Introduction: A detailed understanding of patient demographics, clinical course and outcomes is crucial to optimize care delivery in the ICU. One group that is poorly described are patients who die within 24 hours of ICU admission.
Objectives: Our primary aim was to determine the proportion of ICU patients who die within 24 hours. Our secondary aims were to describe the clinical characteristics of this population and factors associated with death within 24 hours.
Methods: We conducted a retrospective population-based cohort and nested case control study utilizing data from a provincial clinical information system (eCritical Alberta). Cases were patients ≥18 years of age who died within 24 hours of admission to any of the 17 adult ICUs in Alberta between January 1, 2016 and June 30, 2017. Controls were ICU patients who remained alive at 24 hours. Information on demographics, illness severity, diagnosis, and frailty status were obtained. Additional system-level data captured included day and time of admission, location prior to admission, duration of antecedent hospitalization and details of life sustaining treatments received. Variables of clinical interest were identified a priori, and odds ratios calculated using logistic regression analysis.
Results: Of 19,556 patients admitted to ICU in an 18-month period, 649 (3.3%) died within 24 hours of admission, representing 29.8% of ICU deaths. The mean (SD) age was 62.0 (17.2) years, 39.6% were female, mean (SD) APACHE II score was 34.3 (8.4) and 35.6% were frail (Clinical Frailty Scale [CFS] score ≥5). Most were medical admissions (n=417, 64.3%) and median (IQR) ICU stay prior to death was 10 (4.3-16.8) hours. Most received invasive ventilation (n=514, 79.2%) and at least one vasoactive medication (n=545, 84.0%) whereas fewer received non-invasive ventilation and/or renal replacement therapy (5.1% and 7.9%, respectively). Factors associated with death within 24 hours were admission from the Emergency Department (n=319, 49%; adjusted-OR 1.5, 95% CI, 1.1-1.9) and neurologic (adjusted-OR 4.6, 95% CI, 3.1-6.9) or trauma diagnosis (adjusted-OR 6.1, 95% CI, 2.4-15.6). Admission time, day of the week, duration of hospitalization prior to admission and CFS were not associated with death within 24 hours.
Conclusion: Patients who die within 24 hours constitute one third of all ICU deaths. They tend to be middle-aged and have high acuity of illness. Most are admitted from the Emergency Department. Neurologic diagnoses and trauma are most strongly associated with early death. Further studies are needed to determine what proportion of these deaths are expected and how clinician and patient/family decision-making may influence outcomes in this group.
Introduction: A detailed understanding of patient demographics, clinical course and outcomes is crucial to optimize care delivery in the ICU. One group that is poorly described are patients who die within 24 hours of ICU admission.
Objectives: Our primary aim was to determine the proportion of ICU patients who die within 24 hours. Our secondary aims were to describe the clinical characteristics of this population and factors associated with death within 24 hours.
Methods: We conducted a retrospective population-based cohort and nested case control study utilizing data from a provincial clinical information system (eCritical Alberta). Cases were patients ≥18 years of age who died within 24 hours of admission to any of the 17 adult ICUs in Alberta between January 1, 2016 and June 30, 2017. Controls were ICU patients who remained alive at 24 hours. Information on demographics, illness severity, diagnosis, and frailty status were obtained. Additional system-level data captured included day and time of admission, location prior to admission, duration of antecedent hospitalization and details of life sustaining treatments received. Variables of clinical interest were identified a priori, and odds ratios calculated using logistic regression analysis.
Results: Of 19,556 patients admitted to ICU in an 18-month period, 649 (3.3%) died within 24 hours of admission, representing 29.8% of ICU deaths. The mean (SD) age was 62.0 (17.2) years, 39.6% were female, mean (SD) APACHE II score was 34.3 (8.4) and 35.6% were frail (Clinical Frailty Scale [CFS] score ≥5). Most were medical admissions (n=417, 64.3%) and median (IQR) ICU stay prior to death was 10 (4.3-16.8) hours. Most received invasive ventilation (n=514, 79.2%) and at least one vasoactive medication (n=545, 84.0%) whereas fewer received non-invasive ventilation and/or renal replacement therapy (5.1% and 7.9%, respectively). Factors associated with death within 24 hours were admission from the Emergency Department (n=319, 49%; adjusted-OR 1.5, 95% CI, 1.1-1.9) and neurologic (adjusted-OR 4.6, 95% CI, 3.1-6.9) or trauma diagnosis (adjusted-OR 6.1, 95% CI, 2.4-15.6). Admission time, day of the week, duration of hospitalization prior to admission and CFS were not associated with death within 24 hours.
Conclusion: Patients who die within 24 hours constitute one third of all ICU deaths. They tend to be middle-aged and have high acuity of illness. Most are admitted from the Emergency Department. Neurologic diagnoses and trauma are most strongly associated with early death. Further studies are needed to determine what proportion of these deaths are expected and how clinician and patient/family decision-making may influence outcomes in this group.