What Helps People Make Values-Congruent Medical Decisions? Eleven Strategies Tested across 6 Studies.

Background. High-quality health decisions are often defined as those that are both evidence informed and values congruent. A values-congruent decision aligns with what matters to those most affected by the decision. Values clarification methods are intended to support values-congruent decisions, but their effects on values congruence are rarely evaluated. Methods. We tested 11 strategies, including the 3 most commonly used values clarification methods, across 6 between-subjects online randomized experiments in demographically diverse US populations (n1 = 1346, n2 = 456, n3 = 840, n4 = 1178, n5 = 841, n6 = 2033) in the same hypothetical decision. Our primary outcome was values congruence. Decisional conflict was a secondary outcome in studies 3 to 6. Results. Two commonly used values clarification methods (pros and cons, rating scales) reduced decisional conflict but did not encourage values-congruent decisions. Strategies using mathematical models to show participants which option aligned with what mattered to them encouraged values-congruent decisions and reduced decisional conflict when assessed. Limitations. A hypothetical decision was necessary for ethical reasons, as we believed some strategies may harm decision quality. Later studies used more outcomes and covariates. Results may not generalize outside US-based adults with online access. We assumed validity and stability of values during the brief experiments. Conclusions. Failing to explicitly support the process of aligning options with values leads to increased proportions of values-incongruent decisions. Methods representing more than half of values clarification methods commonly in use failed to encourage values-congruent decisions. Methods that use models to explicitly show people how options align with their values offer more promise for helping people make decisions aligned with what matters to them. Decisional conflict, while arguably an important outcome in and of itself, is not an appropriate proxy for values congruence.

Primary Care Providers' Preferences and Concerns Regarding Specific Visual Displays for Returning Hemoglobin A1c Test Results to Patients.

Purpose. Patient portals of electronic health record systems currently present patients with tables of laboratory test results, but visual displays can increase patient understanding and sensitivity to result variations. We sought to assess physician preferences and concerns about visual display designs as potential motivators or barriers to their implementation. Methods. In an online survey, 327 primary care physicians (>50% patient care time) recruited through the online e-community/survey research firm SERMO compared hemoglobin A1c (HbA1c) test results presented in table format to various visual displays (number line formats) previously tested in public samples. Half of participants also compared additional visual formats displaying target goal ranges. Outcome measures included preferred display format and whether any displays were unacceptable, would change physician workload, or would induce liability concerns. Results. Most (85%-89%) respondents preferred visual displays over tables for result communications both to patients tested for diagnosis purposes and to diagnosed patients, with a design with color-coded categories most preferred. However, for each format (including tables), 11% to 23% rated them as unacceptable. Most respondents also preferred adding goal range information (in addition to standard ranges) for diagnosed patients. While most physicians anticipated no workload changes, 19% to 32% anticipated increased physician workload while 9% to 28% anticipated decreased workload. Between 32% and 40% had at least some liability concerns. Conclusions. Most primary care physicians prefer visual displays of HbA1c test results over table formats when communicating results to patients. However, workload and liability concerns from a minority of physicians represent a barrier for adoption of such designs in clinical settings.

Improving the Understanding of Test Results by Substituting (Not Adding) Goal Ranges: Web-Based Between-Subjects Experiment.

BACKGROUND: Most displays of laboratory test results include a standard reference range. For some patients (eg, those with chronic conditions), however, getting a result within the standard range may be unachievable, inappropriate, or even harmful. OBJECTIVE: The objective of our study was to test the impact of including clinically appropriate goal ranges outside the standard range in the visual displays of laboratory test results. METHODS: Participants (N=6776) from a demographically diverse Web-based panel viewed hypothetical glycated hemoglobin (HbA1c) test results (HbA1c either 6.2% or 8.2%) as part of a type 2 diabetes management scenario. Test result visual displays included either a standard range (4.5%-5.7%) only, a goal range (6.5%-7.5%) added to the standard range, or the goal range only. The results were displayed in 1 of the following 3 display formats: (1) a table; (2) a simple, two-colored number line (simple line); or (3) a number line with diagnostic categories indicated via colored blocks (block line). Primary outcome measures were comprehension of and negative reactions to test results. RESULTS: While goal range information did not influence the understanding of HbA1c=8.2% results, the goal range only display produced higher levels of comprehension and decreased negative reactions to HbA1c=6.2% test results compared with the no goal range and goal range added conditions. Goal range information was less helpful in the block line condition versus the other formats. CONCLUSIONS: Replacing the standard range with a clinically appropriate goal range could help patients better understand how their test results relate to their personal targets.

