Risk and Recovery in Complex Medical Domains: Computational and Graphical Models of Performance

Abstract

Medicine is a complex, safety-critical and highly interactive system. Despite efforts to provide safe, effective care, adverse events still occur — clinicians make diagnostic and therapeutic errors, system constraints impact the coordination and delivery of care and patients suffer unexpected complications and injuries. Our current understanding of the nature of medical adverse events, and our ability to develop durable preventative or mitigating strategies has been hampered by somewhat out-dated and inadequate models of clinical risk, which focus almost exclusively on patient factors, provider factors or the specific clinical procedure being performed. Absent from such models are the interactions and inter-dependencies between different system components (staffing, instrumentation, protocols, procedures, access to and quality of information, communication modes, and scheduling cycles, system-wide volume and acuity, throughput pressures), an understanding of the reliability of such components under different operating conditions and the significance of failure during different phases of clinical care. This paper describes the use of computational and graphical models to improve our understanding of medical risk by modeling complex interactions between clinical subsystems and variations in risk across different phases of clinical care. Many of the submodels developed here are generalizable across a broad range of other healthcare settings.