By Eric Walter, MD, MSc

Pulmonary and Critical Care Medicine, Northwest Permanente and Kaiser Sunnyside Medical Center, Portland

Dr. Walter reports no financial relationships relevant to this field of study.

SYNOPSIS: Family members, clinicians, and experts agree on many key themes regarding how prognostic information should be conveyed to patient surrogates in the ICU. However, important differences of opinion exist.

SOURCE: Anderson WG, et al. A multicenter study of key stakeholders’ perspectives on communicating with surrogates about prognosis in ICUs. ANNALSATS, 2014; [Epub ahead of print]

Effectively communicating a prognosis to a patient’s surrogates in the ICU can be challenging. Clinical guidelines exist but are based on low-level evidence and expert opinion. Family and friends, members of the care team, and others may have different perspectives on how prognosis should be conveyed to surrogates. In this article, Anderson and colleagues solicited input from these groups as well as “experts” in an effort to understand how participants thought prognosis should be communicated to surrogate decision makers.

The authors performed semi-structured interviews with surrogate-decision makers for adult ICU patients at high risk of death or severe disability. All surrogates had already had at least one discussion regarding prognosis with the patients’ physicians. The authors also interviewed the patients’ physicians and non-physician clinicians (nurses, social workers, and chaplains). To be eligible, physicians had to have already had a prognostic discussion with an eligible surrogate. Non-physician clinicians had to have witnessed or participated in such as discussion. Finally, they also interviewed people they considered to be experts in fields such as health communications, ethics, geriatrics, health care disparities, palliative care, psychology, critical care, and others.

One-hundred forty-two interviews were conducted with 118 participants. Most participants were surrogates (47). There were 17 physicians, 28 non-physician clinicians, and 26 experts. There was agreement between participants from all groups with respect to the importance of a truthful prognosis, showing compassion, tailoring the discussion depending on family needs, and checking for understanding. These themes are all part of current published guidelines regarding how prognosis should be communicated. In addition to these topics, some new themes arose from these interviews. Nearly all participants (96%) discussed the importance of ensuring that all members of the treatment team were aware of, and conveying, the same prognosis. Helping families see the prognosis for themselves was discussed by 71% of participants (79% of surrogates). The use of drawings, pictures, or sharing radiographs was recommended and bedside discussions of the treatments patients were receiving was described as helpful. However, families often felt uncomfortable discussing prognosis at the bedside where patients may overhear the discussion.

The authors did find some themes where different groups disagreed. Most participants agreed that an early discussion of the possibility of a poor outcome was important. However, only 43% of surrogates agreed that the prognosis should be discussed regularly over the course of the ICU stay. This contrasted with > 75% of participants in the other groups. Also, most surrogates, non-physician clinicians, and experts supported the use of numeric estimates of prognosis, while only 12% of physicians supported this idea.


Communicating a poor prognosis is something all of us do every day in the ICU. It a skill, not unlike placing a central line, honed over years of practice. However, procedural skills improve with the benefit of nearly instant feedback, whereas we rarely receive feedback on how well we are communicating a poor prognosis. Anderson and colleagues should be commended for helping to describe some of this feedback by eliciting recommendations from surrogates and non-physician clinicians in this study.

A strength of the study was the establishment of new key themes agreed upon by all stakeholders as elements of effective prognostic communication. As important, however, was the identification of themes that were not agreed upon as being effective communication (i.e., the use of numeric estimates).

A limitation of this study was generalizability. These results came from a relatively small sample of participants drawn from three large academic medical centers. It is not clear if these results are generalizable to prognostic discussions in other ICU settings. Furthermore, these medical centers have all previously led research on ICU communication. It seems very likely that many of the themes discussed by physicians and clinicians at these hospitals were based on results of previous research at that institution. For this reason, it is not surprising that there was such widespread support for published practice guidelines. Another potential limitation was the inclusion of national experts. A significant proportion of participants interviewed (22%) were experts. Since one motivation for the study was that many current guidelines are based on expert opinion only, it seems somewhat circular to have solicited expert opinions in this study.

Despite some limitations, results from this study should be welcomed. They add to our understanding of the many layers of effective prognostic communication. Future studies should examine these discovered themes in a more prospective manner and in a wider patient population.