Monitor, anticipate, respond, and learn: developing and interpreting a multilayer social network of resilience abilities

Resilient performance is influenced by social interactions of several types, which may be analysed as layers of interwoven networks. The combination of these layers gives rise to a “network of networks”, also known as a multilayer network. This study presents an approach to develop and interpret multilayer networks in light of resilience engineering. Layers correspond to the four abilities of resilient systems: monitor, anticipate, respond, and learn. The proposal is applied in a 34-bed intensive care unit. To map relationships between actors in each layer, a questionnaire was devised and answered by 133 staff members, including doctors, nurses, nurse technicians, and allied health professionals. Two multilayer networks were developed: one considering that actors are 100% available and reliable (work-as-imagined) and another considering suboptimal availability and reliability (work-as-done). The multilayer networks were analysed through actor-centred (Katz centrality, degree deviation, and neighbourhood centrality) and layer-centred metrics (inter-layer correlation, and assortativity correlation). Strengths and weaknesses of social interactions at the ICU are discussed based on the adopted metrics.