Effect of Harm Anchors in Visual Displays of Test Results on Patient Perceptions of Urgency About Near-Normal Values: Experimental Study.

BACKGROUND: Patient-facing displays of laboratory test results typically provide patients with one reference point (the "standard range"). OBJECTIVE: To test the effect of including an additional harm anchor reference point in visual displays of laboratory test results, which indicates how far outside of the standard range values would need to be in order to suggest substantial patient risk. METHODS: Using a demographically diverse, online sample, we compared the reactions of 1618 adults in the United States who viewed visual line displays that included both standard range and harm anchor reference points ("Many doctors are not concerned until here") to displays that included either (1) only a standard range, (2) standard range plus evaluative categories (eg, "borderline high"), or (3) a color gradient showing degree of deviation from the standard range. RESULTS: Providing the harm anchor reference point significantly reduced perceived urgency of close-to-normal alanine aminotransferase and creatinine results (P values <.001) but not generally for platelet count results. Notably, display type did not significantly alter perceptions of more extreme results in potentially harmful ranges. Harm anchors also substantially reduced the number of participants who wanted to contact their doctor urgently or go to the hospital about these test results. CONCLUSIONS: Presenting patients with evaluative cues regarding when test results become clinically concerning can reduce the perceived urgency of out-of-range results that do not require immediate clinical action.

Graphics help patients distinguish between urgent and non-urgent deviations in laboratory test results.

OBJECTIVE: Most electronic health record systems provide laboratory test results to patients in table format. We tested whether presenting such results in visual displays (number lines) could improve understanding. MATERIALS AND METHODS: We presented 1620 adults recruited from a demographically diverse Internet panel with hypothetical results from several common laboratory tests, first showing near-normal results and then more extreme values. Participants viewed results in either table format (with a "standard range" provided) or one of 3 number line formats: a simple 2-color format, a format with diagnostic categories such as "borderline high" indicated by colored blocks, and a gradient format that used color gradients to smoothly represent increasing risk as values deviated from standard ranges. We measured respondents' subjective sense of urgency about each test result, their behavioral intentions, and their perceptions of the display format. RESULTS: Visual displays reduced respondents' perceived urgency and desire to contact health care providers immediately for near-normal test results compared to tables but did not affect their perceptions of extreme values. In regression analyses controlling for respondent health literacy, numeracy, and graphical literacy, gradient line displays resulted in the greatest sensitivity to changes in test results. DISCUSSION: Unlike tables, which only tell patients whether test results are normal or not, visual displays can increase the meaningfulness of test results by clearly defining possible values and leveraging color cues and evaluative labels. CONCLUSION: Patient-facing displays of laboratory test results should use visual displays rather than tables to increase people's sensitivity to variations in their results.

Effect of Public Deliberation on Attitudes toward Return of Secondary Results in Genomic Sequencing.

The increased use of genomic sequencing in clinical diagnostics and therapeutics makes imperative the development of guidelines and policies about how to handle secondary findings. For reasons both practical and ethical, the creation of these guidelines must take into consideration the informed opinions of the lay public. As part of a larger Clinical Sequencing Exploratory Research (CSER) consortium project, we organized a deliberative democracy (DD) session that engaged 66 participants in dialogue about the benefits and risks associated with the return of secondary findings from clinical genomic sequencing. Participants were educated about the scientific and ethical aspects of the disclosure of secondary findings by experts in medical genetics and bioethics, and then engaged in facilitated discussion of policy options for the disclosure of three types of secondary findings: 1) medically actionable results; 2) adult onset disorders found in children; and 3) carrier status. Participants' opinions were collected via surveys administered one month before, immediately following, and one month after the DD session. Post DD session, participants were significantly more willing to support policies that do not allow access to secondary findings related to adult onset conditions in children (Χ 2 (2, N = 62) = 13.300, p = 0.001) or carrier status (Χ 2 (2, N = 60) = 11.375, p = 0.003). After one month, the level of support for the policy denying access to secondary findings regarding adult-onset conditions remained significantly higher than the pre-DD level, although less than immediately post-DD (Χ 2 (1, N = 60) = 2.465, p = 0.041). Our findings suggest that education and deliberation enhance public appreciation of the scientific and ethical complexities of genome sequencing.

