Effect of using pediatric emergency department virtual observation on inpatient admissions and lengths of stay

Objective To determine whether using emergency department (ED) virtual observation for select pediatric conditions decreases admission rates for these conditions, and to examine effects on length of stay. Methods The option of ED virtual observation care for 9 common pediatric conditions was introduced in 2009; associated order sets were developed. Retrospective secondary analyses of administrative data from our tertiary care pediatric ED and children's hospital were performed for the year before (year 0) and after (year 1) this disposition option was introduced. The proportion of visits admitted to the inpatient unit and length of stay (LOS) were determined for all visits considered eligible for ED virtual observation care on the basis of diagnosis codes for both study years. Results There were 1614 observation-eligible visits in year 0 and 1510 in year 1. In year 1, 18% (n = 266) of observation-eligible visits received ED virtual observation care. Admission rates for observation-eligible visits were similar after this model of care was introduced (25% year 0, 29% year 1, P =.02). Median LOS for ED virtual observation visits was 8.8 hours (interquartile range 6.5-12.4). ED LOS was shorter for ED discharges (5.6 hours year 0, 5.1 hours year 1, P <.001) and unchanged for admissions (6.0 hours year 0, 5.8 hours, year 1, P =.41) after introducing ED virtual observation. Conclusions Admission rates for observation-eligible visits were not lower in the year after ED virtual observation care was introduced. LOS decreased for ED discharges and was unchanged for admissions. Reevaluation of the effects of pediatric ED virtual observation on admission rates and LOS after longer periods of use is indicated.