I am more likely to have this happen to me than to have that happen to me. Categorical I am a person who has a high chance of this happening. Relative Probability I have a risk that is higher to this degree. Absolute Probability My risk is this. Comparative My (group s) risk is this, which is higher than another s Probability (group s) risk. OR My risk is this if I do X, which is higher than my risk if I do Y which is that. Incremental Probability My risk will change that much if I do this. Needs What Robert wanted: I am a person who has a high risk What Robert got: My risk is this Need-Congruent Types of Risk Knowledge Need Avoid Surprise and Regret Recognize Dominant Options What Patients Care About Care that this could happen Care that this is most / least Motivate to Act or Not Act Care that this is good / bad Make Multi-Attribute Tradeoff Decisions Make Magnitude- Dependent Decisions Care about this more than that Care that this is X% not Y% Congruent Types of Risk Knowledge Relative / Comparative Categorical Comparative and/or Comparative Probability Precise Comparative or Incremental Probabilities Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research & Review, 2013 Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research & Review, 2013 Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research & Review, 2013 Need-Congruent Types of Risk Knowledge Need Avoid Surprise and Regret Recognize Dominant Options What Patients Care About Care that this could happen Care that this is most / least Motivate to Act or Not Act Care that this is good / bad Make Multi-Attribute Tradeoff Decisions Make Magnitude- Dependent Decisions Care about this more than that Care that this is X% not Y% Congruent Types of Risk Knowledge Relative / Comparative Categorical Comparative and/or Comparative Probability Precise Comparative or Incremental Probabilities What does Robert need? Context! Context Via Side by Side Icon Arrays Created at clinician.iconarray.com Zikmund-Fisher BJ. The right tool is what they need, not what we have: A taxonomy of appropriate precision in patient risk communication. Medical Care Research & Review,

9 vizhealth.org Visual Displays That Aid Categorization Questions People Ask Questions People Ask Does Robert need a number? What is my number? What is my number? What is normal? What is dangerous? Am I high? Is the difference important? Labels? E.g., above average, high levels, above threshold BUT: When statistics and evaluative labels are presented together, the label, NOT the number, dictates what people do. Zikmund-Fisher, et al., 2007, Am. J. Obstet. Gynecol. Non-Numerical Risk Displays Consumer Reports Patient Needs ALWAYS consider the congruence of data type and format with patients immediate and specific needs The Right Tool at the Right Time 9

10 might want need later might want need later NO Take Away Messages We must recognize why we are providing data before we provide numbers Improve functional numeracy via Context to create evaluability Need-congruent formats Providing the right number does not guarantee the right message 10