Delayed Admissions to the Pediatric Intensive Care Unit: Progression of Disease or Errors in Emergency Department Management.

ABSTRACT

OBJECTIVES: Few studies have evaluated impact of emergency department (ED) management on delayed transfers to the pediatric intensive care unit (PICU). Our study objectives were to describe patient characteristics of PICU transfers less than or equal to 12 hours of admission and determine the reason for transfer. METHODS: We conducted a retrospective chart review of patients transferred to PICU less than or equal to 12 hours of admission. We extracted patient demographics, emergency severity index category, ED, floor and PICU length of stay (LOS), and PICU "significant" interventions. Charts were reviewed independently by the study principal investigator and a PICU attending who classified transfers as secondary to progression of disease or error in ED management. Furthermore, errors were classified as diagnostic, management, or disposition errors. RESULTS: A total of 164 patients met inclusion criteria. Most were male (86/164, 52.4%), with emergency severity index category 2 (116/164, 70.7%) and respiratory diagnosis (98/164, 59.8%). Most transfers (136/164, 82.9%) resulted from progression of illness. No significant interventions were performed in 48.8% (80/164) of patients. Of 164 transfers, 28 (17.1%) resulted from ED error, and half of these were management errors. Compared with disease progression, the ED error group had a significantly shorter median floor LOS {3.45 [interquartile range (IQR) 2.15, 7.56] vs 6.58 (IQR 3.70, 9.20); P = 0.005}, more PICU interventions [1.5 (IQR 0, 4) vs 0 (IQR 0, 2); P = 0.006], and longer PICU LOS [2.50 (IQR 1.09, 4.25) vs 1.36 (IQR 0.80, 2.50); P = 0.013]. CONCLUSIONS: Most PICU transfers less than or equal to 12 hours after admission result from illness progression. Half of these do not require significant interventions. The PICU transfers after ED management error had significantly shorter floor LOS, longer PICU LOS, and more interventions.