Patient Education and Counseling 94 (2014) 43–49
Contents lists available at ScienceDirect
Patient Education and Counseling journal homepage: www.elsevier.com/locate/pateducou
Medical Education
Development and evaluation of a risk communication curriculum for medical students Paul K.J. Han a,b,*, Katherine Joekes c, Glyn Elwyn d,e, Kathleen M. Mazor f, Richard Thomson g, Philip Sedgwick c, Judith Ibison c, John B. Wong b a
Center for Outcomes Research and Evaluation, Maine Medical Center, Portland, USA Tufts University School of Medicine, Boston, USA Centre for Medical and Healthcare Education, St George’s, University of London, London, UK d Dartmouth Center for Health Care Delivery Science, Hanover, USA e Cochrane Institute for Primary Care & Public Health, Cardiff University School of Medicine, Cardiff, UK f Meyers Primary Care Institute, University of Massachusetts Medical School, Worcester, USA g Institute of Health and Society, Newcastle University, Newcastle upon Tyne, UK b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 July 2013 Received in revised form 4 September 2013 Accepted 7 September 2013
Objective: To develop, pilot, and evaluate a curriculum for teaching clinical risk communication skills to medical students. Methods: A new experience-based curriculum, ‘‘Risk Talk,’’ was developed and piloted over a 1-year period among students at Tufts University School of Medicine. An experimental study of 2nd-year students exposed vs. unexposed to the curriculum was conducted to evaluate the curriculum’s efficacy. Primary outcome measures were students’ objective (observed) and subjective (self-reported) risk communication competence; the latter was assessed using an Observed Structured Clinical Examination (OSCE) employing new measures. Results: Twenty-eight 2nd-year students completed the curriculum, and exhibited significantly greater (p < .001) objective and subjective risk communication competence than a convenience sample of 24 unexposed students. New observational measures of objective competence in risk communication showed promising evidence of reliability and validity. The curriculum was resource-intensive. Conclusion: The new experience-based clinical risk communication curriculum was efficacious, although resource-intensive. More work is needed to develop the feasibility of curriculum delivery, and to improve the measurement of competence in clinical risk communication. Practice implications: Risk communication is an important advanced communication skill, and the Risk Talk curriculum provides a model educational intervention and new assessment tools to guide future efforts to teach and evaluate this skill. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Risk communication Medical education Communication skills
1. Introduction Risk information—information about the probability of future outcomes—is a cornerstone of evidence-based medicine (EBM), and the communication of risk information to patients is an essential prerequisite for shared decision making (SDM) [1,2]. Yet risk information, whether pertaining to the probability of benefits or harms, is extremely challenging to understand and communicate. Past research has documented not only substantial deficits in numeracy—the ability to understand and use numerical information—among both lay persons and physicians [3–6], but numerous
* Corresponding author at: Center for Outcomes Research and Evaluation, Maine Medical Center, 509 Forest Avenue, Portland, ME 04101, USA. Tel.: +1 207 661 7619; fax: +1 207 662 3110. E-mail address: [email protected] (Paul K.J. Han). 0738-3991/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.pec.2013.09.009
cognitive biases that limit people’s ability to process and interpret risk information and to effectively use this information in decision making [7–9]. Accordingly, a growing body of research has been devoted to developing more effective methods for communicating risk information. To date, this work has focused primarily on the textual or visual representation of risk information—e.g., using alternative summary statistics or graphical representations—and the application of these methods in decision support tools. Evidence-based best practices for representing risk information have begun to be identified [9–11], and the International Patient Decision Aids Standards (IPDAS) Collaboration has recently summarized key recommendations for the representation of risk information in decision support tools [12]. Relatively little attention, however, has been devoted to the verbal communication of risk information to patients in face-toface encounters and real-life clinical settings, and little is known or
44
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
published about how to improve physicians’ competence in this endeavor. In the UK, Elwyn, Edwards, and colleagues developed a brief workshop-based training program in shared decision making (SDM) and risk communication, which utilized visual aids (tabulated data, histograms, bar charts) designed to facilitate risk communication [13]. A randomized trial of this intervention among general practitioners in the UK demonstrated significant improvements in patient involvement in, and satisfaction with, decision making, and in clinicians’ satisfaction and use of multiple representational formats in risk communication [13]. Elsewhere in the UK, educators at St. Georges, University of London (SGUL) in 2001 developed a risk communication curriculum for undergraduate medical students, consisting of a mix of didactic and experiential workshop-based training undertaken during the pre-clinical years of medical school, and skills assessment using Observed Structured Clinical Examination (OSCE) [14,15]. Other UK medical schools devote varying amounts of curricular time to this task (KJ—personal communication); however, their experiences have not been published. The overall objective of the current study, a collaborative effort between investigators at Tufts University School of Medicine (TUSM) in Boston, MA (USA) and SGUL, was to develop, pilot, and evaluate a new curriculum, named ‘‘Risk Talk,’’ to teach risk communication skills to undergraduate medical students. Its specific objectives were to obtain proof-of-concept evidence on the feasibility and efficacy of this new curriculum, and to develop new measures to assess students’ competence in clinical risk communication. 2. Methods 2.1. Study population and setting The study population consisted of 2nd-year undergraduate medical students at TUSM. The curriculum was piloted in a cohort of ‘‘Maine Track’’ (MT) students (approximately 36 of 200 total TUSM students in each class) who choose to receive some preclinical (Years 1–2) and most clinical training (Years 3–4) at the Maine Medical Center rather than Boston campus of TUSM. MT and Boston-based students have similar academic characteristics, including Medical College Admissions Test (MCAT) scores (32.03 vs. 32.15, p = .85, for 2nd-year MT and Boston students, respectively) and college grade point average (GPA) (3.54 vs. 3.61, p = .13). 2.2. Conceptual framework Development of the Risk Talk curriculum was guided by a conceptual model of key communication processes in SDM, developed by Elwyn et al. (Fig. 1) [16]. This model translates the
Fig. 1. Conceptual model of shared decision making and risk communication. Adapted from Elwyn et al. [16].
