Helping Doctors and Patients Make Sense of Health Statistics

The value of personalised risk information: a qualitative study of the perceptions of patients with prostate cancer

OBJECTIVE

To explore the experiences of patients with prostate cancer with risk information and their perceptions of the value of personalised risk information in treatment decisions.

DESIGN

A qualitative study was conducted using focus groups. Semistructured interviews explored participants' experiences with using risk information, and their perceptions of the potential value of personalised risk information produced by clinical prediction models.

PARTICIPANTS

English-speaking patients, ages 54-82, diagnosed with prostate cancer within the past 3 years, residing in rural and non-rural geographic locations in Maine (USA), and attending prostate cancer patient support groups.

SETTING

6 focus groups were conducted with 27 patients; separate groups were held for patients with low-risk, medium-risk and high-risk disease defined by National Comprehensive Cancer Network guidelines.

RESULTS

Several participants reported receiving risk information that was imprecise rather than precise, qualitative rather than quantitative, indirect rather than direct and focused on biomarker values rather than clinical outcomes. Some participants felt that personalised risk information could be useful in helping them make better informed decisions, but expressed scepticism about its value. Many participants favoured decision-making strategies that were heuristic-based and intuitive rather than risk-based and deliberative, and perceived other forms of evidence-emotions, recommendations of trusted physicians, personal narratives-as more reliable and valuable in treatment decisions.

CONCLUSIONS

Patients with prostate cancer appear to have little experience using personalised risk information, may favour heuristic-based over risk-based decision-making strategies and may perceive personalised risk information as less valuable than other types of evidence. These decision-making approaches and perceptions represent potential barriers to the clinical use of personalised risk information. Overcoming these barriers will require providing patients with greater exposure to risk information, education about the nature and value of personalised risk information and training in deliberative decision-making strategies. More research is needed to confirm these findings and address these needs.

How do physicians provide statistical information about antidepressants to hypothetical patients?

BACKGROUND

Little is known about how physicians provide statistical information to patients, which is important for informed consent.

METHODS

In a survey, obstetricians and gynecologists (N = 142) received statistical information about the benefit and side effects of an antidepressant. They received information in various formats, including event rates (antidepressant v. placebo), absolute risks, and relative risks. Participants had to imagine 2 hypothetical patients, 1 for whom they believed the drug to be safe and effective and 1 for whom they did not, and select the information they would give those patients. We assessed whether the information they selected for each patient was complete, transparent, interpretable, or persuasive (i.e., to nudge patients toward a particular option) and compared physicians who gave both patients the same information with those who gave both patients different information.

RESULTS

A similar proportion of physicians (roughly 25% each) selected information that was 1) complete and transparent, 2) complete but not transparent, 3) not interpretable for the patient because necessary comparative information was missing, or 4) suited for nudging. Physicians who gave both patients the same information (61% of physicians) more often selected at least complete information, even if it was often not transparent. Physicians who gave both patients different information (39% of physicians), in contrast, more often selected information that was suited for nudging in line with the belief they were asked to imagine. A limitation is that scenarios were hypothetical.

CONCLUSIONS

Most physicians did not provide complete and transparent information. Clinicians who presented consistent information to different patients tended to present complete information, whereas those who varied what information they chose to present appeared more prone to nudging.

Challenges in conducting community-driven research created by differing ways of talking and thinking about science: a researcher's perspective

Increasingly, health scientists are becoming aware that research collaborations that include community partnerships can be an effective way to broaden the scope and enhance the impact of research aimed at improving public health. Such collaborations extend the reach of academic scientists by integrating a variety of perspectives and thus strengthening the applicability of the research. Communication challenges can arise, however, when attempting to address specific research questions in these collaborations. In particular, inconsistencies can exist between scientists and community members in the use and interpretation of words and other language features, particularly when conducting research with a biomedical component. Additional challenges arise from differing perceptions of the investigative process. There may be divergent perceptions about how research questions should and can be answered, and in expectations about requirements of research institutions and research timelines. From these differences, misunderstandings can occur about how the results will ultimately impact the community. These communication issues are particularly challenging when scientists and community members are from different ethnic and linguistic backgrounds that may widen the gap between ways of talking and thinking about science, further complicating the interactions and exchanges that are essential for effective joint research efforts. Community-driven research that aims to describe the burden of disease associated with Helicobacter pylori infection is currently underway in northern Aboriginal communities located in the Yukon and Northwest Territories, Canada, with the goal of identifying effective public health strategies for reducing health risks from this infection. This research links community representatives, faculty from various disciplines at the University of Alberta, as well as territorial health care practitioners and officials. This highly collaborative work will be used to illustrate, from a researcher's perspective, some of the challenges of conducting public health research in teams comprising members with varying backgrounds. The consequences of these challenges will be outlined, and potential solutions will be offered.

Calculating alcohol risk in a visualization tool for promoting healthy behavior

OBJECTIVE

To investigate effective methods for communicating the personalized risks of alcohol consumption, particularly to young people.

METHODS

An interactive computerized blood alcohol content calculator was implemented in Flash based on literature findings for effectively communicating risk. Young people were consulted on attitudes to the animation features and visualization techniques used to display personalized risk based on disclosed alcohol consumption.

RESULTS

Preliminary findings reveal the calculator is relatively enjoyable to use for its genre. However, the primary aims of the visualization tool to effectively communicate personalized risk were undermined for some users by technical language. Transparency of risk calculations might further enhance the tool for others. Worryingly, user feedback revealed a tension between accurate presentation of risk and its consequent lack of sensationalism in terms of personal risk to the individual.

CONCLUSION

Initial findings suggest the tool may provide a relatively engaging vehicle for exploring the link between action choices and risk outcomes. Suggestions for enhancing risk communication include using intelligent techniques for selecting data presentation formats and for demonstrating the effects of sustained risky behavior.

PRACTICE IMPLICATIONS

Effective communication of risk contributes only partially to effecting behavior change; the role of the tool in influencing contributing attitudinal factors is also discussed.

A practical educational tool for teaching child-care hospital professionals attending evidence-based practice courses for continuing medical education to appraise internal validity in systematic reviews

RATIONALE, AIMS AND OBJECTIVES

Having a quick, practical, educational tool designed for busy child-care professionals to check whether systematic reviews (SRs) contain valid information would help them regularly update their evidence-based knowledge and apply it to their patients. Continuing our annual workshop courses encouraging paediatric hospital professionals to use evidence-based information, in a preliminary study, we compared the commonly used Critical Appraisal Skill Programme (CASP) questionnaire for appraising overall internal validity in SRs with a new, practical tool designed to check internal validity quickly.

