Harrison F. Kay, Silky Chotai, Joseph B. Wick, David P. Stonko, Matthew J. McGirt, Clinton J. Devin

February 2016, Volume 25, Issue 3, pp 843 - 849 Original Article Read Full Article 10.1007/s00586-015-4175-8

First Online: 27 August 2015


Evaluate the factors associated with postoperative ICU admission in patients undergoing surgical management of degenerative lumbar spine disease.


Patients undergoing surgery for degenerative lumbar spine disease were enrolled into a prospective registry over a 2-year period. Preoperative variables (age, gender, ASA grade, ODI %, CAD, HTN, MI, CHF, DM, BMI, depression, anxiety) and surgical variables (instrumentation, arthrodesis, estimated blood loss, length of surgery) were collected prospectively. Postoperative ICU admission details were retrospectively determined from the electronic medical record. Student’s t test (continuous variables) and Chi-square test (categorical variables) were used to determine the association of each preoperative and surgical variable with ICU admission.


808 Patients (273 laminectomy, 535 laminectomy and fusion) were evaluated. Forty-one (5.1 %) patients were found to have postoperative ICU admissions. Reasons for admission included blood loss (12.2 %), cardiac (29.3 %), respiratory (19.5 %), neurologic (31.7 %), and other (7.3 %). For preoperative variables, female gender (P < 0.001), history of CAD (P = 0.003), history of MI (P = 0.008), history of CHF (P = 0.001), age (P = 0.025), and ASA grade (P = 0.008) were significantly associated with ICU admission. For surgical variables, estimated blood loss (P < 0.001) and length of surgery (P < 0.001) were significantly associated with ICU admission.


Age, female gender, ASA grade, cardiac comorbidities, intraoperative blood loss, and length of surgery were associated with increased risk of postoperative ICU admission. Knowledge of these factors can aid surgeons in patient selection and preoperative discussion with patients about potential need for unexpected admission to the ICU.

Read Full Article