Hospitals can use data to develop tools that accurately predict emergency department visits, according to a study in Emergency Medicine Journal.
Researchers studied five years of historical data from two different hospitals to create a model for predicting ED visits. They found the model's predicted monthly ED visits and actual ED visits had only a 2 percent mean average percentage error. When predicting daily admissions, the model and actual data had an 11 percent MAPE. As the time interval decreased, the percentage error increased, indicating the increasing difficulty of making predictions for short periods of time.
The authors concluded that healthcare providers can use data to create a valid ED prediction tool, and this tool may help EDs with scheduling and bed management.
Keys to an Efficient, Cost-Effective Emergency Department: Q&A with Jeff Wood of TeamHealth
Washington State Approves Budget With $8.8M for New Medicaid ED Plan
Researchers studied five years of historical data from two different hospitals to create a model for predicting ED visits. They found the model's predicted monthly ED visits and actual ED visits had only a 2 percent mean average percentage error. When predicting daily admissions, the model and actual data had an 11 percent MAPE. As the time interval decreased, the percentage error increased, indicating the increasing difficulty of making predictions for short periods of time.
The authors concluded that healthcare providers can use data to create a valid ED prediction tool, and this tool may help EDs with scheduling and bed management.
More Articles on Capacity Management:
Ohio ED Utilization Outpaced Other States by 34% in 2010Keys to an Efficient, Cost-Effective Emergency Department: Q&A with Jeff Wood of TeamHealth
Washington State Approves Budget With $8.8M for New Medicaid ED Plan