METHOD

During a course in 2010, two 'teacher-brokers' taught experienced paediatric hospital professionals to use and compare the CASP and the new practical tool to appraise a Cochrane SR on beclomethasone for asthma in children by assessing internal validity only from the two most weighted randomized controlled trials in the forest plot. At 15 days and 6 months, participants then answered questionnaires designed to assess qualitative data including feelings about working together, memorization and possibly provide feedback for Cochrane reviewers.

RESULTS

Using the CASP, participants agreed that the Cochrane SR analysed contained overall valid results. Conversely, using the new quick tool, they found poor internal validity. Participants worked well together in a group, took less time to apply the new tool than the CASP (1 vs. 2.5 hours) and provided Cochrane feedback.

CONCLUSIONS

Our quick practical tool for teaching critical appraisal encourages busy child-care hospital professionals to work together, carefully check validity in SRs, apply the findings in clinical practice and provide useful feedback for Cochrane reviewers.

Establishing a program for individuals at high risk for breast cancer

Our need to create a program for individuals at high risk for breast cancer development led us to research the available data on such programs. In this paper, we summarize our findings and our thinking process as we developed our own program. Breast cancer incidence is increasing worldwide. Even though there are known risk factors for breast cancer development, approximately 60% of patients with breast cancer have no known risk factor, although this situation will probably change with further research, especially in genetics. For patients with risk factors based on personal or family history, different models are available for assessing and quantifying risk. Assignment of risk levels permits tailored screening and risk reduction strategies. Potential benefits of specialized programs for women with high breast cancer risk include more cost -effective interventions as a result of patient stratification on the basis of risk; generation of valuable data to advance science; and differentiation of breast programs from other breast cancer units, which can result in increased revenue that can be directed to further improvements in patient care. Guidelines for care of patients at high risk for breast cancer are available from various groups. However, running a high-risk breast program involves much more than applying a guideline. Each high-risk program needs to be designed by its institution with consideration of local resources and country legislation, especially related to genetic issues. Development of a successful high-risk program includes identifying strengths, weaknesses, opportunities, and threats; developing a promotion plan; choosing a risk assessment tool; defining "high risk"; and planning screening and risk reduction strategies for the specific population served by the program. The information in this article may be useful for other institutions considering creation of programs for patients with high breast cancer risk.

Is a picture worth a thousand words? The interaction of visual display and attribute representation in attenuating framing bias

The attribute framing bias is a well-established phenomenon, in which an object or an event is evaluated more favorably when presented in a positive frame such as “the half full glass” than when presented in the complementary negative framing. Given that previous research showed that visual aids can attenuate this bias, the current research explores the factors underlying the attenuating effect of visual aids. In a series of three experiments, we examined how attribute framing bias is affected by two factors: (a) The display mode—verbal versus visual; and (b) the representation of the critical attribute—whether one outcome, either the positive or the negative, is represented or both outcomes are represented. In Experiment 1 a marginal attenuation of attribute framing bias was obtained when verbal description of either positive or negative information was accompanied by corresponding visual representation. In Experiment 2 similar marginal attenuation was obtained when both positive and negative outcomes were verbally represented. In Experiment 3, where the verbal description represented both positive and negative outcomes, significant attenuation was obtained when it was accompanied by a visual display that represented a single outcome, and complete attenuation, totally eliminating the framing bias, was obtained when it was accompanied by a visual display that represented both outcomes. Thus, our findings showed that interaction between the display mode and the representation of the critical attribute attenuated the framing bias. Theoretical and practical implications of the interaction between verbal description, visual aids and representation of the critical attribute are discussed, and future research is suggested.

Genetic data and electronic health records: a discussion of ethical, logistical and technological considerations

Objective

The completion of sequencing the human genome in 2003 has spurred the production and collection of genetic data at ever increasing rates. Genetic data obtained for clinical purposes, as is true for all results of clinical tests, are expected to be included in patients’ medical records. With this explosion of information, questions of what, when, where and how to incorporate genetic data into electronic health records (EHRs) have reached a critical point. In order to answer these questions fully, this paper addresses the ethical, logistical and technological issues involved in incorporating these data into EHRs.

Materials and methods

This paper reviews journal articles, government documents and websites relevant to the ethics, genetics and informatics domains as they pertain to EHRs.

Results and discussion

The authors explore concerns and tasks facing health information technology (HIT) developers at the intersection of ethics, genetics, and technology as applied to EHR development.

Conclusions

By ensuring the efficient and effective incorporation of genetic data into EHRs, HIT developers will play a key role in facilitating the delivery of personalized medicine.

Statistical literacy of obstetrics-gynecology residents

BACKGROUND

Residents' ability to interpret statistics is important for scholarly pursuits and understanding evidence-based medicine. Yet there is limited research assessing residents' statistical literacy and their training in statistics.

METHODS

In 2011 we surveyed US obstetrics-gynecology residents participating in the Council for Resident Education in Obstetrics and Gynecology In-Training Examination about their statistical literacy and statistical literacy training.

RESULTS

Our response rate was 95% (4713 of 4961). About two-thirds (2980 of 4713) of the residents rated their statistical literacy training as adequate. Female respondents were more likely to rate their statistical literacy training poorly, with 25% (897 of 3575) indicating inadequate literacy compared with 17% (141 of 806) of the male respondents (P < .001). Respondents performed poorly on 2 statistical literacy questions, with only 26% (1222 of 4713) correctly answering a positive predictive value question and 42% (1989 of 4173) correctly defining a P value. A total of 51% (2391 of 4713) of respondents reported receiving statistical literacy training through a journal club, 29% (1359 of 4713) said they had informal training, 15% (711 of 4713) said that they had statistical literacy training as part of a course, and 11% (527 of 4713) said that they had no training.

CONCLUSIONS

The findings suggest that statistical literacy training for residents could still be improved. A total of 37% (1743 of 4713) of obstetrics-gynecology residents have received no formal statistical literacy training in residency. Fewer residents answered the 2 statistical literacy questions correctly compared with previous studies.

