Objective: To examine associations between mortality, length of stay, and the sources of admission to tertiary pediatric intensive care.
Design: A retrospective analysis of prospectively collected data.
Setting: A tertiary medical center with a 16-bed medical-surgical intensive care unit and a 15-bed cardiac pediatric intensive care unit (PICU).
Patients: All admissions from July 1, 1998, through June 30, 2004. Multivariable regression methods compared length of stay and mortality between the sources of PICU admission, controlling for multiple variables, including severity of illness.
Measurements and Main Results: Of 8,897 eligible admissions, 74% were directly from the study hospital’s emergency department or operating rooms, while 26% were from indirect sources, including the study hospital’s wards (11%) or interhospital transfer from either non-PICU (12%) or PICU settings (3%). Compared with emergency department admissions, ward admissions had higher odds of mortality (odds ratio 1.65, 95% confidence interval 1.08–2.51), transfer admissions from non-PICU settings did not have elevated odds of mortality (odds ratio 0.80, 95% confidence interval 0.51–1.25), and inter-PICU transfer admissions had higher odds of mortality (odds ratio 1.43, 95% confidence interval 0.80–2.56), although not reaching statistical significance. Compared with emergency department admissions, ward admissions stayed almost 4 days longer in the PICU, while interhospital transfer admissions from non-PICU and PICU settings stayed 2 and 6 days longer, respectively.
Conclusions: Outcomes of tertiary pediatric intensive care vary significantly by source of admission. Strategies aimed at reduction of mortality at the tertiary PICU should target transfer admissions from the hospital’s wards and from PICUs of other hospitals.
From the Department of Pediatrics and Communicable Diseases, Division of Pediatric Critical Care Medicine (FOO, TPS) and Child Health Evaluation and Research Unit (FOO, MMD, SJC); Department of Anesthesiology and Critical Care (ALR); Department of Internal Medicine, Division of General Internal Medicine, and Gerald R. Ford School of Public Policy (MMD); and Critical Care Support Services (RED), University of Michigan Health System, Ann Arbor, MI.
*See also p. 118.
The authors have not disclosed any potential conflicts of interest.
For information regarding this article, E-mail: firstname.lastname@example.org