Even if care seems right from a medical perspective, if care doesn’t fit for each individual, we may ‘deliver care’ without actually ‘caring’. The recently published Making Care Fit Manifesto (1) states that for care to fit, care should be maximally responsive to patients’ unique situation and supportive to their priorities. Care should also be minimally disruptive to patients’ lives, loved ones, and social network. Making care fit requires patients (and their caregivers) and clinicians to collaborate, both in content and manner, and it is an ongoing and iterative process where care plans should continuously be evaluated and modified.
This is especially pertinent for young adults living with type 1 diabetes. Previous research showed that young adults with type 1 diabetes have relatively poor biomedical and psychosocial outcomes (2,3). For example, HbA1c levels are higher in younger adults compared to other age groups, and strikingly, they are also higher now than they were a decade ago (3).
At the same time, very little is known about what young adults do to implement diabetes care in their lives, and what price they have to pay in terms of negative effects on themselves and their surroundings. Also, how do we bridge what happens in their personal environment (‘point of life’) and what happens during clinical encounters (‘point of care’)? Because whatever is left undiscussed with their clinicians, is also left unconsidered when designing care plans.
We explored experiences of young adults with type 1 diabetes trying to make care fit into their lives. First, we asked 62 young adults with type 1 diabetes (Median age: 27, IQR 24 to 27, 80% women) from the French ‘Community of Patients for Research’ (ComPaRe) for their experiences with the burden of treatment. They reported a high burden4 of diabetes treatment (76.5 out of 150, IIQR 59 to 94). Importantly, 3 of every 4 young adults (74%) reported that their investment of time, energy, and efforts in healthcare is unsustainable over time. This is about twice as high as for other people with chronic conditions.
Second, the Dutch ééndiabetes foundation (for and by young adults with type 1 diabetes) asked its members for their experiences in making diabetes care fit into their lives. Nine of 25 young adults (36%) indicated that their diabetes care regularly or very often hinders their education, work, hobbies, leisure or social lives. When asked to describe their biggest efforts to fit diabetes care into their daily life, they responded:
“You want to live a fun and spontaneous life, but you have to nonstop keep an eye on your sugars.”
“Food. Everything you have to do then. That’s why I sometimes skip my meals.”
“Planning. Not just the hospital appointments and changing needles, but also planning with the energy I will or will not have.”
“Regulating hypo’s and then compensating for the time I couldn’t function well due to a hypo.”
“On time and constant planning ahead. What do I need? Do I have all my stuff before I leave? When do I have to place new orders to make sure I don’t run out?”
Additionally, we asked young adults what they do to fit diabetes care into their lives, but what they don’t discuss with their clinician. Some indicated they discuss “everything” or “nothing” with their clinician. Others said:
“I use a DIY loop. (I did tell my clinician but he wasn’t interested)”
“I delay my hospital appointments as long as I can.”
“I discuss everything about my diabetes with my clinician. All the better they can help me.”
“I bolus less to prevent hypos.”
Communication is key to bridge efforts of making care fit at the point of life and at the point of care. Our future work will focus on uncovering better ways to improve the quality of diabetes care through improving conversations. And to help young adults with their lifelong, daily and ongoing endeavor of making diabetes care fit.
Kunneman M, Griffioen IPM, Labrie NHM, Kristiansen M, Montori VM, van Beusekom MM, Making Care Fit Working G. Making care fit manifesto. BMJ Evid Based Med. 2021.
Johnson B, Elliott J, Scott A, Heller S, Eiser C. Medical and psychological outcomes for young adults with Type 1 diabetes: no improvement despite recent advances in diabetes care. Diabet Med. 2014;31(2):227-231.
Redondo MJ, Libman I, Maahs DM, Lyons SK, SaracoM, Reusch J, Rodriguez H, DiMeglio LA. The Evolution of Hemoglobin A1c Targets for Youth With Type 1 Diabetes: Rationale and Supporting Evidence. Diabetes Care. 2021;44(2):301-312.
Tran VT, Harrington M, Montori VM, Barnes C, Wicks P, Ravaud P. Adaptation and validation of the Treatment Burden Questionnaire (TBQ) in English using an internet platform. BMC Med. 2014;12:109.
Cardiovascular disease (CVD) continues to be a leading cause of mortality and disease burden worldwide. There are several approaches to prevent CVD and new ones continue to emerge. The American College of Cardiology/American Heart Association guidelines for primary cardiovascular (CV) prevention recommend shared decision making (SDM) to co-create individualized plans for preventive care.
As part of an NIH-funded project which aims to identify implementation approaches that promote high-quality SDM about CV prevention for patients in primary care settings, designers and researchers at the KER Unit created Heart Health: A CV Prevention Choice Tool, an SDM tool for use by patients and clinicians during the clinical encounter to co-create a plan of care. The tool:
Offers an individualized 10-year estimate of the patient’s atherosclerotic cardiovascular disease (ASCVD) risk.
Offers lifestyle interventions – Mediterranean diet, physical activity, smoking cessation – and medication – e.g., statins, aspirin, other lipid, blood pressure, glucose lowering drugs – options and demonstrates how starting/continuing/stopping them can affect jointly and cumulatively the patient’s ASCVD risk.
Supports a conversation by which patient and clinician create an evidence-based plan that is desirable, useful, and feasible for both.