Patient autonomy in multiple sclerosis--possible goals and assessment strategies

Patient autonomy has been increasingly acknowledged as prerequisite for successful medical decision making in Western countries. In medical decisions with a need to involve a health professional, patient autonomy becomes apparent in the extent of patients' participation in the communication as described in the concept of shared decision making. Patient autonomy can be derived from different perspectives or goals and the focus of evaluation approaches may vary accordingly. Multiple sclerosis (MS) is a paradigmatic disease to study patient autonomy mainly because MS patients are highly disease competent and due to ambiguous evidence on many aspects of disease-related medical decision making. This review gives an overview on measurement issues in studying decision making in MS, categorized according to prerequisites, process measures and outcomes of patient autonomy. As relevant prerequisites role preferences, risk attribution, risk tolerance, and risk knowledge are discussed. Regarding processes, we distinguish intra-psychic and interpersonal aspects. Intra-psychic processes are elucidated using the theory of planned behavior, which guided development of a 30-item scale to capture decisions about immunotherapy. Moreover, a theory of uncertainty management has been created resulting in the development of a corresponding measurement concept. Interpersonal processes evolving between physician and patient can be thoroughly analyzed from different perspectives by use of the newly developed comprehensive MAPPIN'SDM inventory. Concerning outcomes, besides health related outcomes, we discuss match of preferred roles during the decision encounters (preference match), decisional conflict as well as an application of the multidimensional measure of informed choice to decisions of MS patients. These approaches provide an overview on patient-inherent and interpersonal factors and processes modulating medical decision making and health behavior in MS and beyond.

Impact of selective evidence presentation on judgments of health inequality trends: an experimental study

Reducing health inequalities is a key objective for many governments and public health organizations. Whether inequalities are measured on the absolute (difference) or relative (ratio) scale can have a significant impact on judgments about whether health inequalities are increasing or decreasing, but both of these measures are not often presented in empirical studies. In this study we investigated the impact of selective presentation of health inequality measures on judgments of health inequality trends among 40 university undergraduates. We randomized participants to see either a difference or ratio measure of health inequality alongside raw mortality rates in 5 different scenarios. At baseline there were no differences between treatment groups in assessments of inequality trends, but selective exposure to the same raw data augmented with ratio versus difference inequality graphs altered participants’ assessments of inequality change. When absolute inequality decreased and relative inequality increased, exposure to ratio measures increased the probability of concluding that inequality had increased from 32.5% to 70%, but exposure to difference measures did not (35% vs. 25%). Selective exposure to ratio versus difference inequality graphs thus increased the difference between groups in concluding that inequality had increased from 2.5% (95% CI −9.5% to 14.5%) to 45% (95% CI 29.4 to 60.6). A similar pattern was evident for other scenarios where absolute and relative inequality trends gave conflicting results. In cases where measures of absolute and relative inequality both increased or both decreased, we did not find any evidence that assignment to ratio vs. difference graphs had an impact on assessments of inequality change. Selective reporting of measures of health inequality has the potential to create biased judgments of progress in ameliorating health inequalities.

The effectiveness of a graphical presentation in addition to a frequency format in the context of familial breast cancer risk communication: a multicenter controlled trial

Background

Inadequate understanding of risk among counselees is a common problem in familial cancer clinics. It has been suggested that graphical displays can help counselees understand cancer risks and subsequent decision-making. We evaluated the effects of a graphical presentation in addition to a frequency format on counselees’ understanding, psychological well-being, and preventive intentions.

Design: Multicenter controlled trial.

Setting: Three familial cancer clinics in the Netherlands.

Methods

Participants: Unaffected women with a breast cancer family history (first-time attendees).

Intervention: Immediately after standard genetic counseling, an additional consultation by a trained risk counselor took place where women were presented with their lifetime breast cancer risk in frequency format (X out of 100) (n = 63) or frequency format plus graphical display (10 × 10 human icons) (n = 91).

Main outcome measures: understanding of risk (risk accuracy, risk perception), psychological well-being, and intentions regarding cancer prevention. Measurements were assessed using questionnaires at baseline, 2-week and 6-month follow-up.

Results

Baseline participant characteristics did not differ between the two groups. In both groups there was an increase in women’s risk accuracy from baseline to follow-up. No significant differences were found between women who received the frequency format and those who received an additional graphical display in terms of understanding, psychological well-being and intentions regarding cancer prevention. The groups did not differ in their evaluation of the process of counseling.

Conclusion

Women’s personal risk estimation accuracy was generally high at baseline and the results suggest that an additional graphical display does not lead to a significant benefit in terms of increasing understanding of risk, psychological well-being and preventive intentions.

Trial registration

Current Controlled Trials http://ISRCTN14566836

Conflicting and changing breast cancer screening recommendations: survey study of a national sample of ob-gyns after the release of the 2009 USPSTF guidelines

OBJECTIVE

To assess obstetrician-gynecologists' (ob-gyns') use of multiple conflicting guidelines assess after the release of the 2009 U.S. Preventive Services Task Force (USPSTF) breast cancer screening recommendations.

STUDY DESIGN

A nationally representative sample of American College of Obstetricians and Gynecologists (ACOG) Fellows were invited to complete a survey.

RESULTS

A total of 235 of 399 ob-gyns responded (59% response rate). Twenty percent and 89% indicated that USPSTF and ACOG guidelines influence their practice, respectively, 84% are influenced by more than one guideline. The plurality of respondents was able to correctly identify ACOG and USPSTF guidelines on 10 of 12 questions. One-third agreed with both ACOG's and USPSTF's recommendations regarding mammography screening for women 40-49 years old. A total of 42% of the sample made at least one change in their practice after the release of the 2009 USPSTF breast cancer screening guidelines.

CONCLUSION

Some ob-gyns made changes to their practices after the release of the USPSTF guidelines. When multiple guidelines exist, as in the case with breast cancer screening, physicians utilize multiple, and at times conflicting, guidelines. More research will be needed to better understand the impact (negative or positive) of multiple guidelines on the quality of healthcare.

Expressing death risk as condensed life experience and death intensity

Some risk exposures, including many medical and surgical procedures, typically carry hazards of death that are difficult to convey and appreciate in absolute terms. I propose presenting the death risk as a condensed life experience (i.e., the equivalent amount of life T that would carry the same cumulative mortality hazard for a person of the same age and sex based on life tables). For example, if the risk of death during an elective 1-hour procedure is 0.01%, and same-age and same-sex people have a 0.01% death risk over 1 month, one can inform the patient that "this procedure carries the same death risk as living 1 month of normal life." Comparative standards from other risky activities or from a person with the same disease at the same stage and same predictive profile could also be used. A complementary metric that may be useful to consider is the death intensity. The death intensity λ is the hazard function that shows the fold-risk estimate of dying compared with the reference person. The death intensity can vary substantially for different phases of the event, operation, or procedure (e.g., intraoperative, early postoperative, late postoperative), and this variability may also be useful to convey. T will vary depending on the time window for which it is computed. I present examples for calculating T and λ using literature data on accidents, ascent to Mount Everest, and medical and surgical procedures.

