Center-level geographic calculations

Author

Affiliation

Vagish Hemmige

Montefiore Medical Center/ Albert Einstein College of Medicine

Source code

The full R script is available at:

• R/geographical_computations_for_centers.R

This R script file is itself reliant on the following helper files:

• R/setup.R
• R/functions.R

Creating Distance-Based Buffers Around Transplant Centers

This section generates geographic buffers around transplant centers at predefined distances (e.g., 50, 100, 200 miles).

Four buffer objects are created:

• Transplant_center_all_buffer
• Transplant_centers_active_buffer
• Transplant_centers_HIV_buffer
• Transplant_centers_HOPE_buffer

Each object is organized hierarchically by:

1. Organ type
2. Distance
3. Year (when applicable)

This nested structure allows consistent iteration across organs, time points, and buffer distances in later analyses.

---

Step 1: Initialize Empty Buffer Lists

Empty lists are created to store buffered sf objects.
These will hold polygon geometries representing the catchment area around each transplant center.

---

Step 2: Loop Over Organs and Distances

For each organ in organ_list
and each distance in distance_list:

• Only distances expressed in miles are processed in this block.
• Mile distances are converted to meters, since the spatial data are projected in EPSG:5070 (which uses meters).

Distance conversions:

• 50 miles → 80,467.2 meters
  
• 100 miles → 160,934.4 meters
  
• 200 miles → 321,868 meters

---

Step 3: Create Buffers for All Centers

For each organ and distance:

• st_buffer() generates circular buffer polygons around each center.
• The result represents the geographic area within the specified radius of the center.

These are stored as:

Transplant_center_all_buffer[[organ]][[distance]]

---

Step 4: Create Year-Specific Buffers

Within each organ–distance combination, the code then loops over year_list and generates buffers for:

• Active transplant centers
  
• HIV R+ transplant centers
  
• HOPE (HIV D+/R+) transplant centers

Each of these is buffered using the same distance in meters to ensure comparability.

The resulting structure is:

• Transplant_centers_active_buffer[[organ]][[distance]][[year]]
• Transplant_centers_HIV_buffer[[organ]][[distance]][[year]]
• Transplant_centers_HOPE_buffer[[organ]][[distance]][[year]]

---

Why This Structure Matters

This approach ensures:

• Consistent geometry projection (EPSG:5070)
• Identical buffer distances across center categories
• A clean, iterable nested structure for downstream population overlays
• Straightforward generation of paired comparisons (e.g., 2017 vs 2022)

These buffered polygons are later intersected with census tract–level HIV population estimates to quantify geographic access to transplant programs.

# Create buffer objects for transplant centers based on distance

Transplant_center_all_buffer<-list()
Transplant_centers_active_buffer<-list()
Transplant_centers_HIV_buffer<-list()
Transplant_centers_HOPE_buffer<-list()


for (organ_loop in organ_list){
  for (distance_loop in distance_list){
    
    if (str_detect(distance_loop, "mile"))
    {
      
      distance<-case_when(
        distance_loop=="50mile" ~ 80467.2,
        distance_loop=="100mile" ~ 160934.4,
        distance_loop=="200mile" ~ 321868
      )
      
      Transplant_center_all_buffer[[organ_loop]][[distance_loop]]<-st_buffer(Transplant_centers_all_sf [[organ_loop]],dist = distance)
      
      for (year_loop in year_list)
      {
        
        Transplant_centers_active_buffer[[organ_loop]][[distance_loop]][[year_loop]]<-
          st_buffer(Transplant_centers_active_SF[[organ_loop]][[year_loop]], 
                    dist = distance)
        
        Transplant_centers_HIV_buffer[[organ_loop]][[distance_loop]][[year_loop]]<-
          st_buffer(Transplant_centers_HIV_sf[[organ_loop]][[year_loop]], 
                    dist = distance)
        
        Transplant_centers_HOPE_buffer[[organ_loop]][[distance_loop]][[year_loop]]<-
          st_buffer(Transplant_centers_HOPE_sf[[organ_loop]][[year_loop]], 
                    dist = distance)
        
      }
      
      
    }
    
  }
}

Other portions of the analysis

• Setup: Defines global paths, data sources, cohort inclusion criteria, and analysis-wide constants.
• Functions: Reusable helper functions for cohort construction, matching, costing, and modeling.
• Tables: Summary tables and regression outputs generated from the final models.
• Figures:Visualizations of costs, risks, and model-based estimates.
• About: methods, assumptions, and disclosures