principles of SDM into discrete, fundamental, goal-oriented communication tasks: choice talk, option talk, and decision talk. Choice talk aims to make sure patients know that reasonable options are available. Option talk aims to provide detailed information about options. Decision talk aims to support values clarification and patient preferences when deciding what is best. The broad goal of these discrete communication tasks is to facilitate deliberation—the consideration of information about pros and cons of alternative options, the range of possible futures, and clarification about individual preferences for those possible futures [17]. Addressing a vital component of option talk, Risk Talk was designed to communicate information on the benefits, harms, and uncertainties of available choice options, thus bridging the discussion of options and decisions. 2.3. Curriculum structure Risk Talk drew on the overall structure and approach at SGUL of classroom-based teaching during the pre-clinical years of medical school; Appendix A provides an overview. The focus of the current project was an experiential teaching workshop provided to secondyear students. Held in the first half of the year, this 3-h workshop consisted of a 45-min didactic lecture covering theoretical principles and practical approaches to risk communication, followed by a 2-h experiential learning session in which small groups of three students practiced communicating risk information with trained professional standardized patients (SPs), facilitated by a faculty preceptor. Students took turns leading, observing, and providing feedback on the discussions. To evaluate the efficacy of the 2nd-year curriculum, we developed an OSCE with a single clinical case designed to assess risk communication competence (described below). Additional learning experiences for first- and third-year students were developed and piloted but not formally evaluated. First-year students received a 1-h didactic lecture designed to introduce theoretical principles in risk communication. A small group of third-year students participated in a 45-min experiential learning session, involving role-play in dyads, offered during Internal Medicine clerkship rotations at Maine Medical Center. 2.4. Risk Talk workshop The Risk Talk workshop focused on two main content areas. The first was theoretical knowledge regarding both the role of risk communication in SDM, and the challenges involved in delivering it. The second content area was practical skills, the conceptual domains of which were derived from the existing SGUL curriculum and augmented through further literature review to identify best practices in risk communication [9–11].
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
The practical skills content of the curriculum was further augmented by the use of verbal ‘‘scripts’’—sample words and phrases designed to give students exemplary language to facilitate learning and practice of verbal risk communication skills. This approach was modeled on clinical skills training for SDM developed within the Health Foundation supported Making Good Decisions in Collaboration (MAGIC) program, developed by Elwyn et al. in the UK to teach SDM skills to physicians (http:// www.health.org.uk/areas-of-work/programmes/shared-decision-making/) [16]. MAGIC scripts were designed to model the key SDM content elements of choice talk, option talk, and decision talk (Fig. 1) [16]. In similar fashion, we created Risk Talk scripts to address key content and process elements of risk communication; examples are in Fig. 2. The scripts addressed essential functions of effective risk communication identified in the literature, including improving conceptual understanding of risk and uncertainty [18,19], increasing the evaluability of risk information [20], and reducing cognitive biases in the interpretation of risk information [21]. Improving conceptual understanding involves strategies to help patients understand the fundamental nature of risk estimates—their origin and derivation as well as their inherent uncertainty—and includes educating patients about the time frame and reference class of risk estimates. Improving evaluability involves strategies to make risk information easier to process and understand, including using natural frequencies with consistent denominators [3,22]. 1.
