Measuring interaction and influence among health professionals within a huntington's disease clinic: a social network approach.

Abstract

Background Interactions within multidisciplinary teams can underpin improved care integration and outcomes for people living with Huntington's disease (HD). Social network analysis, based on graph theory, provides a tool to measure interaction and influence within these inter-professional networks. Aim(s): To measure interaction and influence among health professionals within a HD clinic using social network analysis. Methods This study was conducted at a publicly funded multidisciplinary HD clinic in Victoria, Australia. Re-development in 2016-2017 led to transition from a neurologist-led to a more collaborative service delivery. We adopted a retrospective before (Year 2014-2015; n=339) and after (Year 2018-2019; n=346) cohort design. Patient-related interaction data was manually extracted from medical records, then exported for analysis in R. The network analysis included interaction correlations, Walktrap community detection (based on the idea that short random walks tend to stay in the same community), and local (degree, closeness, betweenness) and global (PageRank) centrality. Results There were more of stronger positive interaction correlations in the After cohort (between physiotherapist and occupational therapist (OT) (r=0.57); speech pathologist (SP) and dietician (r=0.47); SP and OT (r=0.45); physiotherapist and SP (r=0.41); OT and dietician (r=0.41)) compared from the Before cohort (SP and dietician (r=0.53); neurologist and clinical liaison (r=0.38)). Walktrap algorithm detected three unconnected communities in the Before cohort, while the After cohort had two communities (one larger-sized) connected through the neurologist (Figure 1). The most influential professionals in the After cohort were neurologist (highest degree=7 and betweenness=22) [most social connections and control over information flow, respectively], clinical liaison (lowest closeness=0.27) [closest to other professionals], and dietician (highest PageRank=0.15) [most in-links from important professionals]. Conclusion This study showed a social network approach to measure interaction and influence among health professionals within a HD clinic..