This document describes the development process we followed to produce Heart Health.
Steps for the design, evaluation, and prototyping of the tool
A. Literature review
Clinicians and researchers thoroughly searched the literature for relative risk reduction estimates of major adverse cardiovascular events for the interventions of interest. The relative risk (RR) is the probability of an outcome in the exposed group compared to the probability of an outcome in the unexposed group. The RR estimates for smoking cessation and exercise come from observational cohort studies. All other estimates of the effect of these interventions on CV risk come from randomized trials. When possible, we sought high quality systematic reviews of these primary studies. For each intervention we also sought evidence on other (nonCV) benefits, medication administration routine, adverse effects, and patient out-of-pocket costs.
Among the interventions of interest were nutrition and diet, antiplatelet medication, glucose-lowering medications, lipid-lowering therapy, blood pressure lowering medications, smoking cessation, and weight loss.
B. Summarizing the evidence
What to include?
After completing this extensive review, and discussing our findings with preventive cardiologists, we decided to include in the final tool:
Statins- medium and high dose
Blood pressure lowering medications
Decisions made on content
In addition to the factors needed to compute a 10-year ASCVD risk, we also included 2 additional parameters: lipoprotein (a) and coronary calcium score. We assumed these could be helpful in patients who were unsure as to how intensely to pursue preventive care (e.g., patients at so-called intermediate CV risk). When evaluating the use of the tool in clinic, it was evident that these parameters were rarely available in primary care settings and rarely used in discussions. Thus, we decided to exclude them from the current version.
Another parameter that was added after discussing with experts was a question on family history of premature (males <55 years, females <65 years) myocardial infarction, stroke, or sudden death in a first degree relative. This question did not affect the risk calculations per se but, if selected, a disclaimer would be displayed when calculating risk: “Your family history of heart disease means that your risk may be higher than shown. Consider further discussion with a preventive cardiologist.”
For the activities included (smoking cessation, Mediterranean diet, and exercise), it was decided to include links to patient education websites created by the Mayo Clinic for each activity, where patients and clinicians could obtain more detailed information and suggestions on how to make these changes to their lifestyle.
For Diabetes medications, we included GLP-1 agonists and SGLT-2 inhibitors. If the patient has diabetes, these two medications are part of the medications table. If the patient does not have diabetes, they are not initially included. However, the option of adding them to the table is available.
This tool uses the ASCVD risk calculator to estimate the patient’s 10-year ASCVD risk and a 100-person pictograph to display this risk. Then we use best estimates of risk reduction against this risk estimation to propose a revised ASCVD risk given the interventions chosen, assuming independence. This is based on the approach used in the highly popular and effective Statin Choice tool.
To calculate the current risk of having a coronary event (described as “heart attack” in the tool) in the next 10 years, the tool uses the ASCVD risk calculator equation and data such as: age, sex at birth (M/F), African American (Y/N), smoker (Y/N), Diabetes (Y/N), treated blood pressure (Y/N), total cholesterol (100-350 mg/dL), HDL cholesterol (10-120 mg/dL), and systolic blood pressure (90-250 mmHg) that has to be completed by the clinician or auto-populated from the electronic health record. For further detail on how to calculate the current risk using the ASCVD risk calculator, please refer to pages 32-34 on the 2013 Report on the Assessment of Cardiovascular Risk: Full Work Group Report Supplement. 
For each intervention, we used their RR estimates to calculate the future risk of having a coronary event in the next 10 years if the patient started, or stopped using, that intervention. The future risk is calculated by multiplying the current risk by the RR overall. The RR overall is calculated using the RR estimates shown in the following table, which are different depending on the use status of the intervention (i.e. if a medicine and/or an activity is started or stopped). A patient’s RR overall is calculated by multiplying together the RRs of each intervention that the patient chooses to start using, and then further multiplying by the RRs of each intervention that the patient is currently using that they and their clinician elect to stop using. The RR of an intervention that is stopped is the inverse of the RR of an intervention that is started
RR overall = RR interventions started * RR interventions stopped
This means for example, that if a patient switches from medium to high dose statins, then: RR overall = RRstatins high * 1/RRstatins medium
Any interventions that the patient is currently using and will continue to use are not included in the calculation of RR overall and therefore do not contribute to the estimate of future risk. This reflects the difficulty of determining any further risk reduction that may be achieved by continuing current interventions beyond that which is reflected in the patient’s current risk estimate. (For example, for a patient currently taking a blood pressure lowering medicine, the benefits of the intervention are reflected in the estimate of the patient’s current risk and it is unknown to what extent continued use of the medication will lower that risk further.)
Future risk= current risk * RR overall.
Activity or medicine option
RR activity/med started
RR activity/med stopped
Blood Pressure Lowering Medications
As described above, in order to overcome evidence limitations, a number of methodological compromises were made in calculating future risk. This reflects principles that are used in developing all our shared decision-making tools:
Patients and clinicians need support in making decisions even when optimal evidence does not exist.
Risk is only a device that in some circumstances may be helpful in decision making.
It is more important that any risk presented is a useful approximation that can help people make reasonable decisions than that it is precise, particularly when imprecision is unlikely to affect the final decision.
The most appropriate method for calculating risk should be based on the quality of the reasonably applicable evidence and its ability to contribute usefully and feasibly to patient and clinician decision making.
