The Relationship Between Remoteness and Outcomes in Critically Ill Children
A significant number of children live in remote geographic areas without direct access to tertiary care PICU. Our objective was to explore the relationship between remoteness and outcomes of critically ill children in Canada.Design:
Retrospective cohort study of patients admitted to the PICU from February 1, 2015, to January 31, 2016.Setting:
Pediatric tertiary care PICU in Canada.Patients:
All children admitted to PICU during the study period.Interventions:
NoneMeasurements and Main Results:
Four hundred fifty-five unique PICU admissions were included. One hundred sixty-nine patients were transported from another center of whom 28 lived in remote areas. For transported patients, remoteness (hazard ratio, 2.76, p < 0.001; hazard ratio, 2.22, p = 0.006), admission Pediatric Risk of Mortality (hazard ratio, 1.11; p = 0.001; hazard ratio, 1.05, p = 0.016), and transport by a noncritical care trained team (hazard ratio, 0.61, p = 0.021; hazard ratio, 0.66, p = 0.045) were associated with increased PICU and hospital lengths of stay, respectively. PICU mortality increased with duration of transport (odds ratio, 1.46; 95% CI, 1.09–1.97; p = 0.012). The odds of a remote-area patient being refused admission during the winter were significantly higher (odds ratio, 8.2; 95% CI, 3.0–22.3; p < 0.001) than a patient not requiring transport. Admission Pediatric Risk of Mortality score (4, interquartile range, 1–8 vs 2, interquartile range, 0–5; p = 0.001) and mortality rate (7.1%, 12/169 vs 0%, 0/286; p < 0.001) were significantly higher for transported than for nontransported patients.Conclusions:
Remoteness was associated with increased PICU and hospital length of stay, and duration of transport was associated with higher admission Pediatric Risk of Mortality (PRISM) scores and mortality rates. Patients requiring transport had a significantly higher PICU mortality rate than those presenting directly to a tertiary care center. Further studies are needed to explore potential policy and healthcare resource implications of these findings.