Effects of Design Features of Explicit Values Clarification Methods: A Systematic Review.

BACKGROUND: Diverse values clarification methods exist. It is important to understand which, if any, of their design features help people clarify values relevant to a health decision. PURPOSE: To explore the effects of design features of explicit values clarification methods on outcomes including decisional conflict, values congruence, and decisional regret. DATA SOURCES: MEDLINE, all EBM Reviews, CINAHL, EMBASE, Google Scholar, manual search of reference lists, and expert contacts. STUDY SELECTION: Articles were included if they described the evaluation of 1 or more explicit values clarification methods. DATA EXTRACTION: We extracted details about the evaluation, whether it was conducted in the context of actual or hypothetical decisions, and the results of the evaluation. We combined these data with data from a previous review about each values clarification method's design features. DATA SYNTHESIS: We identified 20 evaluations of values clarification methods within 19 articles. Reported outcomes were heterogeneous. Few studies reported values congruence or postdecision outcomes. The most promising design feature identified was explicitly showing people the implications of their values, for example, by displaying the extent to which each of their decision options aligns with what matters to them. LIMITATIONS: Because of the heterogeneity of outcomes, we were unable to perform a meta-analysis. Results should be interpreted with caution. CONCLUSIONS: Few values clarification methods have been evaluated experimentally. More research is needed to determine effects of different design features of values clarification methods and to establish best practices in values clarification. When feasible, evaluations should assess values congruence and postdecision measures of longer-term outcomes.

One-Sided Social Media Comments Influenced Opinions And Intentions About Home Birth: An Experimental Study.

As people increasingly turn to social media to access and create health evidence, the greater availability of data and information ought to help more people make evidence-informed health decisions that align with what matters to them. However, questions remain as to whether people can be swayed in favor of or against options by polarized social media, particularly in the case of controversial topics. We created a composite mock news article about home birth from six real news articles and randomly assigned participants in an online study to view comments posted about the original six articles. We found that exposure to one-sided social media comments with one-sided opinions influenced participants' opinions of the health topic regardless of their reported level of previous knowledge, especially when comments contained personal stories. Comments representing a breadth of views did not influence opinions, which suggests that while exposure to one-sided comments may bias opinions, exposure to balanced comments may avoid such bias.

Design Features of Explicit Values Clarification Methods: A Systematic Review.

BACKGROUND: Values clarification is a recommended element of patient decision aids. Many different values clarification methods exist, but there is little evidence synthesis available to guide design decisions. PURPOSE: To describe practices in the field of explicit values clarification methods according to a taxonomy of design features. DATA SOURCES: MEDLINE, all EBM Reviews, CINAHL, EMBASE, Google Scholar, manual search of reference lists, and expert contacts. STUDY SELECTION: Articles were included if they described 1 or more explicit values clarification methods. DATA EXTRACTION: We extracted data about decisions addressed; use of theories, frameworks, and guidelines; and 12 design features. DATA SYNTHESIS: We identified 110 articles describing 98 explicit values clarification methods. Most of these addressed decisions in cancer or reproductive health, and half addressed a decision between just 2 options. Most used neither theory nor guidelines to structure their design. "Pros and cons" was the most common type of values clarification method. Most methods did not allow users to add their own concerns. Few methods explicitly presented tradeoffs inherent in the decision, supported an iterative process of values exploration, or showed how different options aligned with users' values. LIMITATIONS: Study selection criteria and choice of elements for the taxonomy may have excluded values clarification methods or design features. CONCLUSIONS: Explicit values clarification methods have diverse designs but can be systematically cataloged within the structure of a taxonomy. Developers of values clarification methods should carefully consider each of the design features in this taxonomy and publish adequate descriptions of their designs. More research is needed to study the effects of different design features.

Risk Communication, Values Clarification, and Vaccination Decisions.