Reporting of other outcomes
For each medication stated above, we created a table with key information, the most common adverse effects, and other benefits of the medication. We selected the most discussed in practice and the most relevant to the clinical context of primary prevention. Also, we gathered information on average cost of these medications per month according to the GoodRx service, recognizing that these estimates vary greatly depending on the patient’s insurance.
Statins medium dose
Prevents strokes by 29%
Muscle aches (0-5 in 100)
Statins high dose
Prevents strokes up to 48%
Muscle aches (0-10 in 100)
Prevents strokes up to 14% alone or in combination with statins.
Muscle/joint aches, flu-like symptoms.
Prevents colorectal cancer by 20%.
Easy bruising, bleeding (3 in 1000)
Blood pressure medications
Prevents strokes up to 40%; other benefits depend on medicine used.
Depends on medicine used
Prevents strokes up to 20%. 
Prevents death by 11%, loss of >5% of body weight, prevents kidney failure by 20%. [21, 22]
Nausea-vomiting (2-3 in 10), diarrhea (1 in 10).
Prevents death by 20%, loss of 2% of body weight, prevents kidney failure by 30%.
Urinary and genital infections (200 in 1000 over 5 years), DKA (4 in 1000 over 5 years).[24, 25]
Design process: Prototyping and refinement of the tool
Designers at the KER Unit start the process of designing the tool by reviewing how relevant conversations take place in medical encounters in usual practice. After obtaining patient and clinician informed consent, we video recorded 5 preventive cardiology visits which were then reviewed by members of the team. These observations were fundamental to creating the first version of the Heart Health tool.
We invited 4 clinicians working in preventive cardiology, consultative internal medicine, and primary care at Mayo Clinic to test this tool prototype with their patients. We tested two versions of the tool with a total of 8 patients. All encounters were video recorded and reviewed. Each version of the tool was improved considering user experience, observed misuses, and clinician recommendations for adding, removing, or modifying the tool’s content.
One example of the kind of refinements done to the tool (that came from the observation of its use in clinical practice and from expert feedback) is the change to the order of display of the intervention screens. Initially, our tool showed the medications tab first and the activities tab second. The order of these seemed to imply that the best approach to discussing CV risk prevention was by focusing on medications first.
However, when observing the videos, we saw that the conversation always started with changes in lifestyle given their large impact on CV health. We also recognized that the issues presented in the tool that pertained to medicines weren’t particularly informative when considering lifestyle changes (e.g. side effects). We wanted our tool to support as much as possible the usual conversation that patients have with their clinicians. For this reason, we made the very simple but relevant decision to order the tool so that activities were available for discussion before medicines.
The third version of the prototype tool was considered the final one and was sent to a software development company that later released the tool to be implemented within the electronic workflow and online as a standalone webapp. The team at KER Unit was in constant communication with the software development team, and together we worked on improving the tool’s visual display while making sure the purpose and logic of each screen was maintained.
1. Roth Gregory, A., et al., Global Burden of Cardiovascular Diseases and Risk Factors, 1990–2019. Journal of the American College of Cardiology, 2020. 76(25): p. 2982-3021.
2. Arnett, D.K., et al., 2019 ACC/AHA Guideline on the Primary Prevention of Cardiovascular Disease: A Report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. J Am Coll Cardiol, 2019. 74(10): p. e177-e232.
3. Tenny S, H.M., Relative Risk. [Updated 2021 Mar 30]. 2021: StatPearls [Internet]. Treasure Island (FL): StatPearls Publishing.
4. Goff David, C., et al., 2013 ACC/AHA Guideline on the Assessment of Cardiovascular Risk. Journal of the American College of Cardiology, 2014. 63(25_Part_B): p. 2935-2959.
5. Duncan, M.S., et al., Association of Smoking Cessation With Subsequent Risk of Cardiovascular Disease. Jama, 2019. 322(7): p. 642-650.
6. Martínez-González, M.A., A. Gea, and M. Ruiz-Canela, The Mediterranean Diet and Cardiovascular Health. Circulation Research, 2019. 124(5): p. 779-798.
7. Wahid, A., et al., Quantifying the Association Between Physical Activity and Cardiovascular Disease and Diabetes: A Systematic Review and Meta-Analysis. Journal of the American Heart Association, 2016. 5(9): p. e002495.
8. Taylor, F., et al., Statins for the primary prevention of cardiovascular disease. Cochrane Database Syst Rev, 2013. 2013(1): p. Cd004816.
9. Zhan, S., et al., Ezetimibe for the prevention of cardiovascular disease and all‐cause mortality events. Cochrane Database of Systematic Reviews, 2018(11).
10. Gelbenegger, G., et al., Aspirin for primary prevention of cardiovascular disease: a meta-analysis with a particular focus on subgroups. BMC Medicine, 2019. 17(1): p. 198.
11. Sakima, A., et al., Optimal blood pressure targets for patients with hypertension: a systematic review and meta-analysis. Hypertens Res, 2019. 42(4): p. 483-495.
12. Schmidt, A.F., et al., PCSK9 monoclonal antibodies for the primary and secondary prevention of cardiovascular disease. Cochrane Database of Systematic Reviews, 2017(4).
