#509 Blood Loss in Spine Tumour Surgery - How Much Can the Surgeon Expect?

General Session: Cutting Edge Technology - Imaging Guidance

Presented by: K.H.Z. Chua


Y. Chen (1)
K.H.Z. Chua (1)
R. Goy (2)
R.W.M. Lam (1)
N.S. Kumar (1)

(1) National University Health System, Department of Orthopaedic Surgery, Singapore, Singapore
(2) National University Health System, Department of Anaesthesia, Singapore, Singapore


Study Design: Meta-analysis.

Background: The vertebral column is the commonest site of bony metastasis, accounting for 18,000 new cases in North America yearly. Spinal metastasis often represents an advanced stage in the patient's primary disease. Such patients are often elderly, have compromised cardiovascular status, poor physiological reserve and altered immune status, all of which render them more susceptible to the complications of intraoperative blood loss and associated transfusion.

Currently no consensus exists regarding the expected volume of blood lost during metastatic spine tumour surgery with various papers quoting anywhere between 1L to 6L. Knowledge of the expected blood loss prior to surgery however is important as it facilitates pre-operative planning, informed consent for surgery, intra- and post-operative management of fluid balance and blood transfusion.

We conducted a meta-analysis of published literature on spine tumour surgery to answer the following research question: “What is the expected blood loss in major spinal tumour surgery for metastatic spinal disease?”

Materials and Methods: A comprehensive online search of the English literature using Medline, Embase, the Cochrane Central Register of Controlled Trials and the Web of Science was performed.

We included articles published in the English language from 31 January 1992 until 31 January 2012. This initial online search yielded 98 relevant articles. Two senior investigators independently reviewed all abstracts. The full text of articles that were deemed eligible for further consideration obtained and reviewed. Eighty articles were excluded at this stage, largely due to lack of clear blood loss data, leaving 18 eligible articles. A hand search of the reference lists of relevant articles was also performed to identify further articles for analysis, yielding 5 articles. Thorough analysis of the full texts of these 23 eligible articles led to a further exclusion of 5 articles, leaving 18 articles for the final meta-analysis of blood loss data. Disagreements regarding eligibility of articles for analysis were resolved by consensus.

Selected articles for final analysis were independently graded according to the Centre for Evidence-Based Medicine (CEBM) Levels of Evidence.

We evaluated the possibility of publication bias by obtaining a funnel plot (created by plotting the sample size against the effect estimate). The Eger's regression asymmetry test was used to assess the existence of publication bias.

Results: 18 selected articles had a total of 785 patients who had undergone major spine tumour surgery for metastatic spinal disorders. The pooled estimate of the blood loss occurring during spinal tumour surgeries was calculated to be 2180ml (95%CI 1805, 2554).

The risk of publication bias was minimal overall as shown in the funnel plot, with most observations falling within the funnel and forming a symmetrical pattern on either side of the mean.

Conclusions: The expected blood loss of a patient undergoing major surgery for spinal tumour constitutes more than a third of the circulating blood volume in a typical cancer patient with significantly impaired physiological reserve. Moreover, cases of catastrophic blood loss exceeding 5L exist in almost every series evaluated in this paper, with some reaching as much as 17-18L. Blood loss is a significant problem during spine tumour surgery and concerted effort is needed to address it.