Family presence during cardiopulmonary resuscitation

Jabre P, Belpomme V, et al
N Engl J Med. 2013 Mar 14;368(11):1008-18


This is the first study to ever methodically examine of the effect of our actions on the families of cardiac arrest victims. This is a prospective cluster-randomized control study that wanted to see if actively encouraging the family to be present for the resuscitation of their loved one would be helpful to reduce symptoms of PTSD and depression at 90 days.

  • Studied the effect of our actions on surviving family members of cardiac arrest victims

  • French ambulances contain a physician and a nurse

  • PTSD and Depression of the family members were measured at 90 days

This study was done in France and the ambulance crew consisted of a physician, nurse, and a driver. There were fifteen ambulance units that were assigned to either the intervention group or usual care. The intervention included having one of the team members systematically ask the family members if they would like to be present during the resuscitation.
The family members were contacted by phone 90 days after the initial incident and were asked about PTSD and depressive symptoms using two validated scales (Impact of Event Scale and the Hospital Anxiety and Depression Scale).
An intent-to-treat analysis was performed on the outcomes. Keep in mind that not everyone who was offered a viewing option did so. Also, some of the usual care group also did view the resuscitation.

The frequency of PTSD was significantly higher than the control group. (OR 1.7; CI 1.2 to 2.5, P=0.004)
The frequency of PTSD was significantly higher among those who did not witness the arrest as compared to those that did. (OR 1.6; CI 1.1 to 2.5 p=0.02)
  • When offered, more family members observed resuscitation (79% vs 43%).

  • Family Members were 60-70% less likely to suffer symptoms of PTSD if they witnessed the cardiac arrest.

  • There were similar lower rates of anxiety and depression among those who witnessed the cardiac arrest.

There was difference in the functional outcomes of the cardiac arrests or the stress that the health care providers experienced.

Wonky Research Points

Cluster randomized controlled study
Most randomized studies are randomized by the individual patient. Once a patient is identified, then randomization occurs. This is considered superior to other types of randomization. Cluster randomization means that some other type of organization, geographical etc. is used. In this study, half of the ambulances were randomized to intervention group. Thus one set of ambulances always enrolled family members into this study while a comparable group of other ambulances did not.

Intent to Treat Analysis

Certain treatments don’t always stick or work. For example, a particular antibiotic might work well but cause 25% of the patients to vomit the active drug. An interested pharmaceutical company would like to remove these 25% from analysis and demonstrate better outcomes. But most current studies would use an intent-to-treat analysis and since we intended to treat those patients that vomited, then they would be included.
This study offered the option to view the resuscitation in the intervention group but not all of the family members took them up on the offer. This intent-to-treat analysis demonstrated an effect on rates of PTSD suggesting that it was a very robust difference.

Good Points about this study

  1. Prespecitve, randomized

  2. Use of validated scales for measuring of PTSD and Depression

  3. Intent-to-treat analysis

Problems with this study

  1. Cluster Randomization

  2. Is this generalizable to a paramedic system? (remember there were MD’s an a nurse on the ambulance.)

  3. Non validated scale to measure stress of the medical team.

Encouraging family members to be present for the resuscitation will decrease their PTSD and depression symptoms.

Dr. Sporer is the EMS Medical Director for Alameda County

Tagged with:

Leave a Reply