Reducing cognitive biases involves strategies to overcome well-known psychological biases in the interpretation of risk information, including the use of multiple frames (e.g., both positive and negative, absolute and relative risk reduction). Scripts aimed at operationalizing all of these functions served as the basis for the experiential learning workshops and were provided to all students. 2.5. Evaluation and measures Consisting of a non-randomized experimental study comparing 2nd year MT students exposed to the curriculum, and a convenience sample of unexposed Boston students who served as the control group, the evaluation focused on the efficacy of the 2nd-year workshop. All students had otherwise identical exposure to training in evidence-based medicine, clinical reasoning, and clinical communication as part of the TUSM curriculum to which all students were exposed on the Boston campus during their 1st year. An incentive ($15 gift card) was provided to all participating students. Students’ risk communication competence was evaluated using both objective (observational) and subjective (self-report) measures; the objective measures were completed by both standardized patients (SPs) and faculty observers. The study protocol was reviewed and approved by the Institutional Review Board (IRB) of both Tufts University School of Medicine and Maine Medical Center.
Assess preferences for information “Would you like me to go into more detail about the pros and cons of the treatment choices, including numbers and statistics?”
2.
Explain the magnitude of risks, using multiple frames
•
Both percentages and natural frequencies “Based on who you are, medical experts estimate that your risk of developing colon cancer in your lifetime is 10%.” “This is the same as saying that 10 out of 100 people like you will develop colon cancer during their lifetimes.”
•
Both negative and positive frames “Medical experts estimate that about 10 out of 100 people like you will develop colon cancer during their lifetimes.” “You can also look at this in another way: 90 out of 100 people like you [with your characteristics] will stay free of colon cancer.”
•
Emphasize absolute risk reduction “Studies show that taking DRUG X lowers people’s risk of a heart attack by 25%, or about 1/4. But to see if it is worth taking DRUG X, we need to count the number of people who are actually helped by DRUG X...” “The studies showed that DRUG X lowered the total number of people having heart attacks from 4 out of 100 to 3 out of 100.” So DRUG X helped about 1 out of every 100 people who took it.”
3.
Explain meaning of the reference class “One way to think about this is to imagine a group of [100] people like you, who have the following things in common: [sex, age, family history of cancer …etc]. Of this group of [100] similar people, 10 will develop colon cancer in their lifetime.”
4.
Acknowledge uncertainty
•
Meaning of chance / inability to predict single events “We can’t predict the future of any one person. Estimates of risk, of the chances, only tell us how many people in some group are likely to get cancer. They can’t tell us who will get the disease or not.” “Even if the risk of a complication from surgery is 10 out of 100, we don’t know whether you will be one of the unlucky 10 who will suffer the complication, or the lucky 90 who will not.”
•
45
General scientific uncertainty “Estimates of risk, of the chance of something happening, are not perfect because science is not perfect. We don’t know everything we need to know to predict the future.” Fig. 2. Risk talk scripts.
46
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
Objective risk communication competence was the primary study outcome, measured through an OSCE conducted 3 weeks after the 2nd-year workshop, and in which both exposed and unexposed students participated. The OSCE utilized three trained professional SPs who enacted a case resembling those used in the workshops; the same SPs evaluated all students (in both the exposed and unexposed cohorts) (Appendix B). The OSCEs were conducted over 4 days, and scheduling for individual exposed and unexposed students was random and driven entirely by students’ own availability and preferences. Students’ objective (observed) risk communication competence was then assessed in two ways. First, SPs themselves rated the students immediately after each encounter, using a brief risk communication competence measure developed by adapting items from assessment tools used at SGUL (Table 1). This measure consisted of a 5-item SP Risk Communication Process (SP-RCP) measure assessing generic patient-centered communication skills (i.e., not specific to the domain of risk communication per se), and a 5-item SP Risk Communication Content (SP-RCC) assessing key content elements. SPs received 4 h of training on simulating patient encounters and applying the measures. The second way in which objective risk communication competence was assessed was through independent and asynchronous ratings by two faculty (based at SGUL and thus unacquainted with students)—a study co-investigator (KJ) and an experienced medical educator without specific expertise in risk communication—who each conducted blinded evaluations of video recordings of all OSCEs. For these evaluations, the faculty raters used versions of the SP rating instruments that were revised and expanded to improve their precision and comprehensiveness (Table 2). The expanded Risk Communication Process (RCP) measure included 13 items, while the expanded Risk Communication Content (RCC) measure included nine items. Development of the expanded RCC was informed by input from 18 international risk communication experts (Acknowledgments) recruited from the membership of the IPDAS Background Update group on Presenting Probabilities (http://ipdas.ohri.ca/who.html), and additional suggestions by IPDAS members and the study team. These experts reviewed an initial pool of 13 candidate items and completed a survey in which they were asked to rank, on a 1–5
Table 1 Standardized patient (SP) risk communication competence measure. Risk communication processa 1. Set the stage for the discussion 2. Assessed preferences for information 3. Checked understanding, elicited questions 4. Gave information clearly using plain language, avoided jargon 5. Demonstrated responsiveness, empathy, respect, professionalism Showed respect, care, and concern by words, tone, pace, eye contact, or posture Listened carefully Risk communication contentb 1. Explained risk using multiple frames (e.g., disease vs. health, mortality vs. survival) 2. Discussed both absolute and relative risk reduction 3. Explained the reference class, applicability of risk information to this patient 4. Acknowledged uncertainty in risk estimates Meaning of chance/inability to predict single events Imprecision in risk estimates General scientific uncertainty Meaning of the reference class, applicability of risk information to this patient 5. Distilled the ‘‘gist’’ or ‘‘bottom line’’ of the risk information a Rated on 3-point Likert scale: ‘‘Does not do/does poorly,’’ ‘‘Does fairly well,’’ ‘‘Does well’’. b Rated on 4-point Likert scale: ‘‘Poor (no attempt),’’ Fair (not clear),’’ ‘‘Good (fairly clear),’’ ‘‘Excellent (very clear)’’.
