CPUs can provide “teachable moments” that result in quantifiable improvements in cardiac risk factors. Katz et al20 prospectively evaluated the effect of CPU admission on modifiable cardiac risk factors. The authors measured predetermined outcomes immediately and 3 months later. Following CPU evaluation, patients reported lower susceptibility to ischemic heart disease. Patients also reported improvements in the following behaviors compared to baseline:
These results were not compared to prior inpatients with comparable complaints and evaluations.
CPUs: Strategic Considerations
See Box 42-1 and reference 21.
Box 42-1 ACEP's Keys to a Successful Observation Unit |Favorite Table|Download (.pdf)
Box 42-1 ACEP's Keys to a Successful Observation Unit
- Clearly defined mission criteria
- Well-planned policies and procedures
- Proper staffing, location, and equipment
- Carefully developed programs for quality assurance and utilization review
Effect of CPU on Cost and Charges
If the decision is made to develop a CPU, the literature consistently identifies decreased cost and charges after CPU implementation. Three prospective randomized controlled trials compared either costs or charges for CPU patients compared to chest pain inpatients. In 1996, Gomez identified a $624.00 per patient decrease in hospital charges for CPU patients compared to inpatients.22 In 1997, Roberts measured a $567.00 decrease in cost per patient assigned to a CPU compared to inpatient evaluation.13 In 2010, Miller et al23 published cost outcomes for CPU versus inpatient status. However, this article described the use of cardiac magnetic resonance imaging (CMR) as part of CPU evaluation. Miller reported a mean decrease in costs (favoring CPU with CMR) of $1641.00.
Additional published studies of CPU financial impact used varying methodologies. Mikhail et al performed a comparative study with retrospective controls.24 The authors identified a $1470.00 decrease in cost for patients admitted to CPUs compared to inpatients. In 1994, Hoekstra et al25 compared charges per patient of those assigned to a CPU to inpatients. This study was not randomized and evaluated patients at 2 hospitals. Decreases in charges per CPU patient (compared to controls) for the 2 hospitals were $2030.00 and $1160.00.
Published studies on the financial impact of CPUs use 2 outcomes: cost and charges. These outcomes are in 1994 to 1997 US dollars. Given the heterogeneity of this data, the authors calculated percent decrease in charges or cost for the 4 studies cited previously.13,22,24,25 The range for percent decrease in charges or cost per patient for the 4 studies is 27% to 62%.
Effect of CPU on Length of Stay
The 2 prospective RCTs of CPU versus inpatient evaluation13,22 reported 10.9 and 11.7 hours of decreased evaluation times favoring CPU evaluation. The weighted average for this improvement in length of stay (LOS) is 11.4 hours (performed by authors).