13. Jia, X., et al., GLP-1 Receptor Agonists and Cardiovascular Disease: a Meta-Analysis of Recent Cardiac Outcome Trials. Cardiovasc Drugs Ther, 2018. 32(1): p. 65-72.
14. Zou, C.-Y., et al., Effects of SGLT2 inhibitors on cardiovascular outcomes and mortality in type 2 diabetes: A meta-analysis. Medicine, 2019. 98(49): p. e18245.
15. Chou, R., et al., Statins for Prevention of Cardiovascular Disease in Adults: Evidence Report and Systematic Review for the US Preventive Services Task Force. Jama, 2016. 316(19): p. 2008-2024.
16. Watson, K.E., The JUPITER trial: How will it change clinical practice? Rev Cardiovasc Med, 2009. 10(2): p. 91-6.
17. Ouchi, Y., et al., Ezetimibe Lipid-Lowering Trial on Prevention of Atherosclerotic Cardiovascular Disease in 75 or Older (EWTOPIA 75). Circulation, 2019. 140(12): p. 992-1003.
18. Brar, K.S., Ezetimibe (Zetia). Medical journal, Armed Forces India, 2004. 60(4): p. 388-389.
19. Cao, Y., et al., Population-wide Impact of Long-term Use of Aspirin and the Risk for Cancer. JAMA oncology, 2016. 2(6): p. 762-769.
20. Sabatine, M.S., et al., Evolocumab and Clinical Outcomes in Patients with Cardiovascular Disease. N Engl J Med, 2017. 376(18): p. 1713-1722.
21. Kristensen, S.L., et al., Cardiovascular, mortality, and kidney outcomes with GLP-1 receptor agonists in patients with type 2 diabetes: a systematic review and meta-analysis of cardiovascular outcome trials. Lancet Diabetes Endocrinol, 2019. 7(10): p. 776-785.
22. Khera, R., et al., Association of Pharmacological Treatments for Obesity With Weight Loss and Adverse Events: A Systematic Review and Meta-analysis. Jama, 2016. 315(22): p. 2424-34.
23. Filippatos, T.D., T.V. Panagiotopoulou, and M.S. Elisaf, Adverse Effects of GLP-1 Receptor Agonists. The review of diabetic studies : RDS, 2014. 11(3-4): p. 202-230.
24. Halimi, S. and B. Vergès, Adverse effects and safety of SGLT-2 inhibitors. Diabetes & Metabolism, 2014. 40(6, Supplement 1): p. S28-S34.
25. Musso, G., et al., Diabetic ketoacidosis with SGLT2 inhibitors. BMJ, 2020. 371: p. m4147.
Dr. Margaret Schwarze, a surgeon from the University of Wisconsin, and her colleagues published a proof of concept study “A Framework to Improve Surgeon Communication in High-Stakes Surgical Decisions—Best Case/Worst Case.”1 (https://www.ncbi.nlm.nih.gov/pubmed/28146230) This article was recently the topic of discussion during our bi-weekly Shared Decision Making working group.
Schwarze and colleagues described that hospitalized elderly adults who have urgent surgical conditions may receive unwanted burdensome surgical care at the end of life. Routine discussions between surgeons and elderly patients may not result in a care plan that authentically honors the goals, values, and preferences of patients.
To improve these discussions, they developed a “Best Case/Worst Case” framework to discuss high stakes surgical decisions (https://www.youtube.com/watch?v=FnS3K44sbu0). Surgeons were instructed to draw two lines on a paper. One line represented the option of pursuing surgical treatment and the other line represented the option of choosing supportive care. At the top of each line (or option), the surgeon would write and write and describe the “best case scenario” (or outcome) of that option. At the bottom of each line, the surgeon would write and describe the “worst case scenario” of each option. Somewhere in the middle, the surgeon would describe the “most likely scenario” of each treatment option. Surgeons were allowed to describe each best and worst case scenario as they best saw fit according to the individual patient circumstances. Thirty cardiac, vascular, and general surgeons at the University of Wisconsin completed a two hour training on the communication framework.
In this pre/post study, investigators enrolled 32 elderly hospitalized patients with urgent, but not emergent, surgical conditions with a high risk of adverse outcome (≥40% risk for serious surgical complication or ≥8% risk of death). In the pre-intervention group, usual care conversations were audiotaped. In the post-intervention group, conversations using the “best case/worst case” framework were audiotaped.
The primary outcome was the OPTION 5 score (https://www.ncbi.nlm.nih.gov/pubmed/25956069), which allows a rater to rate the decision making process on 1) presentation of multiple options, 2) establishment of a partnership with the patient, 3) description of the treatment differences in each option, 4) elicitation of patient preferences, and 5) integration of patient preferences into the plan.
Prior to the intervention, the median OPTION 5 score of audiotaped conversations was 41 (on a 0-100 scale)—and improved to 74 in the post-intervention group. Surgeons in the intervention group were more likely to involve patients and families in decision making, were more likely to present various treatment choices, and were more likely to describe outcomes rather than isolated procedural risks.
During the discussion at our SDM working group, several strengths of this approach were noted:
This method was easily adoptable by surgeons and can be used in high stakes decisions in the acute hospital setting.