Table 2 Observer risk communication process and content measures. Risk communication process (RCP)a 1. Greeted patient appropriately and introduced self 2. Set the stage for discussion 3. Assessed preferences for information 4. Checked understanding, elicited questions 5. Gave information clearly using plain language, avoided jargon 6. Demonstrated responsiveness, empathy, respect, professionalism 7. Elicited patient concern and responded appropriately 8. Gave a well-paced explanation 9. Maintained a dialog 10. Demonstrated empathy 11. Actively listened 12. Built a therapeutic relationship throughout the consultation 13. Closed consultation appropriately Risk communication content (RCC)b 1. Discussed the quality/strength/weakness (e.g., validity, reliability, credibility) of the risk evidence 2. Specified the reference class (patient population) for whom the risk estimates apply 3. Specified the time period over which the risk estimates apply 4. Explained the magnitude of risk using both negative and positive frames 5. Explained risk estimates using both proportions (e.g., ‘‘9 out of 100’’) and percentages (e.g., 9%) 6. Discussed differences between baseline risk and modified risk in absolute terms (absolute risk reduction) or both absolute and relative terms 7. Acknowledged general uncertainty in all risk estimates, using qualitative terms 8. Acknowledged uncertainty due to chance or randomness (inability to predict single events) 9. Placed the magnitude of risks in context by comparing to risks of other outcomes (e.g., other diseases, treatments, familiar events) a Rated on 3-point Likert scale anchored by ‘‘Poor/no attempt,’’ and ‘‘Excellent/ Very clear’’. b Rated on 3-point Likert scale anchored by ‘‘The behavior is not observed, or erroneous information is given,’’ and ‘‘Excellent/Very clear’’.
scale, the five elements most important to include in a brief physician–patient discussion of risk information. Four unranked items were deleted or integrated with other items to produce a final 9-item RCC measure (Table 2). The two SGUL faculty raters each evaluated all OSCEs for students from both cohorts using the expanded RCP and RCC. To explore the relationship between competence in risk communication and SDM, the faculty raters also coded all OSCEs using the OPTION-SF (Observing Patient Involvement-Short Form) scale, an observational measure of SDM (Table 3) [23]. Prior to coding the OSCEs, two faculty raters first familiarized themselves with the measures, and then each coded a random sample of four videorecorded student-SP encounters from the 2nd-year workshop. They then compared their coding, discussed discrepancies, identified sources of disagreement, and reached an agreed level of understanding of each item.
Table 3 Observing patient involvement-short form (OPTION-SF) measure. 1. The clinician is observed to draw attention to an identified problem where choices exist, and which requires a decision making process to be initiated. 2. The clinician introduces (or validates) the options, which can include the choice of ‘no action’ 3. The clinician explains the pros of options to the patient (taking ‘no action’ is an option) 4. The clinician explains the cons of options to the patient (taking ‘no action’ is an option) 5. The clinician supports patients to explore (construct) their personal preferences in response to the options that have been described. Rated on 5-point Likert scale; detailed anchors listed on website http:// www.optioninstrument.com/.
