A system analysis of a suboptimal surgical experience

BACKGROUND: System analyses of incidents that occur in the process of health care delivery are rare. A case study of a series of incidents that one of the authors experienced after routine urologic surgery is presented. We interpret the sequence of events as a case of cascading incidents that resulted in outcomes that were suboptimal, although fortunately not fatal. METHODS: A system dynamics approach was employed to develop illustrative models (flow diagrams) of the dynamics of the patient's interaction with surgery and emergency departments. The flow diagrams were constructed based upon the experience of the patient, chart review, discussion with the involved physicians as well as several physician colleagues, comparison of our diagrams with those developed by the hospital of interest for internal planning purposes, and an iterative process with one of the co-authors who is a system dynamics expert. A dynamic hypothesis was developed using insights gained by building the flow diagrams. RESULTS: The incidents originated in design flaws and many small innocuous system changes that have occurred incrementally over time, which by themselves may have no consequence but in conjunction with some system randomness can have serious consequences. In the patient's case, the incidents that occurred in preoperative assessment and surgery originated in communication and procedural failures. System delays, communication failures, and capacity issues contributed largely to the subsequent incidents. Some of these issues were controllable by the physicians and staff of the institution, whereas others were less controllable. To the system's credit, some of the more controllable issues were addressed, but systemic problems like overcrowding are unlikely to be addressed in the near future. CONCLUSION: This is first instance that we are aware of in the literature where a system dynamics approach has been used to analyze a patient safety experience. The qualitative system dynamics analysis was useful in understanding the system, and contributed to learning on the part of some components of the system. We suggest that further data collection and quantitative analysis would be highly informative for identification of system changes to improve quality and safety.