Training clinicians to deliver devastating news in emergency departments may help staff feel more confident. But a broad review of the available research found a major gap: no training evaluation measured what happened to patients or relatives afterward.
The 2026 scoping review examined 156 publications about communicating serious diagnoses, terminal illness and patient deaths in emergency settings. It found that the field has produced more guidance and training reports than rigorous studies of real conversations and their effects on families.
Confidence was the most common measure
Researchers identified 36 evaluations of bad-news training. Of these, 23 measured staff confidence or comfort, mostly through self-report questionnaires. Fourteen reported improvement in at least one outcome.
Only five evaluations assessed communication skills through observers, and none measured outcomes for patients or relatives. No training evaluation observed actual emergency-department conversations.
That distinction matters. Feeling more comfortable may be useful for a clinician facing a difficult conversation, but confidence is not the same as communication skill. It also cannot show whether a family understood what happened, felt supported or experienced fewer longer-term psychological effects.
Simulation appeared in 23 training evaluations. Nineteen reported positive results, two were neutral and two provided no outcome data. However, simulated conversations cannot fully reproduce the time pressure, uncertainty, interruptions and emotional intensity of an emergency death notification.
A literature dominated by guidance and education
Of the 156 publications, only 60 were empirical studies. The rest were mainly training-program reports and educational articles. Among the empirical work, surveys and interviews were the most common methods.
The review team searched four medical databases in July 2025, along with gray literature. It found 7,985 database records and 285 additional records before identifying the 156 publications that met its criteria.
The review did not rate study quality or combine results into a single estimate. It therefore maps the research landscape but cannot establish which communication approach works best.
Most publications came from the United States and the United Kingdom, although the review covered work from 26 countries. The findings may not transfer directly to emergency systems organized differently. The review also included only English-language publications.
Emergency departments need approaches designed for emergencies
SPIKES was the most frequently cited of 25 communication frameworks, with 37 mentions. Developed in 2000 for delivering cancer diagnoses, it remains influential, but the review found no study showing whether it is routinely used in emergency care or improves outcomes for patients or relatives there.
Only three frameworks identified in the review were developed specifically for emergency departments: GRIEV_ING, 8S and EMTalk.
A 2005 evaluation of GRIEV_ING found that a two-hour session combining small-group teaching and role-play improved residents’ confidence and competency. Communication scores, however, did not change significantly.
That result illustrates why training research needs more than one measure. A program may improve how prepared clinicians feel without demonstrating a measurable change in observed communication.
What families say they need
Across the literature, patients and relatives most often emphasized four features: honesty, clear language without jargon, a clinician who stays with them and emotional support.
Reported barriers included time pressure, interruptions, noise, a lack of privacy, insufficient training and clinical situations that were sudden or uncertain. These conditions can make it difficult to apply a structured framework exactly as taught.
Some related studies suggest that the way families are supported during critical events can matter. In a randomized trial across 22 French intensive-care units, relatives of 126 dying patients assigned to a longer end-of-life conference with more time to talk and a bereavement brochure reported fewer PTSD-related symptoms, anxiety and depression at 90 days than relatives receiving the usual conference.
A 2013 trial involving 570 relatives across 15 French emergency medical service units also found fewer PTSD-related symptoms when families were routinely offered the chance to watch resuscitation. These studies examined family support and involvement, not the delivery of bad news itself, and the review states that emergency-department research has not yet measured those outcomes.
What the next research should test
The review identifies three priorities: observe real emergency-department conversations, develop and assess frameworks designed for emergency care, and measure patient and relative outcomes alongside staff performance.
Future studies will also need to account for the possibility that positive training findings are more likely to be published than disappointing results. For now, the evidence supports training as a promising area, but not a conclusion that greater staff confidence necessarily leads to better experiences or psychological outcomes for families.
AI tools were used to assist with the preparation of this article.
Leave a Reply