Prediction of blood loss during surgery on the lumbar spine with the patient supported prone on the Jackson table

• Published in: The spine journal 2012-12, Vol.12 (12), p.1103-1110
• Main Authors: Mathai, Koshy M., MD, Kang, James D., MD, Donaldson, William F., MD, Lee, Joon Y., MD, Buffington, Charles W., MD
• Format: Article
• Language: English
• Subjects: Adult; Aged; Blood loss; Blood Loss, Surgical; Female; Humans; Jackson table; Lumbar spine surgery; Lumbar Vertebrae - surgery; Male; Middle Aged; Models, Anatomic; Orthopedics; Predictive Value of Tests; Prone Position
• ISSN: 1529-9430; 1878-1632
• DOI: 10.1016/j.spinee.2012.10.027

Abstract Background context To date, many studies have examined the effects of one or several factors on blood loss during lumbar spine surgery. The nature and extent of the operation, patient position, blood pressure, and a variety of factors related to patient size have been touted as predictors of blood loss. Purpose To measure multiple factors implicated as determinants of blood loss and develop a multivariable statistical model capable of predicting blood loss. Study design An observational study of patients undergoing lumbar spine surgery in the prone position on the Jackson table. Patient sample A total of 71 healthy adult men and women undergoing lumbar spine surgery in a university hospital setting. Outcome measure Blood loss during surgery. Methods We observed 35 surgeries and recorded demographic and body habitus data on each patient as well as surgical variables, blood pressure, and peripheral venous pressure. We measured bladder pressure intermittently as a surrogate for intra-abdominal pressure. We constructed a statistical model with the results and validated that model in a separate set of 36 subjects. Results The Jackson table supported all our patients regardless of body dimensions without causing an increase in bladder pressure. Blood loss during surgery averaged 1,167±998 mL (mean±1 standard deviation, range 32–3,745). The statistical model was able to account for about 75% of the variability in blood loss using four variables: the number of laminectomies, whether bone was harvested from the iliac crest, experience of the surgeon doing the initial exposure and closure, and distension of the epidural veins. Data on these variables that were collected in the validation study found a multiple correlation coefficient (R2 ) of 0.66 between predicted and observed blood loss. Conclusions This is the first study to build a successful multivariable predictive model of blood loss during spine surgery. The Jackson table was effective in supporting patients with different body sizes and shapes, thus removing raised intra-abdominal pressure as an important factor.