Communicating Uncertainty to Your Clients
(Estimated reading time 5 min)
Chelsea Greenberg, DVM, MS, Diplomate ACVIM (oncology)
Uncertainty in veterinary medicine is inevitable and causes stress for both you and your client. For veterinarians, uncertainty may create self-doubt, feelings of inferiority, and emotional insecurity. 1 For clients, uncertainty may spur anxiety, worry, fear, and the tendency to catastrophize. Clients facing uncertainty may express denial, avoidance, selective ignoring, selective misinterpretation, and minimization. 2
Many veterinarians feel uncomfortable addressing situational uncertainty with clients. Internalized misconceptions can quickly turn situational uncertainty into self-doubt. This may manifest into thoughts like: “I need to acquire more knowledge about the disease” or “I’m not good enough/smart enough,” etc. 1 Knowledge only goes so far in addressing feelings of uncertainty, and learning to cope with, and communicate, uncertainty is paramount for the psychological well-being of both you and your client.
Although there are many diseases in our patients that are easily treated, uncertainty is common. Veterinarians are often comfortable communicating prognostic uncertainty for certain disease states, such as acute pancreatitis, diabetes mellitus, or automobile injuries. When cancer is diagnosed, however, these same veterinarians may hesitate to communicate the prognosis even though the required communication skills are similar to those used to discuss a non-cancer prognosis. Sometimes clinicians feel they must provide a specific prognostic point estimate for a cancer patient when that is not required to provide your client with realistic and hopeful expectations. 3
Communicating prognostic uncertainty
A basic understanding of statistics is useful for understanding the prognosis and communicating it effectively to your clients. Cancer survival information is often visually expressed as a Kaplan-Meier (KM) survival curve (see below). The KM survival curve presents the probability of survival during a certain time interval. The KM survival curve below shows the outcomes of 3 groups of dogs with metastatic hemangiosarcoma. The dogs were treated with surgery alone (dashed line), surgery and chemotherapy (dotted line), or surgery and Torigen’s autologous vaccine (solid line). The x axis is the time in days and the y axis is the proportion of cancer patients alive:
Each “drop” in the survival curve represents the death of one or more individuals at that time point. The median survival can also be determined by the KM curve. The median is the middle number (halfway point) where 50% of patients are deceased and 50% of them are still alive. From the KM example above, the median survival is 41 days for dogs treated with surgery alone (black oval), and 142 days for dogs treated with either surgery and chemotherapy or surgery and Torigen’s autologous vaccine (black open oval).
Although the median is a useful metric in discussing survival for a population of cancer patients (because it is not influenced by individuals who do very poorly or very well with treatment), it is not the whole story. Distribution of the survival variation is also important. In other words, what happened to the patients 50% before the median survival, and the 50% of the population after? This next example depicts 2 different KM curves with the same median survival (black oval):
The black line depicts patients receiving cancer treatment and the blue line depicts no treatment. Both lines have a very similar steep drop in survival up to the median (180 days), where they dramatically diverge. The dogs that were treated (black line) have higher chance of long-term survival (~ 40% of treated patients are alive at >600 days), versus very few of the untreated dogs were alive at 200 days (blue line). Despite identical median survival time values, the treatment benefit for dogs that survive past the median is clear. If only the median survival time is communicated, versus an explanation of the survival distribution, your client may not understand the advantage for choosing treatment for their pet.
The problem is not prognostic uncertainty, but how it is communicated.3 Research reveals that human patients and their relatives understand and accept a prognosis is uncertain.4 The best way to convey prognostic uncertainty is to communicate the full variation in survival time. This includes the probability for the longest survival, shortest survival, and median (50%) survival. 5 The downward curve of the survival graphs is sometimes interpreted as highly negative by human patients6, demonstrating the importance of educating your clients about what these survival curves represent rather than sending them home as a handout. Highlighting that some patients can survive for a long time can be important, emphasizing the possibility of their pet becoming a long-term survivor (when true).
Clients want both realism and hope when learning about their pet’s prognosis. The following strategies can help achieve that goal. First, explain the median survival in a positive, hopeful manner:
“Half of all dogs will live longer than 180 days.”
Next, describe all of the different variations in survival time. The probability of the longest survival and shortest survival (Kiely 2011). If possible, give personal examples of patients who have been treated in your practice or by your referral partners:
“Some dogs unfortunately do not respond well to treatment, yet close to 40% of dogs will live >600 days with treatment. I have had 3 other patients that have been referred to Dr. Smith for treatment. One of them lived happily for 450 days and the other 2 are still alive 2+ years later.”
Communicating uncertainty with treatment response
Conversations surrounding uncertainty also happen with the availability of new treatment options (such as cancer immunotherapy). Even though there may not be outcome data for a new therapy, or a specific indication, that doesn’t mean you can’t help the client understand the possible benefits a new therapy may provide.
The same prognostic communication techniques can also be applied to the discussion of potential benefits from newer therapies. Helpful information can be available to you even when more concrete data methods are not available. Information can come from a variety of sources. You can extrapolate data from human and animal studies or from clinical pets treated for a different indication. Also, anecdotal information may be available from treated clinical pets. First, describe the available information along with anecdotal data (make sure the client understands what type of information you are discussing) and then communicate the possibility of their pet responding well and/or becoming a long-term survivor (if known):
“I know that we discussed chemotherapy for Fluffy’s cancer and that you have decided against that option due to possible side effects. There is a new cancer treatment that may help Fluffy. The treatment is still undergoing clinical trials to gather more survival data, and they do have some feedback on a few cases similar to Fluffy. This form of treatment has been studied for decades in rats, rabbits, and dogs, with encouraging results. There is even an FDA approved form of this cancer treatment for people. Decreasing cancer size has been reported in some cases similar to Fluffy’s and decreasing the size of her cancer may help improve her comfort for a longer period of time. Side effects appear to be mild, and the side effect rate appears to be low (5%).”
In summary, realistic, and positive communication helps to guide your client through uncertainty. This communication approach allows your client to make an informed decision for their pet and improves the psychological well-being of both you and your client.
Faulkner B. The psychological impact of uncertainty: Why a bigger IQ isn't the answer. Veterinarian's Money Digest 2017;1:20-21.
Stoewen DL, Coe JB, MacMartin C, et al. Identification of Illness Uncertainty in Veterinary Oncology: Implications for Service. Front Vet Sci 2019;6:147.
Kirkeboen G. "The median isn't the message": How to communicate the uncertainties of survival prognoses to cancer patients in a realistic and hopeful way. Eur J Cancer Care (Engl) 2019;28:e13056.
Evans LR, Boyd EA, Malvar G, et al. Surrogate decision-makers' perspectives on discussing prognosis in the face of uncertainty. Am J Respir Crit Care Med 2009;179:48-53.
Kiely BE, Soon YY, Tattersall MHN, et al. How Long Have I Got? Estimating Typical, Best-Case, and Worst-Case Scenarios for Patients Starting First-Line Chemotherapy for Metastatic Breast Cancer: A Systematic Review of Recent Randomized Trials. Journal of Clinical Oncology 2011;29:456-463.
Davey HM, Butow PN, Armstrong BK. Cancer patients' preferences for written prognostic information provided outside the clinical context. British Journal of Cancer 2003;89:1450-1456.