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49 Table 4 Objective risk communication competence scores in Maine (exposed) vs. Boston (unexposed) cohorts.a Measure
Maine cohort (N = 28)
Boston cohort (N = 24)
t
p
SP-RCP SP-RCC RCP RCC OPTION-SF
1.31 1.68 1.13 0.72 1.17
1.02 0.81 0.83 0.26 0.60
3.28 7.69 3.16 7.79 5.36
0.001
Contents lists available at ScienceDirect
Patient Education and Counseling journal homepage: www.elsevier.com/locate/pateducou
Medical Education
Development and evaluation of a risk communication curriculum for medical students Paul K.J. Han a,b,*, Katherine Joekes c, Glyn Elwyn d,e, Kathleen M. Mazor f, Richard Thomson g, Philip Sedgwick c, Judith Ibison c, John B. Wong b a
Center for Outcomes Research and Evaluation, Maine Medical Center, Portland, USA Tufts University School of Medicine, Boston, USA Centre for Medical and Healthcare Education, St George’s, University of London, London, UK d Dartmouth Center for Health Care Delivery Science, Hanover, USA e Cochrane Institute for Primary Care & Public Health, Cardiff University School of Medicine, Cardiff, UK f Meyers Primary Care Institute, University of Massachusetts Medical School, Worcester, USA g Institute of Health and Society, Newcastle University, Newcastle upon Tyne, UK b c
A R T I C L E I N F O
A B S T R A C T
Article history: Received 17 July 2013 Received in revised form 4 September 2013 Accepted 7 September 2013
Objective: To develop, pilot, and evaluate a curriculum for teaching clinical risk communication skills to medical students. Methods: A new experience-based curriculum, ‘‘Risk Talk,’’ was developed and piloted over a 1-year period among students at Tufts University School of Medicine. An experimental study of 2nd-year students exposed vs. unexposed to the curriculum was conducted to evaluate the curriculum’s efficacy. Primary outcome measures were students’ objective (observed) and subjective (self-reported) risk communication competence; the latter was assessed using an Observed Structured Clinical Examination (OSCE) employing new measures. Results: Twenty-eight 2nd-year students completed the curriculum, and exhibited significantly greater (p < .001) objective and subjective risk communication competence than a convenience sample of 24 unexposed students. New observational measures of objective competence in risk communication showed promising evidence of reliability and validity. The curriculum was resource-intensive. Conclusion: The new experience-based clinical risk communication curriculum was efficacious, although resource-intensive. More work is needed to develop the feasibility of curriculum delivery, and to improve the measurement of competence in clinical risk communication. Practice implications: Risk communication is an important advanced communication skill, and the Risk Talk curriculum provides a model educational intervention and new assessment tools to guide future efforts to teach and evaluate this skill. ß 2013 Elsevier Ireland Ltd. All rights reserved.
Keywords: Risk communication Medical education Communication skills
1. Introduction Risk information—information about the probability of future outcomes—is a cornerstone of evidence-based medicine (EBM), and the communication of risk information to patients is an essential prerequisite for shared decision making (SDM) [1,2]. Yet risk information, whether pertaining to the probability of benefits or harms, is extremely challenging to understand and communicate. Past research has documented not only substantial deficits in numeracy—the ability to understand and use numerical information—among both lay persons and physicians [3–6], but numerous
* Corresponding author at: Center for Outcomes Research and Evaluation, Maine Medical Center, 509 Forest Avenue, Portland, ME 04101, USA. Tel.: +1 207 661 7619; fax: +1 207 662 3110. E-mail address: [email protected] (Paul K.J. Han). 0738-3991/$ – see front matter ß 2013 Elsevier Ireland Ltd. All rights reserved. http://dx.doi.org/10.1016/j.pec.2013.09.009
cognitive biases that limit people’s ability to process and interpret risk information and to effectively use this information in decision making [7–9]. Accordingly, a growing body of research has been devoted to developing more effective methods for communicating risk information. To date, this work has focused primarily on the textual or visual representation of risk information—e.g., using alternative summary statistics or graphical representations—and the application of these methods in decision support tools. Evidence-based best practices for representing risk information have begun to be identified [9–11], and the International Patient Decision Aids Standards (IPDAS) Collaboration has recently summarized key recommendations for the representation of risk information in decision support tools [12]. Relatively little attention, however, has been devoted to the verbal communication of risk information to patients in face-toface encounters and real-life clinical settings, and little is known or
44
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
published about how to improve physicians’ competence in this endeavor. In the UK, Elwyn, Edwards, and colleagues developed a brief workshop-based training program in shared decision making (SDM) and risk communication, which utilized visual aids (tabulated data, histograms, bar charts) designed to facilitate risk communication [13]. A randomized trial of this intervention among general practitioners in the UK demonstrated significant improvements in patient involvement in, and satisfaction with, decision making, and in clinicians’ satisfaction and use of multiple representational formats in risk communication [13]. Elsewhere in the UK, educators at St. Georges, University of London (SGUL) in 2001 developed a risk communication curriculum for undergraduate medical students, consisting of a mix of didactic and experiential workshop-based training undertaken during the pre-clinical years of medical school, and skills assessment using Observed Structured Clinical Examination (OSCE) [14,15]. Other UK medical schools devote varying amounts of curricular time to this task (KJ—personal communication); however, their experiences have not been published. The overall objective of the current study, a collaborative effort between investigators at Tufts University School of Medicine (TUSM) in Boston, MA (USA) and SGUL, was to develop, pilot, and evaluate a new curriculum, named ‘‘Risk Talk,’’ to teach risk communication skills to undergraduate medical students. Its specific objectives were to obtain proof-of-concept evidence on the feasibility and efficacy of this new curriculum, and to develop new measures to assess students’ competence in clinical risk communication. 2. Methods 2.1. Study population and setting The study population consisted of 2nd-year undergraduate medical students at TUSM. The curriculum was piloted in a cohort of ‘‘Maine Track’’ (MT) students (approximately 36 of 200 total TUSM students in each class) who choose to receive some preclinical (Years 1–2) and most clinical training (Years 3–4) at the Maine Medical Center rather than Boston campus of TUSM. MT and Boston-based students have similar academic characteristics, including Medical College Admissions Test (MCAT) scores (32.03 vs. 32.15, p = .85, for 2nd-year MT and Boston students, respectively) and college grade point average (GPA) (3.54 vs. 3.61, p = .13). 2.2. Conceptual framework Development of the Risk Talk curriculum was guided by a conceptual model of key communication processes in SDM, developed by Elwyn et al. (Fig. 1) [16]. This model translates the
Fig. 1. Conceptual model of shared decision making and risk communication. Adapted from Elwyn et al. [16].