Whereas many patients undergoing potentially risky surgical procedures may not be aware of potential complications, this method formally allowed for patients and surgeons to at least consider a “worst case scenario.” This has the potential to spark discussion about what a patient values most in determining a treatment plan.
This method allowed surgeons the flexibility to tailor the treatment options as well as the outcomes of those options to the individual patient. This may therefore represent a universal, non-disease and non-context specific method to improve shared decision making discussions in general.
We also noted several questions and limitations:
What constitutes a “best case” or “worst case” outcome may considerably vary between patients—as patients value different things when faced with high stakes, end of life decisions. Some people at our working group thought that the example descriptions of the “most likely” outcome actually seemed worse than the example descriptions of the “worst case” outcomes. Who determines what the best and worst case scenarios were? Was this left up to the individual surgeon? Were the descriptions standardized in any way? Were questions asked to assess if description of best and worst outcomes rang true to the individual patient? How much were patients influenced by a potentially biased presentation of one treatment option versus the other? What do we know about the patients’ perspectives and interpretations of the best case and worst case scenarios?
To our knowledge, the likelihood of the outcomes was not specifically disclosed in a salient manner. If one were to apply the best case/worst case methodology to a “lower stakes” decision, the worst case scenario may be very rare—and the most common outcome a particular decision may be that nothing changes.
Even though the OPTION 5 score was higher in the post intervention group, does this really mean that a better decision was made? While we agree that the OPTION 5 (https://www.ncbi.nlm.nih.gov/pubmed/25956069 ) and OPTION 12 (https://www.ncbi.nlm.nih.gov/pubmed/15713169) scores represent a good attempt to measure a certain quality of shared decision making, there are still various aspects of decision making that are overlooked. Tools to better measure the quality of decision making are needed.
While we congratulate the authors on having a high inter-observer agreement regarding ratings on the OPTION 5 score (.8), this is much higher than what most other groups (including our group and the group validating the instrument) (https://www.ncbi.nlm.nih.gov/pubmed/25956069) have been able to achieve (.6 to .7). In addition, both the pre-intervention and post-intervention OPTION 5 scores were quite a bit higher than what we have seen in other trials, including ours. Additional information about the process of training observers and measuring inter-observer reliability is desirable.
Overall, Dr. Schwarze and colleagues (http://www.surgery.wisc.edu/research/researchers-labs/schwarze/) showed that a framework for formally presenting the best case outcome, worst case outcome, and most likely outcome of various treatment options increased shared decision making as measured by the OPTION 5 score. We congratulate Dr. Schwarze and colleagues for developing and testing a framework to try to improve decision making for high stakes surgical decisions for hospitalized elderly adults!
Submitted by Michael Wilson, M.D. Dr. Wilson studies end-of-life decision-making in the hospital and intensive care unit (ICU). He aims to improve individualized prognostication, shared decision-making and the delivery of quality palliative care to patients and their family members in the hospital setting.
Taylor LJ, Nabozny MJ, Steffens NM, et al. A Framework to Improve Surgeon Communication in High-Stakes Surgical Decisions: Best Case/Worst Case. JAMA Surg 2017.
Submitted by Marleen Kunneman, PhD; Michael R. Gionfriddo, PharmD, PhD; Victor M. Montori, MD, MSc
Metaphors are common in clinical medicine and can be helpful in discussing and understanding the complexities of health and illness. Blood vessels are like plumbing, the brain is like a computer, and when facing illness we use all weapons available to fight the disease. The creativity of the human mind is boundless. Metaphors can help communicate and retain complex concepts between clinicians and between clinicians and patients, with clinicians who use more metaphors considered better communicators.1 Yet, these metaphors can be unhelpful when they become so internalised that we don’t recognize them anymore, and, unconsciously, they shape how we think and act. 2
When it comes to medical decision making, the relationship between the clinician and the patient is often compared to a pilot that takes a passenger to his destination, a plumber that fixes the leak, or a mechanic that fixes your car. We need to accept that the pilot, the plumber and the mechanic are the experts and that they are therefore able to make decisions about how to address the problems. We, the ordinary people, have not studied and/or gained sufficient experience to understand these issues, let alone to be meaningfully involved in making such decisions. Such metaphors are often used by opponents of shared decision making to illustrate that the expertise necessary to understand the complex issues of health and illness is not easily translated in the limited time frame of an encounter, and therefore, patients should respect and trust clinicians’ expertise and delegate to them the difficult task of deciding what to do.
In shared decision making (SDM), clinicians and patients work together to figure out how to best address the patient’s situation. It is a conversation between the clinician and patient, a way to craft care, and a way to fundamentally care for this patient, not just for people like this patient. 3 This characteristic makes it inappropriate to use metaphors like mechanics fixing a car. Mechanics take care of cars, not of the owners. It is rare, exceptional, for a mechanic (or pilot, or plumber) to see the owner’s situation in high definition. At best, in fulfilling their duties – fixing the car – they can honor the relationship between the ‘object’ and the ‘owner’. In fundamentally caring for this patient, however, clinicians must take care of both the object – the body – and the owner. This is because, as Hitchens said, patients don’t have a body, they are a body.4
A serious illness that disrupts a person’s hopes and dreams should not be compared to a bump in the road which causes your car to break down. The car does not ‘feel’. The car does not experience side effects. Having an issue or needing maintenance does not change the cars experience of being a car or how it views itself, or it’s ‘carness’. Conversely, humans do feel, they experience side effects, and illness can affect how people view themselves and their place and relationship with society. Furthermore, if the patient’s situation is not addressed in a way that fits their life, they cannot just go back to the shop and undo the repair. Or just replace the broken parts, or, for that matter, get a new ‘object’ and replace the old one altogether. If only health were that simple. Indeed, in a service industry like automobile repair, you don’t co-create an oil change.5 But when it comes to care, clinicians and patients co-create ways to address the patient’s situation. It is this patient’s situation that should shape how care is decided on and delivered, and the method behind care and decisions about health care is the deeply human activity of having meaningful conversations between clinicians and patients.
