Design: Multicentre randomised controlled trial
Population: 449 seriously ill ICU patients and 505 family members at three Seattle hospitals
Primary outcome: Family HADS-depression score over 6 months
Effect: Difference −0.07 (95% CI −0.70 to 0.55)

This randomised trial, published in July 2026, enrolled 449 ICU patients at three Seattle hospitals with chronic life-limiting illness or acute illness carrying more than 15% hospital mortality risk, and 505 family members. Nurse facilitators trained in communication supported families in the ICU and for three months afterwards, aiming to help care match the patient's values.

Family depression scores on the Hospital Anxiety and Depression Scale fell over time in both groups, from 7.2 at baseline to 5.9 at six months, with no difference between them (difference −0.07, 95% CI −0.70 to 0.55). Families in the intervention group rated their relationship with the team slightly better, and goal-concordant care was borderline, both with small effects.

Communication matters, but this specific model did not ease depression. Distress after an ICU stay may need more than better conversations.

Hold structured family meetings early in any ICU admission with high risk of death or disability
Ask families what the patient would value, not only what they want done
Screen family members for depression and anxiety and signpost support after discharge or death
Do not expect a communication programme alone to prevent family distress

Why it matters

It tempers expectations for communication-only interventions in a field that has hoped for a lot from them.

The statistics, in plain English

A difference of −0.07 points on a 21-point scale, with an interval from −0.70 to 0.55, is effectively zero. The secondary results reached significance but were small, and with several outcomes tested some will do so by chance.