To characterize social networks in stroke survivors by comparing two established social network scales to a newly developed personal network mapping tool.

Social connectivity is a known determinant of health and disease, and it has been shown to influence prompt access to care and outcomes in stroke patients. We assessed patients using two traditional social network scales, as well as PERSNET, a newly developed social network survey that allows for detailed mapping of social network dimensions around an index patient. Dimensions include quantitative network structure parameters (e.g., size of the network) and network composition (e.g, percentage of contacts who are kin).

67 stroke patients underwent, in random order, the Stroke Lubben Social Network Scale-Revised (LSNS-R), Social Network Scale (SSNS), and PERSNET. We compared the SSNS scores, LSNS-R scores, and PERSNET-derived network metrics using correlation matrices and regression. The PERSNET structural parameters that we studied were: total network size, constraint, effective size, maximum degree and mean degree. PERSNET compositional parameters included diversity of sex, diversity of race, percentage of contacts who are kin, percentage who do not exercise regularly, and percentage who have poor diet.

We found significant correlation between SSNS, LSNS-R and PERSNET-derived indices of network structure. LSNS-R directly correlated with total network size (Spearman’s R 0.45, p<0.001) and inversely correlated with network constraint (Spearman’s R -0.32, p<0.05). SNSS inversely correlated with PERSNET constraint (Spearman’s R -0.25, p<0.05). We found no correlation between SSNS, LSNS-R and PERSNET-derived network composition metrics.

PERSNET appears to be a valid tool for assessing social support, and also captures a novel range of health behavioral data within patients’ social networks that has not been well characterized by previous network tools. PERSNET may help identify patients with social networks that are high risk for adverse stroke outcomes.