Development of a Descriptive Model of Flow Disruption Effect on Bypass Time in a Cardiovascular Operating Room

The current paper proposes utilizing extant flow disruption data, collected from two cardiovascular operating rooms, and utilizing that to develop a mathematical model. Of particular interest is bypass time which is a critical surgical phase, the length of which has been linked to post-operation complications such as hospital-borne infections and readmission. By developing a model that predicts bypass time, the present research can determine the impact by types of flow disruptions – coded based on the RIPCHORD-TWA taxonomy, team member disrupted, and where in the surgery the FD occurs. After a review of literature, a variety of models were explored to support the idea that FDs are impacting bypass time. Largely correlational in nature, these various models demonstrate that there is some relationship between FD count, FD duration, and bypass time. Because of this, hypotheses were generated and more in-depth mathematical modeling options are explored. Based on the nature of the data, an adjacent possible methodology was selected to further analyze this data set. The findings of this study were unable to establish causality, however a number of unique relationships between FDs and the length of bypass time were explored. For some types of FDs, such as those impacting the anesthesia and circulating nurse role, longer bypass time was associated with longer and more numerous – yet less frequent – flow disruptions. Others (perfusionist, communication, interruptions) had similar patterns of FD length and frequency, with the differences being just how many occurred over the length of the operation. Because of the varying nature of these FDs, this research found support for the hypotheses that FDs show differences across the length of bypass time. Most importantly, this project brings with it a new understanding of the nature of the CVOR and the nature of flow disruptions occurring within it. The FDs are not themselves Markovian, and they are not just noise within the system either. By utilizing the adjacent possible methodology and applying that to the CVOR, this project helps to work past barriers in analyzing the system. The process itself, the surgical procedure, can be considered Markovian. Flow disruptions therefore must fit into this system to be better understood. The final conclusions of this project are that FDs represent adjacent possible clouds nested within the Markovian process, demonstrated through the presented data