Abstract
Introduction: As an example of possibilities of Spine Tango we extracted data on dural tears, one of the most frequent types of complications in posterior spinal fusion. Little is known about their predictors. This study examined which factors predict the occurrence of dural tears in posterior spinal fusion.
Methods: Prospective consecutive documentation of hospital based interventions with an evidence level 2++. Between 05/2005 and 11/2006 data of 3437 patients were documented in the registry. 929 patients, who had been treated with posterior spinal fusion after opening of the spinal canal, were included in this study. Median age was 62.7 yrs (min 12.5, max 90.5 yrs) with a female to male ratio of 2:1. In 18 of 929 cases a dural tear occurred. Dural tears being the most frequent type of complications in the registry were chosen as dependent outcome variable (3–6). Multiple linear regression with stepwise elimination was performed on potential predictor-variables of the occurrence of dural tears. Benchmarking compared the performance of single hospitals with international peers.
Results: Hospital (p=0.02) and number of segments of fusion (p=0.018) were found to be predictors of the occurrence of dural tears in posterior spinal fusion. Number of fusions per hospital (min 25, max 526) and academic status of hospital had no influence on the rate of dural tears. Fusions of four and more segments showed an increase of the rate of dural tears by a factor of three compared to fusions of less than four segments. There was no significant difference between fusions of one segment and fusions of two or three segments (1.3 vs. 1.9%) as well as between fusions of four or five segments and fusions of more than five segments (4.6 vs. 4.2%). Differences between hospitals remained when benchmarking dural lesions with case mix.
Discussion: The feasibility of data analysis and benchmarking from the International Spine Registry Spine Tango could be demonstrated. Predictors of dural tears in posterior spinal fusion are
-
hospital and
-
number of segments of fusion.
In fusions of four and more segments a threefold higher risk of dural tears in comparison to fusions of less than four segments should be taken into consideration. A subgroup analysis on the predictor-variable hospital should be performed assessing further covariates. However, this goes beyond the possibilities of documentation in this international spine registry.
Correspondence should be addressed to Dr Owen Williamson, Editorial Secretary, Spine Society of Australia, 25 Erin Street, Richmond, Victoria 3121, Australia.
