Send to

Choose Destination
See comment in PubMed Commons below
Int J Gynecol Cancer. 2011 Nov;21(8):1336-42. doi: 10.1097/IGC.0b013e31822d0ed0.

Predictors of intensive care unit utilization in gynecologic oncology surgery.

Author information

  • 1Division of Gynecologic Oncology, Department of Obstetrics, Gynecology, and Reproductive Sciences, Helen Diller Family Cancer Center, University of California, CA, USA.



The objectives of the study were to examine factors predicting intensive care unit (ICU) admission after surgery for gynecologic cancer and to determine the impact of ICU admission on survival.


This was a retrospective study of women undergoing laparotomy for staging and debulking of gynecologic cancer at an academic hospital with tertiary ICU facilities from July 2000 through June 2003. Data on clinicopathologic factors, comorbidities, operative outcomes, and survival were obtained from medical records and institutional cancer registry. The χ analysis, Kaplan-Meier analysis, and Cox regression methods were used for analyses.


Two hundred fifty-five patients met our inclusion criteria, 43 of whom had a postoperative admission to the ICU. Factors predicting ICU admission on univariate analysis included age 60 years or older, hematocrit of 30% or less, albumin of 3.5 g/dL or less, and Charlson Comorbidity Index (CCI) score greater than 8; after multivariate analysis, CCI score of greater than 8 (hazard ratio, 2.5; confidence interval, 1.11-5.69) and albumin of 3.5 or less (hazard ratio, 3.8; confidence interval, 1.66-8.85) were associated with an increased risk of ICU admission. After adjusting for albumin and CCI score, ICU admission did not have a significant effect on survival.


The ability to predict ICU admission helps in appropriate counseling of patients and identification of institutional resource utilization.

[PubMed - indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Full text links

    Icon for Lippincott Williams & Wilkins
    Loading ...
    Write to the Help Desk