Pediatric observation status: Are we overlooking a growing population in children's hospitals?

Background: Inpatient administrative datasets often exclude observation stays, as observation is considered to be outpatient care. The extent to which this status is applied to pediatric hospitalizations is not known. Objective: To characterize trends in observation status code utilization and 1-day stays among children admitted from the emergency department (ED), and to compare patient characteristics and outcomes associated with observation versus inpatient stays. Design: Retrospective longitudinal analysis of the 2004-2009 Pediatric Health Information System (PHIS). SETTING: Sixteen US freestanding children's hospitals contributing outpatient and inpatient data to PHIS. Patients: Admissions to observation or inpatient status following ED care in study hospitals. Measurements: Proportions of observation and 1-day stays among all admissions from the ED were calculated each year. Top ranking discharge diagnoses and outcomes of observation were determined. Patient characteristics, discharge diagnoses, and return visits were compared for observation and 1-day stays. Results: The proportion of short-stays (including both observation and 1-day stays) increased from 37% to 41% between 2004 and 2009. Since 2007, observation stays have outnumbered 1-day stays. In 2009, more than half of admissions from the ED for 6 of the top 10 ranking discharge diagnoses were short-stays. Fewer than 25% of observation stays converted to inpatient status. Return visits and readmissions following observation were no more frequent than following 1-day stays. Conclusions: Children admitted under observation status make up a substantial proportion of acute care hospitalizations. Analyses of inpatient administrative databases that exclude observation stays likely result in an underestimation of hospital resource utilization for children.