Current issues of ACP Journal Club are published in Annals of Internal Medicine


A communication strategy and brochure reduced the burden of bereavement on relatives of patients dying in the intensive care unit


ACP J Club. 2007 May-Jun;146:69. doi:10.7326/ACPJC-2007-146-3-069

Clinical Impact Ratings

Mental Health: 5 stars

Hospitalists: 4 stars

Critical Care: 7 stars

Source Citation

Lautrette A, Darmon M, Megarbane B, et al. A communication strategy and brochure for relatives of patients dying in the ICU. N Engl J Med. 2007;356:469-78. [PubMed ID: 17267907]



Does an end-of-life family conference held according to specific guidelines reduce the burden of bereavement on relatives of patients dying in the intensive care unit (ICU)?


Design: Randomized controlled trial.

Allocation: Unclear concealment.*

Blinding: Unblinded.*

Follow-up period: 90 days.

Setting: 22 ICUs in France.

Patients: 126 patients ≥ 18 years of age (median age range 68 to 74 y, 56% men) in the ICU who were expected to die within the next few days. Each patient had 1 family member (median age 54 y, 73% women) who participated in the interview at 90 days after the patient's death. The family member was either designated by the patient or ranked highest in the hierarchy for surrogate decision-making, according to French law.

Intervention: Family members were allocated to a proactive communication intervention consisting of a structured family end-of-life conference and a 15-page bereavement information brochure (n = 63) or to standard end-of-life communication (n = 63). The interventional conference had 5 objectives derived from VALUE-based guidelines for end-of-life conferences: value and appreciate what family members said, acknowledge family members' emotions, listen, understand who the patient was as a person, and elicit questions from family members.

Outcomes: Score assessing symptoms related to posttraumatic stress disorder (PTSD) on the Impact of Event Scale (IES) (range 0 [no symptoms] to 75 [severe symptoms]). A threshold of 30 was used to determine high and low scores. Secondary outcomes were symptoms of anxiety and depression using the Hospital Anxiety and Depression Scale (HADS) (range 0 [no distress] to 21 [severe distress]). A threshold of 8 was used to determine clinically significant and nonsignificant symptoms of anxiety and depression.

Patient follow-up: 86% of family members.

Main results

Family members in the proactive communication group had lower IES scores (median score 27 vs 39, P = 0.02) and lower HADS scores (median score 11 vs 17, P = 0.004) than did those in the standard communication group (Table).


A proactive communication strategy involving a VALUE-based end-of-life family conference plus a bereavement brochure reduced symptoms of posttraumatic stress disorder, anxiety, and depression in relatives of patients dying in the ICU.

*See Glossary.

Sources of funding: Assistance Publique-Hôpitaux de Paris and French Society for Critical Care Medicine.

For correspondence: Dr. E. Azoulay, Service de Réanimation Médicale, Hôpital Saint-Louis, Paris, France. E-mail

Table. Proactive communication plus brochure vs standard end-of-life communication for relatives of patients who died in the ICU†

Outcomes at 90 d Proactive strategy Standard communication RRR (95% CI) NNT (CI)
PTSD-related symptoms (IES score > 30)‡ 45% 69% 36% (9.9 to 55) 5 (3 to 18)
Anxiety symptoms (HADS score > 8)§ 45% 67% 34% (6.8 to 54) 5 (3 to 27)
Depression symptoms (HADS score > 8)§ 29% 56% 49% (19 to 69) 4 (3 to 12)

†PTSD = posttraumatic stress disorder; IES = Impact of Event Scale; HADS = Hospital Anxiety and Depression Scale. Other abbreviations defined in Glossary; RRR, NNT, and CI calculated from data in article.
‡IES score range 0 (no PTSD-related symptoms) to 75 (severe PTSD-related symptoms).
§HADS score range 0 (no distress) to 21 (severe distress).


Although life is said to prepare us for death (1), research into end-of-life care is still in its infancy. Many persons die in hospital despite technological support; thus, the ICU setting lends itself to randomized trials of strategies to improve care of the dying.

In this multicenter trial, Lautrette and colleagues randomized critically ill patients expected to die, 90% of whom were mechanically ventilated. Family members of these patients received either usual end-of-life care or a bereavement brochure and a structured, VALUE-based family conference (lasting a median 10 min longer than standard conferences in the control group). The primary outcome of PTSD symptoms and secondary outcomes of anxiety and depression among relatives were all reduced in the intervention group. Patients in the intervention group also received fewer nonbeneficial supports.

Families in both groups had 3 meetings with the ICU team before randomization. This was standard practice in participating ICUs, making it unlikely that sparse communication in the control group explains the positive effects of this trial. VALUE-based conferences and brochures were not used before the trial. In other fields, dissemination of written material is generally not as effective as active strategies in improving processes of care. However, as with most multifaceted interventions, the interaction and relative contribution of each facet is uncertain. Candidate interventions to improve the handling of a family member's death should be selected based on preexisting evidence of potential effect, family-centered as well as patient-centered care, and religious and cultural context. This trial marks the beginning of a new phase of research in which end-of-life interventions will be rigorously analyzed in randomized trials, not just for potential benefit, but also for unanticipated adverse effects (2).

Deborah Cook, MD
McMaster University
Hamilton, Ontario, Canada

Graeme Rocker, MD
Dalhousie University
Halifax, Nova Scotia, Canada


1. Smith R. A good death. An important aim for health services and for us all [Editorial]. BMJ. 2000;320:129-30. [PubMed ID: 10634711]

2. Ventres W, Nichter M, Reed R, Frankel R. Limitation of medical care: an ethnographic analysis. J Clin Ethics. 1993;4:134-45. [PubMed ID: 8334277]