On avoiding framing effects in experienced decision makers

The present study aimed to (a) demonstrate the effect of positive–negative framing on experienced criminal justice decision makers, (b) examine the debiasing effect of visually structured risk messages, and (c) investigate whether risk perceptions mediate the debiasing effect of visual aids on decision making. In two phases, 60 senior police officers estimated the accuracy of a counterterrorism technique in identifying whether a known terror suspect poses an imminent danger and decided whether they would recommend the technique to policy makers. Officers also rated their confidence in this recommendation. When information about the effectiveness of the counterterrorism technique was presented in a numerical format, officers' perceptions of accuracy and recommendation decisions were susceptible to the framing effect: The technique was perceived to be more accurate and was more likely to be recommended when its effectiveness was presented in a positive than in a negative frame. However, when the information was represented visually using icon arrays, there were no such framing effects. Finally, perceptions of accuracy mediated the debiasing effect of visual aids on recommendation decisions. We offer potential explanations for the debiasing effect of visual aids and implications for communicating risk to experienced, professional decision makers.

[HTA-Perspective: Challenges in the early assessment of new oncological drugs]

Oncologic drug therapies have gained wide attention in the context of health policy priority setting for serious and socially significant diseases with high human and monetary costs. Due to uncertainties and scepticism about the actual therapeutic importance of newly approved oncology products, an early assessment programme was already established in Austria in 2007. The assessment of new oncology products is thereby faced with special challenges, since study populations are frequently not representative or the study design is laid out in such a manner that a definitive assessment of patient-relevant endpoints is not possible (cross-overs after interim assessments, surrogate parameters as primary endpoints, uncontrolled studies or those with unrealistic comparators, invalidated post-hoc identified biomarkers). On account of these major uncertainties, even the European Medicines Agency (EMA) is already contemplating multi-stage, "adaptive" approvals, and national reimbursement institutions are increasingly working with outcome-oriented, conditional reimbursement. (As supplied by publisher).

Impact of effectiveness information format on patient choice of therapy and satisfaction with decisions about chronic disease medication: the "Influence of intervention Methodologies on Patient Choice of Therapy (IMPACT)" cluster-randomised trial in general practice

Background

Risk communication is an integral part of shared decision-making in health care. In the context of interventions for chronic diseases it represents a particular challenge for all health practitioners. By using two different quantitative formats to communicate risk level and effectiveness of a cholesterol-lowering drug, we posed the research question: how does the format of risk information influence patients’ decisions concerning therapy, patients’ satisfaction with the communication as well as confidence in the decision. We hypothesise that patients are less prone to accept therapy when the benefits of long-term intervention are presented in terms of prolongation of life (POL) in months compared to the absolute risk reduction (ARR). We hypothesise that patients presented with POL will be more satisfied with the communication and confident in their decision, suggesting understanding of the time-related term.

Methods/Design

In 2009 a sample of 328 general practitioners (GPs) in the Region of Southern Denmark was invited to participate in a primary care-based clinical trial among patients making real-life clinical decisions together with their GP. Interested GPs were cluster-randomised to inform patients about cardiovascular disease (CVD) risk and the effectiveness of statin therapy using either POL or ARR. The GPs attended a training session before informing their patients. Before training and after the trial period they received a questionnaire about their attitudes to risk communication and the use of numerical information. Patients’ redemptions of statin prescriptions will be registered in a regional prescription database to evaluate a possible association between redemption rates and effectiveness format. The Combined Outcome Measure for Risk Communication And Treatment Decision Making Effectiveness (COMRADE) questionnaire will be used to measure patients’ confidence and satisfaction with the risk communication immediately after the conversation with their GPs.

Discussion

This randomised clinical trial compares the impact of two effectiveness formats on real-life risk communication between patients and GPs, including affective patient outcomes and actual choices about acceptance of therapy. Though we found difficulties in recruiting GPs, according to the study protocol we have succeeded in engaging sufficient GPs for the trial, enabling us to perform the planned analyses.

Trial registration

ClinicalTrials.gov Protocol Registration System http://ww.clinicaltrials.gov/NCT01414751

Heuristic decision making in medicine

Can less information be more helpful when it comes to making medical decisions? Contrary to the common intuition that more information is always better, the use of heuristics can help both physicians and patients to make sound decisions. Heuristics are simple decision strategies that ignore part of the available information, basing decisions on only a few relevant predictors. We discuss: (i) how doctors and patients use heuristics; and (ii) when heuristics outperform information-greedy methods, such as regressions in medical diagnosis. Furthermore, we outline those features of heuristics that make them useful in health care settings. These features include their surprising accuracy, transparency, and wide accessibility, as well as the low costs and little time required to employ them. We close by explaining one of the statistical reasons why heuristics are accurate, and by pointing to psychiatry as one area for future research on heuristics in health care.

Insights from a conference on implementing comparative effectiveness research through shared decision-making

For decades, investigators have conducted innovative research on shared decision-making (SDM), helping patients and clinicians to discuss health decisions and balance evidence with patients' preferences for possible outcomes of options. In addition, investigators have developed and used rigorous methods for conducting comparative effectiveness research (CER), comparing the benefits and risks of different interventions in real-world settings with outcomes that matter to patients and other stakeholders. However, incorporating CER findings into clinical practice presents numerous challenges. In March 2012, we organized a conference at Washington University in St Louis (MO, USA) aimed at developing a network of researchers to collaborate in developing, conducting and disseminating research about the implementation of CER through SDM. Meeting attendees discussed conceptual similarities and differences between CER and SDM, challenges in implementing CER and SDM in practice, specific challenges when engaging SDM with unique populations and examples of ways to overcome these challenges. CER and SDM are related processes that emphasize examining the best clinical evidence and how it applies to real patients in real practice settings. SDM can provide one opportunity for clinicians to discuss CER findings with patients and engage in a dialog about how to manage uncertainty about evidence in order to make decisions on an individual patient level. This meeting highlighted key challenges and suggested avenues to pursue such that CER and SDM can be implemented into routine clinical practice.

