In this interview, we speak to Dr. Joshua E. Rosen about his latest research into treatment options and what role a surgeon’s language plays in people's perception of its risks.
Please could you introduce yourself and tell us what inspired your career in surgery?
I am a general surgery resident at the University of Washington currently completing a two-year research fellowship at the Surgical Outcomes Research Center (SORCE) focusing on improving how surgeons and patients work together to make treatment decisions.
I was drawn to surgery as a career because I liked providing tangible interventions that could dramatically improve a patient's disease process. I was also drawn to the difficult and often high-stakes decision-making that occurs in surgery and find it very rewarding to work with patients during difficult times in their lives.
Why is it important that treatments are described effectively when talking to patients? What is the importance of good communication between surgeons and patients?
The ultimate goal of any decision around medical treatments is to maximize the chance of achieving the outcomes and goals that are most important to the patient. This can often best be accomplished through the process of shared decision-making, which emphasizes that clinicians and patients are partners in the decision-making process.
However, in order for patients to be effective partners in the decision-making process, we as surgeons need to provide them with the information they need to engage in a productive conversation. So we need to both listen very carefully to what patients are telling us about their needs, values, and preferences, and also be able to effectively describe different treatment options so that we can work with patients and figure out which one is best for them as a unique individual.
Image Credit: didesign021/Shutterstock.com
Can you describe how you carried out your latest research that you presented at the virtual American College of Surgeons (ACS) Clinical Congress 2021 on perceptions of treatment? What did you discover?
We conducted a series of online surveys using Amazon’s Mechanical Turk (MTurk) crowdsourcing platform. In the surveys, we described appendicitis, which is one of the most common acute surgical conditions, and some options for treating it including surgery and antibiotics. We then described some of the complications that can happen with each treatment, such as a surgical site infection or an abscess in the belly.
Different participants saw the probability of these complications described in different ways: either verbally such as “uncommon”, with a single number such as “3%” or as a range of numbers such as “1-5%”. We then asked participants to estimate the likelihood that a “typical patient with appendicitis” would experience the various complications and they responded on a scale of 0-100%.
What we found was that when we communicated the risk of complications using verbal terms, such as “uncommon” there was a very large amount of variation in how participants interpreted the risk, compared to using a single number or range of numbers.
For example, when we said that an abscess in the belly after taking antibiotics was “uncommon”, the majority of participants estimated the risk as being anywhere from 10-48%, but when we said the risk was “3%”, the range of responses was much lower (the majority between 0 and 14%). They also often estimated the chance of a complication as being much higher when verbal descriptions were used compared to using numbers.
You discovered that using percentages is more effective than using qualitative terms such as ‘sometimes’. Why do you believe this is?
I think that there are a number of mechanisms that may underly this and there have been studies in other fields that try to work this out. In this case, I think one of the most important reasons is that when a person interprets a verbal probability description, they do so within the context of their own life experiences.
For example, if you talk to someone who just won the lottery yesterday, their concept of what “rare” or “unlikely” means, would probably be very different than someone who has been buying tickets but losing for the last five years. Similarly, if you have a friend who had appendicitis and had a complication afterward, you might interpret the term “uncommon” to be a much higher chance than another person whose friend had a totally uneventful recovery.
How will your research help surgeons to make more informed decisions when discussing treatment options? What advice would you give to surgeons when counseling patients?
I think that this research will help surgeons realize how much the language we use to communicate risks and probabilities matters in what our patients hear and interpret. My advice to surgeons would be to carefully consider the ways in which they communicate risk to patients and make sure that they are actually hearing and taking away what you intend to communicate.
The most effective communication technique will likely be different for different patients, but even the simple act of checking in with a patient and clarifying their understanding of what was discussed may be very beneficial.
What impact will your research also have on patients that may be looking at undergoing surgery?
This research is part of a broader series of projects we are undertaking to improve the way we counsel patients about treatment for acute surgical conditions in particular. These decisions often have to take place in a time-pressured environment (such as the emergency room) and often involve clinicians and patients who may not have a long history together (compared to, for example, a discussion of starting a new blood pressure medication with a primary care provider).
Our hope is that the information from this study, and those from other studies like it, will help us build better tools and processes to facilitate effective surgeon-patient communication and help surgeons and patients work together to make the best decision for each unique patient.
Image Credit: Juice Verve/Shutterstock.com
Do you believe that if all surgeons adopted a quantitive approach when discussing treatment options, patients will be able to make better-informed decisions?
I think that each patient is different and there are certainly some patients for whom a totally quantitative approach may not work well. This is why the relationship between a surgeon and their patient is still so important because these conversations and interactions need to be modified for what works best for that individual patient.
I do think though that when possible, surgeons should try to be more quantitative in their discussions and help patients contextualize the numbers for their own life and the decision that is being made. For me, the most important thing is for surgeons to be aware of the powerful effect of our communication practices on how patients interpret risk, and to make sure that what they are intending to communicate is what is being received.
What are the next steps for you and your research?
We are using insights from this and other studies to design tools that can help surgeons and patients make treatment decisions more effectively. For example, we recently launched a new tool to help patients choose between antibiotics or surgery if they have appendicitis that the results from this study. You can view it at www.appyornot.org. Personally, I am interested in continuing to understand the barriers to effective decision-making for both patients and surgeons and developing tools and processes to help overcome them.
Where can readers find more information?
You can find more about the work we are doing at the Surgical Outcomes Research Center here (www.becertain.org) and at the Values and System Sciences lab here (https://www.vsslab.org/)
About Dr. Joshua E. Rosen
Dr. Rosen is a general surgery resident at the University of Washington and an NIH T-32 funded research fellow at the Surgical Outcomes Research Center (SORCE). After studying Nanotechnology Engineering at the University of Waterloo in Ontario, Canada he completed medical school at Yale School of Medicine where he completed a research year studying the outcomes of lung cancer treatment in large databases.
His current research work focuses on decision-making in acute care surgery. He eventually plans to pursue fellowship training in trauma surgery and surgical critical care.
Study reveals the impact of prompt design on ChatGPT's health advice accuracy