Pooling multisite administrative data for economic analysis.

Area of Science:

• Health economics
• Health services research
• Data management

Background:

• Economic evaluations are increasingly important for US health insurers and payers.
• Administrative data from health maintenance organizations (HMOs) offer efficient utilization and cost measures for multisite economic analyses.
• Pooling HMO administrative data is challenging due to differences in care systems, costing, and coding.

Purpose of the Study:

• To describe issues encountered when pooling HMO administrative cost data for multisite economic evaluations.
• To discuss the implications of administrative data attributes relevant to costing.
• To offer suggestions for researchers working with pooled HMO administrative data.

Main Methods:

• The study describes attributes of administrative data relevant to costing.
• It discusses the implications of these attributes.
• It shares experiences and lessons learned from pooling HMO administrative data.

Main Results:

• Pooling HMO administrative data for economic evaluations is problematic due to inherent variations in data sources.
• Understanding data attributes related to costing is crucial for accurate analysis.
• Practical experience highlights key challenges and potential solutions.

Conclusions:

• Pooling HMO administrative data requires careful consideration of data heterogeneity.
• Researchers must address variations in systems of care, costing, and coding for valid economic evaluations.
• Lessons learned can guide future efforts in multisite economic analyses using administrative data.