Investigating a pediatric hospital's response to an inpatient census surge during the 2009 H1N1 influenza pandemic

On November 4, 2009, the 250-bed Seattle Children's Hospital (SCH) identified a surge in its census--245 inpatients, well above the average midnight census of 207. In response, SCH activated its pandemic influenza surge plan in an effort to decrease the inpatient census. Within 16 hours, 51 patients (20.4% of total bed capacity) had been discharged, and inpatient census at SCH decreased to 222 patients.

As part of a quality improvement project, SCH's response to the surge was investigated, with data drawn from interviews, a review of records created in the course of the surge plan implementation, an e-mail survey of attending physicians responsible for patient discharges, and models examining predictors of hospital discharges.

Analysis of three years of hospital data (2007-2009) indicated that the high census on November 4 was an uncommon but not unprecedented occurrence. In addition, there was a clear positive association between an evening's census and the number of discharges during the following 24 hours. SCH discharged essentially the same number of patients on November 4 as on previous high-census days when the surge plan was not activated, suggesting that the surge plan did not succeed in creating excess discharges.

Increasingly, evidence indicates that care quality depends on the degree to which hospital resources are sufficient to meet demand. Reverse triage, at least as implemented by SCH on November 4, 2009, is unlikely to represent an effective solution to surge outside of a disaster setting because of its requirement for centralized decision making. SCH has incorporated the results of this review into the way that it collects and analyzes data, manages flow, and responds to inpatient surges.