Many health-related decisions require choosing between two options, each with risks and benefits. When presented with such tradeoffs, people often make choices that fail to align with scientific evidence or with their own values. This study tested whether risk communication and values clarification methods could help parents and guardians make evidence-based, values-congruent decisions about children's influenza vaccinations. In 2013-2014 we conducted an online 2×2 factorial experiment in which a diverse sample of U.S. parents and guardians (n = 407) were randomly assigned to view either standard information about influenza vaccines or risk communication using absolute and incremental risk formats. Participants were then either presented or not presented with an interactive values clarification interface with constrained sliders and dynamic visual feedback. Participants randomized to the risk communication condition combined with the values clarification interface were more likely to indicate intentions to vaccinate (ß = 2.10, t(399) = 2.63, p < 0.01). The effect was particularly notable among participants who had previously demonstrated less interest in having their children vaccinated against influenza (ß = -2.14, t(399) = -2.06, p < 0.05). When assessing vaccination status reported by participants who agreed to participate in a follow-up study six months later (n = 116), vaccination intentions significantly predicted vaccination status (OR = 1.66, 95%CI (1.13, 2.44), p < 0.05) and rates of informed choice (OR = 1.51, 95%CI (1.07, 2.13), p < 0.012), although there were no direct effects of experimental factors on vaccination rates. Qualitative analysis suggested that logistical barriers impeded immunization rates. Risk communication and values clarification methods may contribute to increased vaccination intentions, which may, in turn, predict vaccination status if logistical barriers are also addressed.

Numeracy and literacy independently predict patients' ability to identify out-of-range test results.

BACKGROUND: Increasing numbers of patients have direct access to laboratory test results outside of clinical consultations. This offers increased opportunities for both self-management of chronic conditions and advance preparation for clinic visits if patients are able to identify test results that are outside the reference ranges. OBJECTIVE: Our objective was to assess whether adults can identify laboratory blood test values outside reference ranges when presented in a format similar to some current patient portals implemented within electronic health record (EHR) systems. METHODS: In an Internet-administered survey, adults aged 40-70 years, approximately half with diabetes, were asked to imagine that they had type 2 diabetes. They were shown laboratory test results displayed in a standard tabular format. We randomized hemoglobin A1c values to be slightly (7.1%) or moderately (8.4%) outside the reference range and randomized other test results to be within or outside their reference ranges (ie, multiple deviations). We assessed (1) whether respondents identified the hemoglobin A1c level as outside the reference range, (2) how respondents rated glycemic control, and (3) whether they would call their doctor. We also measured numeracy and health literacy. RESULTS: Among the 1817 adult participants, viewing test results with multiple deviations increased the probability of identifying hemoglobin A1c values as outside the reference range (participants with diabetes: OR 1.47, 95% CI 1.12-1.92, P=.005; participants without diabetes: OR 1.50, 95% CI 1.13-2.00, P=.005). Both numeracy and health literacy were significant predictors of correctly identifying out-of-range values. For participants with diabetes, numeracy OR 1.32 per unit on a 1-6 scale (95% CI 1.15-1.51, P<.001) and literacy OR 1.59 per unit of a 1-5 scale (95% CI 1.35-1.87, P<.001); for participants without diabetes, numeracy OR 1.36 per unit (95% CI 1.17-1.58, P<.001) and literacy OR 1.33 per unit (95% CI 1.12-1.58, P=.001). Predicted probabilities suggested 77% of higher numeracy and health literacy participants, but only 38% of lower numeracy and literacy participants, could correctly identify the hemoglobin A1c levels as outside the reference range. Correct identification reduced perceived blood glucose control (mean difference 1.68-1.71 points on a 0-10 scale, P<.001). For participants with diabetes, increased health literacy reduced the likelihood of calling one's doctor when hemoglobin A1c=7.1% (OR 0.66 per unit, 95% CI 0.52-0.82, P<.001) and increased numeracy increased intention to call when hemoglobin A1c=8.4% (OR 1.36 per unit, 95% CI 1.10-1.69, P=.005). CONCLUSIONS: Limited health literacy and numeracy skills are significant barriers to basic use of laboratory test result data as currently presented in some EHR portals. Regarding contacting their doctor, less numerate and literate participants with diabetes appear insensitive to the hemoglobin A1c level shown, whereas highly numerate and literate participants with diabetes appear very sensitive. Alternate approaches appear necessary to make test results more meaningful.

Animated randomness, avatars, movement, and personalization in risk graphics.

