Using Social Network Analysis to Map Infectious Risk Across Care Settings
Abstract
Background: Long-term care facilities (LTCFs) and hospitals form an interconnected system in which patient mobility shapes the transmission of multidrug-resistant organisms. Facility-level approaches may fail to capture the system-level structure of infectious risk. We applied Social Network Analysis to map carbapenem-resistant Enterobacterales (CRE) circulation in Tuscany.
Methods: Retrospective study of 2024 regional administrative healthcare flows for residents aged ≥65 years in LTCFs in Tuscany. Three directed weighted networks linking LTCFs, hospitals and emergency departments were built: all-cause transfers, transfers of CRE-colonised and CRE-infected residents. Node-level centrality metrics (degree, weighted degree, and betweenness) were compared between networks using Spearman’s rank correlation stratified by node type, with case studies on top-ranking nodes.
Results: The cohort included 11,914 residents in 325 LTCFs (mean age 85, 70% female); CRE incidence was 4.1/100,000 person-days, with hospital-acquired (79.7) far exceeding LTCF-acquired (3.1) rates. The CRE-colonised network was sparser than the all-cause network (107 vs 411 nodes). Centralities in the CRE-infected network correlated strongly with all-cause centralities (weighted in-degree ρ = 0.91; betweenness ρ = 0.71; p < 0.001), whereas the CRE-colonised network showed only moderate correlations (ρ = 0.59 and ρ = 0.28; p < 0.001); betweenness showed the lowest agreement. Case studies revealed a tertiary hospital with high all-cause betweenness (0.109) but limited CRE bridging role (0.004), and an LTCF channelling 271 CRE-positive transfers exclusively to acute care.
Conclusions: Infectious risk is shaped not only by healthcare volume but by the structural position of facilities within patient-sharing networks. Bridging roles diverge most between all-cause and CRE networks; network-informed indicators may identify high-risk nodes invisible to traditional surveillance and support coordinated prevention.
Key messages: • Patient-sharing networks reveal infectious risk patterns that facility-level surveillance may miss, especially across LTCFs, hospitals and emergency departments.
• Bridging roles differ markedly between all-cause and CRE networks, supporting network-informed targeting of IPC and screening strategies.
Related
- Workforce Cost Absorption among Community Health Promoters in Fragmented Maternal Nutrition and Social Protection Systems: A Qualitative Study Across Three Kenyan Settings
- Effect of local versus global risk and normative information on voluntary quarantine adoption: protocol for a 2×2 randomised controlled trial leveraging app-based epigames across diverse congregate settings (The Epi-Q Study).
- Nature-based social prescribing for children across care settings: baseline findings from CONNECT
- CO065 | THROMBOTIC RISK IS NOT UNIFORM ACROSS APHERESIS SETTINGS: INSIGHTS FROM CAR-T LYMPHOCYTAPHERESIS AND AUTOLOGOUS STEM CELL COLLECTION
- Genetics in Context: Social Factors Modify Polygenic Risk Across Multiple Cardiometabolic Diseases
BibTeX
@article{casigliani2026using,
title = {{Using Social Network Analysis to Map Infectious Risk Across Care Settings}},
author = {V Casigliani and S Forni and F Collini and F Gemmi and A Bedini and F Ruffini and A Bechini and F Marcelloni and C Rizzo and M Falcone},
journal = {The European Journal of Public Health},
year = {2026},
month = oct,
doi = {10.1093/eurpub/ckag169.422},
url = {https://doi.org/10.1093/eurpub/ckag169.422},
note = {CC BY-NC 4.0},
}
RIS
TY - JOUR
TI - Using Social Network Analysis to Map Infectious Risk Across Care Settings
AU - V Casigliani
AU - S Forni
AU - F Collini
AU - F Gemmi
AU - A Bedini
AU - F Ruffini
AU - A Bechini
AU - F Marcelloni
AU - C Rizzo
AU - M Falcone
PY - 2026
DA - 2026/10/01
JO - The European Journal of Public Health
DO - 10.1093/eurpub/ckag169.422
UR - https://doi.org/10.1093/eurpub/ckag169.422
AB - Background: Long-term care facilities (LTCFs) and hospitals form an interconnected system in which patient mobility shapes the transmission of multidrug-resistant organisms. Facility-level approaches may fail to capture the system-level structure of infectious risk. We applied Social Network Analysis to map carbapenem-resistant Enterobacterales (CRE) circulation in Tuscany. Methods: Retrospective study of 2024 regional administrative healthcare flows for residents aged ≥65 years in LTCFs in Tuscany. Three directed weighted networks linking LTCFs, hospitals and emergency departments were built: all-cause transfers, transfers of CRE-colonised and CRE-infected residents. Node-level centrality metrics (degree, weighted degree, and betweenness) were compared between networks using Spearman’s rank correlation stratified by node type, with case studies on top-ranking nodes. Results: The cohort included 11,914 residents in 325 LTCFs (mean age 85, 70% female); CRE incidence was 4.1/100,000 person-days, with hospital-acquired (79.7) far exceeding LTCF-acquired (3.1) rates. The CRE-colonised network was sparser than the all-cause network (107 vs 411 nodes). Centralities in the CRE-infected network correlated strongly with all-cause centralities (weighted in-degree ρ = 0.91; betweenness ρ = 0.71; p < 0.001), whereas the CRE-colonised network showed only moderate correlations (ρ = 0.59 and ρ = 0.28; p < 0.001); betweenness showed the lowest agreement. Case studies revealed a tertiary hospital with high all-cause betweenness (0.109) but limited CRE bridging role (0.004), and an LTCF channelling 271 CRE-positive transfers exclusively to acute care. Conclusions: Infectious risk is shaped not only by healthcare volume but by the structural position of facilities within patient-sharing networks. Bridging roles diverge most between all-cause and CRE networks; network-informed indicators may identify high-risk nodes invisible to traditional surveillance and support coordinated prevention. Key messages: • Patient-sharing networks reveal infectious risk patterns that facility-level surveillance may miss, especially across LTCFs, hospitals and emergency departments. • Bridging roles differ markedly between all-cause and CRE networks, supporting network-informed targeting of IPC and screening strategies.
SN - 1101-1262
SN - 1464-360X
ER -