Joint spatial modeling to identify shared patterns among chronic related potentially preventable hospitalizations

Area of Science:

• Health Services Research
• Spatial Epidemiology
• Public Health

Background:

• Potentially Preventable Hospitalizations (PPH) assess ambulatory care access for populations.
• Geographic patterns in PPH can indicate healthcare provider performance.
• Understanding spatial variation in PPH is crucial for targeted interventions.

Purpose of the Study:

• To jointly model geographical variations in six chronic PPH conditions.
• To identify common and discrepant spatial patterns among PPH conditions.
• To assess the influence of a common pattern on individual PPH conditions.

Main Methods:

• Utilized data from 39,970 PPH admissions (2007-2009) across 240 Basic Health Zones.
• Estimated PPH rates and Standardized Hospitalization Ratios per geographic unit.
• Applied Shared Component Models (SCM) for joint spatial analysis of PPH conditions.

Main Results:

• A common latent component explained 36% of the variability across the six PPH conditions.
• The common pattern identified territorial clusters with elevated PPH risk.
• Specific risk patterns varied, with some conditions showing significant unique spatial variations.

Conclusions:

• A single latent pattern accounts for 36% of the geographical risk distribution for these PPH conditions.
• Shared Component Models offer a valuable tool for assessing healthcare system performance.
• Spatial analysis of PPH can reveal disparities in healthcare access and quality.