BACKGROUND: Risk communication involves conveying two inherently difficult concepts about the nature of risk: the underlying random distribution of outcomes and how a population-based proportion applies to an individual. OBJECTIVE: The objective of this study was to test whether 4 design factors in icon arrays-animated random dispersal of risk events, avatars to represent an individual, personalization (operationalized as choosing the avatar's color), and a moving avatar-might help convey randomness and how a given risk applies to an individual, thereby better aligning risk perceptions with risk estimates. METHODS: A diverse sample of 3630 adults with no previous heart disease or stroke completed an online nested factorial experiment in which they entered personal health data into a risk calculator that estimated 10-year risk of cardiovascular disease based on a robust and validated model. We randomly assigned them to view their results in 1 of 10 risk graphics that used different combinations of the 4 design factors. We measured participants' risk perceptions as our primary outcome, as well as behavioral intentions and recall of the risk estimate. We also assessed subjective numeracy, whether or not participants knew anyone who had died of cardiovascular causes, and whether or not they knew their blood pressure and cholesterol as potential moderators. RESULTS: Animated randomness was associated with better alignment between risk estimates and risk perceptions (F1,3576=6.12, P=.01); however, it also led to lower scores on healthy lifestyle intentions (F1,3572=11.1, P<.001). Using an avatar increased risk perceptions overall (F1,3576=4.61, P=.03) and most significantly increased risk perceptions among those who did not know a particular person who had experienced the grave outcomes of cardiovascular disease (F1,3576=5.88, P=.02). Using an avatar also better aligned actual risk estimates with intentions to see a doctor (F1,3556=6.38, P=.01). No design factors had main effects on recall, but animated randomness was associated with better recall for those at lower risk and worse recall for those at higher risk (F1,3544=7.06, P=.01). CONCLUSIONS: Animated randomness may help people better understand the random nature of risk. However, in the context of cardiovascular risk, such understanding may result in lower healthy lifestyle intentions. Therefore, whether or not to display randomness may depend on whether one's goal is to persuade or to inform. Avatars show promise for helping people grasp how population-based statistics map to an individual case.

Blocks, ovals, or people? Icon type affects risk perceptions and recall of pictographs.

BACKGROUND: Research has demonstrated that icon arrays (also called "pictographs") are an effective method of communicating risk statistics and appear particularly useful to less numerate and less graphically literate people. Yet research is very limited regarding whether icon type affects how people interpret and remember these graphs. METHODS: 1502 people age 35-75 from a demographically diverse online panel completed a cardiovascular risk calculator based on Framingham data using their actual age, weight, and other health data. Participants received their risk estimate in an icon array graphic that used 1 of 6 types of icons: rectangular blocks, filled ovals, smile/frown faces, an outline of a person's head and shoulders, male/female "restroom" person icons (gender matched), or actual head-and-shoulder photographs of people of varied races (gender matched). In each icon array, blue icons represented cardiovascular events and gray icons represented those who would not experience an event. We measured perceived risk magnitude, approximate recall, and opinions about the icon arrays, as well as subjective numeracy and an abbreviated measure of graphical literacy. RESULTS: Risk recall was significantly higher with more anthropomorphic icons (restroom icons, head outlines, and photos) than with other icon types, and participants rated restroom icons as most preferred. However, while restroom icons resulted in the highest correlations between perceived and actual risk among more numerate/graphically literate participants, they performed no better than other icon types among less numerate/graphically literate participants. CONCLUSIONS: Icon type influences both risk perceptions and risk recall, with restroom icons in particular resulting in improved outcomes. However, optimal icon types may depend on numeracy and/or graphical literacy skills.

Belief in numbers: When and why women disbelieve tailored breast cancer risk statistics.

OBJECTIVE: To examine when and why women disbelieve tailored information about their risk of developing breast cancer. METHODS: 690 women participated in an online program to learn about medications that can reduce the risk of breast cancer. The program presented tailored information about each woman's personal breast cancer risk. Half of women were told how their risk numbers were calculated, whereas the rest were not. Later, they were asked whether they believed that the program was personalized, and whether they believed their risk numbers. If a woman did not believe her risk numbers, she was asked to explain why. RESULTS: Beliefs that the program was personalized were enhanced by explaining the risk calculation methods in more detail. Nonetheless, nearly 20% of women did not believe their personalized risk numbers. The most common reason for rejecting the risk estimate was a belief that it did not fully account for personal and family history. CONCLUSIONS: The benefits of tailored risk statistics may be attenuated by a tendency for people to be skeptical that these risk estimates apply to them personally. PRACTICE IMPLICATIONS: Decision aids may provide risk information that is not accepted by patients, but addressing the patients' personal circumstances may lead to greater acceptance.

