Mixed retrospective–prospective cohort study.Objective.
To characterize practice patterns for the use of Cell Saver at our institution, investigate its cost-effectiveness, and propose a new tool for patient selection.Summary of Background Data.
Blood loss is an exceedingly common complication of spine surgery, and Cell Saver intraoperative cell salvage has been used to decrease reliance on allogeneic blood transfusions for blood volume replacement. The cost-effectiveness of Cell Saver has not been established for lumbar spinal surgery, and no universal guidelines exist for clinicians to decide when to utilize this tool. Other authors have proposed cutoffs for anticipated blood loss volumes which indicate that Cell Saver should be used.Methods.
Five hundred and eight patients undergoing lumbar laminectomy in 3 or fewer levels were reviewed from our prospective spinal outcomes registry. Cost information for Cell Saver and allogeneic transfusions was collected from our institution's billing and collections department. Logistic regression was used to identify patient characteristics associated with use of Cell Saver. An incremental cost effectiveness ratio was calculated based on transfusion and cost data. A clinical prediction score was derived using logistic regression.Results.
Use of Cell Saver correlated with increased age, higher body mass index, diabetes, greater American Society of Anesthesiologists classification, and greater number of previous spine surgeries. Outcomes for patients who did and did not have Cell Saver set up intraoperatively were equivocal. Cell Saver was not cost effective based on current usage patterns, but may become cost effective if used for patients with high expected blood loss. A simple clinical prediction rule is proposed which may aid in selection of patients to have Cell Saver present intraoperatively.Conclusion.
Cell Saver is not a cost-effective intervention but may become cost effective if a threshold of expected intraoperative blood loss is used to select patients more judiciously.Conclusion.
Level of Evidence: 3