Abstract: PUB044
Geographic Disparities in Dialysis Outcomes
Session Information
Category: Artificial Intelligence, Digital Health, and Data Science
- 300 Artificial Intelligence, Digital Health, and Data Science
Author
- Ashkar, Ziad Maurice, University of Louisiana at Lafayette, Lafayette, Louisiana, United States
Background
Despite advancements in dialysis, outcomes remain poor, highlighting the interplay of clinical, demographic and socioeconomic factors. One approach to understand these complexities is unsupervised machine learning .The objective of this study is to analyze dialysis facilties using cluster analysis, with a particular focus on geographic disparities .
Methods
Dialysis facilities dataset was accessed from https://data.cms.gov.Sociodemographic data at the ZIP Code Tabulation Area (ZCTA) level were obtained via the Census Bureau API. These data were linked to dialysis facilities by ZIP code. PCA transformation was applied to standardized data ,then K means clustering was applied.Geographic distribution of clusters was examined across states and US Census regions.Python 3.11.6 with pandas, scikit-learn, and geopandas packages were used for analyses.
Results
We identified 2 distinct clusters . Cluster 0 has 4609 clinics and cluster 1 has 2857 clinics. Cluster 1 facilities have higher mean mortality rate and hospitalization rate. After mapping by US census regions, we found that the South and West have the highest percent of cluster 1. Facilities in the South are located in areas with highest percentage of blacks,and the West has the highest percentage of hispanics. Clinics in the South are also in regions with lower household income and lower college education.
Conclusion
Using data from dialysis compare and unclassified machine learning we identified a cluster of clinics with lower outcomes that are more concentrated in the South and West. Facilities in the South are also in regions with lower median household income and lower college education.Further research should explore patient factors, comorbitidies,and regional policy differences to further understand outcomes disparities in dialysis.
Clusters Distribution
| Cluster 0 | Cluster 1 | |
| Mortality Rate | 21.89 | 22.57*** |
| Hospitalization Rate | 138.2 | 146.44** |
| Readmission Rate | 25.47 | 27.6** |
| Midwest | 77.53% | 22.47% |
| Northeast | 70.64% | 29.36% |
| South | 54.01% | 46%* |
| West | 55.43% | 44.57%* |
***p<0.05, *South and West have highest % cluster 1