Anticipating the arrival of low-penetrance genetic testing to primary care medicine.

Primary prevention is a pillar of primary care medicine. Furthermore, the identification of commonly occurring genetic mutations that confer only modest increases in disease risk (i.e., low-penetrance mutations or LPMs) is expanding our conception of how genetic testing supports prevention goals. To date, most predictive genetic testing has focused on identifying the minority of patients who carry mutations that significantly increase their risk for developing future disease (i.e., high-penetrance mutations or HPMs). Genetic tests for LPMs are more similar in structure and purpose to commonly used biomarker tests like lipid testing than to HPM testing. In the primary care setting, LPM testing will likely be presented to patients as one part of a multifactorial risk assessment that contains only a small amount of genetics-specific information. Consequently, preparing primary care clinicians for the anticipated use of LPM genetic tests will not require development of a completely new skill set but rather a re-conceptualization of both genetic testing and biomarker evaluation for primary prevention.

Animated graphics for comparing two risks: a cautionary tale.

BACKGROUND: The increasing use of computer-administered risk communications affords the potential to replace static risk graphics with animations that use motion cues to reinforce key risk messages. Research on the use of animated graphics, however, has yielded mixed findings, and little research exists to identify the specific animations that might improve risk knowledge and patients' decision making. OBJECTIVE: To test whether viewing animated forms of standard pictograph (icon array) risk graphics displaying risks of side effects would improve people's ability to select the treatment with the lowest risk profile, as compared with viewing static images of the same risks. METHODS: A total of 4198 members of a demographically diverse Internet panel read a scenario about two hypothetical treatments for thyroid cancer. Each treatment was described as equally effective but varied in side effects (with one option slightly better than the other). Participants were randomly assigned to receive all risk information in 1 of 10 pictograph formats in a quasi-factorial design. We compared a control condition of static grouped icons with a static scattered icon display and with 8 Flash-based animated versions that incorporated different combinations of (1) building the risk 1 icon at a time, (2) having scattered risk icons settle into a group, or (3) having scattered risk icons shuffle themselves (either automatically or by user control). We assessed participants' ability to choose the better treatment (choice accuracy), their gist knowledge of side effects (knowledge accuracy), and their graph evaluation ratings, controlling for subjective numeracy and need for cognition. RESULTS: When compared against static grouped-icon arrays, no animations significantly improved any outcomes, and most showed significant performance degradations. However, participants who received animations of grouped icons in which at-risk icons appeared 1 at a time performed as well on all outcomes as the static grouped-icon control group. Displays with scattered icons (static or animated) performed particularly poorly unless they included the settle animation that allowed users to view event icons grouped. CONCLUSIONS: Many combinations of animation, especially those with scattered icons that shuffle randomly, appear to inhibit knowledge accuracy in this context. Static pictographs that group risk icons, however, perform very well on measures of knowledge and choice accuracy. These findings parallel recent evidence in other data communication contexts that less can be more-that is, that simpler, more focused information presentation can result in improved understanding. Decision aid designers and health educators should proceed with caution when considering the use of animated risk graphics to compare two risks, given that evidence-based, static risk graphics appear optimal.

'I'll do what they did": social norm information and cancer treatment decisions.

OBJECTIVE: Using a cancer-treatment scenario, we tested whether descriptive norm information (e.g., the proportion of other people choosing a particular treatment) would influence people's hypothetical treatment choices. METHODS: Women from an Internet sample (Study 1 N=2238; Study 2 N=2154) were asked to imagine deciding whether to take adjuvant chemotherapy following breast cancer surgery. Across participants, we varied the stated proportion of women who chose chemotherapy. This descriptive norm information was presented numerically in Study 1 and non-numerically in Study 2. RESULTS: The descriptive norm information influenced women's decisions, with higher interest in chemotherapy when social norm information suggested that such chemotherapy was popular. Exact statistics about other people's decisions had a greater effect than when norms were described using less precise verbal terms (e.g., "most women"). CONCLUSION: Providing patients with information about what other people have done can significantly influence treatment choices, but the power of such descriptive norms depends on their precision. PRACTICE IMPLICATIONS: Communication of descriptive norms is only helpful if prevailing decisions in the population represent good clinical practice. Strategic presentation of such statistics, when available, may encourage patient outliers to modify their medical decisions in ways that result in improved outcomes.