Communicating clinical trial outcomes: Effects of presentation method on physicians’ evaluations of new treatments

Physicians expect a treatment to be more effective when its clinical outcomes are described as relative rather than as absolute risk reductions. We examined whether effects of presentation method (relative vs. absolute risk reduction) remain when physicians are provided the baseline risk information, a vital piece of statistical information omitted in previous studies. Using a between-subjects design, ninety five physicians were presented the risk reduction associated with a fictitious treatment for hypertension either as an absolute risk reduction or as a relative risk reduction, with or without including baseline risk information. Physicians reported that the treatment would be more effective and that they would be more willing to prescribe it when its risk reduction was presented to them in relative rather than in absolute terms. The relative risk reduction was perceived as more effective than absolute risk reduction even when the baseline risk information was explicitly reported. We recommend that information about absolute risk reduction be made available to physicians in the reporting of clinical outcomes. Moreover, health professionals should be cognizant of the potential biasing effects of risk information presented in relative risk terms.

A proposal for an additional clinical trial outcome measure assessing preventive effect as delay of events

Many effect measures used in clinical trials are problematic because they are differentially understood by patients and physicians. The emergence of novel methods such as accelerated failure-time models and quantile regression has shifted the focus of effect measurement from probability measures to time-to-event measures. Such modeling techniques are rapidly evolving, but matching non-parametric descriptive measures are lacking. We propose such a measure, the delay of events, demonstrating treatment effect as a gain in event-free time. We believe this measure to be of value for shared clinical decision-making. The rationale behind the measure is given, and it is conceptually explained using the Kaplan–Meier estimate and the quantile regression framework. A formula for calculation of the delay of events is given. Hypothetical and empirical examples are used to demonstrate the measure. The measure is discussed in relation to other measures highlighting the time effects of preventive treatments. There is a need to further investigate the properties of the measure as well as its role in clinical decision-making.

Competence training in evidence-based medicine for patients, patient counsellors, consumer representatives and health care professionals in Austria: a feasibility study

BACKGROUND

Informed and shared decision-making require competences for both partners - healthcare professionals and patients. There is a lack of training courses in evidence-based medicine for patients and counsellors.

OBJECTIVE

We investigated feasibility, acceptability and the potential effects of a 2 x 2.5 days training course on critical health competences in patients, patient counsellors, consumer representatives and healthcare professionals in Austria.

METHODS

We adapted a previously developed curriculum for patient and consumer representatives. The adaptation comprised the specific needs of our target group in Austria and was founded on Carl Rogers' theory of person-centred education. For the formative evaluation a questionnaire was applied to address the domains: 1) organisational conditions (time and duration of the course, location, and information given in advance, registration); 2) assistance outside the courses; 3) teaching methods (performance of lecturers, teaching materials, structure of modules and blocks) and 4) satisfaction; 5) subjective assessment of competences. Participants evaluated the course, using a 5-point Likert scale. Long-term implementation was assessed using semi-structured interviews three to six months after the course. To estimate the increase in critical health competences we used the validated Critical Health Competence Test (CHC test).

RESULTS

Eleven training courses were conducted including 142 participants: patients (n=21); self-help group representatives (n=17); professional counsellors (n=29); healthcare professionals (n=10); psychologists (n=8); teachers (n=10) and others (n=29). 97 out of 142 (68 %) participants returned the questionnaire. On average, participants strongly agreed or agreed to 1) organisational conditions: 71 % / 23 %; 2) assistance outside the courses: 96 % / 10 %; 3) teaching methods: 60 % / 28 %; and 4) satisfaction: 78 % / 20 %, respectively. Interviews showed that the training course raised awareness, activated and empowered participants. Participants passed the CHC test with mean person parameters of 463±111 (pre-test, n=120) and 547±135 (post-test, n=91). For participants who returned both tests (n=71) person parameters were comparable: pre-test 466±121 versus post-test 574±100, p<0,001.

CONCLUSION

Training in evidence-based medicine for patients, patient counsellors, consumer representatives and healthcare professionals is feasible. For a broad implementation, train-the trainer courses and further research are needed.

Assessing the Effect of Visualizations on Bayesian Reasoning through Crowdsourcing

People have difficulty understanding statistical information and are unaware of their wrong judgments, particularly in Bayesian reasoning. Psychology studies suggest that the way Bayesian problems are represented can impact comprehension, but few visual designs have been evaluated and only populations with a specific background have been involved. In this study, a textual and six visual representations for three classic problems were compared using a diverse subject pool through crowdsourcing. Visualizations included area-proportional Euler diagrams, glyph representations, and hybrid diagrams combining both. Our study failed to replicate previous findings in that subjects' accuracy was remarkably lower and visualizations exhibited no measurable benefit. A second experiment confirmed that simply adding a visualization to a textual Bayesian problem is of little help, even when the text refers to the visualization, but suggests that visualizations are more effective when the text is given without numerical values. We discuss our findings and the need for more such experiments to be carried out on heterogeneous populations of non-experts.

It's for your benefit: exploring patients' opinions about the inclusion of textual and numerical benefit information in medicine leaflets

Objectives

To explore consumers' beliefs and preferences for benefit information in medicine leaflets and examine their understanding and reaction to treatment benefits.

Methods

Forty-five participants were recruited to eight focus groups, run concurrently in Australia (23 participants in four groups) and the UK (22 participants in four groups). Participants were provided with amended leaflets based on the medicine clopidogrel, containing textual and numerical benefit information presented using numbers needed to treat (NNT). A topic guide which explored use of leaflets, preferences and opinions was used to direct discussion. Focus group discussions were recorded, transcribed verbatim and content analysed using adapted cross-case study analysis.

Key findings

The consensus was that the inclusion of benefit information was a positive factor. Many participants felt that textual benefit information offered an incentive to take a medicine, although some Australian participants had concerns that included benefit information could create anxiety. The presentation of numerical benefit information provoked strong feelings of disbelief and shock. Participants were surprised that so few people would benefit. Some participants struggled to understand and interpret the NNT and others found it difficult to comprehend the magnitude of the benefit information, instead operating on initial and often crude assumptions of what the data meant. In both countries the provision of numerical benefit information appeared to shake participants' faith in drug treatments. Participants were concerned about how this might affect the ‘less-informed’ patient. However, in the UK, participants stated that their adherence to treatment was also reinforced by their doctor's advice.