Using de-humanizing comparisons can be problematic in shaping how we think and act, and in how we are understood and perceived. Most importantly, when using such metaphors, a fundamental aspect of medicine – caring – gets lost, forgotten, or neglected. Metaphors are common and they can support a complex conversation about health or illness, but we must be careful that these metaphors do not distract us from caring.
Casarett D, Pickard A, Fishman JM, et al. Can metaphors and analogies improve communication with seriously ill patients? J Palliat Med. 2010;13(3):255-260.
Sat May 7th. All set, ready to go! Excited to visit the KER Unit for a few weeks and to join them at the SAEM SDM Consensus Conference in New Orleans. This will be my first visit to the Mayo Clinic, and one I’ve been looking forward to since I became a research collaborator last winter.
Wed May 11th. We just returned from the Consensus Conference. It was inspiring and motivating to see so many participants (most of them clinicians) trying to find ways to make SDM work in practice and to improve care for their patients. Victor presented his keynote lecture ‘What is SDM? (and what it is not)’ and we worked on writing a paper on this keynote for Academic Emergency Medicine.
Thu May 12th. First day at the KER Unit. What a day! I attended a course on EBM, discussed grants and ongoing research projects with Juan Pablo, Mike and Aaron, and had a braindump on SDM (old and new thinking) with Victor and Ian. Note to self: replace ‘yes, but…’ by ‘yes, and…’.
Sun May 15th. Friday, I finished the AEM paper with Ana and Erik. Gaby presented her study on the effects of social networks in management of diabetes on Saturday. In the evening, we got together for drinks and laughs (with bubbles, cheese and chocolates) at Annie’s place. Today, I’m going out to meet Nilay for brunch.
Mon May 16th. Started with the weekly huddle this morning: what a great way to get an overview of what each member of the team is working on right now. I worked on our Choice Awareness project* and attended the Patient Advisory Group to discuss Juan Pablo’s project on SDM in Thyroid cancer treatment. Amazing how this group of patients manages to come together every month (for over 10 years!), to improve the work of the researchers and to make sure that researchers don’t lose the connection with ‘the real world’.
Tue May 17th. Trying to see whether the Choice Awareness project can take us to the moon! Maybe. Also met with Kasey to learn more about the ICAN tool.
Wed May 18th. No trip to the moon (yet), we will have to find other methods to make this journey. I worked with Victor to build my Apollo II. Juan Pablo and Ian joined, which led to a conversational dance of thoughts, (crazy) ideas, hypotheses, and approaches. Best day ever! In spite of, as well as because of the challenges we faced this morning. In the afternoon we came together with a group of clinicians and researchers interested in SDM in diagnostics to see how to take this field forward.
Fri May 20th. Yesterday, I discussed the progress and challenges around the Choice Awareness project in the SDM journal club. We went for dinner and drinks afterwards to continue our discussion on SDM old and new thinking. I continued with the project today, focusing on capturing the differences in SDM between a mechanical approach and a human connection. It takes two to tango, but we have no way to measure that dance. Speaking of dancing (and of mechanical approach versus human connection), in the evening we had a birthday party at the local salsa place.
May 22nd. BBQ with the KER Unit team at Aaron’s place yesterday and smores at the river with Gaby, Mike and the Montori family today.
May 25th. Worked on the Choice Awareness project for the past few days. Met with the department of Neurology yesterday to discuss possible collaboration. Kasey received good news (scholarship), as did Laura (residency). Maggie arrived, and Ana said goodbye. Sara had her last day before her maternity leave. I worked on Aaron’s manuscript and discussed a second paper for AEM on SDM/informed consent with Rachel.
May 26th. Last day at the KER Unit. Overwhelmed by how much I learned about the team, the work, the collaborations. And, to be honest: about myself and about my work as a researcher. I’m impressed how a team that advocates kind and careful care manages to practice what they preach and welcome guests in such a warm and friendly way. After saying goodbye to Kirsten, this kind and careful visit ended with a road trip with Ben to the airport. What an experience.
With love, Marleen
Marleen Kunneman, PhD. Research fellow at the department of Medical Psychology of the Academic Medical Center, University of Amsterdam (the Netherlands), and research collaborator of the KER Unit.
*Note: Results of our Choice Awareness project will be presented at the European Association for Communication in Healthcare (EACH) Conference in Heidelberg (September 7th-10th, 2016). Oral presentation on September 10th: ‘Choice Awareness as Pre-requisite for Shared Decision Making in Videos of Clinical Encounters’.
