Incidents of sudden death during restraint of agitated individuals in Sweden between 1992 and 2024

Abstract

Restraint-related sudden deaths in agitated individuals raise complex questions at the intersection of medicine and law. Hyperactive delirium with extreme agitation as well as positional asphyxia due to restraint have been proposed to account for these deaths. However, the exact physiological mechanisms responsible and to what extent restraint contributes to the lethal outcome remain debated. In this nationwide, 32-year retrospective study between 1992 and 2024, we examined circumstances surrounding sudden deaths during restraint of agitated individuals in Sweden. A total of 52 cases were identified, with an average of 0.17 deaths per million inhabitants annually. Ninety percent of cases involved prone restraint and 69% showed evidence of stimulant use. In 15 cases from 2005 onward, peri-arrest arterial blood gas data revealed profound metabolic and respiratory acidosis, with a mean blood pH of 6.52 (range: 6.30–6.95; median: 6.50), mean lactate concentration of 26.3 mmol/L (range: 8.6–41.0; median: 30), and mean pCO₂ of 14.8 kPa (range: 6.4–22.3; median: 15.3). Based on these findings, we propose a two-phase pathophysiological model of restraint-related cardiac arrest. The initial “priming phase” involves extreme physical exertion, creating a critically acidotic state that requires full respiratory and cardiovascular function to maintain homeostasis. If the “priming phase” is followed by restraint that restricts ventilatory function and hampers venous return, e.g., restraint in the prone position, an unstable “tipping phase” is initiated, that may culminate in cardiac arrest. This model builds on previous hypotheses and emphasizes the potentially lethal consequences of inhibiting ventilatory function in acutely agitated individuals.