principles of SDM into discrete, fundamental, goal-oriented communication tasks: choice talk, option talk, and decision talk. Choice talk aims to make sure patients know that reasonable options are available. Option talk aims to provide detailed information about options. Decision talk aims to support values clarification and patient preferences when deciding what is best. The broad goal of these discrete communication tasks is to facilitate deliberation—the consideration of information about pros and cons of alternative options, the range of possible futures, and clarification about individual preferences for those possible futures [17]. Addressing a vital component of option talk, Risk Talk was designed to communicate information on the benefits, harms, and uncertainties of available choice options, thus bridging the discussion of options and decisions. 2.3. Curriculum structure Risk Talk drew on the overall structure and approach at SGUL of classroom-based teaching during the pre-clinical years of medical school; Appendix A provides an overview. The focus of the current project was an experiential teaching workshop provided to secondyear students. Held in the first half of the year, this 3-h workshop consisted of a 45-min didactic lecture covering theoretical principles and practical approaches to risk communication, followed by a 2-h experiential learning session in which small groups of three students practiced communicating risk information with trained professional standardized patients (SPs), facilitated by a faculty preceptor. Students took turns leading, observing, and providing feedback on the discussions. To evaluate the efficacy of the 2nd-year curriculum, we developed an OSCE with a single clinical case designed to assess risk communication competence (described below). Additional learning experiences for first- and third-year students were developed and piloted but not formally evaluated. First-year students received a 1-h didactic lecture designed to introduce theoretical principles in risk communication. A small group of third-year students participated in a 45-min experiential learning session, involving role-play in dyads, offered during Internal Medicine clerkship rotations at Maine Medical Center. 2.4. Risk Talk workshop The Risk Talk workshop focused on two main content areas. The first was theoretical knowledge regarding both the role of risk communication in SDM, and the challenges involved in delivering it. The second content area was practical skills, the conceptual domains of which were derived from the existing SGUL curriculum and augmented through further literature review to identify best practices in risk communication [9–11].
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
The practical skills content of the curriculum was further augmented by the use of verbal ‘‘scripts’’—sample words and phrases designed to give students exemplary language to facilitate learning and practice of verbal risk communication skills. This approach was modeled on clinical skills training for SDM developed within the Health Foundation supported Making Good Decisions in Collaboration (MAGIC) program, developed by Elwyn et al. in the UK to teach SDM skills to physicians (http:// www.health.org.uk/areas-of-work/programmes/shared-decision-making/) [16]. MAGIC scripts were designed to model the key SDM content elements of choice talk, option talk, and decision talk (Fig. 1) [16]. In similar fashion, we created Risk Talk scripts to address key content and process elements of risk communication; examples are in Fig. 2. The scripts addressed essential functions of effective risk communication identified in the literature, including improving conceptual understanding of risk and uncertainty [18,19], increasing the evaluability of risk information [20], and reducing cognitive biases in the interpretation of risk information [21]. Improving conceptual understanding involves strategies to help patients understand the fundamental nature of risk estimates—their origin and derivation as well as their inherent uncertainty—and includes educating patients about the time frame and reference class of risk estimates. Improving evaluability involves strategies to make risk information easier to process and understand, including using natural frequencies with consistent denominators [3,22]. 1.