Policymakers fashionably prescribe shared decision making for patients who face fateful decisions. These patients have two or more medically reasonable courses of action that differ in important aspects. The extent to which these aspects differ in ways that matter to each individual patient justifies patient involvement in the decision-making process. Similarly, the extent to which clinicians can accurately predict the values and preferences of informed patients reduces the value of shared decision making. Only in circumstances where the distribution of patient preferences is very narrow can clinicians correctly deduct patient preferences (e.g., analgesics vs. no intervention for moderate to severe pain). This is often the case when the pros and cons of alternative courses of action are well known, their likelihood estimates are based on highly reliable research evidence, and difference between the benefits and the potential harms and inconveniences is large and clear. In such situations the distribution of patient preferences will be narrow enough that most clinicians can assume correctly what most patients will want. At the extreme, these decisions will seem purely technical, where the right course of action is apparent to those with a good understanding of the situation. This would include professionals with pertinent training. In situations that cannot be resolved by the application of technical knowledge, patients, when informed, will exhibit a range of preferences. It seems appropriate then that patients and clinicians partner to share information, deliberate, and arrive at a decision together. We call this process shared decision making.
Proponents of shared decision making assume that most clinicians and patients, when given the tools, time, and supportive setting necessary, will be able to implement shared decision making. Reality seems to behave differently: surveys suggest that patients are not universally inclined toward shared decision making, clinicians are often portrayed as barriers to this process, and environments have electronic medical records, phone calls, time pressures, competing demands, and noise that conspire to interfere with shared decision making. What’s going on if patients and clinicians aren’t adhering to the shared decision making prescribed on their behalf?
Our group, the KER UNIT, characterizes shared decision making as a conversation – an activity in which patients and clinicians turn with one another (the etymology of conversation—versare turn; con with). In conversation, the options with their attributes or issues are in dynamic interaction as the patient and clinician consider them and experimentally try them on. This highly interactive dynamic requires the active engagement and involvement of the patient and clinician. This turning-with of patients and clinician is the dance of shared decision making.
The clinician is used to contemplating the situations of patients and making tough decisions routinely; but for this patient, at this time, the task is anything but routine. Thus, it is natural to delegate to the more experienced and emotionally detached of the two the task of organizing the decision-making conversation. The clinician, leading the dance, will identify that a decision needs to be made, the relevant options and their relative desirable and undesirable features, and will invite the patient to consider these options and features. But, to what extent are patients willing and able to engage in deliberation?
We propose that the adequate way of answering this question is through empathy. In suggesting empathy we do not mean that clinicians should empathically divine the right decision for the patient; quite the opposite. We are suggesting that the co-creation of decision also involves the co-creation of the patient-clinician relationship and the conversational environment in which each decision is made. Empathy directs attention to the clinician’s active role in finding the right relationship and stance to join this patient at this time in decision making. Clinicians are trained and are expected to exhibit empathy when interviewing and examining patients, responding to patient concerns, and delivering bad news. The role of empathy in supporting decision making has not been fully discussed, to our knowledge. In this case, empathy requires attention to the situation of the patient and to the cues, verbal and nonverbal, the patient offers as the clinician invites the patient into the deliberative process. Some patients may be able to partner fully and co-create the decision; others may engage with the information, but delegate the rest of the tasks of deliberation and decision taking to the clinician. This is the expression of a preference that is being constructed on the spot (it follows that this preference cannot be adequately assessed with a survey tool, before the encounter and therefore out of context). The appropriate stance in the conversation is available to the clinician in subtle signs that the clinician can pick up through empathic attention to the patient. Focus on who the patient and clinician are, and can be, for each other in this conversation allows us to respect that the same patient may be willing to co-create one decision while preferring a lesser role for the next. The challenge for the clinician is to correctly respond, in real time, to these emerging preferences.
Shared decision-making tools produced for use during the clinical encounter need to account for this clinical task and be designed to support empathic decision making. When encounter tools offer too much information or script a step-by-step decision process, they may inadvertently limit the ability of the clinician to empathically guide the process. When tools are used in preparation for the visit, clinicians may assume that completion of the tool and associated worksheets signals that patients are fully engaged and ready to make decisions. That a tool should enable and support empathic decision making is not currently a requirement for their design of decision aids, or a metric for their impact.
In summary, shared decision making is one of an infinite set of ways in which patients and clinicians can engage in conversation about fateful decisions without a technically correct answer. To create the environment in which patients and clinicians co-create decisions, clinicians must actively invite and support patients in the process, empathically “reading” the patient to match their evolving preference for participation. Tools to support this process need to be designed to facilitate and not interfere with empathic decision making, and this may form the basis for new measures of decisional quality.
Thus, we are not just for shared decision making. We are for empathic decision making.
Oncology encounters are highly complex. Communication is suboptimal and there is evidence that patients and clinicians often fail to “get on the same page.” Shared decision making is being promoted as a means of facilitating effective and patient-centered communication in oncology. Here, Dr. Aaron Leppin and colleagues survey patients and clinicians immediately after an oncology encounter to determine the extent to which they agree on whether a cancer care decision was made during that encounter. The extent of agreement is impressively low. These findings have implications for the way we think about shared decision making and the validity of its measurement in oncology. (click here for abstract)
With a new interface that includes versions in English, Spanish, and Chinese, the Statin Choice decision aid (http://statindecisionaid.mayoclinic.org) is out. With over 70,000 uses worldwide year-to-date and new policy endorsements for its use (JAMA Article), the Statin Choice decision aid is helping patients and their clinicians have meaningful conversations about whether to use statins to reduce cardiovascular risk. It helps them adhere to the new guidelines, in a patient-centered manner. And with new work to integrate the tool into all major EHR providers, it may be the best demonstration of meaningful use.
