The “Surprise Question” May Help Stimulate Palliative Care Discussions Interview with:

Joshua R. Lakin, MD Instructor in Medicine Harvard Medical School Dana Farber Cancer Institute

Dr. Joshua Lakin

Joshua R. Lakin, MD
Instructor in Medicine
Harvard Medical School
Dana Farber Cancer Institute What is the background for this study? What are the main findings?

Response: Research has increasingly shown the benefits of early palliative care interventions, especially in those around communication about patient goals and preferences in serious illness. These benefits include improved quality of life and psychological outcomes for patients as well as eased bereavement and decision making for loved ones. We have a large gap to fill in initiating early goals and values conversations with our patients and there are a myriad of systems failures and clinician barriers that do not allow us to do this work in a timely and effective way. Doing so with limited resources, both in specialty palliative care and in the many frontline clinicians doing this work, requires targeting our resources carefully.

Doing these conversations earlier means identifying patients upstream, before they are in the last days of life. The Surprise Question – “Would you be surprised if this patient died in the next year?” – has emerged as an attractive option for screening for early palliative care interventions. It has been studied primarily in dialysis and cancer patients and has been demonstrated to have a strong association with risk of death. We set out to test it in a more diverse primary care population. What should readers take away from your report?

Response: We studied the Surprise Question in a primary care population and found that, while it does help predict death, it missed a majority of patients who could have benefited from palliative care. Given prior literature showing that physicians are not great at prognosticating end of life, these results aren’t too surprising. However, I do find it interesting that this single question clinician screen did have a stronger predictive ability, in multivariate regression, than a tested composite comorbidity score. What recommendations do you have for future research as a result of this study?

Response: I think that the Surprise Question has two key uses when identifying patients for palliative care needs.

First, as a clinician screening tool that, I believe, should be used along with other predictive models to help locate people who may benefit from earlier palliative care interventions. I am looking forward to seeing more studies of the Surprise Question paired with other predictive tools aimed at minimizing false positive results – those patients who died that the Surprise Question missed.

Second, I think this question has significant value in activating clinicians. While it has yet to be studied or discussed much in the literature, to my knowledge, I believe that physicians saying, “No, I wouldn’t be surprised if this patient died” activates engagement in palliative care work for their patients. It helps doctors to move into a more contemplative state about what their patients might need as their illnesses advance. I think this is a possibly quite powerful effect that other algorithmic based screening tools lack and I would love to see research look into this as well – not only how the test functions as a screening tool but also its effect on following action. Is there anything else you would like to add?

Response: I think that the other key point is that this study focuses on prediction of death, but there are many other reasons why we should be thinking about earlier palliative care for patients – intense symptom burden from disease and our treatments, existential distress, suffering in the context of living with serious illness. Better vision of who might die from their illness is only one part of getting proven palliative care interventions to patients. This study is only focusing on one component of identifying patients to drive earlier and better palliative care. Thank you for your contribution to the community.


Note: Content is Not intended as medical advice. Please consult your health care provider regarding your specific medical condition and questions.

More Medical Research Interviews on

[wysija_form id=”5″]