Reducing cognitive biases involves strategies to overcome well-known psychological biases in the interpretation of risk information, including the use of multiple frames (e.g., both positive and negative, absolute and relative risk reduction). Scripts aimed at operationalizing all of these functions served as the basis for the experiential learning workshops and were provided to all students. 2.5. Evaluation and measures Consisting of a non-randomized experimental study comparing 2nd year MT students exposed to the curriculum, and a convenience sample of unexposed Boston students who served as the control group, the evaluation focused on the efficacy of the 2nd-year workshop. All students had otherwise identical exposure to training in evidence-based medicine, clinical reasoning, and clinical communication as part of the TUSM curriculum to which all students were exposed on the Boston campus during their 1st year. An incentive ($15 gift card) was provided to all participating students. Students’ risk communication competence was evaluated using both objective (observational) and subjective (self-report) measures; the objective measures were completed by both standardized patients (SPs) and faculty observers. The study protocol was reviewed and approved by the Institutional Review Board (IRB) of both Tufts University School of Medicine and Maine Medical Center.
Assess preferences for information “Would you like me to go into more detail about the pros and cons of the treatment choices, including numbers and statistics?”
2.
Explain the magnitude of risks, using multiple frames
•
Both percentages and natural frequencies “Based on who you are, medical experts estimate that your risk of developing colon cancer in your lifetime is 10%.” “This is the same as saying that 10 out of 100 people like you will develop colon cancer during their lifetimes.”
•
Both negative and positive frames “Medical experts estimate that about 10 out of 100 people like you will develop colon cancer during their lifetimes.” “You can also look at this in another way: 90 out of 100 people like you [with your characteristics] will stay free of colon cancer.”
•
Emphasize absolute risk reduction “Studies show that taking DRUG X lowers people’s risk of a heart attack by 25%, or about 1/4. But to see if it is worth taking DRUG X, we need to count the number of people who are actually helped by DRUG X...” “The studies showed that DRUG X lowered the total number of people having heart attacks from 4 out of 100 to 3 out of 100.” So DRUG X helped about 1 out of every 100 people who took it.”
3.
Explain meaning of the reference class “One way to think about this is to imagine a group of [100] people like you, who have the following things in common: [sex, age, family history of cancer …etc]. Of this group of [100] similar people, 10 will develop colon cancer in their lifetime.”
4.
Acknowledge uncertainty
•
Meaning of chance / inability to predict single events “We can’t predict the future of any one person. Estimates of risk, of the chances, only tell us how many people in some group are likely to get cancer. They can’t tell us who will get the disease or not.” “Even if the risk of a complication from surgery is 10 out of 100, we don’t know whether you will be one of the unlucky 10 who will suffer the complication, or the lucky 90 who will not.”
•
45
General scientific uncertainty “Estimates of risk, of the chance of something happening, are not perfect because science is not perfect. We don’t know everything we need to know to predict the future.” Fig. 2. Risk talk scripts.
46
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49
Objective risk communication competence was the primary study outcome, measured through an OSCE conducted 3 weeks after the 2nd-year workshop, and in which both exposed and unexposed students participated. The OSCE utilized three trained professional SPs who enacted a case resembling those used in the workshops; the same SPs evaluated all students (in both the exposed and unexposed cohorts) (Appendix B). The OSCEs were conducted over 4 days, and scheduling for individual exposed and unexposed students was random and driven entirely by students’ own availability and preferences. Students’ objective (observed) risk communication competence was then assessed in two ways. First, SPs themselves rated the students immediately after each encounter, using a brief risk communication competence measure developed by adapting items from assessment tools used at SGUL (Table 1). This measure consisted of a 5-item SP Risk Communication Process (SP-RCP) measure assessing generic patient-centered communication skills (i.e., not specific to the domain of risk communication per se), and a 5-item SP Risk Communication Content (SP-RCC) assessing key content elements. SPs received 4 h of training on simulating patient encounters and applying the measures. The second way in which objective risk communication competence was assessed was through independent and asynchronous ratings by two faculty (based at SGUL and thus unacquainted with students)—a study co-investigator (KJ) and an experienced medical educator without specific expertise in risk communication—who each conducted blinded evaluations of video recordings of all OSCEs. For these evaluations, the faculty raters used versions of the SP rating instruments that were revised and expanded to improve their precision and comprehensiveness (Table 2). The expanded Risk Communication Process (RCP) measure included 13 items, while the expanded Risk Communication Content (RCC) measure included nine items. Development of the expanded RCC was informed by input from 18 international risk communication experts (Acknowledgments) recruited from the membership of the IPDAS Background Update group on Presenting Probabilities (http://ipdas.ohri.ca/who.html), and additional suggestions by IPDAS members and the study team. These experts reviewed an initial pool of 13 candidate items and completed a survey in which they were asked to rank, on a 1–5
Table 1 Standardized patient (SP) risk communication competence measure. Risk communication processa 1. Set the stage for the discussion 2. Assessed preferences for information 3. Checked understanding, elicited questions 4. Gave information clearly using plain language, avoided jargon 5. Demonstrated responsiveness, empathy, respect, professionalism Showed respect, care, and concern by words, tone, pace, eye contact, or posture Listened carefully Risk communication contentb 1. Explained risk using multiple frames (e.g., disease vs. health, mortality vs. survival) 2. Discussed both absolute and relative risk reduction 3. Explained the reference class, applicability of risk information to this patient 4. Acknowledged uncertainty in risk estimates Meaning of chance/inability to predict single events Imprecision in risk estimates General scientific uncertainty Meaning of the reference class, applicability of risk information to this patient 5. Distilled the ‘‘gist’’ or ‘‘bottom line’’ of the risk information a Rated on 3-point Likert scale: ‘‘Does not do/does poorly,’’ ‘‘Does fairly well,’’ ‘‘Does well’’. b Rated on 4-point Likert scale: ‘‘Poor (no attempt),’’ Fair (not clear),’’ ‘‘Good (fairly clear),’’ ‘‘Excellent (very clear)’’.
