[Applying social network analysis to the health system]

Social network analysis is a method for describing and contextualising the structural relationships amongst social entities. The aim of this study was to assess the relationships between hospitals through the use of an example of a hospital that takes care of myocardial infarction in the region of Provence Alpes Côte d'Azur (PACA).

This retrospective study included all patients hospitalised for acute coronary syndrome in the PACA region between January and June 2000. The description analysis consisted of making a graph describing the transfer of patients between hospitals to calculate indicators for the centrality of the hospital and its prestige. The graph was also used to determine and measure the sub-group cohesion (cliques) and the density of the hospital network.

A total of 2049 patients suffering from myocardial infarction were admitted to 78 hospitals during the study period. There were 867 transfers from one hospital to another. 51 cliques were highlighted; 29 hospitals do not belong to a clique. Among the 49 hospitals who do belong to sub-groups, 50% are linked to two cliques. Eight hospitals belong to more than 8 sub-groups, and the maximum number of sub-groups to which a hospital is linked is 12. The density of the hospital network is 0.12.

Social network analysis is a field of research open and adaptable to the public health scene. This method could be applied to the evaluation of health and hospital administration and management networks.