Research collaboration in pediatric critical care: a social network analysis of coauthorship
Duffett M, Brouwers M, Meade MO, Cook DJ
Aims & Objectives
Clinical research is a collaborative enterprise — researchers benefit from the expertise, experience, and resources of their collaborators. Our objectives were to describe the extent and patterns of collaboration among pediatric critical care researchers and to identify prominent individuals, centres, and countries.
We conducted a social network analysis of coauthorship. We used PICUtrials.net to identify published RCTs in pediatric critical care and extracted author names and affiliations from the publications. We used the number of publications, citations, and measures of network centrality to assess prominence.
We included 1934 individuals and 27 groups who published a total of 372 RCTs between 1986 and 2017. A minority (312 [16%]) coauthored more than 1 RCT (maximum 15). Researchers published with a median (IQR) of 7 (5, 11) coauthors. We identified a large cluster of 742 (38%) researchers publishing 128 (34%) RCTs. There were 25 smaller disconnected clusters of researchers publishing a median of 3 RCTs each (range 2–15). There were also 140 RCTs (38%) published by groups of researchers who published a single RCT and were not connected by co-authorship to any other researchers. Groups with at least one member who previously published an RCT accounted for 206 (55%) of the RCTs. The most prominent researchers varied with the metric; 13 (93%) of the 14 who ranked in the top 50 for all metrics were from the United States or Canada.
The research enterprise in pediatric critical care is highly clustered and fragmented. Most prominent individuals were from the United States and Canada.