Conclusions

Participants wanted to receive information about the benefits of their medicines. However, they may misinterpret the numerical information provided.

Communicating and Distorting Risks with Graphs: An Eye-Tracking Study

Graphs can represent complex information in accessible ways. Unfortunately, many graphs are poorly designed and lead to errors in judgment and decision-making. Here, we examine the influence of distorted graphs used by advertisers and major news organizations to communicate risks. Results indicated that the distorted graphs were associated with large judgment errors and that cognitive abilities (e.g., numeracy, graph literacy, cognitive reflection) predicted differences in error rates. Eye-tracking results revealed a strong link between elaborative information search and stimuli-memory, which mediated the ability-judgment relationship. Discussion focuses on cognitive mechanisms (e.g., elaborative encoding), implications for HFES graph design guidelines, and emerging opportunities for personalized decision support.

Medication effectiveness may not be the major reason for accepting cardiovascular preventive medication: a population-based survey

BACKGROUND

Shared decision-making and patients' choice of interventions are areas of increasing importance, not least seen in the light of the fact that chronic conditions are increasing, interventions considered important for public health, and still non-acceptance of especially risk-reducing treatments of cardiovascular diseases (CVD) is prevalent. A better understanding of patients' medication-taking behavior is needed and may be reached by studying the reasons why people accept or decline medication recommendations. The aim of this paper was to identify factors that may influence people's decisions and reasoning for accepting or declining a cardiovascular preventive medication offer.

METHODS

From a random sample of 4,000 people aged 40-59 years in a Danish population, 1,169 participants were asked to imagine being at increased risk of cardiovascular disease and being offered a preventive medication. After receiving 'complete' information about effectiveness of the medication they were asked whether they would accept medication. Finally, they were asked about reasons for the decision.

RESULTS

A total of 725 (67%) of 1,082 participants accepted the medication offer. Even quite large effects of medication (up to 8 percentage points absolute risk reduction) had a smaller impact on acceptance to medication than personal experience with cardiovascular disease. Furthermore, increasing age of the participant and living with a partner were significantly associated with acceptance. Some 45% of the respondents accepting justified their choice as being for health reasons, and they were more likely to be women, live alone, have higher income and higher education levels. Among those who did not accept the medication offer, 56% indicated that they would rather prefer to change lifestyle.

CONCLUSIONS

Medication effectiveness seems to have a moderate influence on people's decisions to accept preventive medication, while factors such as personal experience with cardiovascular disease may have an equally strong or stronger influence, indicating that practitioners could do well to carefully identify the reasons for their patients' treatment decisions.

[Rationing health care by thresholds for clinical benefit and its acceptance by the German population]

Fair rationing in publicly accessible health care has become a subject of current international debate. One suggestion is to cut reimbursement for any medical intervention below some threshold of small clinical benefit. One can further differentiate between thresholds of small expectable clinical benefit as such and thresholds of low chances for clinical success. Public acceptance of both types of thresholds has been tested in a population survey. Results are presented and discussed in this paper. (As supplied by publisher).

Semantically enabling pharmacogenomic data for the realization of personalized medicine

Understanding how each individual's genetics and physiology influences pharmaceutical response is crucial to the realization of personalized medicine and the discovery and validation of pharmacogenomic biomarkers is key to its success. However, integration of genotype and phenotype knowledge in medical information systems remains a critical challenge. The inability to easily and accurately integrate the results of biomolecular studies with patients' medical records and clinical reports prevents us from realizing the full potential of pharmacogenomic knowledge for both drug development and clinical practice. Herein, we describe approaches using Semantic Web technologies, in which pharmacogenomic knowledge relevant to drug development and medical decision support is represented in such a way that it can be efficiently accessed both by software and human experts. We suggest that this approach increases the utility of data, and that such computational technologies will become an essential part of personalized medicine, alongside diagnostics and pharmaceutical products.

What do web-use skill differences imply for online health information searches?

BACKGROUND

Online health information is of variable and often low scientific quality. In particular, elderly less-educated populations are said to struggle in accessing quality online information (digital divide). Little is known about (1) how their online behavior differs from that of younger, more-educated, and more-frequent Web users, and (2) how the older population may be supported in accessing good-quality online health information.

OBJECTIVE

To specify the digital divide between skilled and less-skilled Web users, we assessed qualitative differences in technical skills, cognitive strategies, and attitudes toward online health information. Based on these findings, we identified educational and technological interventions to help Web users find and access good-quality online health information.

METHODS

We asked 22 native German-speaking adults to search for health information online. The skilled cohort consisted of 10 participants who were younger than 30 years of age, had a higher level of education, and were more experienced using the Web than 12 participants in the less-skilled cohort, who were at least 50 years of age. We observed online health information searches to specify differences in technical skills and analyzed concurrent verbal protocols to identify health information seekers' cognitive strategies and attitudes.

RESULTS

Our main findings relate to (1) attitudes: health information seekers in both cohorts doubted the quality of information retrieved online; among poorly skilled seekers, this was mainly because they doubted their skills to navigate vast amounts of information; once a website was accessed, quality concerns disappeared in both cohorts, (2) technical skills: skilled Web users effectively filtered information according to search intentions and data sources; less-skilled users were easily distracted by unrelated information, and (3) cognitive strategies: skilled Web users searched to inform themselves; less-skilled users searched to confirm their health-related opinions such as "vaccinations are harmful." Independent of Web-use skills, most participants stopped a search once they had found the first piece of evidence satisfying search intentions, rather than according to quality criteria.

CONCLUSIONS

Findings related to Web-use skills differences suggest two classes of interventions to facilitate access to good-quality online health information. Challenges related to findings (1) and (2) should be remedied by improving people's basic Web-use skills. In particular, Web users should be taught how to avoid information overload by generating specific search terms and to avoid low-quality information by requesting results from trusted websites only. Problems related to finding (3) may be remedied by visually labeling search engine results according to quality criteria.

Using visual displays to communicate risk of cancer to women from diverse race/ethnic backgrounds

OBJECTIVE

This study evaluated how well women from diverse race/ethnic groups were able to take a quantitative cancer risk statistic verbally provided to them and report it in a visual format.

METHODS

Cross-sectional survey was administered in English, Spanish or Chinese, to women aged 50-80 (n=1160), recruited from primary care practices. The survey contained breast, colorectal or cervical cancer questions regarding screening and prevention. Women were told cancer-specific lifetime risk then shown a visual display of risk and asked to indicate the specific lifetime risk. Correct indication of risk was the main outcome.