Enhancements from the first version also include two options for printing in the office: color and black-and-white, in addition to the existing option to emailing the tool after its use to the patient, a family member, or another clinician. In terms of new content, the biggest difference is the exclusion of the aspirin component (see below). We have also beefed up the Documentation tab, an copy-and-paste interim solution before full integration into EHR to enable documentation of shared decision making, a key step toward advancing these conversations as a measure of quality of care.
This version is the result of hundreds of notes suggesting changes and enhancements that result form the experience of using it in practice. We hope to have responded properly. And thank you.
Why was aspirin removed from the latest version of the Statin Choice decision aid?
In response to the new AHA/ACC guidelines for cardiovascular prevention, there has been renewed interest in using the Statin Choice decision aid to translate the recommendations in a patient-centered way. With this attention, there has been interest from preventive cardiologists in using this tool. They brought to our attention that indeed the evidence about efficacy of aspirin for the primary prevention of cardiovascular disease is inconsistent: clearer effect in men in relation to heart attacks but not stroke, in women about preventing strokes but not so much heart attacks and a series of negative trials in patients with diabetes and peripheral vascular disease have made it difficult to provide a simple message to all at-risk patients: a baby aspirin can reduce your risk of cardiovascular events. Also, emerging evidence suggests that the risk of bleeding with aspirin goes up as the risk of cardiovascular events, such that those who may benefit the most are also most likely to be harmed (although most aspirin bleeds are relatively inconsequential compared to a heart attack or a stroke).
It is telling when experts are talking more about using aspirin to prevent colon cancer than to prevent cardiovascular events (to our knowledge no one is yet recommending it for this purpose).
We will continue to monitor this evidence as we, the producers of Statin Choice, thought the evidence was good enough to add to and keep in the tool, and we will have a low threshold to put it back in as new evidence emerges, both of its efficacy and harm.
Written by Victor M. Montori, MD and Jon C. Tilburt, MD
Lee and Emanuel raise the profile of the shared decision making (SDM) provisions in the Patient Protection and Affordable Care Act. We concur that those provisions should spur research and development in SDM. However, their claims, that we already know how to implement SDM and that it is time for pay-for-performance for use of certified decision aids, are both premature and misguided.
Studies of decision aids implemented outside clinical visits show improvements in patient knowledge about the available options and about their risks and benefits, but not in actual sharing of decision making. Decision aids for use by patients and clinicians during the visit may work better. Video data from hundreds of recorded visits show a stark difference: patients in decision aid visits are better informed and participate more in making decisions. Patients and clinicians end up more comfortable with decisions they made together. In-visit decision aids galvanize patients and clinicians around a shared a purpose – to make the best possible evidence-based decision given the patient’s values, preferences, goals, and context. Yet, getting this degree of patient engagement does not happen with the flip of a switch and routine implementation remains untested.
Lee and Emanuel rightly point out the potential utilitarian benefits of SDM, particularly about surgical decisions. However, in so doing they jeopardize the patient-centered vision at the core of SDM. Moreover, their economic claim of cost-savings overreaches the current state of the evidence, making their SDM-linked pay for performance proposal premature. Research on SDM implementation is green, clinicians and patients are not ready, training and tools are just evolving. Thus, we support the law’s push for research and development. Their proposal is also dangerous. A focus on cost containment and pay for performance can corrupt the journey toward implementing SDM for all: we fear that the next time a clinician pulls out a decision aid, the clinician will be thinking about reimbursement while the patient wonders whether the clinician has her back.
There is quite a bit of evidence about best ways to convey risk information to help with policy or clinical decision making. Pictographs and bar graphs along with numbers and descriptions are considered best. Some emerging research suggests that some elements will help some patients more than others (for instance people with low numeracy).
Recently, Fagerlin, Zikmund-Fisher and Ubel published their decalogue of risk communication in the Journal of the National Cancer Institute. Their ten steps to better risk communication were:
Use plain language to make written and verbal materials more understandable.
Present data using absolute risks.
Present information in pictographs if you are going to include graphs.
Present data using frequencies.
Use an incremental risk format to highlight how treatment changes risks from preexisting baseline levels.
Be aware that the order in which risks and benefits are presented can affect risk perceptions.
Consider using summary tables that include all of the risks and benefits for each treatment option.
Recognize that comparative risk information (eg, what the average person’s risk is) is persuasive and not just informative.
Consider presenting only the information that is most critical to the patients’ decision making, even at the expense of completeness.
Repeatedly draw patients’ attention to the time interval over which a risk occurs.
Online software to create pictographs can therefore be quite handy. Some do so without resorting to giving each “person like you” an anthropomorphic shape .
Our favorite however, is one that shows the outcomes showing visual cues that are easily relatable, based on the iconic smiley face. We are impressed by Dr. Chris Cates’ Visual Rx tool, which is a free online tool that creates “smiley face plots” to depict the impact of a treatment on 100 people.