Table 2 Observer risk communication process and content measures. Risk communication process (RCP)a 1. Greeted patient appropriately and introduced self 2. Set the stage for discussion 3. Assessed preferences for information 4. Checked understanding, elicited questions 5. Gave information clearly using plain language, avoided jargon 6. Demonstrated responsiveness, empathy, respect, professionalism 7. Elicited patient concern and responded appropriately 8. Gave a well-paced explanation 9. Maintained a dialog 10. Demonstrated empathy 11. Actively listened 12. Built a therapeutic relationship throughout the consultation 13. Closed consultation appropriately Risk communication content (RCC)b 1. Discussed the quality/strength/weakness (e.g., validity, reliability, credibility) of the risk evidence 2. Specified the reference class (patient population) for whom the risk estimates apply 3. Specified the time period over which the risk estimates apply 4. Explained the magnitude of risk using both negative and positive frames 5. Explained risk estimates using both proportions (e.g., ‘‘9 out of 100’’) and percentages (e.g., 9%) 6. Discussed differences between baseline risk and modified risk in absolute terms (absolute risk reduction) or both absolute and relative terms 7. Acknowledged general uncertainty in all risk estimates, using qualitative terms 8. Acknowledged uncertainty due to chance or randomness (inability to predict single events) 9. Placed the magnitude of risks in context by comparing to risks of other outcomes (e.g., other diseases, treatments, familiar events) a Rated on 3-point Likert scale anchored by ‘‘Poor/no attempt,’’ and ‘‘Excellent/ Very clear’’. b Rated on 3-point Likert scale anchored by ‘‘The behavior is not observed, or erroneous information is given,’’ and ‘‘Excellent/Very clear’’.
scale, the five elements most important to include in a brief physician–patient discussion of risk information. Four unranked items were deleted or integrated with other items to produce a final 9-item RCC measure (Table 2). The two SGUL faculty raters each evaluated all OSCEs for students from both cohorts using the expanded RCP and RCC. To explore the relationship between competence in risk communication and SDM, the faculty raters also coded all OSCEs using the OPTION-SF (Observing Patient Involvement-Short Form) scale, an observational measure of SDM (Table 3) [23]. Prior to coding the OSCEs, two faculty raters first familiarized themselves with the measures, and then each coded a random sample of four videorecorded student-SP encounters from the 2nd-year workshop. They then compared their coding, discussed discrepancies, identified sources of disagreement, and reached an agreed level of understanding of each item.
Table 3 Observing patient involvement-short form (OPTION-SF) measure. 1. The clinician is observed to draw attention to an identified problem where choices exist, and which requires a decision making process to be initiated. 2. The clinician introduces (or validates) the options, which can include the choice of ‘no action’ 3. The clinician explains the pros of options to the patient (taking ‘no action’ is an option) 4. The clinician explains the cons of options to the patient (taking ‘no action’ is an option) 5. The clinician supports patients to explore (construct) their personal preferences in response to the options that have been described. Rated on 5-point Likert scale; detailed anchors listed on website http:// www.optioninstrument.com/.
P.K.J. Han et al. / Patient Education and Counseling 94 (2014) 43–49 Table 4 Objective risk communication competence scores in Maine (exposed) vs. Boston (unexposed) cohorts.a Measure
Maine cohort (N = 28)
Boston cohort (N = 24)
t
p
SP-RCP SP-RCC RCP RCC OPTION-SF
1.31 1.68 1.13 0.72 1.17
1.02 0.81 0.83 0.26 0.60
3.28 7.69 3.16 7.79 5.36
0.001
