Publicação
Dissecting medical referrals in health services
| dc.contributor.author | Duarte, Regina de Brito | |
| dc.contributor.author | Han, Qiwei | |
| dc.contributor.author | Soares, Claudia | |
| dc.contributor.institution | NOVA School of Business and Economics (NOVA SBE) | |
| dc.contributor.institution | Faculdade de Ciências e Tecnologia (FCT) | |
| dc.contributor.pbl | BioMed Central (BMC) | |
| dc.date.accessioned | 2026-04-27T18:20:01Z | |
| dc.date.available | 2026-04-27T18:20:01Z | |
| dc.date.issued | 2026-04-13 | |
| dc.description | Publisher Copyright: © The Author(s) 2026 | |
| dc.description.abstract | Background: Medical referrals between primary care (PC) and specialist care(SC) physicians are essential to integrated healthcare delivery but often encounterinefficiencies arising from communication gaps, coordination issues, and organizational structure. While prior studies have explored patient-sharing networks,limited quantitative evidence exists on how physicians’ professional relationships,such as co-training, co-affiliation, or network embeddedness, relate to referralbehavior. This study develops a data-driven framework combining network analysis and graph neural networks (GNNs) to examine how professional-networkstructure characterizes referral formation. Methods: The study analyzed five years of consultation data from a largePortuguese private healthcare system. Two complementary networks were constructed: a directed referral network linking PC to SC physicians based on patienttransitions within a 30-day window, and an undirected professional network representing shared institutional and educational ties. Graph-based embedding models (GraphSAGE, Attri2Vec, Node2Vec) were trained to predict referral links, using physician demographic and network-based inputs. Feature importance was evaluated using Shapley Additive Explanations (SHAP), and model embeddings were visualized using UMAP to assess structural coherence.Results: Incorporating professional-network information improved referral prediction accuracy across all models. SHAP analyses identified degree and eigenvector centrality as dominant predictors of referral probability, indicating thatphysicians with broader professional connectivity and greater embeddednesswere more likely to receive referrals. Age and gender effects were minor andsecondary to structural network factors. Visualization of learned embeddingsrevealed that professional-network information produced coherent physician clusters aligned with hospital co-affiliations and cross-role interactions, demonstrating that professional networks capture referral proximity and organizational cohesion.Conclusions: Physicians’ professional-network positions are strongly associated with referral behavior and provide incremental predictive value beyond demographics alone. Embeddedness and centrality reflect relational mechanismsof trust, reputation, and coordination that relate to access to specialist care.Incorporating network analytics into referral management can inform policyinterventions aimed at improving transparency, balancing referral loads, andpromoting equitable access to specialist services. | en |
| dc.description.version | preprint | |
| dc.description.version | publishersversion | |
| dc.description.version | published | |
| dc.format.extent | 17 | |
| dc.format.extent | 1975873 | |
| dc.format.extent | 2555486 | |
| dc.identifier.doi | 10.1186/s12913-026-14099-9 | |
| dc.identifier.issn | 1472-6963 | |
| dc.identifier.other | PURE: 160476377 | |
| dc.identifier.other | PURE UUID: aab64d65-a8e6-44d9-a4e6-dcee54e9691b | |
| dc.identifier.other | crossref: 10.1186/s12913-026-14099-9 | |
| dc.identifier.other | Scopus: 105044557921 | |
| dc.identifier.other | ORCID: /0000-0003-3071-6627/work/221983198 | |
| dc.identifier.other | WOS: 001820726600001 | |
| dc.identifier.uri | http://hdl.handle.net/10362/202606 | |
| dc.identifier.url | https://www.webofscience.com/wos/woscc/full-record/WOS:001820726600001 | |
| dc.identifier.url | https://www.scopus.com/pages/publications/105044557921 | |
| dc.language.iso | eng | |
| dc.peerreviewed | yes | |
| dc.subject | Medical referrals | |
| dc.subject | Professional networks | |
| dc.subject | Social network analysis | |
| dc.subject | Graph Neural Networks | |
| dc.subject | Link prediction | |
| dc.subject | Health services research | |
| dc.subject | Healthcare coordination | |
| dc.title | Dissecting medical referrals in health services | en |
| dc.title.subtitle | Role of physician professional networks | en |
| dc.type | journal article | |
| degois.publication.firstPage | 1 | |
| degois.publication.lastPage | 17 | |
| degois.publication.title | BMC Health Services Research | |
| degois.publication.volume | 26 | |
| dspace.entity.type | Publication | |
| rcaap.rights | openAccess |