Background: This study aimed to characterize initial clinical presentations of patients served by emergency medical services (EMS) who die following injury, with particular attention to patients with occult ("talk-and-die") presentations.
Methods: This was a population-based, multiregion, mixed-methods retrospective cohort study of fatally injured children and adults evaluated by 94 EMS agencies transporting to 122 hospitals in seven Western US regions from 2006 to 2008. Fatalities were divided into two main groups: occult injuries (talk-and-die; Glasgow Coma Scale [GCS] score ≥ 13, no cardiopulmonary arrest, and no intubation) versus overt injuries (all other patients). These groups were further subdivided by timing of death: early (<48 hours) versus late (>48 hours). We then compared demographic, physiologic, procedural, and injury patterns using descriptive statistics. We also used qualitative methods to analyze available EMS chart narratives for contextual information from the out-of-hospital encounter.
Results: During the 3-year study period, 3,358 persons served by 9-1-1 EMS providers died, with 1,225 (37.1%) in the field, 1,016 (30.8%) early in the hospital, and 1,060 (32.1%) late in the hospital. Of the 2,133 patients transported to a hospital, there were 612 (28.7%) talk-and-die patients, of whom 114 (18.6%) died early. Talk-and-die patients were older (median age, 81 years; interquartile range, 67-87 years), normotensive (median systolic blood pressure, 138 mm Hg; interquartile range, 116-160 mm Hg), commonly injured by falls (71.3%), and frequently (52.4%) died in nontrauma hospitals. Compared with overtly injured patients, talk-and-die patients had relatively fewer serious head injuries (13.7%) but more frequent extremity injuries (20.3% vs. 10.6%) and orthopedic interventions (25.3% vs. 5.0%). EMS personnel often found talk-and-die patients lying on the ground with hip pain or extremity injuries.
Conclusion: Patients served by EMS who "talk-and-die" are typically older adults with falls, transported to nontrauma hospitals, with subtle clinical indications of the severity of their injuries. Improving recognition of talk-and-die patients may avoid fatal outcomes in a portion of these patients.
Level of evidence: Epidemiologic study, level III.