RESULTS

Correct responses on icon arrays were 46% for breast, 55% for colon, and 44% for cervical; only 25% correctly responded to a magnifying glass graphic. Compared to Whites, African American and Latina women were significantly less likely to use the icon arrays correctly. Higher education and higher numeracy were associated with correct responses. Lower education was associated with lower numeracy.

CONCLUSIONS

Race/ethnic differences were associated with women's ability to take a quantitative cancer risk statistic verbally provided to them and report it in a visual format.

PRACTICE IMPLICATIONS

Systematically considering the complexity of intersecting factors such as race/ethnicity, educational level, poverty, and numeracy in most health communications is needed.

On defensive decision making: how doctors make decisions for their patients

BACKGROUND

In the past decade, the number of lawsuits for medical malpractice has risen significantly. This could affect the way doctors make decisions for their patients.

OBJECTIVE

To investigate whether and why doctors practice defensive medicine with their patients.

DESIGN

A questionnaire study was conducted in general practice departments of eight metropolitan hospitals in Spain, between January and February 2010.

SETTING AND PARTICIPANTS

Eighty general practitioners (48% men; mean age 52 years) with an average of 15.3 years of experience and their 80 adult patients (42% men; mean age 56 years) participated in the study.

MAIN OUTCOME MEASUREMENTS

Participants completed a self-administered questionnaire involving choices between a risky and a conservative treatment. One group of doctors made decisions for their patients. Another group of doctors predicted what their patients would decide for themselves. Finally, all doctors and patients made decisions for themselves and described the factors they thought influenced their decisions.

RESULTS

Doctors selected much more conservative medical treatments for their patients than for themselves. Most notably, they did so even when they accurately predicted that the patients would select riskier treatments. When asked about the reasons for their decisions, most doctors (93%) reported fear of legal consequences.

DISCUSSION AND CONCLUSIONS

Doctors' decisions for their patients are strongly influenced by concerns of possible legal consequences. Patients therefore cannot blindly follow their doctor's advice. Our study, however, suggests a plausible method that patients could use to get around this problem: They could simply ask their doctor what he or she would do in the patient's situation.

Using visual aids to improve communication of risks about health: a review

Recent research has shown that patients frequently experience difficulties understanding health-relevant numerical concepts. A prominent example is denominator neglect, or the tendency to pay too much attention to numerators in ratios (e.g., number of treated patients who died) with insufficient attention to denominators (e.g., overall number of treated patients). Denominator neglect can lead to inaccurate assessments of treatment risk reduction and thus can have important consequences for decisions about health. Here, we reviewed a series of studies investigating (1) different factors that can influence patients' susceptibility to denominator neglect in medical decision making—including numerical or language-related abilities; (2) the extent to which denominator neglect can be attenuated by using visual aids; and (3) a factor that moderates the effectiveness of such aids (i.e., graph literacy). The review spans probabilistic national U.S. and German samples, as well as immigrant (i.e., Polish people living in the United Kingdom) and undergraduate samples in Spain. Theoretical and prescriptive implications are discussed.

[Do media reports and public brochures facilitate informed decision making about cervical cancer prevention?]

With the introduction and recommendation of the new HPV (human papillomavirus) vaccination in 2007, cervical cancer prevention has evoked large public interest. Is the public able to make informed decisions on the basis of media reports and brochures? To answer this question, an analysis of media coverage of HPV vaccination (Gardasil®) and Pap (Papanicolaou) screening was conducted from 2007-2009, which investigated the minimum requirement of completeness (pros and cons), transparency (absolute numbers), and correctness (references concerning outcome, uncertainty, magnitude) of the information. As a bench mark, facts boxes with concise data on epidemiology, etiology, benefits, harms, and costs were compiled in advance. Although all vaccination reports and brochures covered the impact of prevention, only 41% provided concrete numbers on effectiveness (90/220) and 2% on absolute risk reductions for the cancer surrogate dysplasia (5/220), whereby none of the latter numbers was correct. The prevention potential was correctly presented once. Only 48% (105/220) mentioned pros and cons. With regard to screening, 20% (4/20) provided explicit data on test quality and one expressed these in absolute numbers, while 25% (5/20) reported the prevention potential; all given numbers were correct. Finally, 25% (5/20) mentioned the possibility of false positive results. Minimum requirements were fulfilled by 1/220 vaccination and 1/20 screening reports. At present, informed decision making based on media coverage is hardly possible.

A competing risk approach for the European Heart SCORE model based on cause-specific and all-cause mortality

BACKGROUND

The European Heart SCORE model constitutes the basis for national guidelines for primary prevention and treatment of cardiovascular disease (CVD) in several European countries. The model estimates individuals' 10-year CVD mortality risks from age, sex, smoking status, systolic blood pressure, and total cholesterol level. The SCORE model, however, is not mathematically consistent and does not estimate all-cause mortality. Our aim is to modify the SCORE model to allow consistent estimation of both CVD-specific and all-cause mortality.

METHODS

Using a competing risk approach, we first re-estimated the cause-specific risk of dying from cardiovascular disease, and secondly we incorporated non-CVD mortality. Finally, non-CVD mortality was allowed to also depend on smoking status, and not only age and sex. From the models, we estimated CVD-specific and all-cause 10-year mortality risk, and the expected residual lifetime together with corresponding expected effects of statin treatment.

RESULTS

The modified model provided CVD-specific 10-year mortality risks similar to those of the European Heart SCORE model. Incorporation of non-CVD mortality increased 10-year mortality risks, in particular for older individuals. When non-CVD mortality was assumed unaffected by smoking status, the absolute risk reduction due to statin treatment ranged from 0.0% to 3.5%, whereas the gain in expected residual lifetime ranged from 3 to 11 months. Statin effectiveness increased for non-smokers and declined for smokers, when smoking was allowed to influence non-CVD mortality.

CONCLUSION

The modified model provides mathematically consistent estimates of mortality risk and expected residual lifetime together with expected benefits from statin treatment.

Data display format and hospital ward reports: effects of different presentations on data interpretation

BACKGROUND

Graphs are often used in medical communication, both in clinical practice and health management. They can help the processing of quantitative information but may also contribute to drawing wrong conclusions. The aim of the survey is to study the graphical perception of the data at the management level and its possible effects, showing how some criteria of appraisal of a phenomenon are influenced by the graphical format.

