Study design: Retrospective analysis of the prospectively collected American College of Surgeons National Surgical Quality Improvement database.
Objective: We assessed whether preoperative cigarette smoking and smoking duration predicted adverse, early, perioperative outcomes in patients undergoing elective spine surgery.
Summary of background data: Prior studies have assessed the association of smoking and long-term outcomes for a number of spine surgery procedures, with conflicting findings. The association between smoking and 30-day outcomes for spine surgery is unknown.
Methods: A total 14,500 adults, classified as current (N = 3914), prior (N = 2057), and never smokers. Using propensity scores, current and prior smokers were matched to never smokers. Logistic regression was used to predict adverse postoperative outcomes. The relationship between pack-years and adverse outcomes was tested. Sensitivity analyses were conducted limiting the study sample to patients who underwent spine fusion (N = 4663), and using patient subgroups by procedure.
Results: In unadjusted analyses, prior smokers were significantly more likely to have prolonged hospitalization (1.2, 95% confidence interval [CI]: 1.1-1.3) and major complications (1.3, 95% CI: 1.1-1.6) compared with never smokers. No association was found between smoking status and adverse outcomes in adjusted, matched patient models. Current smokers with more than 60 pack-years were more likely to die within 30 days of surgery (3.0, 95% CI, 1.1-7.8), compared with never smokers. Sensitivity analyses confirmed these findings.
Conclusion: The large National Surgical Quality Improvement population was carefully matched for a wide range of baseline comorbidities, including 29 variables previously suggested to influence perioperative outcomes. Although previous studies conducted in subgroups of spine surgery patients have suggested a deleterious effect for smoking on long-term outcomes in patients undergoing spine surgery, our analysis did not find smoking to be associated with early (30 d) perioperative morbidity or mortality.