Abstract 13832: Perioperative Physician Care Model in Cardiothoracic Surgery: A Pilot Study

Abstract

Background and Methods: Perioperative Medicine is an emerging discipline which seeks to optimise the care of patients during the preoperative and postoperative stages. This is particularly pertinent for older patients who are at an increased risk of a poor postoperative outcome. Evidence from other surgical disciplines demonstrates that Comprehensive Geriatric Assessment (CGA) is a valuable tool for optimising perioperative care. We describe the deployment of a single perioperative physician fellow, supported by consultant perioperative geriatricians, cardiologists and surgeons for a pilot 3-month period. We estimate the upper bounds of the impact that such a deployment could have on capacity and length of stay. Results: The perioperative fellow working 60% whole time equivalent undertook 80 CGAs during the 3 month pilot phase. Mean patient age was 75 ± 10 years, the ratio of female to male was 39:41. We therefore estimate that a perioperative physician could review and follow up 240 new patients per year, representing approximately 10% of the patients treated in our unit per year. However, when the perioperative fellow is directed to those patients with the longest stay lengths, there is substantial amplification of the effect size, as the top decile of patients occupy 34% of all bed-days within the cardiothoracic ward (Figure). Bed occupancy as a function of length of stay is described by non-normal distribution, and this fat-tailed distribution explains the large benefit likely to be accrued from relatively modest input costs in perioperative physicians (Figure b, Shapiro-Wilks p < 0.05). Conclusion: Perioperative fellowship models are likely to be a cost-effective solution to the challenge of escalating demand for complex cardiac care in the setting of constrained supply. Future work will focus on the refinement of a neural network which will seek to allow identification of patient variables which best predict prolonged length of stay.