METHODS

One hundred and five medical doctors and health direction professionals of hospitals in Naples, Rome, Siena and Turin were interviewed. Four different graphs or table related to the same hypothetical data on average hospital stay in the period January 2000 to September 2003 were shown to participants, and their impressions were recorded.

RESULTS

Less than one-fourth of the participants understood that the data set was the same for the different diagrams. The process of understanding is mostly correlated with being a director, having a degree in medicine and working in central-northern cities. The table seems easier for interpretation (98.1%), more suitable (84.8%), more used (92.4%) and more pleasant than other data presentation. On the other hand radar format had worse results in all questions.

CONCLUSIONS

The choice of a graphical format may influence the understanding of data. Further research is needed in order to sustain the improvement of medical and health professionals' knowledge in the display data format.

Presenting evidence-based health information for people with multiple sclerosis: the IN-DEEP project protocol

Background

Increasingly, evidence-based health information, in particular evidence from systematic reviews, is being made available to lay audiences, in addition to health professionals. Research efforts have focused on different formats for the lay presentation of health information. However, there is a paucity of data on how patients integrate evidence-based health information with other factors such as their preferences for information and experiences with information-seeking. The aim of this project is to explore how people with multiple sclerosis (MS) integrate health information with their needs, experiences, preferences and values and how these factors can be incorporated into an online resource of evidence-based health information provision for people with MS and their families.

Methods

This project is an Australian-Italian collaboration between researchers, MS societies and people with MS. Using a four-stage mixed methods design, a model will be developed for presenting evidence-based health information on the Internet for people with MS and their families. This evidence-based health information will draw upon systematic reviews of MS interventions from The Cochrane Library. Each stage of the project will build on the last. After conducting focus groups with people with MS and their family members (Stage 1), we will develop a model for summarising and presenting Cochrane MS reviews that is integrated with supporting information to aid understanding and decision making. This will be reviewed and finalised with people with MS, family members, health professionals and MS Society staff (Stage 2), before being uploaded to the Internet and evaluated (Stages 3 and 4).

Discussion

This project aims to produce accessible and meaningful evidence-based health information about MS for use in the varied decision making and management situations people encounter in everyday life. It is expected that the findings will be relevant to broader efforts to provide evidence-based health information for patients and the general public. The international collaboration also permits exploration of cultural differences that could inform international practice.

Do the media provide transparent health information? A cross-cultural comparison of public information about the HPV vaccine

The media is a powerful tool for informing the public about health treatments. In particular, the Internet has gained importance as a widely valued source for health information for parents and adolescents. Nonetheless, traditional sources, such as newspapers, continue to report on health innovations. But do websites and newspaper reports provide balanced information? We performed a systematic media analysis to evaluate and compare media coverage of the human papillomavirus (HPV) vaccine on websites and in newspapers in Germany and Spain. We assessed to what extent the media provide complete (pros and cons), transparent (absolute instead of relative numbers), and correct information about the epidemiology and etiology of cervical cancer as well as the effectiveness and costs of the HPV vaccine. As a basis for comparison, a facts box containing current scientific evidence about cervical cancer and the HPV vaccine was developed. The media analysis included 61 websites and 141 newspaper articles in Germany, and 41 websites and 293 newspaper articles in Spain. Results show that 57% of German websites and 43% of German newspaper reports communicated correct estimates of epidemiological data, whereas in Spain 39% of the websites and 20% of the newspaper did so. While two thirds of Spanish websites explicitly mentioned causes of cervical cancer as well as spontaneous recovery, German websites communicated etiological information less frequently. Findings reveal that correct estimates about the vaccine's effectiveness were mentioned in 10% of German websites and 6% of German newspaper reports; none of the Spanish newspaper reports and 2% of Spanish websites reported effectiveness correctly. Only German websites (13%) explicitly referred to scientific uncertainty regarding the vaccine's evaluation. We conclude that the media lack balanced reporting on the dimensions completeness, transparency, and correctness. We propose standards for more balanced reporting on websites and in newspapers.

Revisiting gender differences in somatic symptoms of depression: much ado about nothing?

BACKGROUND

Women have a higher prevalence of Major Depressive Disorder (MDD) and report more severe depressive symptoms than men. Several studies have suggested that gender differences in depression may occur because women report higher levels of somatic symptoms than men. Those studies, however, have not controlled or matched for non-somatic symptoms. The objective of this study was to examine if women report relatively more somatic symptoms than men matched on cognitive/affective symptoms.

METHODS

Male and female patients receiving treatment for MDD in outpatient psychiatric clinics in New Jersey and Pennsylvania, USA were matched on Beck Depression Inventory-II (BDI-II) cognitive/affective symptom scores. Male and female BDI-II somatic symptom scores were compared using independent samples 2-tailed t-tests.

RESULTS

Of 472 male and 1,026 female patients, there were 470 male patients (mean age = 40.1 years, SD = 15.1) and 470 female patients (mean age = 43.1 years, SD = 17.2) successfully matched on BDI-II cognitive/affective symptom scores. Somatic symptoms accounted for 35% of total BDI-II scores for male patients versus 38% for matched female patients. Female patients had somatic symptom scores on average 1.3 points higher than males (p<.001), equivalent to 4% of the total BDI-II scores of female patients. Only 5% of male patients and 7% of female patients scored 2 or higher on all BDI-II somatic symptom items.

CONCLUSIONS

Gender differences in somatic scores were very small. Thus, differences in the experience and reporting of somatic symptoms would not likely explain gender differences in depression rates and symptom severity.

Statistical literacy in obstetricians and gynecologists

The Obstetrician-Gynecologist Statistical Literacy Questionnaire (OGSLQ) was designed to examine physicians' understanding of various number tasks that are relevant to obstetrician-gynecologists (ob-gyns) practice. Forty-seven percent of the nationally representative, practicing ob-gyns responded. Physicians did poorly on the questions about numerical facts (e.g., number of women living with HIV/AIDS), better on questions about statistical concepts (e.g., incidence, prevalence), and best on questions about numerical relationships (e.g., convert frequency to percentage) with 0%, 7%, 36%, answering all correctly, respectively. Only 19% correctly estimated the number of U.S. women with cancer. Sixty-six percent were able to use sensitivity and specificity to choose a test option. Around 90% could translate between frequency and probability formats. Forty-nine percent of respondents were able to calculate the positive predictive value of a mammography screening test. Physicians lack some understanding of statistical literacy. It is important that we monitor physicians' statistical literacy and provide training to students and physicians.