Background:
Pseudoarthrosis following spinal fusion is an important cause of pain, neurologic decline, and reoperation.
Methods:
The HCUP State Inpatient Databases were queried in New York, California, Florida, and Washington for adult patients who underwent new spinal fusion between 2009 and 2011. In accordance with the HCUP methods series and analysis guidelines, generalized linear mixed effects models were employed to estimate the odds of experiencing post-operative pseudoarthrosis as a function of multivariable patient characteristics, comorbidities, and surgical approach.
Results:
1,295 of 107,420 (1.2%) patients undergoing cervical fusion developed pseudoarthrosis requiring reoperation. On multivariable analysis, risk factors included posterior (OR = 4.47, 95CI: 3.92-5.10) and combined approaches (OR = 1.77, 95CI: 1.33-2.36), fusion of 9 or more vertebrae (OR = 2.54, 95CI: 1.38-4.68), smoking (OR = 1.19, 95CI: 1.05-1.34), and long-term steroid use (OR = 1.89, 95CI: 1.18-3.00). 2,665 of 148,081 (1.8%) patients undergoing thoracic or lumbar fusion developed pseudoarthrosis. Posterior (OR = 0.58, 95CI: 0.51-0.56) and Combined approaches (OR = 0.46 (0.40-0.54) had reduced rates, while fusion of 4-8 (OR = 1.52, 95CI: 1.39-1.67) and 9 or more vertebrae (OR = 1.87, 95CI: 1.49-2.34), hypertension (OR = 1.18, 95CI: 1.09-1.28), sleep apnea (OR = 1.48, 95CI: 1.26-1.72), smoking (OR = 1.22, 95CI: 1.12-1.33), and long-term steroid use (OR = 1.53, 95CI: 1.08, 2.18) had increased rates.
Conclusions:
This study strongly associates several diagnoses with the development of pseudoarthrosis; however, further prospective study is warranted to establish causation.