Manner of Death Certification After Significant Emotional Stress: An Inter-Rater Variability Study and Review of the Literature.

Not commonly seen, the risk of sudden death after significant emotional stress has been reported since antiquity and incorporated into folk wisdom, reflected by phrases such as "scared to death" and "broken heart." A typical "victim" suffers from significant and often life-threatening natural diseases, making determination of the manner of death complicated, and at times controversial. The present study is designed to assess inter-rater variability and nonuniformity and controversy seen in manner of death certification in certain cases of death with significant stress involved in the circumstances of death. Members of the National Association of Medical Examiner (NAME) were surveyed to assess differences in manner of death certification for eight sudden unexpected death scenarios in middle-aged men and women with underlying cardiac disease after significant stressful events including: being chased down a lonely road followed by a verbal confrontation, a roll-over motor vehicle collision (MVC) without injuries, a fall from a wheelchair in a MVC, an alleged armed robbery, an involuntary commitment, an arrest by police, sexual intercourse, and a severe panic attack with breathing problems. In all cases, the autopsy examination revealed hypertensive and arteriosclerotic cardiovascular disease. In all cases, natural diseases were identified as significant contributing condtions, including emphysema in six cases and diabetes mellitus in three. Eighty-six responses were collected. The results show wide inter-rater variability, ranging from very good to poor (Kappa ranges from 0.16 to 0.94). One hundred fifty-five comments were collected. Most of the comments addressed more than one topic and were followed by discussions, open questions, and responders' experiences with previous cases and legal proceedings. Our data show that cases of sudden death after significant stress have almost complete agreement in four cases and very high inter-rater variability in the other four. We propose that a detailed analysis of each case and an algorithmic approach could improve the predictability of the outcomes of death investigations for the legal system and for families.