Background context:

Surgical site infections (SSIs) after spinal fusion surgery increase healthcare costs, morbidity and mortality. Routine measures of obesity fail to consider site specific fat distribution.

Purpose:

To assess the association between the spine adipose index and deep surgical site infection and determine a threshold value for spine adipose index that can assist in preoperative risk stratification in patients undergoing posterior instrumented lumbar fusion (PILF).

Study design/setting:

Multicenter retrospective case-control study.

Patient sample:

We reviewed patients who underwent PILF from January 1, 2010 to December 31, 2018.

Outcome measures:

All patients developing a deep primary incisional or organ-space SSI within 90 days of surgery as per US Center for Disease Control and Prevention criteria were identified. We gathered potential pre-operative and intra-operative deep infection risk factors for each patient. A 1:1 match was performed using the following criteria: gender, age (± 3 years), ethnicity, date of surgery (± 1 year), and hospital location of surgery. Spine adipose index was measured on pre-operative mid-sagittal cuts of T2 weighted MRI scans. Each measurement was repeated twice by three authors in a blinded fashion, with each series of measurement separated by a period of at least six weeks.

Methods:

Stepwise binary logistic regression analysis was used to assess the association between SAI and the development of deep SSI. Separate logistic regression models were used for body mass index (BMI) and direct measures of subcutaneous fat thickness. Receiver Operating Characteristic (ROC) analysis was used to determine the optimal value for SAI, and subsequent risk ratios were calculated using the identified threshold. Intra- and inter-observer reliabilities were assessed using intra-class coefficients.

Results:

Forty-two patients were included in final analysis, with twenty-one cases and twenty-one matched controls. The spine adipose index was significantly greater in patients developing deep SSI (p =0.029), and this relationship was maintained after adjusting for confounders (p=0.046). Risk of developing deep SSI following PILF surgery was increased 2.0-fold when the spine adipose index was ≥0.51. The spine adipose index had excellent (ICC >0.9; p <0.001) inter- and intra-observer reliabilities.

Conclusion:

The spine adipose index is a novel radiographic measure and an independent risk factor for developing deep SSI, with 0.51 being the ideal threshold value for pre-operative risk stratification in patients undergoing PILF surgery.