Background
Patients and their relatives often expect to be actively involved in decisions of treatment. Even during resuscitation and acute medical care, patients may want to have their relatives nearby, and relatives may want to be present if offered the possibility. The principle of family presence during resuscitation (FPDR) is a triangular relationship where the intervention of family presence affects the healthcare professionals, the relatives present, and the care of the patient involved. All needs and well‐being must be balanced in the context of FPDR as the actions involving all three groups can impact the others.

Objectives
The primary aim of this review was to investigate how offering relatives the option to be present during resuscitation of patients affects the occurrence of post‐traumatic stress disorder (PTSD)‐related symptoms in the relatives.

The secondary aim was to investigate how offering relatives the option to be present during resuscitation of patients affects the occurrence of other psychological outcomes in the relatives and what effect family presence compared to no family presence during resuscitation of patients has on patient morbidity and mortality.

We also wanted to investigate the effect of FPDR on medical treatment and care during resuscitation. Furthermore, we wanted to investigate and report the personal stress seen in healthcare professionals and if possible describe their attitudes toward the FPDR initiative.

Search methods
We searched CENTRAL, MEDLINE, Embase, PsycINFO, and CINAHL from inception to 22 March 2022 without any language limits. We also checked references and citations of eligible studies using Scopus, and searched for relevant systematic reviews using Epistomonikos. Furthermore, we searched ClinicalTrials.gov, WHO ICTRP, and ISRCTN registry for ongoing trials; OpenGrey for grey literature; and Google Scholar for additional trials (all on 22 March 2022).

Selection criteria
We included randomized controlled trials of adults who have witnessed a resuscitation attempt of a patient (who was their relative) at the emergency department or in the pre‐hospital emergency medical service.

The participants of this review included relatives, patients, and healthcare professionals during resuscitation.

We included relatives aged 18 years or older who have witnessed a resuscitation attempt of a patient (who is their relative) in the emergency department or pre‐hospital. We defined relatives as siblings, parents, spouses, children, or close friends of the patient, or any other descriptions used by the study authors. There were no limitations on adult age or gender.

We defined patient as a patient with cardiac arrest in need of cardiopulmonary resuscitation (CPR), a patient with a critical medical or traumatic life‐threatening condition, an unconscious patient, or a patient in any other way at risk of sudden death. We included all types of healthcare professionals as described in the included studies. There were no limitations on age or gender.

Data collection and analysis
We checked titles and abstracts of studies identified by the search, and obtained the full reports of those studies deemed potentially relevant. Two review authors independently extracted data. As it was not possible to conduct meta‐analyses, we synthesized data narratively.

Main results
The electronic searches yielded a total of 7292 records after deduplication. We included 2 trials (3 papers) involving a total of 595 participants: a cluster‐randomized trial from 2013 involving pre‐hospital emergency medical services units in France, comparing systematic offer for a relative to witness CPR with the traditional practice, and its 1‐year assessment; and a small pilot study from 1998 of FPDR in an emergency department in the UK.

Participants were 19 to 78 years old, and between 56% and 64% were women. PTSD was measured with the Impact of Event Scale, and the median score ranged from 0 to 21 (range 0 to 75; higher scores correspond to more severe disease). In the trial that accounted for most of the included participants (570/595), the frequency of PTSD‐related symptoms was significantly higher in the control group after 3 and 12 months, and in the per‐protocol analyses a significant statistical difference was found in favor of FPDR when looking at PTSD, anxiety and depression, and complicated grief after 1 year. One of the included studies also measured duration of patient resuscitation and personal stress in healthcare professionals during FPDR and found no difference between groups. Both studies had high risk of bias, and the evidence for all outcomes except one was assessed as very low certainty.

Authors' conclusions
There was insufficient evidence to draw any firm conclusions on the effects of FPDR on relatives' psychological outcomes.

Sufficiently powered and well‐designed randomized controlled trials may change the conclusions of the review in future.