Measuring severity of illness: comparisons across institutions

Area of Science:

• Health Services Research
• Hospital Administration
• Medical Economics

Background:

• Traditional patient classification for healthcare resource utilization relies heavily on diagnostic criteria.
• Existing methods like diagnostic-related groups (DRGs) and staging may not fully capture patient complexity.
• There is a need for a more accurate system to predict and compare hospital resource use.

Purpose of the Study:

• To introduce and evaluate a novel severity of illness index.
• To determine if this index creates more homogeneous patient subgroups regarding resource utilization.
• To compare hospital resource use (charges, length of stay) across different hospital settings using the new index.

Main Methods:

• A generic severity of illness index was developed and applied to various medical and surgical conditions.
• Patient subgroups were created based on the severity of illness index.
• Hospital resource utilization metrics, including total charges, length of stay, routine charges, and laboratory charges, were assessed.
• Comparisons of total charges and length of stay were made between academic teaching hospitals and community hospitals (with and without teaching programs), controlling for severity of illness.

Main Results:

• The severity of illness index produced patient subgroups that were more homogeneous in terms of hospital resource use compared to diagnostic-related groups, staging, and generalized patient management paths.
• When controlling for severity of illness, total hospital charges and length of stay were found to be similar between academic teaching hospitals and community hospitals (both with and without teaching programs).

Conclusions:

• A severity of illness index offers a more effective method for classifying patients based on resource utilization than traditional diagnostic criteria.
• Severity of illness is a critical factor that influences hospital resource use and should be considered when comparing healthcare facilities.
• The findings suggest that differences in hospital charges and length of stay may be better explained by patient severity than by hospital type or teaching status alone.