htai 2015 - social network analysis of hta collaborations: the case of rebrats (tazio vanni)
TRANSCRIPT
Social Network Analysis of HTA Collaborations: the case of REBRATS
Tazio Vanni General Coordinator of Health Technology Assessment
Department of Science and Technology Secretariat of Science, Technology and Strategic Inputs
Brazilian Ministry of Health
Research team
• Luciana Leão
• Rodrigo Costa
• Flávia Elias
• Juliana Girardi
• Lucas Felipe
• Oney Araújo
• Elenilda Martins
Introduction
• Health Technology Assessment (HTA) is becoming increasingly multidisciplinary.
• Policy-makers and researchers are relying more on multi-institutional networks to develop strong, intellectually diverse teams that can answer complex research questions.
• National governments have promoted HTA networks in order: – to develop useful HTA studies – to support capacity building among health care workers
and policy-makers – to foster sustainable health systems
Saúde e CT&I: Inovação para o acesso
Social and Economic Development
Health System Science and Technology
System
Objective: To integrate health researchers and decision-makers in the promotion, development and diffusion of health technology assessment to support a sustainable Brazilian National Health System.
Working groups:
1. Research Prioritization and Promotion; 2. Methodological Development and Evaluation; 3. Professional Training and Continuous Education; 4. Information and Communication Management; 5. Technological Horizon Scanning; 6. Health Services Research.
Members: Currently 81 institutions, including research centres, hospitals, universities, federal, regional and municipal health agencies.
Coordination: Joint Committee lead by the Department of Science and Technology, Ministry of Health
Investment: ~US$ 10 millions
Studies: 435 studies
Introduction
• Social network analysis (SNA) can be used to evaluate collaborations between institutions and groups.
• SNA has the potential: – to identify hubs and authorities
– to improve information diffusion and consensus building
– to monitor and evaluate integration among institutions
– to inform expansion and restructuring plans
Objective
• No peer-reviewed publications investigating networks in HTA could be found in Medline and Embase.
• The overall objective of this study was to evaluate patterns of scientific collaboration in HTA networks, using REBRATS as a case study.
Methods
• REBRATS relies on a virtual repository (SISREBRATS), which is the largest repository of HTA studies in Portuguese.
• From the 396 studies included in SISREBRATS from 11/2009 to 09/2014, we extracted data for authors’ names, institutional affiliation and location (state in Brazil).
Methods
• We developed a programme in C++ to structure the data extracted in mixing matrices.
• Gephi open-source network analysis software for visualization and exploration of networks and complex systems.
• Network layout was defined using the Fruchterman-Reingold and Force Atlas 2 algorithms.
SNA metrics
• Nodes metrics included centrality and betweenness.
• Network statistics included density, connected components, diameter, average distance between nodes and clustering coefficient.
Methods
• Individuals were classified according to time since graduation, gender and education.
• Institutions were classified according to REBRATS membership, the region they were based and type of institution.
Top 10 authors Centrality Institution Top 10 authors Betweenness Institution
Francisco Acurcio 52 UFMG Flavia Elias 20.086 Fiocruz
Carisi Polanczky 50 IATS Denizar Vianna 71.312 UERJ
Edina Koga 41 UNIFESP Braulio Luna 16.848 UNIFESP
Jose Jardim 35 UNIFESP Carisi Polanczky 16.630 IATS
Tais Galvao 30 UFAM Francisco Acurcio 15.849 UFMG
Flavia Elias 29 FIOCRUZ Cid Manso 13.689 IMS UERJ
Braulio Luna 28 UNIFESP Claudia Coeli 12.056 UFRJ
Luis Rohde 28 UFRGS Marcus Tolentino 10.484 UFAM
Jorge Ribeiro 28 UFRGS Marisa Santos 10.399 INC RJ
Regis Andriolo 28 UEP Bernardo Tura 9.591 INC RJ
0%
5%
10%
15%
20%
25%
30%
2 - 12 13 - 22 23- 32 33 - 42 43 - 52
Time since first graduation
0%
10%
20%
30%
40%
50%
60%
Male Female
Gender
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
50%
MSc/MA PHD Experts* Graduates
Educational Level
15
Number of institutions: 249 Number of REBRATS members: 78 Average nº collaborators (members): 16 Average nº collaborators (non-members): 7
Network statistics
Number of nodes
Average degree
Density
Average clustering coefficient
Number of connected components
249
9.373
0,037
0.767
1
Institution Centrality Institution Betweenness
FIOCRUZ 87 FIOCRUZ 0.2198
UFBA 63 USP 0.1427
USP 61 UFBA 0.1196
UFMG 52 UFRGS 0.1132
UFRGS 52 UFMG 0.1071
UERJ 49 UFSC 0.0900
HCPA 47 HCPA 0.0776
UFRJ 46 UERJ 0.0757
SVS_MS 42 UNICAMP 0.0556
UFSC 37 FM_USP 0.0433
Discussion
• We have shown that SNA is instrumental to monitor and evaluate, as well as to inform the expansion and restructuring plans for HTA networks, and possibily other research networks.
• There are different dimensions of collaborations that should be further investigated.
• The creation, evolution and integration should be further explored.
Discussion
• More comprehensive datasets with other variables that can explain the structure and performance of these networks.
• Combination of datasets and methods should be employed to further our understanding of these networks.
• Development of impact assessment tools for research networks.
• Define guidelines for the creation and management of research networks.
Discussion
• There are still many questions for which the HTA community has not provided answers.
• The increase in life expectancy and the epidemiological transition coupled with the growth of the knowledge economy are increasing the preassure on the HTA community to develop new assessment tools.
• To met these challenges we have to consider what are the optimal research ecosystems.