Decision aid did not help surrogates, clinicians agree on treatment goals for critically ill patients

Nearly all surrogates chose a more aggressive treatment goal for patients on life support than was suggested by a decision aid, despite the aid being based on surrogates' reports of patients' values.


A personalized, web-based decision aid did not help surrogate decision makers and clinicians agree on treatment goals for patients who were on extended life support, a study found.

Researchers compared a theory-based decision aid that was individualized to the clinical condition of each critically ill patient to usual care using a blinded, randomized trial in 13 medical and surgical ICUs at five hospitals. Family members using the decision aid were guided to think about what their loved ones valued most and which ICU goal of care was best aligned with these values. The decision aid provided personalized prognostic estimates and explanations of treatment options and interactively clarified patient values to inform a family meeting. The control group received information according to usual care practices followed by a family meeting.

The study's primary outcome was improved concordance on one-year survival estimates, measured with the clinician-surrogate concordance scale (range, 0 to 100 percentage points, with higher scores indicating more discordance). Secondary and additional outcomes assessed the experiences of surrogates (psychological distress, decisional conflict, and quality of communication) and patient outcomes (length of stay and six-month mortality rates). The study enrolled 277 patients, 416 surrogates, and 427 clinicians.

Results were published by Annals of Internal Medicine on Jan. 29.

The researchers found that although the decision aid reduced surrogates' decisional conflict and improved their understanding of physicians' prognostic beliefs, it did not change clinician-surrogate concordance about one-year survival estimates, surrogates' distress symptoms, patients' clinical outcomes, or the actual decisions made.

Concordance improvement did not differ between intervention and control groups (mean difference in score change from baseline, −1.7 percentage points [95% CI, −8.3 to 4.8 percentage points]; P=0.60). Surrogates' postintervention estimates of patients' one-year prognoses did not differ between intervention and control groups (median, 86.0% vs. 92.5%; P=0.23) and were substantially more optimistic than results of a validated prediction model (median, 56.0%) and physician estimates (median, 50.0%).

Eighty-two surrogates in the intervention group (43%) favored a treatment option that was more aggressive than their report of patient preferences. Although surrogates in the intervention group had greater reduction in decisional conflict than those in the control group (mean difference in change from baseline, 0.4 points [95% CI, CI, 0.0 to 0.7 points]; P=0.041), other surrogate and patient outcomes did not differ.

According to the authors, the results of their study challenges the idea that decision aids can easily change resource utilization in acute care settings. Instead, they wrote, decision support in acute care may require more individualized attention.

An accompanying editorial noted that surrogates' tendency to override predictions indicates how they process information from physicians and construct their own individual preferences. While seriously ill patients may prefer better function over longer life, surrogates may struggle to acknowledge this, the editorialists said, and decision aids that provide more information and encourage more deliberation may not help.

“Indeed, the value of deliberation in clinical decision making is itself unsubstantiated,” the editorialists wrote. “Inducing deliberation rather than allowing intuitive judgments often leads to lower-quality decisions and may serve primarily to increase the decision maker's confidence in the decision, a form of confirmation bias.”