Share this:
May 2nd, 2017

A case of the best, worst, and most likely

By Victor M. Montori


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 ( 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 (  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 (, 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 ( ) and OPTION 12 ( 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) ( 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 ( 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.



  1. 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.

Tags: Research in SDM, SDM in practice, Shared decision making


Please login or register to post a comment.
Contact Us · Privacy Policy