Study design:

Retrospective case control study.

Objective:

To determine the influence of obesity on risk factors for adverse outcome after lumbar spine fusion (LSF).

Summary of background data:

Obesity is risk factor for complications after LSF and poses unique challenges regarding optimization of care. Nonetheless, this patient population is not well-studied.

Methods:

Adult patients undergoing LSF were identified the State Inpatient Database. Patients were identified as obese or non-obese using ICD-9 codes. Outcome variables were 90-day readmission, major medical complication, infection, and revision rates. Data were queried for demographics, comorbidities, surgery characteristics, and outcome variables. Logistic multivariate regression was utilized, serially testing interactions between obesity and other independent variables in separate models for each outcome. The Benjamini-Hochberg procedure was used to adjust statistical significance for multiple comparisons.

Results:

262,153 patients were included: 31,062 obese and 231, 091 non-obese. For major complications, obese patients had lower odds ratios (ORs) vs. non-obese patients for CVA, diabetes with chronic complications, age ≥65, CHF, history of MI,, renal disease, chronic pulmonary disease, Medicare/Medicaid payor, >2 levels fused, TLIF/PLIF, and female gender, and higher OR for non-white race. For readmission, obese patients had lower OR for age ≥65, history of MI, renal disease, and mental health disease, and higher OR for female gender. For revision, obese patients had higher OR for female gender and TLIF/PLIF. For infection, obese patients had lower OR for diabetes with and without chronic complications, and higher OR for female gender.

Conclusion:

Many medical comorbidities have less impact in obese patients than non-obese patients in predicting adverse outcomes despite increased rates of adverse outcomes in obese patients. These findings reflect the impact of obesity as an independent risk factor and have important implications for preoperative optimization.

Level of evidence:

3.